Mar

INTRODUCTION

The most recent outbreak of severe acute respiratory syndrome (SARS) has been caused by coronavirus-2 (SARS-CoV-2) – a new single-strand, positive-sense-RNA beta-coronavirus first reported in 2019 in Wuhan, China. The virus has spread to nearly all countries across the world.^1-4

SARS-CoV-2 infection, also known as Coronavirus Disease 2019 (COVID-19), replicates mainly in the upper and lower respiratory tract. The transmission of COVID-19 from symptomatic and asymptomatic patients is usually through respiratory droplets, generated by coughing and sneezing or through contact with contaminated surfaces.^4,5 The disease has an incubation period of approximately 5.2 days.⁶

Most infections are mild and uncomplicated.⁴ After one week of the onset of disease, 5-10% of patients tend to develop pneumonia, needing hospitalisation.^4,6 Some of these patients develop further complications, often leading to death.^4,6 The overall case fatality rate is 1.4%, with a noticeably higher rate after the sixth decade of life.⁴

People aged ≥ 60 years, especially with underlying medical conditions – such as cardiovascular disease, hypertension, diabetes mellitus (DM), chronic respiratory disease, cancer, immunodeficiency, obesity – and those of male-sex, have an increased risk of dying.^4,7-12 Risk of severe adverse outcome is also associated with an increased number of associated co-morbidities.¹⁰

The impact of active cancer, endocrine disorders, autoimmune inflammatory rheumatic diseases etc. on COVID-19 outcomes has been investigated widely.^13-18 Divergent views have emerged regarding the role of renin angiotensin aldosterone system (RAAS) inhibitors, steroids, and immunomodulators in COVID-19 mortality.

The objective of our study was to evaluate the risk posed by epidemiological and demographic variables in our local population. We also sought to analyse the impact of co-morbidities on in-hospital mortality in confirmed COVID-19 patients.

METHODS

Study design:

We conducted a retrospective analysis of demographics characteristics (age and sex) and medical co-morbidities – hypertension, chronic heart failure, ischaemic heart disease, DM, thyroid disorders, asthma, chronic obstructive pulmonary disease (COPD), chronic kidney disease (CKD) (eGFR < 60 mL/min/1.73 m2), chronic liver disease, active malignancy, immunosuppression, post-transplant status, chronic inflammatory arthritis and other rheumatic disorders – in all patients with confirmed COVID-19, who were admitted in two peripheral district general hospitals under a single National Health Service (NHS) trust serving primarily the rural population of western England.

Inclusion and Exclusion Criteria:

To determine COVID-19 status, nose and throat-swab specimens were obtained for real-time reverse transcription polymerase chain reactions (rt-PCR) in all adult (≥18 years) patients, attending one of the two district general hospitals (Royal Shrewsbury Hospital, Shrewsbury; and Princess Royal Hospital, Telford) under Shrewsbury & Telford Hospitals NHS Trust (SaTH) in the period from 1^st March to 15^th May 2020.

Patients who tested positive (either by N gene and ORF1ab gene positive / ORF1ab gene positive or N gene positive) and required subsequent in-hospital management were included in the study. Patients who were discharged after initial senior review (usually by a consultant physician), or brought in as a cardiac-respiratory arrest, were excluded. Re-admissions to the hospital beyond 48 hours following hospital discharge due to COVID-19 were excluded from the study. Patients diagnosed solely on radiological or clinical findings without a positive rt-PCR test were not included in our study.

We analysed the data based on the index-admission (including failed-discharge: re-admission within 48 hours following hospital discharge). No follow-up data was collected post-hospital discharge of these patients.

Data collection & analysis:

A list of all confirmed COVID-19 patients over a 76-day period was identified from the trust microbiology database. A search of the electronic patient records was completed by four members of our team. Supplementary data was gleaned from existing hospital paper records. Patient demographics, presenting symptoms, associated co-morbidities, medications, admission and discharge dates, intensive therapy unit (ITU) admissions, renal profile, referral source and outcomes were recorded in the specifically designed electronic datasheet.

Study Outcome:

The impact of epidemiological and demographic characteristics, and pre-existing medical conditions on the mortality of confirmed COVID-19 patients requiring in-hospital treatment was analysed.

RESULTS

A total of 303 confirmed COVID-19 (rt-PCR positive) samples were collected over a 76-day period. Five patients had been tested twice, and this was accounted for. Thirty-five patients were excluded from the study: twenty-four of them discharged after initial senior review without requiring in-hospital treatment, seven brought in with cardio-pulmonary resuscitation (CPR) in progress, three had inadequate data, and one was <18 years old. Of the 263 patients admitted, 70 (26.6%) died in hospital (Figure-1).

Figure-1: Flowchart of sampling and analysis

We stratified the mortality rates among the admitted patients by age (Table-1). A chi-square test of independence revealed that the mortality rate was significantly related to an advanced age (χ² =27.078, p<0.001). The age and sex distributions of admissions and mortality are shown in Figure-2 (a, b, c).

Table-1: Medical admissions and mortality stratified by age

N: number of patients, m: male, f: female.

Figure-2 (a,b,c): Age, Sex, Admission and Mortality pyramids

We considered two age cohorts - below 60 and ≥60 years of age and other relevant demographic parameters (sex and residence in own-home/care-home) to analyse the impact on mortality rates (Table-2). Of the admitted patients, 159 (60.5%) were male, and 104 (39.5%) were female. The mortality rate was strongly associated with advanced age ≥60 years (χ² =17.120, p<0.001) but independent of sex distribution (χ² =1.784, p=0.182). However, it was also affected by the care facility (χ² =18.146, p<0.001) with a higher mortality rate among the group of patients with residence in a long-term care-home.

Table-2: Admission and Mortality stratified by demographic variables

N: Number of patients; Care-home: Long-term care in residential or nursing home.

To identify the strength of the associations, we conducted a univariate logistic regression analysis with mortality as the dependent variable and the demography and presence/absence of the co-morbidities as the independent variable (Table-3). We found that age as a continuous predictor had an odds ratio of 1.058 (p<0.001), which translated to increased odds of dying by 5.8% for every year of advanced age. Using age as a categorical predictor with the other two categories, the odds of death for patients aged below 60 years was found to be 0.195 times the odds of death for the patients aged 60 years or above.

Table-3: Univariate logistic regression analysis of the demographic variables and co-morbidities

^*Cardiovascular: Hypertension, chronic heart failure, ischaemic heart disease, chronic arrhythmias; ^§Respiratory: Asthma, chronic obstructive pulmonary disease, interstitial lung disease, asbestosis; ^$DM & Endocrine: Diabetes Mellitus (DM), thyroid disorders, Addison's; ^£Renal: Chronic Kidney Disease (eGFR < 60 mL/min/1.73 m2), on dialysis, chronic nephritis, renal transplant; ^§§Rheumatic: Rheumatoid arthritis, inflammatory Polyarthritis, chronic vasculitis Syndrome, connective tissue disorders; ^±Liver & hepato-biliary: Fatty liver, non-alcoholic steatohepatitis, liver transplant; ^€Thyroid: Hypothyroidism, hyperthyroidism, post-thyroidectomy; ^€€Long-term oral steroids: Prednisolone, hydrocortisone, dexamethasone; ^$$Immunomodulators: Methotrexate, tacrolimus, sirolimus, mycophenolate, dapsone, sulfasalazine and azathioprine.

Based on the Charlson Comorbidity Index (CCI) score, the severity of co-morbidities was categorised into four cohorts: mild/no co-morbidity (CCI:0), moderate (CCI:1-2), severe (CCI:3-4), and very severe (CCI≥5) [Table-4(4a)].

Table-4: Impact of CCI score and specific medical-conditions on admission and mortality
4a) Admission and mortality stratified by CCI score based cohorts

4b) Admission and mortality stratified by specific medical-conditions

N: Number of patients; DM: Diabetes Mellitus; *RAAS-inhibitors; ^§Non RAAS-inhibitors.

The impact of CCI score-based cohorts on mortality are shown in Figure-3 (a-f). CCI value also predicted significant association with odds ratio 1.255 (p<0.001). If the CCI score was utilised as a categorical predictor with the other two parameters (age and place of primary care), it remained a significant predictor with the odds of death for the patients with CCI-scores between 0-4 turning out to be 44.8% (p=0.005) of the odds of death for the patients with CCI scores ≥5 (Table-3).

Figure-3(a - f): Pie-chart representing impact of CCI score-based cohorts on mortality


a) Overall admitted patients: discharge and mortality; b) CCI score 0: discharge and mortality; c) CCI score 1-2: discharge and mortality; d) CCI score 3-4: discharge and mortality; e) CCI score ≤4: discharge and mortality; f) CCI score ≥5: discharge and mortality.

Interestingly, the eGFR at presentation turned out to be a significant predictor of mortality (OR=0.961, p<0.001). Of the co-morbidities, pre-existing renal disease was found to be an important predictor of mortality with OR=1.996 (p=0.027). Long-term oral steroids were another significant predictor of mortality, with the odds of death for the patients with long-term oral steroids use being 341.2% (p=0.016) of the odds of death for the patients without such medication. Patients with no background medical conditions (OR=0.181, p=0.022) fared better, with significantly lower odds of death compared to patients with at least one known medical condition (Table-3).

We also analysed the mortality of our patients with specific medical condition-based cohorts [Table-4(4b)]. A high mortality of 52.9% [OR (95%CI): 4.053(1.091–15.063), p=0.037] was observed in patients who were on long-term oral steroids. A 33.3% [OR (95%CI):1.510(0.791–2.883), p=0.212] mortality rate was observed among in-patients with known diabetes on pharmacotherapy.

Many of the demographic variables and the co-morbidities were inter-related – the odds of death for a patient coming from their own-home was only 26% (OR=0.263, p<0.001) of the odds for those residing in a long-term care-home (Table-3). To offset the possibility of any confounding effect, we utilised multiple logistic regression analysis with all the important variables taken together (Table-5). Taking consideration of confounding effects, only age, care facility, presence of active malignancy and long-term oral steroids were found to be significant predictors of mortality. Interestingly, the presence of active malignancy was found to have a lower risk of death – this is possibly due to a bias on account of a relatively small number of patients in that subset of our study. Age was the most significant predictor of mortality, followed by a primary area of the care facility and the presence of active malignancy.

Table-5: Multiple logistic regression analysis of the demographic variables and co-morbidities

DM: Diabetes Mellitus

DISCUSSION

COVID-19 has taken 800,000 lives world-wide as reported by the World Health Organisation (WHO) on August 30, 2020. A recent systematic review and meta-analysis have reported the association of COVID-19 with a severe disease course in about 23% of infected patients and has a mortality of about 6%.¹⁹ The mortality rate varies in different geographical areas. In-hospital mortality was significantly higher in the United States of America (USA) (22.23%) and Europe (22.9%) compared to Asia (12.65%) – (p<0.0001).²⁰ However, there was no significant difference when compared to each other (p=0.49).²⁰ Our study showed a 26.6% in-hospital mortality.

The mean age of the patients in our study was 68.74 years (SD:16.89) – 60.5% of them were male and 39.5% female. 70.7% of these patients were aged ≥60 years. Univariate analysis showed that the mortality rate was significantly age-dependent (OR=1.058, p<0.001) – mortality (33.9%) was higher in patients aged ≥60 years, rising sharply ≥80 years to 44.0% (χ² =27.078, p<0.001). Our results were consistent with other studies.²¹

Among the demographic characteristics, mortality-risk was independent of sex distribution (χ² =1.784, p=0.182) in our study. This is in contrast to a meta-analysis, which reported the association between male-sex and COVID-19 mortality (OR =1.81; 95%CI:1.25–2.62).²² Multicentric studies in the United Kingdom (UK) would be warranted to see the trend in the local population.

Long-term care-home residents suffered 50.0% mortality (χ² =18.146, p<0.001). The London School of Economics report on May 14, 2020, estimated that the COVID-19 related deaths of care-home residents contributed to 54% of all excess deaths in England and Wales. Our study findings indicate long-term care-homes as hot-spots requiring shielding and protective measures against COVID-19 – a conclusion corroborating other studies.²³

We aimed to define the predictive-role of co-morbidities on COVID-19 mortality, an aspect that has been probed earlier as well.^7-12 The CCI score remains a reliable method to measure co-morbidity.²⁴ For admission to intensive care, NICE recommended CCI-score ≥ 5 requires critical care advice to help in treatment decision regarding the essential benefit of organ support for seriously unwell COVID-19 patients. We examined the predictive mortality-risk of CCI scores among the admitted patients.

The mortality rate in cohorts with CCI ≤4 and CCI scores ≥5 were 20.3% and 36.2% respectively. The odds of death for CCI ≤4 cohort was less than half (44.8%) compared with CCI scores ≥5 cohort. Based on this finding, we strongly recommend CCI scoring as a clinical risk-stratification tool in COVID-19.

We examined the impact of organ specific co-morbidities on in-hospital mortality in our study as well. Patients with no background medical conditions showed a low mortality rate 6.9% [OR (95%CI): 0.181(0.042–0.782), p=0.022] and had better outcomes with significantly lower odds of death, compared to patients with at least one medical condition on univariate logistic regression analysis (Table-3). The mortality rate was 3.2% in CCI-0 cohort [Table 4(4a)].

The impact of COVID-19 on patients with CKD, glomerulo-nephropathies, on dialysis dependent patients and post renal transplant patients remains unclear. Patients with SARS-CoV-2 infection were frequently found to have renal dysfunction – the latter was associated with greater complications and in-hospital mortality.²⁵ A mortality rate of 3.6%, was reported in patients attending an outpatient haemodialysis centre.²⁶ Another study has concluded 3.07-fold (95%CI:1.43–6.61)mortality among renal failure patients.²⁷ We found, the pre-existing renal disease to be a cause of significant concern with 37.7% mortality [OR(95%CI): 1.996(1.082 – 3.681), p=0.027] with the eGFR at presentation being a significant predictor (OR=0.961, p <0.001) (Table-3).

The use of steroids in COVID-19 continues to be explored.The RECOVERY trial in UK, after evaluation at 28 days, concluded that dexamethasone reduced deaths by one-third in ventilated patients [age-adjusted rate ratio (RR) 0.65; 95% CI: 0.48–0.88; p=0.0003], and by one-fifth in other patients receiving supplemental oxygen with or without non-invasive ventilation (RR 0.80; 95%CI: 0.67 to 0.96; p=0.0021), although no benefit was observed in mild or moderate cases not requiring oxygen support (17.0% vs.13.2%; RR 1.22; 95% CI, 0.93e1.61; p¼0.14). In contrast, a systematic review concluded that the results from retrospective studies are heterogeneous, and it was difficult to assign a definite protective role of corticosteroids in this setting.²⁸ We found long-term oral steroids use to be a significant predictor of mortality – 52.9% [OR(95%CI): 3.412(1.261–9.23), p=0.016] – this was 341.2% of the odds of death for the patients without any long-term oral steroids use (Table-3). The sample size of this cohort was relatively small with 9 deaths out of 17 patients. However, based on our results, it may be safe to suggest that further population-based studies would be required to determine the impact of long-term oral corticosteroid use in COVID-19.

A major proportion of endocrine disorders are of autoimmune aetiology. The impact of thyroid disorders on COVID-19 is yet to be studied widely.^15,16 We found no increased risk of mortality [OR (95%CI): 0.914 (0.285–2.934), p=0.880] in patients with thyroid disorders. However, 33.33% [OR(95%CI): 1.510(0.791–2.883), p=0.212] mortality was seen among the diabetic patients on pharmacotherapy in our study [Table-4(4b)].

Pre-existing hypertension is an accepted risk factor for COVID-19 mortality.^26,27 However, the role RAAS-inhibitors and upregulation of ACE-2 receptors in COVID-19 mortality call for targeted clinical research for further clarification.²⁹ A meta-analysis of four studies showed that patients treated with RAAS-inhibitors had a lower risk of mortality [RR: 0.65(95%CI:0.45–0.94), P=0.20].³⁰ We did not observe any significant mortality-risk difference between RAAS-inhibitors treatment group [OR(95%CI): 0.760(0.365–1.586), p=0.465] and non RAAS-inhibitor treatment groups [OR(95%CI): 0.704(0.253–1.964), p=0.503] [Table-4(4b)]. We recommend the continuation of RAAS-inhibitors during COVID-19 unless there exist other compelling medical reasons for their discontinuation.

A prospective study in the UK concluded that the mortality from COVID-19 in cancer patients appeared to be driven principally by age, gender, and co-morbidities.¹³ The study could not identify evidence suggesting cancer patients on cytotoxic chemotherapy, or other anticancer treatment, were at an increased risk of mortality from COVID-19 compared to the general population.¹³ We also did not detect any increased risk of mortality in patients with active malignancy [OR(95%CI): 0.078(0.008–0.725), p=0.025)] (Table-5).

The impact of various non-specific immunomodulators in COVID-19 outcome remains inconclusive.¹⁴ Our study did not reveal any significant predictive mortality-risk with the use of long-term immunomodulators (methotrexate, tacrolimus, sirolimus, mycophenolate, dapsone, sulfasalazine and azathioprine) on multiple logistic regression analysis. We reached the same conclusion with patients suffering from chronic rheumatic disorders on similar analysis (Table-5).

Our study had some unique characteristics. We analysed all the eligible samples over a consecutive 76-day period at the initial peak of the pandemic. The study was conducted across two district general hospitals, allowing an insight into two differently located rural populations. We conducted univariate and multiple logistic regression analysis of the demographic variables and co-morbidities to examine the predictive-risk of contributing factors in COVID-19 mortality. The association between CCI scores and in-hospital mortality was also analysed in detail. We included demographic characteristics such as age, sex and residence in a long-term care-home while factoring in the associations.

Our study was not without limitations, though. We were unable to study the predictive-risk of obesity, socioeconomic status and ethnicity due to inadequate data. The “White British” group consisted of 80.61% of admitted patients, and no ethnicity was documented in 17.11% of our patients (Table-6, Figure-4).

Table-6: Medical admissions and mortality stratified by ethnicity

N: Number of patients

Figure-4: Bar charts showing Admission and Mortality stratified by Ethnicity

We relied solely on electronic database and hospital records to conduct the study retrospectively. The few subsets of patients such as those on prescribed long-term oral steroids, immunomodulators, thyroid disorders, chronic liver disease, and active malignancy had relatively small sample sizes with possible introduction of bias. We did not categorise diabetic patients into insulin dependent/non-insulin dependent or well/poorly glycaemic control cohorts. We did not aim to split the respiratory group into well or poorly controlled asthma or COPD subsets. Patients on a long-term steroid inhalation treatment were not included in the steroid cohort – a more extensive population-based study may be better suited for such an analysis.

CONCLUSIONS

Patients aged ≥ 60 years, residence in a long-term care-home, pre-existing renal disease, multiple co-morbidities (especially those with CCI ≥ 5), and patients on long-term oral steroids need to be considered as having a high risk of dying from COVID-19, along with other established risk factors such as hypertension, diabetes and chronic respiratory disease. RAAS-inhibitors need not be discontinued due to COVID-19. Further studies are necessary to establish links between long-term oral steroids use, chronic rheumatic disease, non-specific immunomodulators and COVID-19 mortality.

Introduction

The creative process is an enigma; there are conflicting opinions about creativity and creative people. Research studies on creativity have produced contradictory results. The long-standing belief that creativity results from a strange clairvoyant state is still occasionally associated with psychiatric disorders. ¹ Although a decline in creativity with aging indicates that it is biologically based, a relationship between creativity and psychopathology is overstated in both print and media. Reductionism tends to misconstrue creativity as a product of psychopathology. Nonetheless, whilst psychopathology can facilitate creativity, it does not produce creativity. The inspirational characteristics of creativity remain shrouded in mystery.

