Explainable Machine Learning for Real-Time Hypoglycemia and Hyperglycemia Prediction and Personalized Control Recommendations

BACKGROUND

The occurrences of acute complications arising from hypoglycemia and hyperglycemia peak as young adults with type 1 diabetes (T1D) take control of their own care. Continuous glucose monitoring (CGM) devices provide real-time glucose readings enabling users to manage their control proactively. Machine learning algorithms can use CGM data to make ahead-of-time risk predictions and provide insight into an individual's longer term control.

METHODS

We introduce explainable machine learning to make predictions of hypoglycemia (<70 mg/dL) and hyperglycemia (>270 mg/dL) up to 60 minutes ahead of time. We train our models using CGM data from 153 people living with T1D in the CITY (CGM Intervention in Teens and Young Adults With Type 1 Diabetes)survey totaling more than 28 000 days of usage, which we summarize into (short-term, medium-term, and long-term) glucose control features along with demographic information. We use machine learning explanations (SHAP [SHapley Additive exPlanations]) to identify which features have been most important in predicting risk per user.

RESULTS

Machine learning models (XGBoost) show excellent performance at predicting hypoglycemia (area under the receiver operating curve [AUROC]: 0.998, average precision: 0.953) and hyperglycemia (AUROC: 0.989, average precision: 0.931) in comparison with a baseline heuristic and logistic regression model.

CONCLUSIONS

Maximizing model performance for glucose risk prediction and management is crucial to reduce the burden of alarm fatigue on CGM users. Machine learning enables more precise and timely predictions in comparison with baseline models. SHAP helps identify what about a CGM user's glucose control has led to predictions of risk which can be used to reduce their long-term risk of complications.

Toward a common standard for data and specimen provenance in life sciences

Open and practical exchange, dissemination, and reuse of specimens and data have become a fundamental requirement for life sciences research. The quality of the data obtained and thus the findings and knowledge derived is thus significantly influenced by the quality of the samples, the experimental methods, and the data analysis. Therefore, a comprehensive and precise documentation of the pre-analytical conditions, the analytical procedures, and the data processing are essential to be able to assess the validity of the research results. With the increasing importance of the exchange, reuse, and sharing of data and samples, procedures are required that enable cross-organizational documentation, traceability, and non-repudiation. At present, this information on the provenance of samples and data is mostly either sparse, incomplete, or incoherent. Since there is no uniform framework, this information is usually only provided within the organization and not interoperably. At the same time, the collection and sharing of biological and environmental specimens increasingly require definition and documentation of benefit sharing and compliance to regulatory requirements rather than consideration of pure scientific needs. In this publication, we present an ongoing standardization effort to provide trustworthy machine-actionable documentation of the data lineage and specimens. We would like to invite experts from the biotechnology and biomedical fields to further contribute to the standard.

Evaluation of an established oncology triage hotline in a model of emergency department avoidance: assessing the UKONS triage tool and call outcomes

PURPOSE

In response to a demonstrable need for 24/7, specialist oncology advice for patients undergoing systemic anti-cancer therapy, many healthcare institutions have adopted a telephone triage (TT) service. This is true of the Clatterbridge Cancer Centre which uses the UKONS framework to guide its decisions. This study aims to investigate the utilisation and outcomes of this TT service, with a focus on the most unwell call outcomes and factors leading to referrals to accident and emergency departments that could be mitigated with service development and modifications.

METHODS

A retrospective evaluation study was conducted of calls occurring between 1st September 2021 and 31st August 2022. A descriptive analysis of call UKONS grading, triage outcome and primary complaint was performed.

RESULTS

The TT hotline received 23,766 calls of which only 9066 were for clinical advice. Of the clinical calls, 45.2% were UKONS red. The majority of red calls 53.3% were directed to AED. The proportion of red calls going to AED changed drastically depending on the timing of call and the corresponding services available at those times, with 38.3% of reds being sent to AED in hours but 72.3% out of hours. The profile of complaints also showed significant differences in hours versus out of hours.

CONCLUSION

Significant use of the hotline supports a genuine demand for oncology TT services. In order to reduce referrals to AED, this study supports the creation of alternative destinations of emergency care, especially out of hours.