Methodological issues that include both a definition and an evaluation of creativity impede the research into creativity. These challenges make the correlations between the studies problematic, and they deliver opposing outcomes. Although there is no confirmed relationship between psychopathology and creative accomplishment, the search for such a relationship hinders our understanding of human potential and the deeper levels of consciousness. Early detection of creative talents in children might enable providing them with special guidance, thereby averting potential psychiatric problems.

The superficial reductionism of 20th century biological psychiatry compressed all mental phenomena, including creativity, into compact neurobiological compartments, and the only way to achieve this was to medicalise it. Any assumed correlations between creativity and mental disorders will be clarified only when we gain a greater understanding of the creative process. In cases where creativity and mental illness indeed coexist, a psychiatric understanding of creativity may provide insights into patient functioning and assist in defining both normalcy and psychopathology.

Whilst human beings have existed on this planet for millions of years, the technological advancements of the last few centuries transpired without any perceptible changes in the development of the human brain. Historically, our ancestors were drawing two-dimensional pictures until just a few centuries ago. No one has hitherto been able to explain this sudden burst of creativity. An expanded model of brain-mind-consciousness that can appreciate the wonder of creativity is needed.

Defining Creativity

Researchers have long been interested in a potential connection between creativity and mental illness. The major challenge here is to define creativity and establish measurable indicators. Creativity has been described as the process of bringing something new into existence. It involves the capacity to take unrelated structures and combine them harmoniously in different ways for new purposes. The creative mind is alert to unexpected connections. An individual with a rich reservoir of knowledge is regarded as intelligent, whereas an individual who uses that knowledge in an original and constructive way is considered creative. ² The creative process is not fully understood; some even feel that a precise definition is unattainable.³ Nonetheless, creativity can be described as the process of bringing something new into being where the outcome is larger than the input received by the creative mind. ^4,5 Creative individuals are sensitive to gaps in human knowledge and these voids act as catalysts in their search for solutions. This is the highest form of human adaptation; whether to a greater or a lesser degree, it may exist in all people.

The creative process can be likened to a four-stage computer process. ⁶ If information processing and storage is the primary process, the second stage is the incubation or pondering phase, during which ideas germinate at a subconscious level. The third phase involves illumination, or flashes of insight, and the fourth is the period of elaboration during which the new idea is developed and tested. These stages can be additionally likened to the biological rhythm of conception, gestation, birth and infancy. This pattern is not strict. As a rule, the process of illumination is gradual with countless small bursts of insight, such as with Charles Darwin’s elaborations on his theory of evolution.

Dream processes shed some light on creativity. As with poetry, dreams are replete with visual and highly idiosyncratic metaphors. Dreams are the art of the unconscious; whilst dreaming, we tap into a creative source. The dreaming psyche has seemingly unlimited creative potential. An anecdote about Kekule, the chemist, recounts that he conceived the benzene ring after a dream in which he saw a serpent biting its tail.

The Creative Personality

Creative people must be assessed on an individual basis. Not all persons of superior intelligence are creative, and not all creative people have superior intelligence. Although creative potential is dependent on intelligence, actual creative achievement is independent of intelligence (e.g. one does not have to be tall to be a successful basketball player). Highly intelligent people are prone to self-criticism, which has an inhibiting effect on the development of creativity. A combination of high intelligence and special aptitudes appears to promote creativity. Unconventionality, egocentrism, flexibility, tolerance for ambiguity and a preference for complexity are among the attributes of creative individuals. ⁷ Psychological testing has shown that creative individuals are frequently more emotionally troubled than are non-creative individuals; however, they also have more ego strength for dealing with problems. Their personal qualities include imagination, persistence, perseverance, dedication and stamina. Creative children tend to be egotistic and gullible. This egotism provides them with the confidence to believe that they are capable of unique achievements, whilst momentary gullibility enables them to break through scepticism and into creativity.

McClelland illuminated a controversial notion when he described the creative individual as one who is characterised by competition, either with an external standard of excellence or with his or her own internal aspirations. ⁸ Driving absorption, the ability to ignore failure and adversity and tremendous curiosity are noted as a predictive set of personality traits.⁹Although creative individuals are difficult to live with, whether their creativity flourishes or not frequently depends on the support that they receive from others. ¹⁰ Among the characteristics of creative people, Tarlaci (2014) included openness to experimentation and change, rebelliousness, individuality, sensitivity, playfulness, self-assertiveness, curiosity and simplicity. ¹¹

Although there is a compulsion for order, symbolisation and communication are at the core of creativity. Intelligence, domain-specific knowledge/expertise, motivation and adaptive traits such as openness, broad interests and self-confidence are closely associated with creativity (Feist 1999). ¹² Despite the fact that these characteristics of creative people are obviously independent of psychopathology, they point towards better mental health. Research on creativity in neuroscience has revealed that creativity is associated with ‘ordinary’ rather than psychopathological brain processes.¹³

Psychopathology

Since the time of Plato, philosophers have debated a conceivable connection between creativity and psychopathology. He proposed a logical paradox when he stated that a poet does not know what he is going to write, and yet he cannot produce a poem if he has no picture of what he describes. As a Greek philosopher, Plato was a reincarnationist. He obviously solved his own riddle by attributing hidden creative knowledge to remembrances of a previous life and to springing from ‘Divine madness’. Aristotle noted the predisposition of great artists and poets to melancholia, but he perceived creativity as a rational process. Shakespeare repeated the older perspective through one of his characters who states, ‘the lunatic, the lover, and the poet are of imagination, all compact’. During the 20th century, systematic investigations into this relationship were unable to either support or refute this association. Cesare Lombroso failed to clarify this confusion in his book, Genius and Mental Illness. Nonetheless, his influence led to speculation that genius is an ‘ancestral gift’ transmitted in families in the same manner as mental disorders.

Recent empirical research has shown that creative individuals have a higher tendency towards psychopathology than those in non-creative professions. This propensity is expressed in personality traits, behaviours and experiences similar to those identified in clinically ill patients (Jamison 1989). The evidence has not clarified whether the psychopathology linked to creativity relates more closely to features of schizophrenia or affective disorders. Countless novelists and dramatists have family histories of psychiatric disorders. Severe personality deviations have been observed among visual artists and writers and possibly among thinkers and scholars as well. Jamison noticed mood disorders among writers and artists. ¹⁴

Bipolar disorder may be more frequent among creative individuals than in the normal population. One study reported a higher incidence of depression and bipolar disorder among creative people, and especially among writers.¹⁵Another study noted a higher incidence of depression and alcoholism among writers and artists. Following recent epidemiological studies with large samples, Kyaga et al. (2013) argued in favour of an association between professional authors and psychiatric disorders. ¹⁶ They illuminated familial associations between the creative professions and schizophrenia, bipolar disorder, anorexia nervosa and possibly autism. ¹⁶They noted that this association was more evident in cases of self-employed artists and less so in scientific creativity, where the subjects had passed through several professional screening procedures.

In another epidemiological study, Parnas et al. (2019) found that the relatives of academics have a significantly increased risk of suffering from schizophrenia or bipolar disorder. ¹⁷ In another study, they suggested that ‘creativity and an increased risk for mental disorders seem to be linked by a shared vulnerability that is not manifested by clinical mental disorders in the academics.’ ¹⁸ The literature has made significant connections between bipolar disorder and creative accomplishment, with much of the thinking inspired by biographical accounts of poets and musicians who presented with signs of bipolar disorder. ¹⁹ Studies by Burkhardt et al. (2018) suggest that, in persons at-risk for bipolar disorder, their mood swings are strongly associated with creativity, but whilst there is evidence of increased creativity, there is no evidence of higher creative achievement. ²⁰

Observations of the bipolar mood domain identify a high prevalence of changes in intuition, empathy, appreciation of danger and predictive capacity. However, these changes do not necessarily include supra-sensory changes in the primary senses of smell, taste, vision, touch or hearing. Parker et al. (2018) suggested that clinicians should be aware of non-psychotic, supra-sensory phenomena in patients with bipolar disorder and that the identification of such features could explain the increased creativity evident in those with a bipolar condition. ²¹

After examining the life of Charles Dickens, Longworth and Carlson (2018) maintained that there was very little historical evidence for the suggestion that he experienced bipolar disorder. ²² However, they did suggest that he displayed characteristic bipolar symptoms. They also maintained that his childhood was an outstanding example of personal resilience and that his own story was just as fascinating, if not even more intriguing, than any of those that he had created. Their investigations concluded that Dickens’ story confirmed the connection between writers, creativity and mood disorders. Retrospective psychiatric assessment of historical figures and the slotting of these celebrities into biological compartments may be risky. Biographical studies of creative people are criticised for having possible recall, interviewer, selection and cultural sampling bias. ²³

The suicide rate is high among artists, and this has been linked to manic depression. Adverse financial circumstances and disappointments due to the rejection of their artistic productions are sufficient to explain this apparently high rate. In contrast, musicians have a low suicide rate, very likely reflecting the healing effect of music. In addition to alcohol, opium has been a historical favourite addictive drug of writers, of which Charles Dickens is an example; opium addiction was partially responsible for his death. ²⁴ Ludwig’s study on 1000 outstanding individuals found an upsurge in alcohol abuse in artists, especially writers. ²⁵ Post (1994) found a similar result among prose writers and playwrights. ²⁶ Although Ernest Hemingway, the Nobel Prize winner for literature, may be a good example of this phenomenon, he committed suicide later in his life. Creative individuals may be notorious for their alcohol and drug misuse; however, it is not clear whether drug induced psychopathology promotes their creative expression. Whilst it is possible that the disinhibiting influence of mild psychopathology and the judicious use of alcohol or drugs could facilitate creativity, this phenomenon has potentially contributed to the confusion in which psychopathology is described as the ‘producer’ of creativity.

Absence of Psychopathology

Alongside these studies, other reports glorify the mental health of geniuses and eminent individuals. The Stanford 35-year follow-up study of over 1000 geniuses, the MacKinnon study of creativity in architects and Havelock Ellis’s psycho-biographical study of eminent men all emphasised the absence of psychopathology among these creative individuals. ²⁷

In an investigation on the prevalence of psychopathology, in a sample of 291 famous men, Post (1994) noted that they all excelled by virtue of their abilities, originality, drive, perseverance, industry and meticulousness. ²⁶Even though most of them had unusual personality characteristics and minor neurotic abnormalities, all of the subjects in this study were emotionally warm, with a gift for friendship and sociability. Post additionally noted that, among creative individuals, scientists show the fewest psychological abnormalities. Functional psychoses are less frequent than epidemiology would suggest. Depressive conditions, alcoholism and possible psychosexual problems are more prevalent than expected in some professional categories, particularly among writers. Hare (1987) noted that banning stimulant drugs in sports did not lower the achievements significantly, and that the same should be true of creativity. Poetic vision has been equated with psychedelic experiences. ²⁸ Creative activity has been observed to be at its highest level in patients who are moderately ill, and at its lowest level in groups identified as severely ill. ²⁹

Although there is no significant difference in the incidence of psychotic illness among males and females, there is less creativity among the latter. If the hypothetical connection between creativity and psychopathology were valid, the incidence of creativity should be proportional to gender. Historically, unfavourable social pressures and opposing cultural factors have represented major explanations for the lower incidence of creativity among women. This disparity points towards the fact that creativity has to be nurtured and is not automatically generated by psychopathology. Despite an equal incidence of mental illness in men and women, there have been few female geniuses in any culture; this challenge the probability of a clear connection between psychopathology and creativity. The same argument may be used against a pure biological view of creativity; both men and women have the same biological make up, yet fewer geniuses have been identified among female population.

Psychodynamic Perspectives

Psychoanalysts have postulated dynamic psychopathologies for the creative process. Analysts incline towards seeing artists as neurotics and their productions as sublimations of sexuality and regression in the service of the ego. ³⁰ They consider the motives for creative activity as impulses that compensate for dissatisfaction and as defences against depression. Some perspectives differ from traditional psychoanalytical ideas, emphasise the crucial role of synthetic ego operations and draw distinctions between psychopathology and creativity. ³¹Analysts suggest that novel ideas exist in the subterranean regions of the mind. Whilst the conscious mind has no access to these hidden areas in the normal state, it is easier for a disturbed mind to tap information from the unconscious or preconscious. ³² Sims suggests that the psychotic and the creative states are subjectively indistinguishable and that delusions arrive in the minds of the mad in the same manner that ideas drop into the minds of the creative.³³In contrast,Slater and Meyer report only minor psychiatric disorders among creative people. ³⁴Although it would appear that psychopathology does not preclude creative activity, it may release it. In general, the creative person enjoys conflict free intimacy with the preconscious and is a model of psychological health. ³⁵

Orderly Mind

The neurobiological model of schizophrenia suggests that a deficit in the systems involved in information-processing could contribute to its symptomatology. ³⁶It has been hypothesised that such a deficit could favour the creative association between information units.³⁷Psychopathology linked creativity has even been associated with abstract disciplines such as mathematics. If these views were accepted, creativity and schizophrenia would be separated only by a ’neurological difference’. Andreasen challenged the hypothesis of a connection between creativity and schizophrenia.³⁸He argued that the bizarre nature of schizophrenic experiences is far from original, and that the cognitive impairment of such patients inhibits their creativity.

The creative intelligent person experiences an attention surplus, whereas a schizophrenic patient suffers from an attention deficit. As a case in point, a creative child may figure out in two seconds what the teacher is going to say, after which he may be looking around, waiting for the teacher to finish and appearing as if he is not paying attention. In contrast, because of a failure in the normal filtering of stimuli, schizophrenics tend to make unusual associations that result from over-inclusive thinking in which countless disconnected elements are included in their reasoning. ³⁹ Although higher cognitive individuals also demonstrate ‘pseudo over-inclusive thinking’, this is due to their capacity to conceive and utilise two or more contradictory concepts simultaneously.⁴⁰

Bleuler (1950) described intellectual ambivalence as both characteristic of schizophrenia and as superficially similar to the janusian process of oppositional thinking that involves conceiving of two or more opposites simultaneously. ⁴¹ The Kent-Rosanoff word association test has been used to assess this process. ⁴² In contrast to the creative thinker who is fully aware of logical contradictions, the schizophrenic patient is unconscious of the contradictory nature of his or her utterances. For example, when Albert Einstein derived his theory of relativity, derived from the fact that a man falling from the roof of a house was both in motion and at rest, he was fully aware of the contradictory nature of his thinking. ⁴³ Another example is Frank Lloyd Wright’s revolutionary design of Falling water, in which nature and interior space coexist. The janusian process was initially identified in highly creative writers, visual artists and scientists. The fluency of association observed among creative individuals can be mistaken for over-inclusive thinking. ⁴⁴Since their brains process increased sensory input effectively without cognitive overload, creative individuals derive an advantage from their higher levels of associative thinking.

Contrary to popular belief, in their cognitive and conceptual style, creative writers resemble those suffering from the manic phase of affective disorders, rather than schizophrenics. However, whereas the over-inclusiveness of maniacs is based on bizarre associations, that of writers is due to an imaginative recognition of original associations. Whilst writers are capable of controlled flights of fancy, manic imaginations are bizarre and based on personalised reason. The racing thoughts of a creative intellect are productive, whereas those of the manic are destructive. Albert Einstein claimed that he discarded a new idea every two minutes.

Creative thinking is polythetic and should not be confused with flight of ideas. Schuldberg (1990) investigated the overlap between schizotypal and hypomanic traits and suggested that affective symptoms may be more important than primary process thinking in determining creativity within the general population. ⁴⁵ The fluctuation of thoughts experienced by higher cognitive ability individuals can be mistaken for mood swings. Fink et al. (2014) challenged the connection between creativity and psychopathology and proposed that the domains of artistic and scientific creativity should be analysed separately. ⁴⁶

Although the creative potential of autistic people has been recognised, they differ from over perceptive children in many respects. One fundamental difference is that the creative potentials of the latter are polythetic, whereas such potentials of the of autistic individuals are generally monothetic. A key diagnostic criterion for autism—restricted and repetitive behaviours and interests—combined with a small number of research studies, suggest that generating original ideas or artefacts may be challenging for autistic individuals. ⁴⁷Nonetheless, a minority within this population has exceptional artistic gifts, and a wider group embraces activities typically associated with creative expression, including visual art, music, poetry and theatre.

A three-level multilevel meta-analytic approach investigated the relationship between creativity and schizophrenia. The analyses of Acar et al. (2018), with 200 effect sizes gathered from 42 studies, detected a mean effect size of r =−0.324, 95%CI [−0.42, −0.23]. ⁴⁸When the analyses focused on the moderators, they found that the relationship between schizophrenia and creativity was moderated by the type and content of the creativity measure, the severity of the schizophrenia and the patient status. The negative mean effect size was firmer with semantic-category or verbal-letter fluency tasks than the divergent thinking or associational measures. They submitted that when these findings are analysed along with previous meta-analyses on the association between creativity, psychoticism and schizotypy, creativity and psychopathology appear to have an inverted-U relationship. Whilst a mild expression of schizophrenia symptoms may support creativity, a full demonstration of the symptoms challenges it.

Schizophrenia and schizotypy have frequently been associated with above average creativity; nonetheless, empirical studies on the relationship between schizophrenia spectrum disorders and enhanced creativity have generated inconsistent results. ⁴⁹ Even though some mental processes may appear to be similar in creative and psychotic thinking, the current literature challenges this conclusion. ^50,51,52Psychopathology does not play a role in the genesis of higher order creativity; nonetheless, the psychological defence mechanism of overcompensation goes some distance towards explaining the high achievements of mentally or physically disabled individuals. ⁵³

The Myth of Drug Induced Creativity

The belief that brain alone is the source of creativity led to the idea that altering brain chemistry could make people more creative. The truth may be that the gentle psychopathology created in the brain might serve as a facilitator of creativity rather than a producer of creativity. The psychopathology generated by the psychedelic drugs might help to open Aldous Huxley’s ‘doors of perception.’ Huxley (1954) proposed “Doors of Perception” to illustrate the enlightenment induced by LSD etc.⁵⁴ Interestingly, such a proposal is close to Zizzi and Pregnolato’s depiction of ‘very fast switches from the quantum logic of the unconscious to the classical logic of consciousness’ (Zizzi & Pregnolato,2012). ⁵⁵Those who glorify such drug induced creativity are unaware that long term substance misuse can only kill creativity as the ‘switches’ become permanently damaged and lead to psychopathological states.

When one’s sense of self is suspended and space-time sense dissolves, psychedelic experiences occur, and such experiences should not be confused for true mystical experiences. Psychedelic experiences are pseudo-mystical experiences. True mystical perceptions and cognitions relate to what is essentially ineffable, pertaining to the nature of existence rather than being limited to familiar objects that are intrinsic to everyday experience. The hallucinating drug user or alcoholic is functioning at the level of impaired consciousness, while the mystic is operating at a higher level of consciousness. Mystics have full awareness of their altered state of consciousness and they are also in a position to switch back to their ordinary mode of perception, unlike a hallucinating patient. It may be true that psychedelic experience has created an interest in artistic activity and the raw materials obtained in such experience may be useful in eventual artistic creation, but the psychedelic experience as such is not a creative experience because motor functioning is impaired during psychedelic experience and information flow to the hands and fingers are affected. ⁵⁶ The natural state of a relaxed, happy, and well-adjusted person is more creative rather than the perplexed psychedelic state. There may be ‘psychedelic artists,’ but not psychedelic scientists indicating the difference in the creative process of scientific generativity and artistic.

Drug induced creativity is a conundrum that need serious clarification as many young people are trapped in such faulty perceptions. Cannabis is the most widely used illegal substance globally. Schafer et al (2011) suggested that cannabis produces psychotomimetic symptoms, which in turn might lead to connecting seemingly unrelated concepts.⁵⁷Such divergent thinking is considered primary to creative thinking. They argue that a drug induced altered state of mind may indeed lead to breaking free from ordinary thinking and associations, thereby, increasing the likelihood of generating novel ideas or associations. But the harmful effects of cannabis use have been extensively evaluated.^{58,59,60,61,62,63}Cannabis abuse is quite unlikely to generate any sustainable creativity-‘the creative Big Bang’ would soon end up as a big crunch.

If creativity is a neurological phenomenon, creative people should have additional neural pathways, but psychedelic drugs have not been proven to create such new neural pathways. Speculations about specific brain regions promoting creativity is of great scientific interest. Creativity involve an architect and a set of engineers. According to Amit Goswami, quantum unconscious domain is the architect and the real source of creativity if brain does the engineering works. ⁶⁴ Psychoactive substances do not act directly on the quantum consciousness but may help to open the gates to the hidden dimensions of consciousness. When quantum views of creativity are given due significance, the neurologically based psychedelic promotional views of creativity crumble. If not having creative abilities is deemed as a ‘brain deficit,’ use of illegal drugs to promote creativity can be compared to using medications to treat ADHD. But only if we use the ‘brain disease model’ of psychiatry, the argument of ‘brain deficit model’ will hold water. It may be even true that psychedelic drugs may have a quick and transient destressing effect and that could promote a creative mental state, but the production of any direct creativity through the use of such drugs is questionable.

Problematic Childhood

Some children of superior intelligence attempt to mask their creativity by being over-talkative and overactive. Such children run the risk of being misdiagnosed as ADHD. Creative children frequently have a unique sense of humour that their peer group cannot appreciate. Creative children are every so often resented by peers because of crazy or unusual ideas and their forcefulness and passion in presenting them or for pushing their ideas on others. Their divergent thinking is not helpful in school examinations, which require convergent thinking, and this could explain the poor academic achievement of several geniuses. The divergent thought processes of creative children must be differentiated from inattention and underachievement. For example, a highly intelligent child might fathom out what the teacher is going to say next and become inattentive. Although creativity is associated with divergent thinking, this alone does not correlate well with creative achievement.⁶⁵ Creative children overflow with ideas and play with new ideas and concepts. They are not motivated or even concerned about high grades and need individualized attention lest they might fall on the wayside. There is nothing more frustrating than being a creative intelligent and become underachieved.

Creative children demonstrate certain unusual traits such as daydreaming, wanting to work alone, sharing bizarre thoughts and conflicting opinions. These qualities will not please the traditional teachers and bring them in conflict with them and their lack of conformity to the classroom structure can be even confused with attention deficit hyperactivity disorder. Highly critical parents kill creativity; unfortunately, countless creative individuals have chaotic childhoods leading to psychological problems in their adult lives. Mismatching due to variation in I.Q could lead to mismatching with parents and siblings. Mother and father may be of average intelligence, but the child can be above average intelligence, and could cause mismatching leading to behavioural problems.

There is a special group of children around the world who have high intelligence and intuition, healing abilities, and a strong spirituality and they are grouped as Indigo people in appropriate cultures. It can be stated that Indigo is people with high sensitivity level, unique creativity, high intuition ability, healer, and people with their own charisma for those around. ⁶⁶According to the proponents of these new ideas, these children are often mislabelled as having behaviour disorders. Little is known from scientific research about the Indigo phenomenon. Indigenous populations are familiar with Indigo-like children. The purpose of studying these children when they are adults is to better understand these children when they are older and advance behaviour health sciences by increasing awareness of the Indigo phenomenon and learning about their lived experiences. There has been hardly any serious scientific study on the Indigo phenomenon.

A phenomenological study looked at the lived experiences of 10 adult Indigos. The study explored the lived experiences of 10 adult Indigos on the island of Oahu, Hawai'i (> or = 18+ years old-7 females, 3 males; mean age = 52.4 + SD). ⁶⁷ Through in-depth semi-structured personal interviews, the experiences of these adults were analysed and interpreted to identify the common experiences faced during childhood, what worked for their assimilation into society, and recommendations for parents, educators, and health professionals on how to work with Indigos. Bioenergy field photographs of each participant were also taken. Statements related to the phenomenon were placed into themes, coded, and categorized as the investigators reached a consensus of common themes. Seven primary themes and nine secondary themes emerged from the findings.

The primary themes were: grandmother/mother had a similar gift; guided by a higher power to heal self and others; felt 'different' or misunderstood; did not openly share their unique abilities; having challenges with partner relationships; history of abuse/violence or frequently disciplined; and use of intuition at work and/or school. Secondary themes included: Using Hawaiian and cultural healing methods; everyone has a degree of intuition and the use of intuition to know when to see a doctor or not; various unique abilities from body and multiple careers; mental health institutions, and financial struggle. Self-reports on participants' life purpose, their unique abilities, and being misunderstood were also collected. It was concluded that Indigos felt mislabelled or misunderstood throughout their lives despite their belief that their life purpose was to help humankind.

Academic psychologists are sceptical about Indigo phenomenon and argue that the phenomenon is a cover up to normalise the odd behaviour of children who could otherwise be included under the category of attention deficit disorder, attention deficit hyperactivity disorder, autistic spectrum disorders and learning disabilities. Health experts are concerned that labelling a disruptive child an "indigo" may delay proper diagnosis and treatment that could help the child or investigate the parenting style that may be causing the behaviour.

Inspiration and Perspiration

Creativity is regarded as the product of inspiration or creative imagination combined with meticulous, disciplined effort. The Edisonian perception of invention as 1% inspiration and 99% perspiration is explained by the hypothesis of interactive creativity; it assumes that the inspirational aspect has a paranormal component. In his thesis on interactive creativity, Laszlo supported his hypothesis with observations on cultural creativity. ⁶⁸These observations included the collective advance of entire populations through the typical creative activity of their members and by documented incidents in modern science in which different investigators developed scientific insights simultaneously, without any known contact. ⁶⁸ Early cultures developed similar 7tools; calculus was discovered simultaneously by Newton and Leibni and biological evolution was described independently by Darwin and Wallace. Similarly, Graham Bell and Elisha Grey both applied to patent the telephone on the same day. The Rubic’s cube was conceived simultaneously and designed both by Rubic and a Japanese inventor. Nylon was discovered in both New York and London, hence, the name NyLon.

Jung’s research into the phenomenon of creative synchronicity helped him to formulate his concept of the collective unconscious. Psychological disturbances may represent the consequences of creative endeavour and Jung (1973) considered them the price to be paid for persistent exploration of the unknown.⁶⁹ Polayni (1994) suggested that scientific discovery is informed by the imagination and integrated by intuition, and vice versa.⁷⁰ This statement is close to the Edisonian perception of creativity: If imagination is a property of the brain, intuition occurs in the unconscious realm. Whilst Laszlo’s views are not definitive, they indeed supplement our existing knowledge about creativity. The inspirational aspect can be better explained by an expanded model of brain-mind-consciousness, and Xavier suggests a para-psychodynamic.⁷¹

Biological Perspectives

Particularly gifted individuals have determined the evolution of civilisation. Karlsson (1984) commented regarding creative individuals: ‘Without their genes, men might still live in caves’. ⁷²Nonetheless, countless gifted individuals have a very ordinary family background, with no ancestral history of creativity. For example, Newton came from an undistinguished family. Genetics researchers look for the biological roots of creativity, with some believing that the mind is reducible to chemistry. Whilst intelligence may be a trait that can be cloned, creativity may not be attached, and it may prove even more complex than genetic manipulation. Kelly et al. (2007) suggested that creative inspiration is akin to mysticism. ³⁵

Responses to both dopamine inhibiting drugs and to the psychoses triggered by the drugs that increase dopamine activity underlie the dopamine hypothesis of psychosis. However, dopamine over-activity in psychosis should not be confused with dopamine fluctuations in creative individuals. Dopamine diminishes with aging, which may explain the decreasing powers of creativity after the age of twenty.⁷³

The relationship between age and outstanding achievement has captured the attention of researchers into creativity.Whilst Lehman maintained the perspective that creative achievement has a curvilinear single-peak function for age, other researchers have described two separate age-peaks. ⁷⁴Outstanding contributions among mathematicians after the age of fifty are exceptions. The age-related observations support a biological basis for creativity.

Future Directions

Study of creativity is an important area of research where consciousness studies and psychopathology meet each other. Cognitive scientists have pondered over the link between psychopathology and creativity for a long time without making any firm conclusions and appear to be barking at the wrong trees. The very process of creativity ought to be explored before any progress in this area of research could be achieved and the current reductionist model of mind stands as a hindrance. The following suggestions may be helpful for future researchers.

1. Establish the psychopathology of psychotic disorders

2. Creativity linked genetic studies are warranted, biological correlates of creativity need further elucidation.

3. Expand the current model of brain-mind-consciousness complex so as to explain the inspirational element of creativity.

4. More longitudinal, international and multicultural studies recommended.

5. Given the affinity of psychosis-proneness to the artistic creativity domain,

studies should be focussing artistic creativity and scientific creativity separately.

Clinical Implications

The study of creativity has clinical implications: A) Psychiatric understanding of creativity provides a better picture of patient functioning that could assist in clarifying the definitions of both normalcy and psychopathology. B) Early detection of creative talents in children might help to give special guidance for such children and thereby prevent potential psychiatric problems. C) When they coexist, differentiating creativity and mental illness is useful: The former might require nurturing and the latter warrant clinical intervention.

Discussion

Whilst it is true that investigators have observed a high incidence of psychiatric symptomatology of an affective nature among creative writers and artists, it is debatable whether this relationship is causal, an effect or a contributory factor. The increased psychopathological states observed in artistic creative individuals suggest that art and science reflect two different arenas of creativity. The mechanism of generation of novel ideas may be identical in art and scientific creativity, but they are evaluated differently by the two groups resulting in different types of products. Creativity and mental illness can coexist, and the creative impulse has a therapeutic effect on the psychiatric condition. Creativity can be therapeutic for those who are already suffering from mental illness; creative art therapies applied in clinical and psychiatric settings report positive health-related outcomes. ⁷⁵ Even in rare cases of psychopathology induced creativity, the individual will require highly developed intellectual protective factors. It is the disciplined portion of the mind that enables outstanding creative achievements. Creativity of the highest order is a product of laborious intellectual effort. When they coexist, psychopathology is a mediator, not the producer of creativity, and the creativity may be cathartic.

There is no scientific consensus regarding the association between psychopathology and creative achievement. The literature does not substantiate the high reported incidence of mental illness among creative people. It is possible that predispositions to mental illness and creativity tend to co-occur because they reflect an underlying personality and cognitive style predisposed to both creativity and mood disorder. The high reported incidence of mental illness potentially signifies an ‘occupational hazard’ and creativity stands independent of psychopathology. The normal creative person can swing back to reality from a transient ‘creative psychical shift’ (e.g. such as a diver who searches for diamonds in the deep sea and then returns to the shore). The sensitivity and intensity that facilitate creative expression may additionally make highly creative people more susceptible to depression.

Early investigations of geniuses were retrospective. Formal meta-analyses were not considered justifiable. All forms of creativity were mixed in the studies, without distinguishing the different domains of creativity. Most of the studies were confined to English speaking people, whilst creativity is a global phenomenon. A multiplicity of literature does not mean that the ideas expressed are established scientific truth. This is true of creativity, which may be as mysterious as creation itself. The prevailing model of the mind may be inadequate for a full appreciation of the creative process. It would be easier for a ‘camel to pass through the eye of a needle’ than to explain creativity from a reductionist perspective. The inspirational component of creativity continues to be an enigma. It is my contention that creativity is essentially an inner, psychic phenomenon. We do not have even an approximate model of the brain-mind-consciousness complex, let alone know the true aetiology of schizophrenia and bipolar disorder. Therefore, it would be prudent to suspend our psychopathology allied perspectives of creativity until we develop a deeper understanding of consciousness.The association between creativity and psychopathology has soared to the level of cultural myth and this is evident in many films in which mentally ill persons are portrayed as extraordinarily creative. ⁷⁶

Introduction:

Aesophagogastroduodenoscopy (EGD), is a common same-day procedure used for both diagnostic and therapeutic purposes during which a small flexible fibreoptic tubular camera is introduced through the mouth and advanced through the pharynx into the oaesophagus, stomach, and duodenum. EGD procedures are performed under deep sedation, since they can elicit significant pain, discomfort, and anxiety. When pain is not controlled properly, this can lead to an increase use of adjunctive medications such as opioids. This can extend the length of stay and increase adverse outcomes.¹ In a prospective, randomised, double-blinded study, Bedirli et al. found that use of opioid medications such as fentanyl are associated with increased adverse outcomes.²

Since procedural sedation-related complications (such as hypoxia, hypotension, desaturation, and emergent airway intervention) remain one of the biggest challenges in EGD procedures, it is important to select the correct medications to reduce these complications.³ Currently, propofol is the most common and popular main procedural sedation agent used for EGD procedures.^4,5

Propofol is the preferred main intravenous agent in EGD procedures due to its amnestic, sedative, and hypotonic properties.⁶ It has also been favoured due to its ultra-rapid response and duration of effects, usually taking 30-60 seconds for onset of action and lasting up to 4-8 minutes.⁷ However, propofol has been associated with significant adverse effects that include dose-related hypotension, bradycardia, laryngospasms, and apnoea.^8,9 Propofol does not have analgesic properties and will usually be combined with opioids for pain control.¹⁰

Ketamine is classified as a dissociative anaesthetic and is known to provide analgesia and amnesia. Important to note, it causes less respiratory or cardiovascular depression when used alone for children greater than 4 months of age.¹¹ However, ketamine alone can cause such side effects of laryngospasm, increase secretions, and vomiting.^12,13

The mixture of propofol and ketamine in one syringe (coined ketofol) has been shown to be effective in sedation for various procedures, such as spinal anaesthesia,¹⁴ along with orthopaedic¹⁵ and cardiovascular procedures¹⁶ in adults and children. Use of this combination has been favoured in brief but painful emergency room procedures due to the opposing haemodynamic and respiratory effects of both sedative medications.¹⁷ The negative cardiac effects produced by propofol can be attenuated with the use of ketamine, resulting in an increase in mean arterial pressures and cardiac indices.^18,19 The complementary effect of both mediations has enabled the use of lower doses for each medication, thus lowering the toxicity and side effects.^20,21 Although there are studies comparing activities of ketofol sedations,^22-25 there has not been a study published on ketofol compared to propofol use in children during EGD procedures.

The primary goal of this study was to determine if there were differences in the outcomes of adverse cardiac and pulmonary events, vital sign parameters including objective pain, and administration of adjunct pain medication (for which fentanyl was used in this study) during EGD procedures between propofol and ketofol groups. A secondary goal was to determine if there was a difference in site performance metrics for EGD procedures (sedation time, stop of sedation to discharge time, and length of stay) between propofol and ketofol groups.

Methods:

This study protocol was approved by the Naval Medical Center Portsmouth Institutional Review Board in compliance with all applicable federal regulations governing the protection of human subjects. Research data was derived from an approved IRB protocol: number NMCP.2018.0021. Written informed consent was not required by the Naval Medical Center Portsmouth IRB, as this data concerned historical dae-identified patients.

Study Design

This was a single centre, retrospective study of a cohort of children (ranging from 2 to 18 years). Data was collected from March 2011 to September 2013 at Naval Medical Center Portsmouth’s Paediatric Sedation Centre. EGD and sedation protocol was not changed during this time period. EGD procedures were performed by the same paediatric gastroenterologists throughout this time period. Sedation was performed by the same paediatric intensivists throughout this time period. All patients used in this study were involved in a same day EGD procedure. Forty-one patients underwent deep sedation via propofol as main sedation agent from March 2011 to May 2012. Forty-nine patients underwent deep sedation via propofol and ketamine combination as main sedation agent from May 2012 to September 2013. Each main sedation agent was given in a similar fashion as an initial intravenous loading dose based on 1 mg/kg propofol followed by an intravenous infusion of 200-250 mcg/kg/minute that was started less than 10 minutes prior to start of the procedure and stopped at the end of the procedure. Ketofol solution used was a 1:5 ratio of ketamine to propofol (40 mg:200 mg). One mcg/kg of fentanyl was given when the sedationist during the EGD perceived the patient experiencing pain. Exclusion criteria included: patients less than 2 years and greater than 18 years, those less than 10 kg, and patients receiving adjunct medications other than fentanyl, ondansetron, or lidocaine.

Data Collection

Fourteen patients were excluded due to weight less than or equal to 10 kg. In addition, 12 patients were excluded due to age less than 2 years or greater than 18 years. Past medical history, ASA class, procedure indications, age, weight, sex, pain scores, vital signs (mean arterial blood pressure [MAP], heart rate [HR], and respiratory rate [RR]), unplanned events (hypoxia defined as SPO2 less than 85% at any time point, and hypotension defined as mean arterial blood pressure with greater than 20% decrease from baseline blood pressure), emergent airway intervention (defined as apnoea needing bag mask ventilation or CPAP use), and unplanned intubation. Vitals and pain scores were obtained from initial presentation in the sedation suite, start of procedure, every 5 minutes during the procedure to stop of sedation, and at time of discharge. Midway procedure vitals used for statistical analysis were obtained by selecting vitals that were at the half-way point of the patient’s EGD procedure. The Children’s Hospital of Eastern Ontario Pain Scale (CHEOPS) was performed every 5 minutes during the EGD procedure by an independent United States Navy Corpsman and used to evaluate patient’s pain prior, during, at stop of sedation, and after procedure for this study.²⁶

Statistics

All statistics performed in this study were calculated using SPSS Statistics programme (IBM Corp. Released 2017. IBM SPSS Statistics for Windows, Version 25.0. Armonk, NY: IBM Corp.). Mann-Whitney U nonparametric test between independent groups was performed to compare age, weight, total EGD procedure time, total time of sedation during EGD procedure, time from stop of sedation to discharge, total length of stay, and total fentanyl use (in mcg/kg) between the two main groups in this study (propofol and ketofol group). The significance level was set to 0.008 after a Bonferroni was corrected, in order to address the probability of making one or more false discoveries when performing multiple hypotheses tests. A repeated measures ANOVA was run to determine the effect of treatments over time for HR, MAP, and RR. Mauchly's test of sphericity was performed for both age groups. The Fisher Exact test was performed to compare propofol to ketofol in unplanned hypotensive events and unplanned apnoea requiring bag valve mask or CPAP. Pearson Correlation statistics were performed to study possible correlations between propofol or fentanyl and sedation time or length of stay. Statistical significance for Pearson Correlation statistics was considered when two-tailed p<0.001.

Results:

Table 1. Demographic divided into 2 age groups. EE= eosinophilic esophagitis. * indicates statistically significant difference.

Figure 1. Propofol amount total per weight (mg/kg). Bars represent standard error of the mean. There was no significant difference between the propofol and ketofol in all age groups.

Figure 2. Fentanyl amount total per weight (mcg/kg). Bars represent standard error of the mean. In all age groups, there was a significant difference between propofol and ketofol. *p<.008.

Figure 3. Vital signs. Mean arterial pressures for ages A) 2-11 years and B) 12-18 years. Main sedative agents’ MAPs were statistically different for ages 2-11 years (p =0.004), but not for ages 12-18 years (p =0.224) for all time periods. Heart rate for ages C) 2-11 years and D) 12-18 years. For Heart Rate, there was a statistically significant interaction between treatment and time for both age group using the Greenhouse-Geisser correction, p =0.002, p =0.014. Main sedative agents’ HRs showed to be statistically different for both age groups, p <0.001 and p =0.004.

Figure 4. Duration of time. A) total sedation time, B) stop of sedation to discharge, and C) total length of stay. In all age groups, there was no significant difference between propofol and ketofol. Bars represent standard error of the mean. *p<0.008.

Figure 5. Pearson Correlation. Propofol versus total sedation time for ages A) 2-11 years (p<0.01*), and B) 12-18 years (p<0.01*). Propofol versus length of stay for ages C) 2-11 years (p<0.01*), and D) 12-18 years (p<0.01*). * Correlation is significant at the 0.01 level (2-tailed).