P-114 INVESTIGATING THE ROLE OF BOTOX IN THE EMERGENCY MANAGEMENT OF COMPLEX ABDOMINAL WALL HERNIAS

Aims

Botulinum toxin (Botox) has an established role in the elective management of complex abdominal wall hernia, but its role in the emergency setting has not been investigated. Clinical effects of Botox peak at 2–3 weeks after injection, but effects become apparent within 48 hours. We aimed to determine if these early effects can facilitate the acute reduction of irreducible hernia.

Materials and methods

Patients with irreducible abdominal wall hernias over a 30 month period were considered for Botox therapy. Suitable patients were given 300 units of Botox under ultrasound guidance into the anterolateral muscles bilaterally. Data were collected on outcomes.

Results

Twelve patients received Botox. All had incisional hernias and an ASA of 3 or more. Six were on antiplatelets and/or anticoagulation. All received Botox with standard conservative management. In ten patients, the hernia was reduced and symptoms resolved. One patient required surgery at 48 hours due to bowel ischaemia from inter-sac adhesions. A second patient experienced recurrent obstruction despite successful hernia reduction but was palliated due to advanced age and frailty.

Conclusions

This case series is the first study to investigate the role of Botox in the acute management of irreducible hernias. Our findings suggest that in selected patients, Botox can aid hernia reduction and avoid the morbidity of emergency repair. This is of particular value in high risk patients and those with complex defects who benefit from preoperative optimisation techniques. The accurate identification of strangulation is crucial however, as it excludes patients from this approach.

Impact of time from symptom onset on the diagnostic performance of high-sensitivity cardiac troponin for type 1 myocardial infarction

Background

High-sensitivity cardiac troponin has enabled the rapid rule-out and rule-in of myocardial infarction at presentation. However, increases in cardiac troponin may not be detectable early after symptom onset, and uncertainty remains as to how time of symptom onset influences diagnostic performance.

Purpose

To evaluate the impact of time from symptom onset on the diagnostic performance of high-sensitivity cardiac troponin for type 1 myocardial infarction.

Methods

In a secondary analysis of a prospective multicentre randomised controlled trial of consecutive patients with suspected acute coronary syndrome, we evaluated the diagnostic performance of high-sensitivity cardiac troponin I measurements at presentation stratified by time of symptom onset to blood sampling. Diagnostic performance was evaluated in four groups (≤3 hours, 4–6 hours, 7–12 hours and &gt;12 hours from symptom onset) for recommended thresholds to rule-out (sex-specific 99th centile and optimised threshold [64 ng/L]) type 1 myocardial infarction.

Results

This analysis included 41,104 patients (median 60 [interquartile range 49–74] years, 46% female) of which 12,595 (31%), 10,298 (25%), 7,171 (17%) and 11,040 (27%) presented ≤3 hours, 4–6 hours, 7–12 hours and &gt;12 hours, respectively. Type 1 myocardial infarction was the adjudicated diagnosis in 3,692 (9%) patients. For the rule-out of type 1 myocardial infarction, sensitivity was highest in those tested 7–12 hours from symptom onset and lowest in those tested ≤3 hours. In early presenters, a threshold of &lt;2 ng/L had greater sensitivity and negative predictive value (99.4% [95% CI 98.9 to 99.7%] and 99.7% [95% CI 99.5 to 99.9%]) compared to &lt;5 ng/L (96.7% [95% CI 95.7 to 97.6%] and 99.3% [95% CI 99.1 to 99.5%], respectively). In those tested &gt;3 hrs from symptom onset, the sensitivity and negative predictive value for both thresholds were similar, but a threshold of &lt;5 ng/L correctly ruled out more patients (60% [17,056/28,506] versus 29% [8,316/28,506]). For the rule-in of myocardial infarction, the sensitivity of the 99th centile and 64 ng/L was lowest in patients tested within 3 hours (71.7% [95% CI 69.3 to 74.1%] and 46.5% [95% CI 44.1 to 49.2%], respectively), and increased in those tested later from symptom onset. The specificity and positive predictive value were highest when testing was performed 7–12 hours from symptom onset for the sex-specific 99th centile (92.4% [95% CI 91.8 to 93.0%] and 51.3% [95% CI 48.2–54.5%]) and 64 ng/L (96.2% [95% CI 95.7 to 96.7%] and 61.2% [95% CI 57.3 to 65.2%]).

Conclusions

The diagnostic performance of cardiac troponin for myocardial infarction is strongly influenced by the time from symptom onset to testing. In early presenters the limit of detection may facilitate immediate rule-out of myocardial infarction, but otherwise testing at least 3 hours from symptom onset is needed with the optimal time to rule-in myocardial infarction being 7–12 hours from the onset of symptoms.

Funding Acknowledgement

Type of funding sources: Public grant(s) – National budget only. Main funding source(s): British Heart Foundation (BHF)Medical Research council UK (MRC)

What Can Knowledge Bring to Machine Learning?—A Survey of Low-shot Learning for Structured Data

Supervised machine learning has several drawbacks that make it difficult to use in many situations. Drawbacks include heavy reliance on massive training data, limited generalizability, and poor expressiveness of high-level semantics. Low-shot Learning attempts to address these drawbacks. Low-shot learning allows the model to obtain good predictive power with very little or no training data, where structured knowledge plays a key role as a high-level semantic representation of human. This article will review the fundamental factors of low-shot learning technologies, with a focus on the operation of structured knowledge under different low-shot conditions. We also introduce other techniques relevant to low-shot learning. Finally, we point out the limitations of low-shot learning, the prospects and gaps of industrial applications, and future research directions.

Optimal timing and technique for endoscopic management of dysphagia in pediatric aerodigestive patients

INTRODUCTION

The best strategy to manage an interarytenoid defect [Type 1 laryngeal cleft (LC-1) or deep interarytenoid groove (DIG)] in pediatric aerodigestive patients with dysphagia remains uncertain. This study compared benefit of interarytenoid augmentation (IAA) to suture repair or clinical observation alone in pediatric patients.

METHODS

A 3-year retrospective, single-center analysis of children with dysphagia undergoing endoscopic airway evaluation was performed. Physician preference guided treatment plan: suture repair with CO2 laser, IAA (carboxy methylcellulose or calcium hydroxyapatite), or observation. Primary outcome was improved post-operative diet. Significance was assumed at p < 0.05.

RESULTS

449 patients underwent diagnostic endoscopy. Mean age (±SD) at procedure was 21 ± 13 months, with nearly one fourth (28 %) of children ≤ 12 months. Eighty (18 %) had either an LC-1 (n = 55) or DIG (n = 25). Of these, 35 (42 %) underwent suture repair, 22 (28 %) IAA, and 23 (30 %) observation only. Aspiration improved overall in the interventional groups compared to observational controls (58 % vs. 9 %, p < 0.05), with no change in benefit observed by age of intervention. IAA was as effective as suture repair (59 % vs 55 %, p = 0.46). In patients with only a DIG, IAA intervention alone significantly improved swallow function (66.6 % vs. 0 %, p < 0.05).

CONCLUSION

In pediatric aerodigestive patients with dysphagia, 18 % of children have an addressable lesion. IAA or suture repair similarly improves dietary advancement. IAA improves swallow function in patients with DIG. These findings support a novel protocol to intervene in dysphagia patients with LC-1 or DIG via IAA at the initial operative evaluation.

The development and validation of population clusters for integrating health and social care: A protocol for a mixed-methods study in Multiple Long-Term Conditions (Cluster-AIM) (Preprint)

Background

Multiple long-term health conditions (multimorbidity) (MLTC-M) are increasingly prevalent and associated with high rates of morbidity, mortality, and health care expenditure. Strategies to address this have primarily focused on the biological aspects of disease, but MLTC-M also result from and are associated with additional psychosocial, economic, and environmental barriers. A shift toward more personalized, holistic, and integrated care could be effective. This could be made more efficient by identifying groups of populations based on their health and social needs. In turn, these will contribute to evidence-based solutions supporting delivery of interventions tailored to address the needs pertinent to each cluster. Evidence is needed on how to generate clusters based on health and social needs and quantify the impact of clusters on long-term health and costs.

Objective

We intend to develop and validate population clusters that consider determinants of health and social care needs for people with MLTC-M using data-driven machine learning (ML) methods compared to expert-driven approaches within primary care national databases, followed by evaluation of cluster trajectories and their association with health outcomes and costs.

Methods

The mixed methods program of work with parallel work streams include the following: (1) qualitative semistructured interview studies exploring patient, caregiver, and professional views on clinical and socioeconomic factors influencing experiences of living with or seeking care in MLTC-M; (2) modified Delphi with relevant stakeholders to generate variables on health and social (wider) determinants and to examine the feasibility of including these variables within existing primary care databases; and (3) cohort study with expert-driven segmentation, alongside data-driven algorithms. Outputs will be compared, clusters characterized, and trajectories over time examined to quantify associations with mortality, additional long-term conditions, worsening frailty, disease severity, and 10-year health and social care costs.