Figure 6. Pearson Correlation. Fentanyl versus total sedation time for ages A) 2-11 years (p=0.506), and B) 12-18 years (p=0.961). Fentanyl versus length of stay for ages C) 2-11 years (p=0.378), and D) 12-18 years (p=0.352). No significant correlation was not seen for all age groups.

Ninety patients were retrospectively analysed in this study. Baseline demographics were similar between the groups, except for gender proportions (Table 1). Similar amounts of propofol per weight were used in each group (Figure 1). Ketofol significantly reduced fentanyl use in all age groups by ≥.99 mcg/kg compared to propofol alone (p<0.008 for all age groups, Figure 2). In the 2-11 year old age group, the propofol group required a mean fentanyl dose of 1.96 mcg/kg ± 1.24 mcg/kg and the ketofol group required a mean fentanyl dose of 0.44 mcg/kg ± 0.49 mcg/kg; thus ketofol required about 1.52 mcg/kg less fentanyl during EGD procedures. In the 12-18 year old group, the propofol group required a mean fentanyl dose of 1.38 mcg/kg ± 1.05 mcg/kg and the ketofol group required a mean fentanyl dose of 0.39 mcg/kg ± 0.59 mcg/kg; thus ketofol required about 1 mcg/kg less fentanyl during EGD procedures.

Vital signs (HR, RR, and MAP) and CHEOPS pain scores were obtained and analysed for baseline vitals, at the procedure midpoint, at stop of sedation, and at time of discharge. CHEOPS pain scores’ mean was 6 throughout all time periods for both groups, thus there was no statistically significant difference.

A repeated measures ANOVA was run to determine the effect of treatments over time for RR. There was sphericity for the interaction term, as assessed by Mauchly's test of sphericity in both age groups (p =0.070, p =0.762). For RR, there was not a statistically significant interaction between treatment and time for both age groups, p =0.163, p =0.804. The treatments were not found to be statistically different for either age group, p =0.736 and p =0.224.

A repeated measures ANOVA was run to determine the effect of treatments over time for HR. As assessed by Mauchly's test of sphericity, no sphericity was found in either age group (p <0.05). For HR, using the Greenhouse-Geisser correction, a statistically significant interaction was found between treatment and time for both age group, p =0.002, p =0.014. The treatments were found to be statistically different for both age groups, p <0.001 and p =0.004. A repeated measures ANOVA was run to determine the effect of treatments over time for MAP. Sphericity was found in both age groups for the interaction term, as assessed by Mauchly's test of sphericity (p =0.209, p =0.269). For MAP, there was not a statistically significant interaction between treatment and time for either age group, p =0.261, p =0.591. The treatments were statistically different for the 2-11 year old group (p=0.004), but not statistically different for the 12-18 year old group (p=0.224). (Figure 3 A-D)

Unplanned events were analysed for clinically significant hypoxia, hypotension, emergent airway intervention, and unplanned intubations. Ketofol compared to propofol had fewer hypotensive events in the 2-11 year old age group (3.7% versus 15.4%, p=0.192) and in the 12-18 year old age group (0% versus 13.3%, p=0.144). Ketofol compared to propofol had statistically significant fewer apnoea events requiring bag valve mask or CPAP intervention for the 2-11 year old group (3.7% versus 53.8%, p<0.001) and for the 12-18 year old group (0% versus 33.3%, p=0.005). There were no significant hypoxia events. There was one unplanned intubation in the propofol group of a healthy ASA I twelve year old female who had a 20 mL emesis episode after a loading dose of propofol was given on induction. (Table 1)

With propofol leading to significantly more fentanyl usage, more hypotension and emergent airway intervention during EGD procedures, we performed analysis to see if there was an effect on sedation time, time from stop of sedation to discharge, and length of stay (LOS). In all age groups, there was no statistical difference between propofol and ketofol for sedation time (p=0.115 for ages 2-11 years, p=0.124 for ages 12-18 years), time from stop of sedation to discharge (p=0.033 for ages 2-11 years, p=0.511 for ages 12-18 years), and LOS (p=0.026 for ages 2-11 years, p=0.109 for ages 12-18 years) (Figure 4). Based on Pearson correlation test, it was established that propofol was positively correlated with sedation time and LOS (+0.753 for sedation time and +0.611 correlation for length of stay with <0.001 significance) (Figure 5). There was no significant correlation between fentanyl and sedation time or LOS (Figure 6).

Discussion:

This was a retrospective study looking at 90 paediatric patients, ages 2-18 years, undergoing EGD procedures with either propofol or ketofol as the main sedative medication. Patients in this study had similar weights, ages, ASA scores, and EGD indications; however there baseline demographics were not similar with regards to gender proportion (Table 1). Based on the large prospective database on sedation use for procedures outside the operating room obtained by the Paediatric Sedation Research Consortium,²⁷ we separated our patient population into two age risk groups to allow our results to be generalisable. To our knowledge, this study is the first to show that ketofol, when compared to propofol, significantly reduces fentanyl use (≥.99 mcg/kg, Figure 2) and cardiopulmonary adverse outcomes (Table 1) in paediatric EGD procedures.

EGD procedures are known to be invasive and painful. To decrease the pain appreciated by a patient, various adjuncts are often utilised. In our study, we used fentanyl for adjunct pain control. In both main sedation medication groups (propofol and ketofol), there was no statistical significance seen in objective pain based on CHEOPS scores with a mean score of 6 in all age groups (p>0.05). However, the amount of adjunct needed to maintain adequate pain control between both groups statistically and clinically differed (Figure 2). In all age groups, the propofol group compared to the ketofol group required almost 1 mcg/kg more of fentanyl, which leads to the potential for the patient to experience higher incidence of side effects, such as respiratory depression and hypoxia.^28-30

The mechanism that lead to this significant difference in opioid demand between ketofol and propofol is most likely due to ketamine’s ability to stimulate opioid sigma receptors,31 thus leading to opioid sparing effects. It is unclear if propofol and ketamine interaction heightens this opioid sparing effect, because ketamine alone has not been shown to lead to opioid sparing effects in children.³²

In a review of our population’s vital signs, there was a significantly higher MAP in the ketofol group compared to the propofol group for ages 2 to 18 years (Figure 3 A-B). In addition, the ketofol group had statistically higher HR compared to the propofol group for ages 2 to 18 years (Figure 3 C-D). It is our thought that this increase in MAP and HR in the ketofol group mediated by ketamine is a protective factor against major hypotensive changes (31). It is this effect that minimised unplanned hypotensive events in our ketofol group compared to the propofol group by 11.7% in ages 2-11 years and 13.3% in ages 12-18 years (Table 1).

One of the more noticeable differences in our main sedative medication groups was the unplanned apnoea events requiring CPAP or bag mask ventilation intervention. In our 2 to 11 year age group, propofol had 50.1% more apnoea events needing intervention compared to ketofol. In our 12 to 18 year old group, propofol had 33.3% more apnoea events needing respiratory intervention compared to ketofol. This was found to be statistically significant. It is unclear if this is also clinically significant since no emergent airway intubation was needed for these events. However, there was 1 emergent airway intubation in the 12-18 year old propofol group. Based on the review of the medical records, it appears that these apnoea events needing respiratory intervention were mostly associated after the initial loading dose of either propofol or ketofol prior to or at the start of the infusion. Thus, bring into question if not starting with a loading dose bolus would decrease adverse effects.

Despite the propofol group requiring more respiratory intervention, increased fentanyl use, and less haemodynamic stability seen, there was no statistical differences in total sedation time and LOS between the two main sedation medication groups (Figure 4). Yet when we ran correlation statistics, there was a positive correlation to increased propofol dosages, total sedation time, and LOS in ages 2 to 18 years (Figure 5); therefore, if our procedures were longer, there could be a statistically and possibly clinically significant increase in total sedation time and LOS for the propofol only group. Thus, we can infer that propofol compared to ketofol is not the main sedation medication of choice for longer similarly painful procedures in children ages 2 to 18 years, such as EGD procedure followed by a colonoscopy.

A limitation of our study is that it is retrospective. With that, we were not able to blind our sedationists to the main sedation medication that was chosen and could not control the interventions that were performed. Yet, based on strict exclusion criteria, we were able to control the interventions used in our analysis. Also, in our study we did not analyse patients under the age of 2 years. Reviewing the data, we had seven patients, and thus this patient population size was not powered to perform the proper statistical analysis. Another limitation to consider is the time difference in procedure protocols. However, since this study included over 2 years of data, procedure protocols were the same. Additionally, the EGD indications were similar between ketofol and propofol (Table 1). Another consideration of this study is generalisability. This study was conducted at a single centre, but the way we analysed our results makes it easier to perform large scale prospective studies to further investigate our findings. Last limitation is directly correlating fentanyl dose with a significant adverse event. Due to the short length of the EGD procedure, constant sedation medication infusion, and fentanyl dose, it is difficult to directly say that a particular fentanyl dose alone leads to an adverse event. Therefore we can only speculate based on correlation.

Conclusions:

Our study found that ketofol significantly reduced fentanyl use in all age groups by ≥0.99 mcg/kg and cardiopulmonary adverse events compared to propofol alone. In addition, an increase in the amount of propofol was positively correlated to increasing LOS and total sedation time for a child undergoing an EGD procedure. This suggests that increasing amounts of propofol leads to greater LOS. To conclude, our study indicates that ketofol can be a safer alternative to propofol use alone for deep sedation in EGD procedures for children ages 2 to 18 years.

Case Report:

A 33 year-old ASA1, primigravida, presented to our delivery suite with spontaneous onset of labour at 38 weeks of gestation. Epidural analgesia was commenced to alleviate her labour pains. Subsequently, she underwent an assisted vaginal delivery of a live male baby (weighing 4660 gms) using Keiland’s outlet forceps after 90 min second stage of labour. 10 hours postpartum, she complained of dyspnoea & severe central substernal chest pain. She was noted to have an unusual swelling of face and neck with oxygen saturations of 90 % on room air. Ascultation of chest revealed normal bronchovesicular breath sounds, normal heart sounds with absence of added sounds. Arterial blood gases showed an O₂ tension of 11 kpa, CO₂ tension of 5 kpa and pH of 7.34. The diagnosis of subcutaneous emphysema, pneumomediastinum and small left apical pneumothorax (Hamman’s syndrome) was confirmed on chest X-ray (CXR 1). We ruled out differential diagnosis of pulmonary embolism, Tension pneumothorax, angina pectoris, pericarditis, dissection of aortic aneurysm, mediastinitis, cardiac tamponade, chest infection & oesophageal tear. She was managed conservatively by close monitoring for complications, administration of supplemental oxygen and use of simple analgesics. She demonstrated a complete uneventful recovery over the next 24 hours with normalising of chest signs (CXR 2).

CXR 1: shows pneumomediastinum, extensive surgical emphysema & a left apical pneumothorax.

CXR 2: shows small pneumomediastinum, the surgical emphysema & pneumothorax resolved.

Discussion:

Hamman’s syndrome is named after Louis Hamman (1847-1946), the physician who first described it in 1945. The first reference to this condition was in 1618, when Louise Bourgeois, midwife to the Queen of France, wrote, “I saw that she tried to stop crying out and I implored her not to stop for the fear that her neck might swell”³_.

Hamman’s syndrome usually occurs in the 2^ndstage of labour & is associated with prolonged and protracted labour and larger than usual babies ⁴_. However, the clinical presentation is often delayed to the postpartum phase as was clearly seen in our case. The condition seems to be provoked by any valsalva manoeuver such as vigorous coughing/vomiting/sneezing, forced physical activity & enormous efforts during spontaneous vaginal delivery. Its occurrence is usually related to the expulsive phase of labour when ‘pushing down’ actively raised the intraalveolar pressure. This may subsequently increase the intrathoracic pressure up to 50 mm of Hg or higher¹. Rupture of marginal alveoli with air entering along the perivascular sheath into the mediastinum is the most likely mechanism, in our case. It is probable that, the air tracts through the fascial planes into subcutaneous and retroperitoneal tissues. Other reported mechanisms of Hamman’s syndrome include oesophageal rupture during childbirth, or pneumomediastinum related to asthmatic bronchospasm⁵ or chest infection, or dissection of pneumoperitoneum, secondary to epidural catheter placement or caesarean section^1.

Palpable crepitus on face & neck is suggestive of subcutaneous emphysema & appearance of this emphysema in labour is the hallmark of pneumomediastinum. Other features of pneumomediastinum include substernal chest pain, dyspnoea, voice change, cough, sore throat and tachycardia¹. Hamman’s sign, a fine auscultatory crepitation synchronous with the heartbeat, heard along the left sternal border; is sometimes observed in this condition².

Chest X-ray and CT thorax are the diagnostic tests. Majority of the patients with Hamman’s syndrome have pneumomediastinum & subcutaneous emphysema without any pneumothorax and this requires supportive management with strict monitoring. Our patient demonstrated a small pneumothorax, which was managed conservatively. A surgical intervention in the form of subcutaneous air drainage may occasionally be indicated in severe cases.

Overall most cases have a benign, self-limiting course when the aggravating factors are no longer present. Published data indicates that subsequent pregnancies pose no additional risk of recurrence⁵.

Conclusion:

Since Hamman’s syndrome is a potentially dangerous complication of normal childbirth. We propose that every obstetric anaesthetist and obstetrician should be aware of this syndrome.

Introduction

With the increasing use of ultrasound as a standard examination in the first trimester, more incidental adnexal masses are detected. The reported incidence of adnexal masses in pregnancy varies, depending on the criteria used to define the mass. A literature review by Goh W. et al., found that 1% of all pregnancies are diagnosed with an adnexal mass ¹. A more recent article has suggested adnexal masses are diagnosed in 5% of all pregnancies ². Simple and functional cysts are very common, and they usually resolve after the first trimester ³. Mature teratomas are by far the most common persistent adnexal masses found in pregnancy ⁸. It has been estimated that up to 5% of adnexal masses in pregnancy are malignant ⁴.

Ovarian cysts are typically asymptomatic, but they can cause pain due to pressure on adjacent organs, rupture, bleeding or torsion. The latter case is a significant health condition which mainly requires emergency surgical intervention. During pregnancy, surgical management of ovarian cyst complications is more difficult and more challenging. This is mainly because of other differential diagnosis causing similar symptoms related to pregnancy such as ectopic pregnancy and miscarriage. In case of surgical intervention, the second trimester of pregnancy is supposed to be the safest window for surgery as the risk for drug-induced teratogenicity is smaller than in the first trimester, most functional cysts have disappeared by then and it is technically less difficult than operating during the third trimester ¹³.

Antenatally, ultrasound is considered to be the best first-line imaging to evaluate adnexal masses ⁵. Ovarian mass characterization into benign, malignant or borderline can be challenging in pregnancy. This is mainly due to the effect of high levels of gestational hormones which can cause decidualisation of the cystic or solid parts of the ovaries. Benign masses can mimic malignant masses due to this pregnancy related phenomena ¹². One of the largest data in literature on ovarian mass characterization is published by the International Ovarian Tumor Analysis group (IOTA). All IOTA studies excluded pregnant women when they developed and validated the rules and models to characterize ovarian masses ^14-17. This limits our knowledge and ability to use these models in pregnant women. It is known that tumour markers may be raised in pregnancy and should therefore not routinely been done ⁷. An alternative diagnostic tool is Magnetic Resonance Imaging (MRI) which is considered to be safe in pregnancy and can be helpful if the ultrasound imaging is inconclusive in evaluating whether a mass is benign or malignant ^{6; 10}. The American College of Gynecology and Obstetrics recommends that pregnant patients should be reviewed on a case-to-case basis and stated that there are no known biological effects of MRI on fetuses. However, Gadolinium, which help in characterizing ovarian masses, should be avoided when examining a pregnant patient ¹¹.

The aim of this retrospective study was to look into characteristics, size and subsequent management of cases of adnexal masses in early pregnancy.

Methods

This was a retrospective study of data collected between 12/01/2014 and 14/11/2016 in the Early Pregnancy and Gynaecology Unit (EPAGU) of a tertiary referral centre (Guy’s and St Thomas’ NHS Trust, GSTT) in central London. The Ultrasound reporting system (Astraia Software Gmbh, Version 1.24.10, Munich, Germany, 2016) was searched for data. Cases included were consecutive. The study was approved as a service evaluation audit by the Clinical Governance team at Guy’s and St Thomas’ NHS Trust. The study included women who were diagnosed with an adnexal mass while having a transvaginal ultrasound scan TVS at or before 15 weeks of gestation. Pregnancy was confirmed by a positive pregnancy test and an intrauterine gestation on transvaginal ultrasound scan. Women who had the first gestational TVS after 15 weeks of gestation, pregnancies of unknown location, ectopic or trophoblastic pregnancies and patients who had ovarian stimulation treatment were all excluded.

Repeat ultrasound scan reports were retrieved from the Astraia system. Further procedures, tests and imaging results were retrieved using the Electronic Patient Reporting system at GSTT (EPR application, iSOFT Group plc., USA, 2004), PACS (GE Medical Systems, Wisconsin, USA, 2006), Badgernet (Clevermed, Client version 2.9.1.0, Edinburgh, UK). We have used the subjective impression of the examiner as the index test. If surgery was performed the final outcome to identify benignity or malignancy was considered to be the histological diagnosis if any removed tissue. Cytology was used for a reference test in two cases when ovarian cysts were aspirated only. Borderline tumours were classified as malignant for statistical analysis. Tumours were classified using the criteria recommended by the World Health Organisation (WHO) ^{9; 10}. All ultrasound scan images were available and reviewed by author TEG to confirm the US finding. For statistical analysis, the SPSS software package was used (version 24 for Windows, Chicago, IL, USA). A two tailed student’s t test was used to compare means in ovarian masses diameters and a p value of less than 0.05 was considered statistically significant.

Results

7150 patients underwent transvaginal scans for early pregnancy in that period. In total 48 cases of women with adnexal masses in pregnancy and completed data were analysed. Seven women have been excluded; one woman being postpartum at the time of the finding of a large endometrioma, five pregnancies due to assisted conception and one woman was found to have a corpus luteum cyst (Figure 1).

Figure 1: Study flow chart.

41 women with 46 ovarian cysts could be included in the study. Two women had bilateral ovarian cysts, one had two ipsilateral cysts and one woman had three ipsilateral cysts. The mean age at the time of detection of the ovarian mass was 30 (95%CI:28-32) (Fig.2).

Figure 2: Age distribution in the study group.

The mean gestation at the time of first ultrasound was 7.4 weeks (95%CI:6.6-8.3). The mean diameter of ovarian cysts measured 47.7mm (95%CI:39.9-55.4). In 36 women ultrasound alone was performed to reach diagnosis, one woman had an extra MRI scan, two women had tumour markers on top of the TVUS and in two women an MRI scan and tumour markers were performed after the TVUS. The ovarian cyst(s) was on the right ovary in 16/41 women, on the left in 22/41, bilateral in 2/41 and in one case the side of the cyst was not reported. The most common ultrasound subjective impression was mature teratoma (22/46 cysts), followed by simple cysts (12/46 cysts), haemorrhagic cysts (6/46 cysts), endometriomas (5/46 cysts) and one possible mucinous Borderline tumour. The latter was confirmed later on histology as the stage FIGO 1A intestinal type mucinous Borderline tumour (Fig.3).

Figure 3: Distribution of origin of cysts by histology.