Results

The study will commence in October 2021 and is expected to be completed by October 2023.

Conclusions

By studying MLTC-M clusters, we will assess how more personalized care can be developed, how accurate costs can be provided, and how to better understand the personal and medical profiles and environment of individuals within each cluster. Integrated care that considers “whole persons” and their environment is essential in addressing the complex, diverse, and individual needs of people living with MLTC-M.

International Registered Report Identifier (IRRID)

PRR1-10.2196/34405

Anakinra for palmoplantar pustulosis: results from a randomized, double-blind, multicentre, two-staged, adaptive placebo-controlled trial (APRICOT)

BACKGROUND

Palmoplantar pustulosis (PPP) is a rare, debilitating, chronic inflammatory skin disease that affects the hands and feet. Clinical, immunological and genetic findings suggest a pathogenic role for interleukin (IL)-1.

OBJECTIVES

To determine whether anakinra (an IL-1 receptor antagonist) delivers therapeutic benefit in PPP.

METHODS

This was a randomized (1 : 1), double-blind, two-staged, adaptive, UK multicentre, placebo-controlled trial [ISCRTN13127147 (registered 1 August 2016); EudraCT number: 2015-003600-23 (registered 1 April 2016)]. Participants had a diagnosis of PPP (> 6 months) requiring systemic therapy. Treatment was 8 weeks of anakinra or placebo via daily, self-administered subcutaneous injections. Primary outcome was the Palmoplantar Pustulosis Psoriasis Area and Severity Index (PPPASI) at 8 weeks.

RESULTS

A total of 374 patients were screened; 64 were enrolled (31 in the anakinra arm and 33 in the placebo arm) with a mean (SD) baseline PPPASI of 17·8 (10·5) and a PPP investigator's global assessment of severe (50%) or moderate (50%). The baseline adjusted mean difference in PPPASI favoured anakinra but did not demonstrate superiority in the intention-to-treat analysis [-1·65, 95% confidence interval (CI) -4·77 to 1·47; P = 0·30]. Similarly, secondary objective measures, including fresh pustule count (2·94, 95% CI -26·44 to 32·33; favouring anakinra), total pustule count (-30·08, 95% CI -83·20 to 23·05; favouring placebo) and patient-reported outcomes, did not show superiority of anakinra. When modelling the impact of adherence, the PPPASI complier average causal effect for an individual who received ≥ 90% of the total treatment (48% in the anakinra group) was -3·80 (95% CI -10·76 to 3·16; P = 0·285). No serious adverse events occurred.

CONCLUSIONS

No evidence for the superiority of anakinra was found. IL-1 blockade is not a useful intervention for the treatment of PPP.

Metabolomics and a breath sensor identify acetone as a biomarker for heart failure

Introduction

Metabolomics delivers more biological and clinical insight than targeted investigations. We applied metabolomics to patients with heart failure (HF) with reduced ejection fraction (HFrEF).

Methods

46 patients with HFrEF and 20 controls underwent metabolomic profiling, including liquid/gas chromatography mass spectrometry (LC-MS/GC-MS), nuclear magnetic resonance (NMR) metabolomics and solid-phase microextraction (SPME) volatilomics in plasma and urine. HFrEF was defined using left ventricular global longitudal strain (GLS) &lt;18%, ejection fraction &lt;50% and NTproBNP ≥35 pmol/L. A low cost consumer breath acetone (BrACE) sensor validated SPME results in 69 patients.

Results

34 metabolites were identified by GCMS, 33 by LCMS and 2 volatiles by SPME (acetone, 2-pentanone in plasma and urine). Alanine, aspartate and glutamate, citric acid cycle, arginine biosynthesis, glyoxylate and dicarboxylate metabolism were altered in HFrEF. Plasma acetone correlated with NT-proBNP (r=0.59, 95% CI 0.4 to 0.7), triacylglycerol (55:9), 2-oxovaleric and cis-aconitic acid, involved with ketone metabolism and mitochondrial energetics. BrACE &gt;1.5 ppm discriminated HF from other cardiac pathology (AUC 0.88, 95% CI 0.77 to 0.99, P&lt;0.0001).

Conclusion

Breath acetone was detectable in HFrEF patients using a consumer sensor ($1/test) and although not cardiac specific, discriminated HF from other cardiac pathology.

Funding Acknowledgement

Type of funding source: Public grant(s) – National budget only. Main funding source(s): Health Research Council of New Zealand

Jumping to conclusions in persistent pain using a somatosensory modification of the beads task

BACKGROUND

There is theoretical and empirical evidence that persistent pain occurs because of a distortion in top-down perceptual processes. 'Jumping to conclusions' (JTC) tasks, such as the beads task, purportedly capture these processes and have yet to be studied in people with chronic pain. However, the beads task uses visual stimuli, whereas tasks involving processing in the somatosensory domain seem at least more face valid in this population. This study uses a novel somatosensory adaptation of the beads task to explore whether a JTC reasoning style is more common in people with persistent pain compared controls.

METHODS

30 persistent pain patients and 30 age-, gender- and education-matched controls completed the visual beads JTC task and a novel somatosensory version of the JTC task that used tactile stimuli (vibrations to the fingertip).

FINDINGS

Patients with persistent pain showed a 'jumping to conclusions' reasoning style on both tasks compared to the control group and there was no significant difference in the effect sizes on the two tasks.

INTERPRETATION

This preliminarily study demonstrated that individuals with persistent pain show a JTC reasoning style to both visual and somatosensory stimuli. Future research should focus on establishing how or whether this bias directly influences the experience of persistent pain.

The honey bee gut microbiota: strategies for study and characterization

Gut microbiota research is an emerging field that improves our understanding of the ecological and functional dynamics of gut environments. The honey bee gut microbiota is a highly rewarding community to study, as honey bees are critical pollinators of many crops for human consumption and produce valuable commodities such as honey and wax. Most significantly, unique characteristics of the Apis mellifera gut habitat make it a valuable model system. This review discusses methods and pipelines used in the study of the gut microbiota of Ap. mellifera and closely related species for four main purposes: identifying microbiota taxonomy, characterizing microbiota genomes (microbiome), characterizing microbiota-microbiota interactions and identifying functions of the microbial community in the gut. The purpose of this contribution is to increase understanding of honey bee gut microbiota, to facilitate bee microbiota and microbiome research in general and to aid design of future experiments in this growing field.

Impact of social deprivation, demographics and centre on HbA1c outcomes with continuous subcutaneous insulin infusion

AIMS

To assess the impact of social deprivation, demographics and centre on HbA_1c outcomes with continuous subcutaneous insulin infusion (CSII) in adults with Type 1 diabetes.

METHODS

Demographic data, postcode-derived English Index of Multiple Deprivation data and 12-month average HbA_1c (mmol/mol) pre- and post-CSII were collated from three diabetes centres in the north west of England, University Hospital of South Manchester (UHSM), Salford Royal Foundation Hospital (SRFT) and Manchester Royal Infirmary (MRI). Univariable and multivariable regression models explored relationships between demographics, Index of Multiple Deprivation, centre and HbA_1c outcomes.

RESULTS

Data were available for 693 (78%) individuals (UHSM, n = 90; SRFT, n = 112; and MRI, n = 491) of whom 59% were women. Median age at CSII start was 39 (IQR 29.5-49.0) years and median diabetes duration was 20 (11-29) years. Median Index of Multiple Deprivation was 15 193 (6313-25 727). Overall median HbA_1c improved from 69 to 64 mmol/mol (8.5% to 8.0%) within the first year of CSII. In multivariable analysis, higher pre-CSII HbA_1c was significantly associated with higher deprivation (P = 0.036), being female (P < 0.001), and centre (MRI; P = 0.005). Following pre-CSII HbA_1c adjustment, post-CSII HbA_1c or HbA_1c change were not related to demographic factors and deprivation, but remained significantly related to the centre; UHSM and SRFT had larger reductions in HbA_1c with CSII compared with MRI [median -7.0 (-0.6%) vs. -6.0 (-0.55%) vs. -4.5 (-0.45%) mmol/mol; P = 0.005].

CONCLUSIONS

Higher pre-CSII HbA_1c levels were associated with higher deprivation and being female. CSII improves HbA_1c irrespective of social deprivation and demographics. Significant differences in HbA_1c improvements were observed between centres. Further work is warranted to explain these differences and minimize variation in clinical outcomes with CSII.