In total 8/41 women (19.5%) underwent surgical intervention; of these eight cases six underwent major surgery under GA and two had a cyst aspiration under local anaesthesia. Seven out of these eight masses were classified as benign on USS and were subsequently confirmed to be benign by histology or cytology. In only one case a complex adnexal mass was found on USS examination at 9 weeks of gestation and the MRI scan reported possible malignancy. Tumour markers in this 23-year-old woman were normal and a laparotomy was performed at 17 weeks of gestation to remove the mass. Histology showed the mass to be a mucinous borderline tumour, FIGO stage IA. In another patient, an oophorectomy had to be performed at the time of the Caesarean section at term for fetal distress as the ovary was found to be necrotic. In this patient an ultrasound at 10 weeks of pregnancy had demonstrated a 6cm diameter haemorrhagic cyst, which had presumably torted during the pregnancy without any symptoms to prompt the patient to refer herself. Histology in this case had shown an infarcted cyst with fibrosis and calcification. In four of the major surgery cases performed under GA uncomplicated laparoscopies were performed to remove the adnexal mass; in one case a laparoscopic salpingoophorectomy was performed as an emergency for a suspected ovarian torsion at 16/40 weeks. In three cases a laparoscopic procedure for cystectomy was performed electively for ongoing pain. In the first case a cyst was diagnosed in early pregnancy subsequently there was a miscarriage and the cyst was removed 4 months after the diagnosis. In the second case a cyst was found in early pregnancy the woman had a termination at 11 weeks of pregnancy and a cystectomy 5 months later. In the third case a laparoscopic cystectomy was performed 8 weeks after the diagnosis, however the woman suffered a miscarriage at 12 weeks of gestation. Histology confirmed dermoid cysts in all four of these cases. The two cyst aspirations performed under local anaesthesia and ultrasound guidance were both symptomatic for torsion, one at nine weeks and one at ten weeks of pregnancy. In both patients the procedure has been successful. 33/41 (80.5%) with no indication for surgical intervention. There was a significant difference in the mean diameter of ovarian cysts in the expectant management group (41.2mm; 95%CI:34.7-47.7) compared with the mean diameter of cysts in the surgically managed group (74.5mm; 95%CI: 49.2-99.8) (Fig.4).

Figure 4: Mean diameter of the ovarian cysts.

In 33/41 patients no surgical intervention was needed during pregnancy. 13/33 patients had no follow up of their ovarian cyst arranged and no further mentioning of the cysts on routine growth and anomaly scans during pregnancy was found. In 20/33 patients at least one routine follow-up scan was performed 1-2 weeks after the diagnosis and in 12 of these 20 patients a second follow-up had taken place at least 1 month after the diagnosis. In one of the 20 patients with recorded follow-up’s an MRI scan had been arranged 2 months after the initial USS finding of a dermoid cyst.

Discussion

The results of our study confirm findings from previous studies: The vast majority of ovarian masses in pregnancy are benign and invasive cancer in pregnancy is rare; The results show a significant relation between size of adnexal mass and probability for surgery; Ultrasound examination of adnexal masses has been proven to be accurate and safe in pregnancy; Managing ovarian cysts in pregnancy can be challenging. Goh et al. have reported similar outcomes, namely that ovarian torsion is still reported as a rare event in pregnancy and that the management of most adnexal masses in pregnancy can be conservatively managed if asymptomatic and if there are no ultrasound findings suspicious for malignancy ⁸. If a surgical intervention is needed for persistent masses with complications such as torsion Goh et al. have found that laparoscopy during 1st and 2nd trimester can be safely performed ¹. In our cohort two out of six women underwent successful major surgery during the 2nd trimester of pregnancy. One had an emergency laparoscopy for a torsion at 16 weeks of pregnancy and the other had a laparotomy at 17 weeks of pregnancy for a mucinous borderline tumour.

However, to our knowledge currently no evident guidelines exist on how to manage and follow-up ovarian masses during pregnancy. The characteristics and presentation of ovarian mass complications in pregnancy can be mimicked by similar symptoms related to pregnancy, such as ectopic pregnancy. In one of our cases a woman with a known ovarian cyst was found to have a necrotic ovary at the time of Caesarean section, despite no signs of torsion at any time during pregnancy. This only highlights how challenging and difficult the assessment of ovarian masses during pregnancy can be. Additional diagnostic examinations such as tumour markers in suspicious ovarian masses have been found difficult to interpret in pregnancy. However, there has been literature suggesting that if a mass is strongly suspicious for malignancy, it is likely that CA-125 will be severely elevated (1000-10 000) ⁷.

The strength of this study is that the data has been collected using the expertise and facilities of a tertiary referral centre in London (GSTT). The limitations of this study include retrospective data collection, small numbers of cases and loss of follow-up. Although, our study shows the benign nature of most ovarian masses in pregnancy and the ability of ultrasound to safely characterize ovarian masses, a prospective study is required to validate our results. As it is difficult to interpret ovarian cancer tumour markers in pregnancy ^7, other models, such as the IOTA Simple Rules^14,16 or the ADNEX model¹⁷ may play a role for further characterisation of ovarian masses. A prospective trial is required to validate these models in pregnancy.

INTRODUCTION

Adult Still’s disease (AOSD) is an inflammatory disorder of unknown etiology characterized by quotidian (daily) fevers, arthritis, and an evanescent rash and multi-organ involvement [1].First described in children by George Still in 1896, subsequently in 1971 Bywaters described 14 patients with similar presentation [2]. The clinical course of adult Still’s disease (AOSD) can be divided into three main patterns: monophasic (or monocyclic), intermittent, and chronic. Patients with monophasic AOSD have a disease course that typically lasts only weeks to months, completely resolving within less than a year in most patients [3]. Systemic features, including fever, rash, serositis, and hepatosplenomegaly, predominate in this group. The patient we diagnosed as AOSD, with monophasic course, went into remission after proper treatment and is symptom free even after stopping the treatment.

CASE REPORT

46 year old Indian male, non-smoker, married, nondiabetic, normotensive admitted at department of internal medicine in our hospital with history of high grade fever, polyarthritis, and skin rash for the last 4 weeks. The fever was high grade, with maximum temperature reaching 39.2°C. The patient also complained of joint pains involving the knee, ankle, wrist and proximal interphalangeal joints. There was no history of oral ulcers, morning stiffness, ocular symptoms, or contact with infected persons. In the hospital, during the febrile period, he developed macular rash mainly on chest and back [Figure 1]. On examination, the patient was sick looking, febrile-39.2°C. Chest on auscultation was normal, cardiovascular examination was unremarkable. Examination of abdomen revealed mild spleenomegaly. Neurological examination was unremarkable. Investigations revealed hemoglobin 12.7g/dl, erythrocyte sedimentation rate (ESR) 120 mm in 1st hour. Total leukocyte count-12.7 x10 9/L. Liver function showed elevated liver enzymes with Aspartate transaminase-125U/L, Alaninie aminotransferase 60 U/L, low albumin 2.3gms/dl. He was worked on lines of pyrexia of unknown origin and his blood, urine and sputum culture showed no growth. Procalcitonin level was less than 0.5ng/ml. Sputum for AFB was negative for three samples; qunatiferon gold test for tuberculosis was negative. IgM CMV, EBV, HIV, hepatitis B and C were negative malarial parasite, Widal and Brucella serology was negative. CT-chest and abdomen were normal, except for mild spleenomegaly. Echocardiogram was normal. ANA, rheumatoid factor was negative. Lactate dehydrogenase (LDH) 978 U/L, His CRP showed a progressive increase from 82mg/L to 284 mg/L, which decreased after starting steroids. His ferritin levels were 40, 000 (normal range 21.8 -274.6 ng/ml), which were reconfirmed by second sample and he had normal transferrin saturation. On the basis of his history, clinical examination and review of his laboratory investigations, diagnosis of AOSD was made. We started him on prednisolone 60 mgs daily along with Diclofenac potassium 50 mg twice daily, to which he responded and become afebrile. He was discharged with a tapering dose of steroids 5mgs weekly. He is doing well and is completely symptom free.

Figure 1: Skin Rash on the back

DISCUSSION

First described in children by George Still in 1896, “Still’s disease” has become the eponymous term for systemic juvenile idiopathic arthritis [4]. In 1971, the term “adult Still’s disease” was used to describe a series of adult patients who had features similar to the children with systemic juvenile idiopathic arthritis and did not fulfill criteria for classic rheumatoid arthritis.

The etiology of adult Still’s disease (ASD) is unknown; both genetic factors and a variety of infectious triggers have been suggested as important, but there has been no proof of an infectious etiology, and the evidence supporting a role for genetic factors has been mixed. It is uncertain whether all patients with AOSD share the same etiopathogenic factors. Proposed pathogens have included numerous viruses; suspected bacterial pathogens include Yersinia enterocolitica and Mycoplasma pneumoniae [5]. As an example of studies of the immunogenetics of ASD, in a series of 62 French patients, human leukocyte antigen (HLA)-B17, -B18, -B35, and -DR2 were associated with AOSD. However, other studies have not confirmed these findings [6].

Adult Still’s disease is very uncommon. Prevalence of AOSD is estimated to be 1.5 cases per 100, 000-1, 000, 000 people, with an equal distribution between the sexes [6]. There is a bimodal age distribution, with one peak between the ages of 15 and 25 and the second between the ages of 36 and 46. The diagnosis of AOSD is possible only by recognizing the striking constellations of clinical and laboratory abnormalities. It is also to be to remember that AOSD is a diagnosis of exclusion. AOSD has been associated with markedly elevated serum ferritin concentrations in as much as 70 percent of patients. Serum ferritin values above 3000 ng/mL in a patient with compatible symptoms should lead to suspicion of AOSD in the absence of a bacterial or viral infection. Abnormally high serum ferritin values were reported in some case reports and it was suggested that high ferritin levels may be a diagnostic marker of Still's disease [7]. Our patient showed almost all features as laid down in Yamaguchi criteria[Table 1] for the diagnosis of AOSD [8] along with a markedly high ferritin levels.

Table 1 : Diagnostic criteria for AOSD (Yamaguchi)⁸

Exclusions: Infections, malignancy, rheumatological diseases. Five criteria with at least two major criteria. ASOD: Adult onset still’s disease. WBC: White blood cell, ANA: Antinuclear antibody, RF: Rheumatoid factor, PMN: Polymorphonuclea

Non-steroidal anti-inflammatory drugs (NSAIDs), such as aspirin, ibuprofen or naproxen, help to reduce inflammation [9]. Patients with high-fever spikes, severe joint glucocorticoids, such as prednisone (0.5- 1mg/kg/day)Methotrexate has been used successfully in a small series of people to treat adult Still's disease [10]. Some patients are refractory to these conventional therapies. Tumor necrosis factoralpha (TNF) blockers include infliximab, adalimumab, etanercept, anti-interleukin-1, antiinterleukin-6 agents, and most recently anti CD20-expressing B-cell antibodies are also effective in some cases. Other experimental drugs, including cyclosporine and anakinra, have also been successful in small groups of people [9]. Interleukin 6 inhibitors like tocilzumab showed a good result in patients with AOSD resistant to other immunosuppressive agents such as methotrexate, TNF inhibitors and anakinara [11]. Even with treatment, it's difficult to predict the course of adult Still's disease. Some people might only experience a single episode, while for others adult Still's disease may develop occasional flair up or a chronic condition. About one-third of people with the disorder may fall into each of the above groups.

CONCLUSION

A diagnosis of AOSD should be kept in mind in case of pyrexia of unknown origin particularly in a patient who presents with high-grade intermittent fever, polyarthritis and skin rash of more than two weeks duration. However, the patient should be extensively evaluated to rule out other differentials of AOSD like acute or chronic infections, autoimmune disorders, vasculitis and malignant disorders. Serum ferritin values can be powerful adjuncts in making the diagnosis of AOSD [12], where they are usually higher than other inflammatory diseases. Indeed, extreme elevation of serum ferritin up to 75, 500ng/ mL has been reported in AOSD[12]. Several investigators agree that ferritin levels above l, 000 ng/mL are suggestive of AOSD while levels greater than 4, 000ng/mL are very specific for this diagnosis when accompanied by a compatible clinical picture.

Stale and stilted inhales and exhales:
a striking absence of contentedness.

This neuroplastic beast of a brain-
wired with the wonder and wisdom to thrive

and yet, too, the demons
clinging to dendritic branches
choking cellular expression.

I can hear myself screaming
for a serotonergic surge
then goading the glia:

start pruning any circuitry that has died from shame
and find a neurogenesis waiting to be unearthed.

Sometimes, layers upon layers have to be excavated
before the brain’s resuscitation allows for easier breaths.

Introduction

Within the United Kingdom, all doctors are expected to teach.¹ It is assessed throughout their professional career, during annual appraisals for doctors in training and during consultant revalidation. But how are those just embarking on their medical career expected to develop the necessary teaching skills? As three educators at various stages in our clinical careers, we developed and delivered a small course with the aim of addressing this issue.

The General Medical Council within the United Kingdom, suggest that a basic comprehension of teaching should be gained during the undergraduate and postgraduate training of doctors.² Dandavino et al. further suggest that early development of these teaching skills may have additional benefits for the clinician; such as improving communication and assisting undergraduates to develop their own ability to learn.³ Our local training region, Yorkshire and the Humber Deanery (HEYH), have a mandatory post-graduate training day in teaching skills which focuses on generic and clinical teaching skills. This is delivered towards the end of the first foundation year. It is delivered by doctors who have various roles in medical education. Whilst useful in its content, for many it comes late. Doctors have often already been involved in providing teaching to medical students on placement at this time.

AMC recalls from her first postgraduate (foundation) year. One peer was thrilled to have ‘shaken-off’ a final year medical student who was supposed to accompany them on a shift as a learning experience; stating that they were now able to ‘do some work’. She couldn’t understand the desperation to escape one-to-one teaching. On reflection, it was probable that her colleague found it overwhelming to incorporate the additional responsibility of teaching alongside an already stressful clinical workload. Many share these feelings, with new doctors finding time pressures along with competing clinical demands a challenge to implementing clinical teaching.⁴

We thought that giving our graduates simple tools to understand and overcome these challenges may empower them as teachers. It may also improve their confidence in other areas, such as in their own learning and presentation skills.^5-6 This paper proposes a solution; after creating a short course to be delivered immediately following graduation, to empower new doctors as teachers by providing some basic training in clinical teaching. These doctors are then able to use this training as soon as they begin their foundation training, which is ultimately the beginning of their teaching career.

Methods

Two versions of a half-day course, titled ‘Teach the Medic’ were developed in HEYH which ran in successive years. The original course was designed by a surgical trainee (AMC) and a general practitioner running the undergraduate education curriculum (MS). Initial topics were chosen based on experiences of the authors and colleagues. The optional course (see figure 1), was offered to the cohort of Leeds medical students who were in the transition period between finishing their final examinations and commencing their first post as a doctor.

Figure 1: A representation of the initial course structure. Stations were developed as interactive lectures and delivered to participants by doctors of various training levels.

The initial course received encouraging verbal and written feedback from the participants, which was collected on the day of the course. Further feedback was collected a few months into foundation training, allowing enough time to pass for delegates to use this knowledge. This feedback, whilst encouraging, included that delegates were keen for additional workshop style sessions. Subsequently, a modified half-day course ran the following year, with the recruitment of additional postgraduate teachers (including LES). A further 17 newly qualified doctors from various medical schools completed the course, prior to commencing their HEYH foundation post. This modified course (see figure 2) included scenario based sessions around potentially difficult situations for the clinical teacher, and also explored alternative styles of teaching that could be adopted successfully in the workplace.

Figure 2: A representation of the modified course structure. Building on feedback from the initial course, the three co-authors incorporated new small group scenario based discussions, alongside the interactive lectures.

Results

Initial feedback received from evaluation of the day was positive for both courses. For the second course, initial feedback found that all participants found every session very good (71%) or good (29%) overall. 12/17 (71%) thought that the course should be made compulsory to medical students. We also sent a follow-up survey, distributed six months after the course which generated 14 responses. All respondents felt that the course should be run again. All participants either strongly agreed (n=2, 14%) or agreed (n=12, 86%) that they felt more confident in teaching compared with their peers. Regarding individual sessions, 10 participants (71%) had directly incorporated learning from the ‘Teaching Theory’ session, 12 (86%) from the ‘Teaching for your Learning and Portfolio’ session, 11 (79%) from the ‘Teaching your Peers’ session, and 10 (71%) from the ‘Scenarios’ workshop. All participants stated they would still recommend this course to colleagues. They also reported directly incorporating their learning from the sessions into their teaching practice. The responses gathered from the second course implied that participants felt more confident in teaching when compared to their peers.

Discussion

We feel the course content in ‘Teach the Medic’ complements other courses available later in one’s career, such as the Royal College of Surgeons’ ‘Train the Trainer’ course. We propose this course could be run by a junior doctor who has a strong interest in clinical teaching with involvement of a senior colleague with extensive medical education experience. We felt the course was especially beneficial as participants had continued to find it useful long after its delivery.

To expand this project to include a whole year group as a compulsory course is ambitious. It would require development and the use of more resources, but initial feedback suggested participants will find it extremely useful. Bing-You et al.⁶ agree, having found that undergraduate students would be willing to undertake formal instruction in clinical teaching prior to graduation.

As our short course gains momentum within HEYH, this prospect becomes more achievable. When considering a wider delivered course, one must remember that attendance to ‘Teach the Medic’ was optional; suggesting that those who attended had already identified an interest in teaching. This has the potential to bias our data to some degree. However, we still believe that making the session compulsory would allow skill development and empowerment for those who may not consider themselves aspiring medical educators, but who are still in positions to deliver teaching.

Conclusion

Our evolving teaching skills course suggests that close work with both local medical schools and deaneries is important to allow this course to be incorporated into the training of newly qualified doctors. This may be included as a compulsory part of the final year medical school curriculum, an option for a SSC, or as an integrated part of the new starter induction programme delivered by individual hospitals.

Background

This case highlights a rare and interesting medical condition bought about by liquorice ingestion.  While there have been many previous reports of liquorice toxicity secondary to eating confectionary, I have only found one case report of liquorice toxicity secondary to liquorice tea ingestion and the patient there had only a mild case of hypokalaemia and did not require hospital admission unlike patient X.¹

Case presentation

Patient X is a usually fit and well 49-year-old woman who was admitted following a collapse.  Prior to this collapse she had experienced a 3-week history of gradual onset, slowly worsening dull headache and a feeling of tingling in her hands which increased over this timeframe.  On the day of admission, she experienced nausea.  Her past medical history included migraines, for which she self-medicated with paracetamol, ibuprofen and codeine as necessary.  She was also using Cerazette.  She was in full time work, was a lifelong non-smoker and used alcohol very rarely.  She had no family history of note.

On examination, Patient X was a slim individual who was hypertensive with a blood pressure of 170/100 mmHg.  Her other observations were normal.  Her general and neurological examinations were unremarkable.

While initially the medical team felt that Conn’s syndrome was a likely cause of the abnormalities, this was reconsidered on a ward round where, following research on the internet into her abnormalities, patient X brought it to the attention of the team that she had been consuming between six and eight liquorice tea infusions per day for the past two months and had been consuming around three a day for a significant period time prior to this (roughly 18 months).  The diagnosis was made based on her history. 

Investigations

A venous gas demonstrated a metabolic alkalosis with a pH of 7.57.

Blood tests revealed hypokalaemia (K+ = 2.2 mmol/L) and hypophosphataemia (PO4 = 0.35mmol/L).  No other abnormalities were detected. 

More specialist assays were undertaken, and demonstrated that her plasma renin was 2.3 ng/mL/hour (normal range 0.2 – 3.3 ng/mL/hour).  Her morning supine plasma aldosterone level was 29 pg/mL (normal range 30 – 160 pg/mL).  Her morning plasma cortisol level was 612 nmol/L (normal range 138-635 nmol/L).

In addition, the patient also underwent a CT head and a CT renal angiogram, both of which were normal.

DIFFERENTIAL DIAGNOSIS

• Apparent mineralocorticoid excess
• Exogenous mineralocorticoid excess
• Liddle’s syndrome
• Congenital adrenal hyperplasia
• Cushing syndrome
• Liquorice

Treatment

Intravenous potassium and phosphate replacement, cessation of liquorice intake and amlodipine 5mg OD.

Outcome and follow-up 

The patient was discharged with a blood pressure of 132/66 mmHg on amlodipine, a potassium level of 2.9, and a phosphate level of 1.16.  She was given oral potassium supplementation to be taken 3 times per day.  After two weeks, the amlodipine was stopped as she became mildly hypotensive.  Her blood pressure was 120/60 following cessation of amlodipine.  Three months later her plasma potassium level was 4.6 without supplementation.  Mrs X required no follow up although she reports an ongoing feeling of tingling in her hands. 

Discussion

Liquorice is an extract of the roots of the Glycyrrhiza glabra plant and has been used as both a confectionary flavouring agent and a herbal remedy.  It is also commonly used as a laxative, and as a flavouring agent in chewing gums, sweets, and food products.

The active ingredient in liquorice is glycyrrhetinic acid which inhibits the enzyme 11-β-hydroxysteroid dehydrogenase.  This enzyme converts cortisol into inactive cortisone within the distal tubule of the kidney and so in liquorice toxicity there is a build-up of cortisol in distal tubular cells.²  This results in increased mineralocorticoid like activity as there are structural similarities between cortisol and aldosterone, with increased Na+ and water retention in conjunction with increased H+ and K+ excretion.³  Here hyperaldosteronism occurs, but with a low or low-normal plasma aldosterone and renin level.  Serum glycyrrhetinic acid levels can be measured with enzyme-linked immunosorbent assay (ELISA) and high-performance liquid chromatography (HPLC).   Urinary glycrrhetinic acid levels can be measured with gas chromatography-mass spectrometry (GC-MS).⁴

Pseudoaldosteronism secondary to liquorice consumption is a relatively rare occurrence.  Case reports demonstrate a range of clinical manifestations from an asymptomatic patient fortuitously diagnosed to those with more severe presentations such as rhabdomyolysis, hypertensive encephalopathy, asthenia, paralysis, heart failure, and cardiac arrhythmias such as polymorphic ventricular tachycardia and ventricular fibrillation secondary to hypokalaemia.  For these reasons, it has been suggested that the public should be made aware of the potential dangers associated with liquorice consumption. ^5-13

The combination of alkalosis hypokalaemia and hypertension suggests increased mineralocorticoid activity leading to increased renal tubular Na+ reabsorption along with increased k+ and H+ excretion.  Both primary and secondary hyperaldosteronism cause these abnormalities, the former via an appropriate response (renin release) to decreased renal perfusion pressure or decreased sodium concentration in the ultra filtrate, the latter an inappropriate release of aldosterone from the adrenal cortex, often a result of an adrenal adenoma.

Other genetic syndromes, such as Bartter’s or Gitelman’s, cause hypokalaemia with alkalosis but without hypertension.¹⁴

Features of low potassium include generalised weakness and lethargy, ascending paralysis, and rhabdomyolysis.^15-17   Decreased intake is rarely a cause of low potassium as the western diet usually contains significantly more potassium than is needed and because the renal tubular reabsorption mechanism can be extremely effective in limiting potassium excretion.¹⁷

The maximum recommended dose of liquorice is 100mg/day although cases of liquorice toxicity have been reported in association with doses as low as 80mg/day.  Each liquorice tea bag contains approximately 500mg of glycyrrhetic acid, of which approximately 20mg is ingested per infusion.

This is a relatively rare occurrence and it has been suggested that certain groups are more susceptible to toxicity than others – for example those with 11-β-hydroxysteroid dehydrogenase deficiency.¹⁸  It is also thought that those with essential hypertension are also more at risk.¹⁹

LEARNING POINTS/TAKE HOME MESSAGES

• Consumption of liquorice can cause pseudoaldosteronism.
• The clinical picture is similar to that of primary aldosteronism, but is characterised by low levels of both aldosterone and renin.
• While liquorice toxicity can be asymptomatic, clinical manifestations are wide ranging and include cardiac arrhythmias, rhabdomyolysis, weakness and paralysis.
• Pseudohyperaldosteronism caused by liquorice consumption is reversible and generally resolves upon cessation of liquorice consumption.  Prior to resolution, potassium supplements are usually necessary.

Introduction

Spasticity was first described by Lance¹ in 1980, and according to him it was described as; a motor disorder characterised by a velocity-dependent increase in tonic stretch reflexes (muscle tone) with exaggerated tendon jerks, resulting from hyperexcitability of the stretch reflex, as one component of the upper motor neuron syndrome. Spasticity can be a consequence of many neurological conditions including traumatic brain injury, spinal cord injury, stroke and multiple sclerosis. The annual incidence of spasticity in the lower limb following a stroke, traumatic brain injury and spinal cord injury is estimated to be 30 to 485 per 100,000, 100-235 per 100,000 and 0.2 to 8 per 100,000 respectively². Spasticity is characterised by muscle overactivity and can lead to permanent changes in the muscle fibres leading to muscle contractures. Contractures can be very painful and may interfere with seating, posture, mobility and activities of daily living, thus increasing the care cost significantly.

Phenol has been used peripherally and intrathecally for the treatment of spasticity for many years. The botulinum toxin became available in the last decade for treatment of spasticity. Its use has increased since then, and this has led to a decline in the use of phenol. It is still being used in patients who are sensitive to botulinum toxins or have developed antibodies to them. Phenol is both neurolytic and anaesthetic in nature³. The anaesthetic effect of phenol can be seen immediately after the injection where the patient reports an immediate effect. The neurolytic effect takes at least two weeks, and therefore patients should be educated not to expect any significant change in the spasticity before two to four weeks. Phenol can also be used in combination with botulinum toxin to treat multifocal spasticity where the maximum dose of botulinum exceeds the recommended safe dose. This allows several groups of muscles to be treated in a single setting³.

The lethal dosage of phenol has been reported to be greater than eight grams⁴. Phenol in aqueous solution is preferred for peripheral nerve and motor point blocks and is available in 5, 6 and 7% concentration. Injecting botulinum toxins is quite different from performing nerve and motor point blocks. Phenol nerve and motor point blocks take a longer duration of time to perform as compared to botulinum toxins. For motor point blocks, a nerve stimulator with a surface electrode is needed to localise the motor points on the muscles. In the present study, we highlight the importance of management of spasticity in adults with a combination of botulinum toxin and phenol nerve /motor point blocks. A case series of patients who underwent combined phenol and botulinum toxin is presented, describing the diagnosis, number and location of muscles injected, types of phenol nerve and motor point blocks and any complications encountered.

Methods

This is a retrospective study conducted at the Rehabilitation Medicine Department of the University Hospital in Cambridge UK. The study period included from December 2014 to January 2017. The patients were identified from the spasticity clinic database. All patients were assessed in the spasticity clinic, and a plan to inject the botulinum toxin along with phenol nerve block/motor point block agreed with the patient. The patients who decided to have the procedure were appointed to a clinic to perform the agreed injections and blocks. If the patients were on anticoagulants (warfarin, dalteparin or clopidogrel), they were advised to stop the anticoagulation 3 days before the procedure. International Normalised Ratio (INR) was checked before the procedure. The usual dose of anticoagulation was started after the procedure.

Patients were consented and placed on a plinth. Botulinum toxin type A only was used in our study. It was diluted with normal saline, and the muscles were injected either using the surface anatomy or electrical stimulation. Each muscle was either injected at one to two sites, depending on the size of the individual muscle.

Phenol nerve blocks and motor point blocks were performed according to the techniques described by Roy³ and Gaid⁵. Aqueous phenol 5% (phenol in water) was used for all the procedures. The nerves were identified using a nerve stimulator with a surface electrode using 2mA current (Figure 1). The skin was infiltrated with 1% lignocaine, and the nerve was approached with a stimulator needle. The nerve was then ablated with 5% phenol under stimulation guidance. The dose of the phenol was titrated while the nerve was being stimulated. The motor points were located similarly with the help of a surface electrode and marked before ablation with 1 to 2 ml of 5% phenol. The amount of botulinum toxin and phenol was recorded. All patients were reviewed in 6 weeks’ time for any complications.

Figure 1: Nerve Stimulator with surface electrode

Results

Between December 2014 to January 2017, we treated 29 patients with spasticity caused by different neurological conditions with a combination of aqueous phenol and botulinum toxin injections. There were 15 males and 14 females with an age range of 18 to 80 years and a mean age of 49.3 years. The most common diagnosis was multiple sclerosis followed by stroke (Figure 2). A total of 40 phenol nerve or motor point blocks were performed in 29 patients. Nineteen patients (65.5%) received phenol blocks once, 9 (31%) twice and only 1 patient (3.4%) had the phenol block done three times. Where the phenol blocks were repeated, the mean duration between the phenol injection was 14.1 months (range 6-23 months). The procedure was bilateral in 16 (55.2%) and unilateral in 13 (44.8%). The local anaesthetic (trial block) was performed in 6 (20.6%) patients who were ambulatory before the phenol block.

Figure 2: Frequency of Diagnosis

Obturator nerve block was the most common phenol procedure performed (44.8%), followed by posterior tibial nerve block (37.9%). Two (6.9%) patients had both obturator nerve blocks and posterior tibial nerve blocks, whereas 1(3.4%) patient had hamstring motor point blocks, 1(3.4%) patient had gastrocnemius motor point blocks. One patient (3.4%) had bilateral obturator nerve blocks, posterior tibial nerve blocks and rectus femoris motor point blocks (Figure 3).

Figure 3: Frequency of Phenol Nerve/Motor Point Blocks

Botulinum toxin was also injected into various muscles in all 29 patients. The botulinum was repeated every 4 to 6 months in the same muscles. Botulinum toxins were injected bilaterally in 12 (41.4%) and unilaterally in 17 (63.6%) patients. The most common muscles injected with botulinum toxin were hamstrings (44.8%) followed by finger flexors (13.8%). The frequency of botulinum toxins injections is shown in Figure 4.

Figure 4: Muscles Injected with Botulinum Toxins

The most common combination in our series was obturator nerve block and hamstring botulinum toxin injections (34.4%). The combination of posterior tibial nerve block with hamstring botulinum toxins was used in 3 (10.3%), and 2 (6.8%) patients received posterior tibial nerve block with finger flexor botulinum toxin injections. The combination of phenol and botulinum toxin injection is shown in Table 1. There were no complications noted following both phenol as well as botulinum toxin injections.

Table 1: Combination of Phenol and Botulinum Toxins used

Discussion

Perineural injection of aqueous phenol (3 to 7%) can reduce spasticity by blocking the nerve signals to the group of muscles supplied by the nerve. Phenol produces an initial local anaesthetic effect which is followed by neurolysis caused by protein coagulation and inflammation⁶. The neurolysis leaves the nerve with about 25% less function than before but does not disadvantage people with little or no residual function, as a mild progressive denervation can be beneficial in reducing spasticity⁶. Khalili et al⁷first described the technique of phenol nerve blocks and also suggested that the re-growth of most axons is seen with preservation of gamma motor neurons. This means that phenol reduces spasticity without reducing the strength of the muscle significantly.

The use of combining phenol with botulinum toxins injections has been documented in children with cerebral palsy and central nervous degenerative diseases⁸. To date, there are no studies in the literature showing the use of combined phenol and botulinum toxins in the treatment of spasticity in adults. The combination of phenol with botulinum toxin helps to treat multifocal spasticity allowing more spastic areas to be treated. The most frequent pattern used in Gooch et al⁸ study was obturator nerve block and gastrocnemius botulinum toxin injections. In our study, the most common combination was obturator nerve block and hamstring botulinum toxin injections. The possible explanation for this variance is that the majority of our study population suffered from multiple sclerosis and hamstring with hip adductor spasticity is a very common pattern.

The mechanism of action of phenol is different from botulinum toxins. However, the reduction in spasticity with phenol and botulinum toxins is comparable. Manca et al⁹ compared botulinum toxins and phenol nerve blocks to reduce ankle clonus in spastic paresis and concluded that both patient groups showed significant clonus reduction over time with the phenol group effect greater than the botulinum toxins group. They also suggested that the two drugs have a different mechanism of action with phenol reducing the excitability of the alpha motor neuron. A randomised double-blind trial by Kirazli et¹⁰ al compared the effects of botulinum toxins Type A and phenol on post-stroke ankle plantar flexor and invertor spasticity. There was a significant change in Ashworth scores at week 2 and 4 in the group who received botulinum toxins but there was no significant difference between the two groups at week 8 and 12¹⁰. Similarly, the decrease in clonus duration (detected by electromyography) was significant in both groups. However, the group that received botulinum toxins showed significant change at week 2 and 4 compared to phenol group. The reason for this may be the delayed onset of action of phenol as compared to botulinum toxins. Burkel et al⁶ studied the effects of phenol into the peripheral nerves of rats and showed that Wallerian degeneration of the nerves occurs before healing by fibrosis that starts after about 4-6 months following phenol injections. Their study also concluded that following phenol the nerves are left with 25% less function than before and this does not disadvantage the people with little or no residual function⁶.

There is always a risk of deteriorating the mobility or function due to weakness caused by the phenol nerve block. It is our usual practice to perform a local anaesthetic block (trial block) before injecting the phenol in all ambulatory patients or patients who are using spasticity functionally to their advantage. In our series, 20.6% of patients underwent local anaesthetic block before proceeding to the phenol block. There were no adverse effects noted following the local anaesthetic block, and all six patients chose to have the phenol blocks. A recent study by McCrea et al¹¹ looked at the effects of phenol on position and velocity components of spasticity in addition to strength in post-stroke elbow flexor spasticity. The study concluded that phenol paradoxically improved muscle strength in addition to reducing hypertonia¹¹.

In our series, we used phenol mainly for lower limb muscles and botulinum toxins for both the lower and upper limb muscles. For smaller muscles of the upper limb, it is difficult, but not impossible, to find the motor points. The technique for upper limb phenol blocks has been well described in literature³. However, when combining the botulinum toxins with phenol, we find it useful to prefer the phenol block for the lower limb muscles. Gooch et al⁸ also injected larger proximal muscles with phenol, and smaller distal and deeper muscles with botulinum toxins. In our series, the maximum dose of botulinum toxins used was 1000 units of Dysport and the maximum dose of phenol used was 20 ml of 5% aqueous phenol.

Conclusion

The combination of botulinum toxin with phenol injections is effective in treating multi-focal spasticity in clinical settings. The advantage of using phenol in combination with botulinum toxins is cost-reduction and the flexibility of managing various muscle groups at the same time. Further studies are needed to evaluate the long-term cost-effectiveness and complications of combining phenol and botulinum toxins, especially after repeated injections.

INTRODUCTION

Anaesthetic management of a patient with a tracheal tumour is challenging, as the airway is shared with the surgeon and patency must be maintained despite airway manipulation.

Several anaesthetic techniques have been used in patients requiring tracheal resection and reconstruction. Cardiopulmonary bypass standby after femoral artery and vein cannulation and then intravenous/inhalational induction while oxygenating the patient has been considered to be a reasonable approach. Intratracheal tumours are challenging to anaesthetists because of the difficulty in establishment of a patent airway before commencement of surgery. The principal anaesthetic consideration is ventilation and oxygenation in the face of an open airway.

CASE REPORT

A 56 year old male with no other co-morbidities presented to the Thoracic Oncology department with a history of progressive dyspnoea and orthopnoea.On examination he was found to have dyspnoea at rest and could not complete full sentences while talking. Change of position made no difference to his symptoms.

Routine blood investigations which included full blood count, renal and liver functions, coagulation profile, ECG and 2DEcho were within normal limits. PFT showed a typical intrathoracic obstructive picture.

His chest X-ray showed bilateral hyperinflated lungs suggesting airtrapping. CT of the chest showed an intratracheal growth about 4 cm above the carina almost completely obstructing the lumen. An awake flexible bronschoscopy confirmed the CT scan findings. A 2.7mm flexible bronschocope was passed with difficulty beyond the tumour to visualise the carina. Excision of the intratracheal tumour was planned with possible resection and anastomosis of the involved tracheal segment. A careful perioperative plan was  discussed and decided in agreement with the thoracic surgeons, anaesthetist, cardiovascular surgeons and the rest of the team members.

Flexible and rigid bronchoscope, a Sanders venturi, an additional anaesthesia machine and various sizes of reinforced and normal endotracheal tubes and tracheostomy tubes were kept ready.

Preoperatively the patient had incentive spirometry and bronchodilator nebulisation and intravenous steroids. An awake epidural at T9-10 level and radial artery cannulation were done under local anaesthesia without any problems. Two 16 gauge peripheral IV lines were sited under local anaesthesia.

After adequate preoxygenation anaesthesia was induced with IV propofol along with oxygen and sevoflurane with BIS monitoring.  As mask ventilation proved to be easy the patient was paralysed with suxamethonium. There was no difficulty in ventilation after muscle paralysis. An 8.5 number COETT portex tube was placed in  the trachea with the cuff just beyond the cords to avoid possible trauma to the tumour. Since there was preoperative evidence of airtrapping, ventilator settings were set to an I:E ration of 1:3 with a tidal volume of 550ml, respiratory rate of 12-14 per minute and PEEP of 4. At these ventilator settings the airway pressures were reaching up to 22 cm of  H20 and ETCO2 reaching a maximum of 40mmHg. Anaesthesia was maintained with oxygen: air with sevoflurane and atarcurium for muscle paralysis.

Flexible bronchoscopy was done to confirm the position of the endotracheal tube, which showed that the ETT was adequately above the tumour.

A laryngeal drop procedure was done in the supine position with neck extension to facilitate mobilisation of the trachea for resection anastomosis. After the laryngeal drop procedure a right thoracotomy was done in the left lateral position. At this point of the procedure, the patient was ventilated with low tidal volumes of 300 and respiratory rate of 16-20 to keep the ETCO2 at around 40.The right lung was surgically retracted and the trachea  was exposed up to the carina. A repeat bronchoscopy was done through the ETT to help identify the upper and lower extent of the tumour. The trachea was then opened below the tumour, after which a 6.5 reinforced tube was introduced through the left bronchus to aid ventilation of the left lung. This ETT was withdrawn intermittently to help visualisation and aid surgical excision of the tumour and  sleeve resection of the trachea. The left lung was ventilated till partial closure of the trachea. The left-sided tube was then removed. Ventilation resumed through the orotracheal tube with intermittent occlusion of the defect with gauze by the surgeon.  The orotracheal tube was adjusted under vision before closure of the trachea to position it above the anastomotic site. The trachea was sutured and the thoracotomy incision closed without any adverse event. The neck was kept in a flexed position to avoid tension on the tracheal anastomotic area.

The patient was then extubated in the immediate postoperative period without any problems and the recovery was uneventful.

DISCUSSION

Anaesthetic management of a patient with a tracheal tumor is challenging, as the airway is shared with the surgeon, and patency must be maintained despite airway manipulation^{1, 2}. Several anaesthetic techniques have been used in patients requiring tracheal resection and reconstruction ^3–5.

Primary tracheal masses are very rare and mostly malignant, occurring in 0.2 in 1,00,000 persons per year ⁶ and among these squamous cell carcinomas form the main bulk. Cardiopulmonary bypass standby after femoral artery and vein cannulation and then intravenous/inhalational induction while oxygenating the patient with the oxygen inhalation has been considered to be a reasonable approach ⁷. Byrne JG et al (2004) advocated planned use of CPB to facilitate complete resection of thoracic malignancies after careful patient selection ⁸.