Prospective study in bullous pemphigoid: association of high serum anti-BP180 IgG levels with increased mortality and reduced Karnofsky score

BACKGROUND

Bullous pemphigoid (BP) is a subepidermal blistering disease characterized by autoantibodies against the two hemidesmosomal proteins, BP180 (type XVII collagen) and BP230. The multicentre prospective BLISTER (Bullous Pemphigoid Steroids and Tetracyclines) trial randomized 253 patients with BP to compare the benefits and harms between initial treatment with doxycycline or prednisolone.

OBJECTIVES

To analyse distinct autoantibody profiles for the prediction of the disease course in a well-characterized cohort of BP sera.

METHODS

One hundred and forty-three patients of the BLISTER trial consented to participate in this serological study. Sera taken at baseline were analysed by (i) indirect immunofluorescence, (ii) anti-BP180 NC16A (16th noncollagenous domain) and anti-BP230 enzyme-linked immunosorbent assay and (iii) immunoblotting with various substrates. Results were then linked with clinical parameters including age, Karnofsky score, number of blisters, related adverse events and mortality.

RESULTS

Disease activity correlated with immunoglobulin (Ig)G anti-BP180 levels but not with levels of anti-BP230 IgG and anti-BP180 IgE. High levels of both anti-BP180 IgG and anti-BP230 IgG were associated with a low Karnofsky score. The presence of anti-BP230 IgG was more frequent in older patients. Those with higher total IgE serum levels suffered from fewer adverse events. Higher IgG anti-BP180 levels were associated with an increased 1-year mortality rate.

CONCLUSIONS

Analysis of the autoantibody profile is not only of diagnostic relevance but may also be helpful in predicting the course of the disease.

Missing Evidence: Exploring Unpublished Trials in Victoria, Australia

Background: Clinical trial registries were established to improve the transparency and completeness of clinical trial reporting and a number of policies have been introduced to encourage or mandate their use. While prospective trial registration has been endorsed in Australia, there is currently no legal requirement for researchers to register or communicate findings from clinical trials. There has also been, to the best of our knowledge, no analysis previously undertaken on publication rates for clinical trials performed in Australia. Aim: We aimed to determine the proportion of clinical trials that remain 'unpublished' in Victoria, Australia´s second most populous state, between 2009 and 2013. Methods: We used data reported to Cancer Council Victoria's Cancer Trials Management Scheme (CTMS) between 2009 and 2013, to identify trials that had recruited a new patient or recorded any follow-up patient activity in the specified time period. Using this data, we conducted a systematic search of ClinicalTrials.gov , the Australia and New Zealand Clinical Trials Registry (ANZCTR), PubMed and Google for records of the trial. Trial registration numbers, acronyms and scientific titles were used as primary search terms. Results were characterized by type of publication (i.e., whether it was an accredited scientific paper or other) and source location. Results: Of the trials reported to the CTMS between 2009 and 2013, 777 trials were included in this investigation; the majority (58.8%) were randomized controlled trials (RCTs). Compared with previously published findings, communication of trial results in this study was high; 70% of trials published results in an accredited scientific journal and a further 10% in alternate form, such as a conference abstract or media release. Publication rates were higher for trials with a commercial sponsor (85%) compared with trials sponsored by a cooperative group (77%). Nearly 8% of trials in this study had not been registered on an international clinical trials register. Only 39% of unregistered trials had published results. Of the registered trials, those registered on ClinicalTrials.gov were more likely to be published (86%) compared with trials listed on ANZCTR (68%). Between 2009 and 2013 , 8% of trials registered on ClinicalTrials.gov , in our data set, were terminated; 70% of these trials published results. Conclusion: Although the rate at which clinical trial findings were published in Victoria was higher in this investigation compared with equivalent overseas data, trials registered on ClinicalTrials.gov were more likely to publish results than unregistered trials or trials registered on ANZCTR. This suggests a potential need for trial registration and publication guidelines in Australia, similar to that of the United States where the requirements and procedures for submitting registration and summary result information for clinical trials on ClinicalTrials.gov have been compulsory for the last decade.

Abstract P5-21-08: Tolerability of the combination of lapatinib and trastuzumab in older patients with HER2 positive metastatic breast cancer

Background: Older adults are less likely to be included in clinical trials leading to the approval of novel cancer treatments. The Institute of Medicine and ASCO have identified therapeutic phase II trials as a key research priority to increase the evidence base for older adults with cancer. While targeted therapies may represent a less toxic option for older patients, few trials have studied their tolerability and efficacy in older adults. Here, we present a phase II study (NCT01273610) of the combination of trastuzumab and lapatinib in older patients with HER2+ metastatic breast cancer (MBC), incorporating geriatric oncology principles in the study design.