These patients are often mistaken to have asthma and require treatment with inhaled corticosteroids and beta agonists ⁹. They are generally treated for many years for asthma or COPD, unless a CT scan or endoscopic procedure is done for the symptoms ¹⁰. Intratracheal masses usually start getting symptomatic when 75% or more of the tracheal lumen is obstructed. Tracheal lesions present at lower level can have more complicated management of airway, anaesthesia and surgery for successful and safe removal of the mass. ¹⁰

Intratracheal tumours are challenging to anaesthetists because of the difficulty in establishment of a patent airway before commencement of surgery. The principal anaesthetic consideration is ventilation and oxygenation in the face of an open airway. Ventilation can be managed in many ways, including manual jet ventilation, high frequency jet ventilation, distal tracheal intubation, tracheostomy, spontaneous ventilation and CPB.¹¹

Knowledge of various techniques available for management of such cases is vital. In order to have a successful and safe outcome it is extremely important to have good communication between the anaesthetic, surgical and intensive care team.

The challenge in managing such cases lies in establishing and maintaining a patent airway and also preventing seepage of blood and tumour particles distally into the tracheobronchial tree during the surgery.

There is a possibility of total airway obstruction during ventilation attempts using positive pressure because airway obstruction has a fixed and dynamic component. Dislodgement of the tumour, possibly from trauma following intubation causing total obstruction, should also be considered.

Thus an intratracheal tumour was successfully removed without any complications and by avoiding CPB. This case report also highlights the importance of proper planning and good communication between team members to ensure a successful and safe outcome.

Introduction

Increasing number of term deliveries undergo induction of labour (IOL). This figure is as high as 1 in 4 in developed countries, making it one of the most common procedures a woman may experience in pregnancy. ¹ IOL may be achieved with pharmacological, mechanical or surgical methods. ^{1, 2} Mechanical methods were the first methods used to ripen the cervix and induce labour. The National Institute of Clinical Excellence (NICE) does not recommend the routine use of mechanical methods for IOL as only heterogeneous small studies were available at their time of publication more than half a decade ago. ² However, since then there is increasing evidence of safety and efficacy of mechanical IOL. Subsequent publications including those from World Health Organization (WHO) and Cochrane Database of Systematic Reviews support the use of balloon catheter for IOL. ^{1, 3} It is therefore important to revisit the role of mechanical methods of IOL.

The Cochrane Database of Systematic Reviews concluded that mechanical methods of induction of labour have a lower risk of uterine hyperstimulation with similar caesarean section rates and delivery within 24 hours as prostaglandins. Furthermore, mechanical methods reduce the risk of caesarean section when compared with oxytocin induction of labour. ³ This review is consistent with another earlier systematic review. ⁴

Both Pfizer’s Prostin (PGE) and Cook Medical’s Cervical Ripening Balloon (CRB) are licensed for IOL. While the use of Prostin is a standard care in Singapore, the CRB has not been used routinely. We therefore propose a study to evaluate the use of CRB for IOL in Singapore.

Methods

A prospective cohort randomised controlled study was conducted in a tertiary referral maternity unit in Singapore. Pregnant women aged 21 – 40 years old with a singleton pregnancy with no major fetal anomaly who were suitable for vaginal delivery and scheduled for a planned IOL at 37⁺⁰ to 41⁺⁶ weeks gestation were invited for the study. Cases were excluded if at the start of the planned IOL, they were in spontaneous labour, had a cervical dilatation of ³3 cm, had confirmed rupture of membrane, had abnormal cardiotocogram (CTG), had a scarred uterus such as previous caesarean section, had malpresentation in labour, or if caesarean section delivery was indicated. Women who were unable to give or had withdrawn their consent to participate in the trial were also excluded for the study.

All suitable pregnant women receiving team care who require elective IOL were identified in antenatal clinic, antenatal or labour wards by the attending doctor or clinical research coordinator (CRC). Following routine counselling for IOL by the attending doctor, the woman will be offered participation in the study and a member of the research team will counsel and obtain informed consent from her. The woman will be made to understand that participation in the study is voluntary, does not affect her medical care and consent for participation can be withdrawn at any stage of the study. Women who were uncertain in their participation were offered the opportunity to participate during her follow-up or on the day of IOL after further consideration. Patient information leaflet on IOL as well as information of the study were made available to the participants.

On the day of the IOL, the participants were reviewed for the appropriateness of the IOL and participation in the study. A presentation scan, vaginal examination for cervical dilatation and CTG were performed. If they were suitable, they were randomly allocated PGE or CRB IOL in labour ward. Randomization was achieved with third party sealed envelope allocation. A total of 75 envelopes containing a folded paper with the words “Cervical Ripening Balloon” and another 75 identical envelopes containing a folded paper with the word “Prostin” were prepared and shuffled after sealing. These randomized envelopes were then labelled sequentially with their randomization allocation number from 1 to 150. The participants who underwent randomization were allocated to the next randomization allocation numbered envelop which contain either allocation for CRB or PGE IOL.

Participants undergoing CRB IOL will have the CRB inserted after cleaning the vulva and vagina with Cetrimide solution. The uterine and vaginal balloons of the CRB will be gradually inflated with normal saline, initially 40 ml and 20 ml respectively, and a further 20 ml each hour later until each balloon is 80 ml. CTG monitoring was undertaken before and after each inflation for at least 20 minutes. If the participant was not in labour after complete inflation of the balloons, she would be transferred to the antenatal wards for rest before removing the CRB 12 hours after insertion in labour ward when possible.

Participants undergoing PGE IOL will have 3 mg Prostin tablet inserted in the posterior fornix after cleaning the vulva with Cetrimide solution. CTG monitoring was also undertaken for at least 40 minutes after PGE insertion. If the participant was not in labour, she would be transferred to the antenatal wards. If there was no response to the first PGE, a repeat dose was given after 6 hours in labour ward when possible.

Participants will undergo artificial rupture of membrane (ARM) and/or oxytocin infusion augmentation of labour as necessary. If the participant was not in labour or ARM was not possible after removing the CRB or 2 cycles of PGE, the participant would have been considered having a failed IOL and will leave the study protocol with her subsequent management determined by the specialist attending to her. This would typically involve insertion of a third or first PGE in the PGE or CRB arm respectively.

Upon delivery of the pregnancy, a member of the research team will interview the participant and obtain demographics, labour and delivery outcomes data from the clinical notes. Pain and maternal satisfaction scores and comments were also recorded by interviewing the participants in the post-natal period; these findings will however be discussed separately.

The data was collected on a pro forma and entered into an excel spreadsheet. The data was then analysed using IBM SPSS Statistics version 19.

This study was approved by the SingHealth centralised institutional review board with the reference number of 2013/553/D.

Results

A total of 138 women were approached to join the study but 40 (29.0%) women declined. There was no significant difference in maternal age (27.8 ± 5.4 vs 28.7 ± 5.2 years; p = 0.373), ethnicity, proportion of primigravidae (62.5% vs 53.1%; p = 0.349), weight (61.2 ± 15.4 vs 64.4 ± 13.8 kg; p = 0.228), BMI (24.8 ± 5.8 vs 25.3 ± 5.0 kg/m²; p = 0.646) and primary indication for IOL between women who declined and accepted enrolment to the study respectively.

The remaining 98 women were enrolled for the study. Eight-seven women were randomized after excluding 6 women in spontaneous labour, 1 woman with non-cephalic fetal presentation, and 1 woman had confirmed ruptured of membrane on admission for their IOL, as well as 3 other cases in which the women presented for IOL without the availability of the research team (figure 1).

Figure 1. Flow diagram of recruitment, randomisation and completion status

In the CRB arm, one woman withdrew from the study after 8 hours 55 minutes as she felt the discomfort was too unbearable. Another woman was excluded when she was found to have spontaneous version to breech in labour. One woman randomized to PGE did not receive it as she went into spontaneous labour prior to IOL. Another woman in the PGE arm was subsequently found to be only 36⁺³ weeks and was therefore excluded from analysis (figure 1). The remaining 83 cases were analysed and their characteristics are shown in table 1.

The induction to vaginal delivery time, as well as, vaginal delivery rate were similar in both arms of the study (table 2). Compared to PGE arm, participants undergoing CRB IOL were faster in achieving cervical dilatation ≥4 cm (14.4 ± 5.7 vs 23.5 ± 16.6 hr; p = 0.001) and requesting epidural (16.4 ± 5.4 vs 23.2 ± 15.8 hr; p = 0.040), as well as more likely to require oxytocin infusion for augmentation (77.4% vs 50.0%; p = 0.020). Uterine hyperstimulation defined as >5 contractions every 10 minutes was only found in PGE arm. Cervical dilatation from 0 – 2 cm to ≥4 cm was achieved without regular contractions in 2 (6.9%) cases in the CRB arm and 1 (2.4%) case in the PGE arm. The mean frequency of uterine contractions at cervical dilatation ≥4 cm was 2.5 ± 1.4 in 10 minutes for CRB arm compared to 3.8 ± 1.4 in 10 minutes for PGE arm (p <0.001). No case of uterine rupture was observed.

There was 1 (3.2%) case for failed CRB IOL where both uterine and cervical balloons were found in the vagina suggesting that either placement of the uterine balloon was not optimal or it was expelled after placement. The woman went on to have Prostin and delivered vaginally. In the 9 (17.3%) cases in the PGE group that did not respond after 2 cycles, all went on to have the third Prostin successfully except for 2 women who required Caesarean section for persistent failed IOL.

The birth outcomes of both arms of the study were also similar with no case of stillbirth (table 3). There were 2 case of neonatal intensive care unit admission in the PGE arm for continuous positive airway pressure therapy; both were discharged from NICU within 24 hours.

Table 1. Characteristics of participants undergoing cervical ripening balloon (CRB) and Prostin (PGE) induction of labour.

¹ Values are mean ± SD, p calculated with Student t-test; ² Values are percentage (n), p calculated with Pearson chi-square test; ³ Values are percentage (n), p calculated with Fisher’s exact test.

Table 2. Labour outcomes of participants undergoing cervical ripening balloon (CRB) and Prostin (PGE) induction of labour.

¹ Values are mean ± SD, p calculated with Student t-test; ² Values are percentage (n), p calculated with Pearson chi-square test; ³ Values are percentage (n), p calculated with Fisher exact test.

Table 3. Birth outcomes of participants undergoing cervical ripening balloon (CRB) and Prostin (PGE) induction of labour.

¹ Values are mean ± SD, p calculated with Student t-test; ² Values are percentage (n), p calculated with Pearson chi-square test; ³ Values are percentage (n), p calculated with Fisher exact test.

Discussion

To the best of our knowledge, this is the first randomized controlled study to assess the use of CRB for IOL in Singapore. Our study concur with the published literature that both CRB and PGE have similar rate of vaginal deliveries and rate of deliveries within 24 hours. Both methods are effective and safe with PGE having a higher risk of uterine hyperstimulation and need for Caesarean section for failed IOL.

Pharmacological induction of labour using PGE is the most established form of IOL. However, it is important to be able to offer alternative methods to women particularly in cases of hypersensitivity or allergy to PGE. PGE can cause bronchospasm complicating asthma, a medical condition which affects 4 – 12% of pregnant women. ^{5, 6} Similarly, caution should be exercised in the use of PGE in women with other common medical conditions such as hypertension and epilepsy.

In addition, women may not respond to PGE for IOL, or the PGE may only result in uterine tightenings which do not lead to cervical dilatation. In these situations the CRB may be considered as an adjunct for IOL to avoid Caesarean section of ‘failed IOL’.

The risk of uterine hyperstimulation and the need for a repeat dose in 6 to 8 hours for PGE typically require the women to be admitted for IOL. The use of CRB does not require planned intervention until 12 hours later. This potentially allows an outpatient IOL if further studies support its safety in this aspect.

The application of PGE is relatively straightforward and is already performed by both doctors and midwives. The insertion of CRB may however be considered too invasive for midwives thus limiting the type and hence availability of staff to commence IOL. We have explored the learning curve in the insertion of CRB and will discuss this separately.

Conclusion

Both CRB and PGE are effective methods for IOL at term. Each method has its own benefits and limitations. The availability of both methods in an obstetric unit will allow the clinician to choose the most appropriate form of IOL, provide a complementary method of IOL, as well as offer women choice in their IOL.

Introduction

Oesophageal cancer (OC) is the eighth most common cancer affecting an estimated 481,000 people worldwide with a rapidly rising incidence. Due to the poor prognosis of patients with these cancers it is the sixth leading cause of cancer related mortality with 406,000 deaths.^1,2 Although the overall 5-year survival has increased from 4% in the 1970s³ to currently ranging between 15 to 20%⁴, it remains a challenge to treat as clinical presentation is often late and diagnosis is made at advanced stages. Incidence and mortality rates for OCs are two fold higher in males compared to females, however this ratio rises to up to 5-10:1 for oesophageal adenocarcinomas. Cohort studies have shown that the incidence of OC increases with age; the average of onset is between 65 to 70 years. ¹⁴ This article seeks to discuss the epidemiology, diagnosis and staging, prevention and current trends in the management of OC.

Methods

We searched PubMed/MEDLINE, EMBASE and Cochrane Library and Central Register of Controlled Trials (CENTRAL) databases up to November 2014. Our search strategy used a combination of MeSH, textwords, and appropriate words variants of “oesophagus”, “cancer”, “epidemiology”, “adenocarcinoma”, or “squamous cell carcinoma”, and “staging”, “transhiatal oesophagectomy”, “transthoracic oesophagectomy”, “chemotherapy”, “radiotherapy”. This was supplemented with selected systematic reviews, evidence based guidelines and consensus statements.

Epidemiology

There have been major changes in the epidemiology of OC over the last thirty years. The two key histological types of OC are adenocarcinoma and squamous cell carcinoma (SCC) and they differ significantly in their fundamental patterns of incidence and aetiological factors. Oesophageal SCC comprises the majority of cases worldwide and represents 90% of all OCs in most Eastern countries. However the incidence of adenocarcinoma has risen rapidly over the last three decades and it is now the predominant histological type in Western Europe, USA and Australia, particularly amongst white males.^{5, 6} There are other rare histological types, which include lymphoma, leiomyosarcoma, melanoma, rhabdomyosarcoma and small cell carcinoma.⁷ OC accounts for almost 3% of all cancers in the UK and is the ninth most common malignancy in the UK. There were 8,173 new cases in 2008; incidence rates have increased over the last thirty years in the UK and are now one of the highest in Europe.⁸ Incidence rates of OC differ markedly by geographical locations and between ethnic groups; overall, rates are twice as high in less developed regions compared with more-developed regions and the highest rates occur in Asia. In this region, especially in Iran, Turkey, Kazakhstan and China, a very high incidence of oesophageal SCC exists with greater than 100 cases per 100,000 population annually. A similar trend is also seen in South Africa.^9-12 In contrast, the rate of rise in incidence of oesophageal adenocarcinoma in more-developed countries has exceeded that of oesophageal SCC, which has remained the same or decreased. Oesophageal adenocarcinoma now comprises approximately 50% of all OCs in these countries. ¹³

Who gets oesophageal cancer?

The aetiology of OC is multifactorial, with interactions between environmental risk exposures and genetic factors. These can be divided between the two different histological types of OC.

Pathology of oesophageal tumours

Oesophageal tumours are classified as epithelial and non epithelial. Epithelial tumours include papilloma, intraepithelial neoplasia, carcinoma and carcinoid tumours. Non epithelial tumours include leiomyoma, lipoma and gastrointestinal stromal tumours (Table 1).

Table 1: WHO histological classification of oesophageal tumours

Oesophageal adenocarcinoma

Established risk factors for oesophageal adenocarcinoma (Fig. 1) include gastro-oesophageal reflux disease, Barrett’s oesophagus, obesity, male sex, tobacco smoking and a low intake of fruit and vegetables.^{15, 16} There is evidence to suggest that previous infection with Helicobacter Pylori and the use of non-steroidal anti-inflammatory drugs may decrease the risk of OC. ¹⁷

Fig. 1: Adenocarcinoma of the oesophagus (from Lewin et al. Gastrointestinal pathology and its clinical implications)

Barrett’s oesophagus (Fig. 2) occurs when there is metaplastic change in the lining of the oesophagus from normal stratified squamous mucosa to single layered columnar glandular mucosa with variable degrees of goblet cell differentiation.¹⁸ This transition usually occurs in the context of chronic gastro-oesophageal reflux disease, which causes exposure of the epithelium to refluxate. Gastro-oesophageal reflux disease is a major contributory factor and 5% of people with reflux disease develop Barrett’s oesophagus. The estimated prevalence of Barrett’s oesophagus is just under 2% amongst adults in the West and the annual incidence is approximately 0.1%. However, there is evidence to suggest that the rate of diagnosis is increasing by 2% annually.¹⁹ There has been a rise in the incidence of gastro-oesophageal reflux disease, which may be explained by a number of factors. The rise in the prevalence of obesity, specifically central and intra-abdominal obesity has been found to have a link with oesophageal adenocarcinoma. This can be explained by the fact that an increase in adiposity will cause a rise in intra-abdominal pressure thereby increasing reflux that may be asymptomatic. However, studies also suggest that obesity is a strong independent risk factor for oesophageal adenocarcinoma regardless of gastro-oesophageal reflux symptoms implying an underlying link. ^{20, 21} Another factor that may contribute to the rise in reflux disease is the increased use of drugs that relax the lower oesophageal sphincter. There is evidence to suggest that individuals with previous H. Pylori infections are less likely to develop oesophageal adenocarcinoma.²² This might be explained by the gastric atrophy that results from this infection, which will reduce the acidity and quantity of gastric secretions and thus decreasing the chances of gastro-oesophageal reflux. However, the prevalence of H. Pylori infections is decreasing in the Western population, which may contribute to the rising incidence of oesophageal adenocarcinoma. Gastro-oesophageal junction (GOJ) adenocarcinoma was classified by Siewert and Stein into three types. Type I arises from 1 to 5 cm proximal to the GOJ (tumours of the distal oesophagus), type II arises from 1 cm proximal to 2 cm distal to the GOJ (true cardiacarcinoma), and type III arises from 2 to 5 cm distal to the GOJ (subcardial gastric carcinoma). ⁶¹

Fig. 2: Barrett’s oesophagus (adapted from WHO classification of oesophageal tumours)

Oesophageal squamous cell carcinoma

The major risk factors for the development of oesophageal SCC (Fig. 3,4) are tobacco use and alcohol consumption; particularly a combination of both.^{23, 24} Nitrosamine exposure in tobacco smoking and the alcohol metabolite aldehyde, which is a known carcinogen, are probably the underlying reasons for these two risk factors. The high incidence of oesophageal SCC in Northern China, Iran and areas of Southern Africa may be related to a diet rich in nitrosamines and deficient in trace elements and vitamins A & C.

Other risk factors for oesophageal SCC in the Western world include low socioeconomic status, poor oral hygiene, achalasia, history of thoracic radiation, caustic injury, hereditary tylosis and Plummer-Vinson Syndrome.²⁵

Fig. 3: Squamous cell carcinoma of the oesophagus

Fig. 4: Microscopic picture of squamous cell carcinoma (adapted from WHO classification of oesophageal tumours)

How does oesophageal cancer present clinically?

Patients presenting with symptoms of OC almost invariably have advanced disease. The most common presenting symptom is progressive dysphagia with 74% of patients reporting difficulty swallowing.²⁶ This is graded according to the following: ²⁷

• Grade 1: Able to swallow most foods
• Grade 2: Able to swallow soft foods only
• Grade 3: Able to swallow liquids only
• Grade 4: Unable to swallow anything

17% of patients will also report pain on swallowing food and liquids (odynophagia). ²⁶ Typically, patients with oesophageal adenocarcinoma will be white males with a background of gastro-oesophageal reflux disease who have developed dysphagia. On the other hand, patients with oesophageal SCC will present with dysphagia, associated with weight loss and a history of smoking and increased alcohol intake may exist.

Other less common symptoms include dyspnoea, cough, hoarseness, acute haemorrhage and pain which may be retrosternal, back or right upper abdominal. These will usually represent the existence of metastatic disease.

Physical examination is often normal; positive clinical findings may include cachexia, lymphadenopathy and hepatomegaly in the presence of metastases.

How is oesophageal cancer diagnosed?

It is essential to have a low threshold if cancers are to be detected at an early, treatable stage. National Institute for Health and Clinical Excellence (NICE) guidelines state that a patient presenting with symptoms suggestive of upper gastrointestinal cancer should be referred to a team specialising in the management of these cancers. Specifically; patients of any age presenting with dyspepsia in association with alarm symptoms should be urgently referred for endoscopy or to a specialist. The classical ‘alarm’ symptoms associated with OC includes dysphagia, vomiting, anorexia, weight loss and symptoms associated with gastro-intestinal blood loss. Patients aged 55 or more with persistent, recent onset, and unexplained dyspepsia should be referred urgently for an endoscopy.

Diagnosis is usually made by oesophago-gastro-duodenoscopy where multiple biopsy samples must be taken from any mucosal abnormality to exclude early tumours. Suspicious lesions including oesophageal strictures may require repeated biopsies if initial results are negative.

Once diagnosis is made patients should be urgently referred to an Upper Gastro-intestinal team at a specialist centre for investigations to stage disease and further management.

Staging oesophageal cancers

It is essential to accurately stage disease to exclude patients with widespread metastatic disease for whom surgery will not be curative and to identify subgroups of patients who will require neo-adjuvant therapies. Whilst it is difficult to completely eliminate the possibility of ‘open and shut’ cases where tumours are found to be inoperable at the time of surgery; it is important to develop a staging strategy with investigations and procedures that help to minimise this risk. The TNM (Tumour, Node, Metastasis) staging system is used to classify the depth of tumour invasion into or through the oesophageal wall, the status of regional lymph nodes and metastases to distant sites. The TNM7 categories are shown in Tables 2 and the current stage groupings is shown in Table 3. ²⁸

Table 2: TNM7 staging system

Table 3: Stage classification for oesophageal cancer in the 2010 TNM7 staging system

Initial staging assessment includes Computed Tomography (CT) (Fig. 5) of the thorax, abdomen and pelvis and its major role will be in evaluating the T stage to detect local tumour invasion into adjacent structures and determining the presence or absence of metastatic disease. However CT will not be able to determine the depth of tumour invasion. Endoscopic ultrasound (EUS) (Fig. 6) is the main modality used to stage the primary tumour and primarily aids in distinguishing T1 lesions from T2-4 lesions. This method has an accuracy ranging from between 73% to 89% in tumour staging.²⁹ Accurately distinguishing tumour stage will affect treatment as T1 lesions may be treated with endoscopic therapy or with oesophagectomy whereas T2-4 lesions may require neo-adjuvant chemo-radiotherapy prior to surgery. EUS is also used for evaluation of regional lymph nodes however although sensitivity is approximately 80%, the specificity is lower at approximately 70%. ³⁰ It is best to perform a EUS-guided lymph node biopsy for confirmation of involvement.

Fig. 5: CT scan shows irregular wall thickening of the esophagus and enlarged metastatic lymph node.

Fig. 6: Endoscopic ultrasound (EUS) of oesophagus showing T3 tumour

FDG-PET (¹⁸F-fluoroudeoxyglucose PET) (Fig. 7) is a key modality for the detection of distant metastatic disease in OC.^{31, 32} PET may reveal previously occult distant metastases in nodal and non-nodal sites with a sensitivity of 67% and high specificity of 97%. ³³ It can also reveal metastases to bone, which may not be detected using conventional methods. An investigation has shown PET to be the only modality that predicted intended curative resection and it may also be used to prevent unnecessary surgical explorations.³⁴ The use of PET has been shown in a study to change the management of patients from curative to palliative due to detection of previously unknown metastases.³⁵ It has also been used in a prospective study to assess response early after neo-adjuvant chemotherapy to determine the need for urgent surgery or further chemotherapy. The usage of CT and PET in combination has become increasingly available and is useful in selective cases.³⁶

Fig. 7: FDG PET/CT image demonstrating increased uptake at the distal oesophagus and coeliac lymph node in oesophageal cancer case

Minimally invasive surgery is also used as a method to stage OC in many specialist centres.³⁷ A staging laparoscopy can visualise the primary tumour, identify metastases such as hepatic and regional nodal and can accurately detect intraperitoneal dissemination of disease, which may not have been appreciated on other radiological staging investigations. Samples of peritoneal ascites or washings for cytology can also be obtained at this stage if present.

Endoscopic mucosal resection (EMR) can provide accurate histological staging for high grade dysplasia and intramucosal carcinomas. ³⁸ In many cases EMR alone can be a therapeutic intervention depending on the depth of invasion on the specimen.

Treatment

The management of OCs requires a multi-disciplinary team approach involving surgeons, oncologists, radiologists, pathologists, specialist nurses, dietitians and specialists from other specialties if required. Patients considered for surgery or chemo-radiotherapy will require a fitness assessment. In addition to pulmonary function tests, ECG and echocardiogram, cardio-pulmonary exercise testing (CPEX/CPET) is now being increasingly used to assess fitness for major surgery.

OCs can be managed with surgery, chemotherapy or radiotherapy, a combination of the three or palliation in many cases. Disease that is locally advanced without signs of distant metastases is treated with an intention to cure and this will involve a multimodal approach. Metastatic, disseminated and recurrent disease will be treated with palliative intent with chemotherapy to increase survival or measures such as radiotherapy or stent placement for symptomatic relief.

Surgical

Surgical resection can be part of a multimodal approach or alone and is the main option for curative treatment. There are a number of surgical procedures that can be used however it is important to ensure removal of macroscopic and microscopic tumour in association with dissection of lymph nodes with either method as these are vital prognostic factors following surgery.

Open oesophagectomy (OO):

Options for resection include trans-hiatal oesophagectomy and transthoracic approaches and the choice of approach will depend on the location of the tumour, access to lymph nodes and surgeon preference. An Ivor Lewis oesophagectomy (also known as Lewis-Tanner oesophagectomy) involves abdominal mobilization of the stomach and right thoracic approach for resection of the oesophagus. The three-stage modified McKeown oesophagectomy involves a laparotomy, right thoracotomy and neck anastomosis. A resection margin 8-10 cm proximally and 7 cm distally is recommended to achieve an R0 resection (recommendation class IIB, level of evidence C). The next step is to construct a conduit for the anastomosis and this can be achieved by using a gastric tube, large or small bowel. A gastric tube is preferred due to the following factors; ease of use, tension free and longest term conduit survival (recommendation class IIA, level of evidence C). The anastomosis can be performed in the chest or the neck. This relies on multiple factors such as ease of the anastomosis, tension on the repair, ability to diagnose and treat complications and the oncological status. Circular staplers or hand sewn technique usually used with no significant differences in the outcomes. A drainage procedure such as pyloroplasty is recommended to avoid delayed gastric emptying (recommendation class I, level of evidence B). ⁶²

Radical oesophagectomy using either approach has a perioperative mortality of 5-10% and morbidity of 30-40%. ³⁹ Lymph node dissection plays an important role owing to the extensive submucosal lymphatic drainage of the oesophagus. This has meant that nearly 80% of patients undergoing surgery have positive lymph nodes and prognostically this is of importance.^{40, 41} However, there has been controversy with regards to the extent of lymph node dissection required. For optimal staging 10 lymph nodes for T1 and 20-30 lymph nodes for T2 and T3 tumours should be harvested. ⁶² In order to perform a curative resection for carcinoma of the middle and lower third of the oesophagus it is recommended to dissect the abdominal and mediastinal lymph nodes. Three-field lymphadenectomy in the abdomen, chest and neck, is performed in Japan for oesophageal SCC.⁴² Proponents of radical lymphadenectomy argue that it does allow optimal staging, improves loco-regional disease free survival improving the quality of life for these patients.

Minimally invasive oesophagectomy (MIO):

Minimally invasive approaches, which involve laparoscopic mobilisation of the stomach, thoracoscopic mobilisation of the oesophagus and hybrid or robotic approaches, are increasing in many specialist centres. Benefits of this approach include shorter recovery times, decreased post-operative pain and reduced cardiopulmonary complications without jeopardising the oncological outcomes. Luketich et al. reported a mortality rate of 1.7%, leak 5% and empyema 6% following MIO. ⁶³Several randomised controlled trials (RCTs) and comparative studies were conducted to investigate the efficacy and outcomes of MIO. A study by Li et al was conducted on 407 patients underwent MIO and OO found that the overall incidence of complications was lower in the MIO patients. The incidence of pulmonary complications was 20.7% in contrast to 39.7% in the OO group. However, there was no difference in the overall survival among the groups. Another comparative retrospective study by Mu et al. didn’t reveal any difference in the morbidity, anastomotic leak rate, pulmonary complications and length of stay between the approaches and the authors concluded that both techniques are equivalent. ^{63, 64}

Neo-adjuvant chemotherapy

This aims to improve operability; achieving this by shrinking the tumour prior to surgery, down-staging the disease as well as treating occult metastatic disease. Response to treatment can be assessed prior to surgery with repeat radiological investigations. It is now common for patients in the UK with locally advanced disease to undergo neo-adjuvant chemotherapy followed by resection. This is based on the results of a multi-centre study conducted by the Medical Research Council (OEO2), which showed a 9% improvement in two-year survival in patients given 2 cycles of Cisplatin and 5-Fluorouracil chemotherapy compared to those who were not. Five-year survival with surgery alone was 17%, compared with 23% with neoadjuvant chemotherapy.⁴³ The MRC Adjuvant Gastric Infusional Chemotherapy (MAGIC) trial randomized patients to chemotherapy with surgery or to surgery alone and it was found that patients in the chemotherapy group (who received Epirubicin, Cisplatin and infused 5-Fluorouracil, ‘ECF’) had a significant improvement in progression-free survival and a 13% increase in 5-year survival.⁴⁵

In a meta-analysis of neoadjuvant chemotherapy, there was an overall all-cause absolute survival benefit of 7% at 2 years with the addition of chemotherapy. When analysed by subtype, chemotherapy had no significant effect on mortality for patients with squamous cell carcinoma; however, there was a significant survival benefit for patients with oesophageal adenocarcinoma (HR 0.78; p=0.014). ⁴⁷

As a result of this evidence, neoadjuvant chemotherapy is a standard of care for patients with operable mid and lower oesophageal and GOJ adenocarcinoma. The ongoing MRC OEO5 trial is evaluating optimal neoadjuvant chemotherapy regimens: 4 cycles of chemotherapy with ‘ECX’ (Epirubicin, Cisplatin and Capecitabine) compared to two cycles Cisplatin and 5-Fluorouracil, as in OEO2.⁴⁴

Patients who are deemed suitable for surgical management of mid or distal oesophageal (including GOJ) adenocarcinomas are referred to the GI oncology team to assess fitness for chemotherapy. Many of the criteria assessed are similar to those in the pre-operative assessment, particularly performance status and medical comorbidities. Significant history of renal disease or cardiovascular disease, especially ischaemic heart disease would be a relative contraindication to systemic chemotherapy. Toxicities from chemotherapy are wide-ranging and include gastrointestinal upset, hair loss, skin rash, neurotoxicity, renal toxicity, bone marrow suppression (with risk of neutropaenic sepsis, thrombocytopaenia, and anaemia), cardiovascular toxicity, and chemotherapy-related fatigue. In the MAGIC trial, three cycles of epirubicin, cisplatin and capecitabine (ECX) chemotherapy were given both before and after surgery, and approximately one quarter of patients had CTCAE grade 3 or 4 neutropaenia. ⁴⁵

The REAL 2 trial ⁴⁸ was a 2x2 factorial non-inferiority comparison of cisplatin versus oxaliplatin and 5-fluorouracil (5-FU) versus oral capectiabine in patients with oesophageal, gastro-oesophageal junction and gastric tumours. Treatment was given as triplet chemotherapy: epirubicin plus platinum agent (cisplatin or oxaliplatin) plus 5-FU or capecitabine. The trial results showed that oxaliplatin was at least as effective as cisplatin, and oral capectibine was at least as effective as intravenous 5-fluorouracil. There was less grade 3 and 4 neutropaenia with oxaliplatin versus cisplatin, but this was offset by an increase in neuropathy and diarrhoea. As a result of this trial, EOX chemotherapy can be used as an alternative to ECX in both the neoadjuvant and metastatic settings (Table 4).

Table 4: Efficacies of major combination chemotherapy drugs

Neo-adjuvant chemo-radiotherapy

In contrast to the UK, patients in the USA commonly receive neo-adjuvant chemo-radiotherapy (CRT) for locally advanced oesophageal carcinoma. There is evidence that preoperative CRT is superior to surgery alone. A meta-analysis of ten randomised controlled trials showed a hazard ratio for all-cause mortality of 0.81 (95% CI 0.70 to 0.93; p=0.002). This corresponded to a 13% absolute survival benefit at 2 years.⁴⁷ In the subgroup analysis of the Dutch CRT trial (which used paclitaxel and carboplatin combination chemotherapy), the beneficial effect was more pronounced in patients with squamous cell carcinoma (HR 0.34; 95% CI 0.17 to 0.65) compared to adenocarcinoma (HR 0.82; 95% CI 0.58 to 1.16).⁴⁹

There has been no direct head-to-head comparison of neoadjuvant chemotherapy and neoadjuvant CRT in the context of a phase III randomised control trial. Concerns regarding the added morbidity of CRT have meant that chemotherapy alone is the standard neoadjuvant treatment of choice in the UK. However, the role of neoadjuvant CRT is currently being reassessed in the Neo-SCOPE trial.

Definitive chemo-radiotherapy (CRT)

According to current UK consensus guidelines, CRT is the definitive treatment of choice for localised squamous cell carcinoma of the proximal oesophagus. ⁵⁰ Localised squamous cell carcinoma of the middle or lower oesophagus may be treated with CRT alone, or CRT plus surgery. ⁵⁰

In a pivotal study, US Intergroup RTOG-8501 randomised 121 patients with squamous cell carcinoma or adenocarcinoma to receive CRT (cisplatin and 5-Fluorouracil with 50 Gray in 25 fractions), or radiotherapy alone (64 Gray in 32 fractions). This trial ⁴⁶, together with a subsequent systematic review ⁵⁵, demonstrated a survival superiority of CRT over radiotherapy alone (1-year mortality odds ratio 0.61; 95% CI 0.31 to 0.89; p<0.001). This was at the expense of increased toxicity.

This and similar studies ^56-57 have demonstrated a remarkably consistent median survival of 14-18 months and 2 year overall survival of 30-40% with CRT.

CRT practice in the UK is somewhat varied, but within the authors’ multidisciplinary team Cisplatin and 5-Fluorouracil chemotherapy is given in weeks 1 and 5 of a five-and-a-half week course of radiotherapy. The radiation dose used is 50.4 Gray in 28 daily fractions, treating Mondays to Fridays. An alternative radiation dose-fractionation which is supported by the Royal College of Radiologists guidelines is 50 Gray in 25 daily fractions. ⁵⁸

There are few trials directly comparing surgery alone with CRT. A study of 80 patients with squamous cell carcinoma randomised to surgery or CRT failed to show superiority of either strategy in terms of early disease free survival or overall survival. ⁵¹ Adding surgery to CRT can improve local control rates compared with CRT alone, but combined-modality therapy has not been shown to improve survival. It predictably also leads to significantly more treatment-related morbidity.⁵²

The French FFCD 9102 trial recruited 444 patients with potentially resectable OC (90% squamous cell carcinoma) to receive induction CRT. Those patients who showed evidence of response to CRT were then randomised to further CRT or surgery. Median overall survival was 19.3 months in the CRT alone arm, and 17.7 months in those randomised to surgery. The trial met its endpoint of non-inferiority for 2 year overall survival. Again, toxicity was shown to be significantly higher in patients who received both CRT and surgery.⁵³

Although definitive CRT is a current recommended standard of care for localised squamous cell carcinoma of the oesophagus, there is insufficient evidence to to support either a surgical or non-surgical approach ⁵⁰. Surgery should be considered in patients who have histologically-confirmed residual disease at the end of CRT.

For patients deemed unsuitable for surgery with localised adenocarcinoma of the oesophagus, CRT is a valid option for treatment. An American case series of 25 patients with a median age of 77 years showed that CRT using two cycles of mitomycin-C and 5-fluorouracil in combination with radiation (dose 50.4Gy in 28 daily fractions) was effective and tolerable. 68% of these patients had no evidence of residual disease on post-treatment endoscopy. This small series of patients had a two year overall survival of 64%, with a median overall survival of 35 months.⁵⁴

Salvage surgery after definitive CRT

Local recurrence occurs within the first year in 10-30% of patients treated with definitive CRT.⁵⁰ Salvage curative oesophagectomy may be considered within a multidisciplinary team setting. Repeat staging investigations including a CT-PET and EUS are required before a final decision for salvage surgery is made. Survival benefit is limited, and such surgery is associated with an increased in-hospital mortality rate and increased morbidity.⁵⁹ Informing patients of the potential high risks and poor outcomes is an integral part of the decision-making process for salvage surgery.

Palliation

The majority of patients diagnosed with OC are never treated with curative intent as a result of advanced disease or their physical fitness and comorbidities not allowing for radical treatment. It also includes patients who have developed recurrent or metastatic disease following resection. For this group of patients, there are a number of palliative treatments available for relief of symptoms, prolonging and maximising their quality of life. Once again, a multidisciplinary, holistic approach is required to provide the best treatment.

Treatments to provide symptomatic relief such as dysphagia can include intraluminal brachytherapy, endoscopic stenting using self-expanding metal stents or repeated endoscopic dilatations. Dysphagia can also be palliated by chemotherapy or external beam radiotherapy. Laser-thermal Nd-YAG endoluminal tumour destruction and photodynamic therapy can also be administered however this requires a number of treatments and may be more suitable for short exophytic tumours. It is essential to manage pain and nutrition and feeding options through a gastrostomy, jejunostomy or even intravenously can be provided to ensure adequate nutritional status. In addition to providing symptomatic relief it is important to also ensure that these patients receive social and psychological support by identifying and addressing the needs of the patients as well as their carers.

Palliative radiotherapy can be offered to patients with symptomatic primary oesophageal tumours in the context of metastatic or inoperable disease. Palliative dose and fractionation options are varied, but include 27 Gray in 6 fractions treating twice a week for 3 weeks; 30 Gray in 10 fractions treating daily for 2 weeks; 20 Gray in 5 fractions treating daily for 1 week.⁵⁸ The aim of such radiation treatment is to palliate dysphagia. This effect is not immediate, and therefore patients with significant dysphagia are better served initially by endoscopic stenting.

Chemotherapy has been shown to be effective in improving symptoms and overall survival. Patients with good performance status are offered combination chemotherapy. This can be with EOX, as per the MAGIC trial, ⁴⁵or with Cisplatin and 5-Fluorouracil, with or without the addition of Epirubicin (CF or ECF). 5-Fluorouracil can be substituted for oral Capecitabine (i.e. CX or ECX) without any adverse effects on outcomes.⁴⁵

When choosing palliative systemic chemotherapy for patients with incurable OC, the primary aim should be about maximising quality of life. Improvements in outcome with more intensive chemotherapy regimens, such as docetaxel, cisplatin and 5-Fluorouracil, have been shown to be offset by significantly more toxicity.⁶⁰ As a result, Docetaxel containing regimens are not approved in the UK for this indication.⁵⁰

Conclusions

The incidence of oesophageal carcinoma is increasing and despite advances in management and treatment the overall prognosis remains poor. It is essential to recognize and diagnose early, to have a clear pathway for subsequent investigations to ensure accurate staging. This will allow appropriate therapy to be administered to ensure the best possible outcomes are achieved. Treatment of OC is still a challenge however recent advances in surgery, endoscopic treatments and new therapeutic agents will hopefully improve prognosis.