Methods: Patients age ≥ 60 years with MBC and any number of prior chemotherapy (CT) lines received trastuzumab (either 4mg/kg loading dose followed by 2mg/kg weekly or 8mg/kg followed by 6mg/kg q/3 weeks) plus lapatinib 1000 mg/m2 daily in 21-day cycles. Patients completed a pre-treatment geriatric assessment including measures of function, comorbidity, cognition, nutrition, and psychosocial status. A toxicity risk score developed for older adults receiving cytotoxic CT was calculated for each patient (Hurria et al. JCO 2011 & 2016). Relationships between tolerability (dose reductions and grade (G) ≥ 3 toxicity attributed to treatment) and risk score analyzed using a log2 transformation were assessed using generalized linear models, Student's t tests, and Fisher's exact test. Response rate (RR) and progression free survival (PFS) were evaluated.

Results: 40 patients (mean age 72 [60-92]) were accrued from 04/11 to 05/15. 25% (n = 10) were ≥ 75 years of age. 65% of patients (n = 26) had HR+ tumors and 35% (n = 14) were receiving ≥ 3rd line treatment. Median number of cycles was 4 (0-28). RR was 23% (n = 9, 95% CI 11-38%; 1 complete, 8 partial). 23% (n = 9) achieved stable disease. PFS was 2.7 months (95% CI 2.5-12). Based on the toxicity risk score, 21% (n = 8), 54% (n = 21), and 26% (n = 10) were at low, intermediate, and high risk. 70% (n = 28) of patients had G ≥ 2 toxicities and 20% (n = 8) G ≥ 3 toxicities. G 2 and 3 diarrhea occurred in 28% (n = 11) and 5% (n = 2) respectively. 5% (n = 2) were hospitalized due to treatment-related toxicity. No G ≥ 3 cardiac toxicities were observed. 23% of patients (n = 9) had treatment delays, and 43% (n = 17) required a lapatinib dose reduction. The mean toxicity risk score was higher in patients who required dose reductions (Student's t: p = 0.02). No statistically significant relationship was found between toxicity risk scores and the presence of G ≥ 3 treatment toxicity (logistic regression: OR = 3.08, 95% CI [0.54, 21.2], p = 0.22).

Conclusions: Among older patients with MBC (79% at intermediate or high risk of G ≥ 3 cytotoxic CT toxicity), trastuzumab and lapatinib were well tolerated, with only 20% experiencing G3 toxicities. The toxicity risk score was not found to be significantly related with treatment toxicity, which may be explained by the very low incidence of G3 events. Patients with a low toxicity risk score were not likely to require a lapatinib dose reduction.

Citation Format: O'Connor T, Soto-Perez-de-Celis E, Blanchard S, Chapman A, Kimmick G, Muss H, Luu T, Waisman JR, Li D, Mortimer J, Yuan Y, Somlo G, Stewart D, Katheria V, Levi A, Hurria A. Tolerability of the combination of lapatinib and trastuzumab in older patients with HER2 positive metastatic breast cancer [abstract]. In: Proceedings of the 2017 San Antonio Breast Cancer Symposium; 2017 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2018;78(4 Suppl):Abstract nr P5-21-08.

Doxycycline compared with prednisolone therapy for patients with bullous pemphigoid: cost-effectiveness analysis of the BLISTER trial

BACKGROUND

Bullous pemphigoid (BP) is an autoimmune blistering skin disorder associated with significant morbidity and mortality. Doxycycline and prednisolone to treat bullous pemphigoid were compared within a randomized controlled trial (RCT).

OBJECTIVES

To compare the cost-effectiveness of doxycycline-initiated and prednisolone-initiated treatment for patients with BP.

METHODS

Quality-of-life (EuroQoL-5D-3L) and resource data were collected as part of the BLISTER trial: a multicentre, parallel-group, investigator-blinded RCT. Within-trial analysis was performed using bivariate regression of costs and quality-adjusted life-years (QALYs), with multiple imputation of missing data, informing a probabilistic assessment of incremental treatment cost-effectiveness from a health service perspective.

RESULTS

In the base case, there was no robust difference in costs or QALYs per patient at 1 year comparing doxycycline- with prednisolone-initiated therapy [net cost £959, 95% confidence interval (CI) -£24 to £1941; net QALYs -0·024, 95% CI -0·088 to 0·041]. However, the findings varied by baseline blister severity. For patients with mild or moderate blistering (≤ 30 blisters) net costs and outcomes were similar. For patients with severe blistering (> 30 blisters) net costs were higher (£2558, 95% CI -£82 to £5198) and quality of life poorer (-0·090 QALYs, 95% CI -0·22 to 0·042) for patients starting on doxycycline. The probability that doxycycline would be cost-effective for those with severe pemphigoid was 1·5% at a willingness to pay of £20 000 per QALY.

CONCLUSIONS

Consistently with the clinical findings of the BLISTER trial, patients with mild or moderate blistering should receive treatment guided by the safety and effectiveness of the drugs and patient preference - neither strategy is clearly a preferred use of National Health Service resources. However, prednisolone-initiated treatment may be more cost-effective for patients with severe blistering.

Generating new evidence, improving clinical practice and developing research capacity: the benefits of recruiting to the U.K. Dermatology Clinical Trials Network's STOP GAP and BLISTER trials

Clinical trials may benefit clinical practice in three ways: firstly, clinicians may change their practice according to the new trial evidence; secondly, clinical processes can improve when working on a trial; and thirdly, research capacity is increased. We held a meeting to present and discuss the results of two large multicentre randomized controlled trials delivered through the U.K. Dermatology Clinical Trials Network. Investigators gave reflections on how the trials had changed their clinical practice. The STOP GAP trial showed that prednisolone and ciclosporin are equally effective as first-line systemic treatment for pyoderma gangrenosum. The final decision of which treatment to use should be based on the different adverse event profiles of the two drugs in relation to comorbidities, along with age, disease severity and patient preference. The BLISTER trial showed that starting people with pemphigoid on doxycycline produces acceptable short-term effectiveness and a superior safety profile to oral corticosteroids. Recruiting to these trials has led to the development of new specialist clinics with improved documentation. It has increased the profile of participating departments and embedded research in the department's activities. Helping to design and run these trials has also allowed trial staff to develop new skills in research design, which has been beneficial for career development. These and other benefits of recruiting to the trials are summarized here. We hope that these reflections will inspire wider involvement in clinical research.

Comparison of different insulin pump makes under routine care conditions in adults with Type 1 diabetes

AIMS

To compare long-term HbA_1c changes associated with different insulin pumps during routine care in a large cohort of adults with Type 1 diabetes representative of other clinic populations.

METHODS

Observational, retrospective study of 508 individuals starting pump therapy between 1999 and 2014 (mean age, 40 years; 55% women; diabetes duration, 20 years; 94% Type 1 diabetes; median follow-up, 3.7 years). Mixed linear models compared covariate-adjusted HbA_1c changes associated with different pump makes.

RESULTS

The pumps compared were: 50% Medtronic, 24% Omnipod, 14% Roche and 12% Animas. Overall HbA_1c levels improved and improvements were maintained during a follow-up extending to 10 years (HbA_1c : pre-continuous subcutaneous insulin infusion (pre-CSII) vs. 12 months post CSII, 71 (61, 82) vs. 66 (56, 74) mmol/mol; 8.7 (7.7, 9.6) vs. 8.2 (7.3, 8.9)%; P < 0.0001). The percentage of individuals with HbA_1c ≥ 64 mmol/mol (8.0%) reduced from a pre-CSII level of 68% to 55%. After adjusting for baseline confounders, there were no between-pump differences in HbA_1c lowering (P = 0.44), including a comparison of patch pumps with traditional catheter pumps (P = 0.63). There were no significant (P < 0.05) between-pump differences in HbA_1c lowering in pre-specified subgroups stratified by pre-pump HbA_1c , age or diabetes duration. HbA_1c lowering was positively related to baseline HbA_1c (P < 0.001) and diabetes duration (P = 0.017), and negatively related to the number of years of CSII use (P = 0.024).

CONCLUSIONS

Under routine care conditions, there were no covariate-adjusted differences in HbA_1c lowering when comparing different pump makes, including a comparison of patch pumps vs. traditional catheter pumps. Therefore, the choice of CSII make should not be influenced by the desired degree of HbA_1c lowering.

Building a sustainable clinical academic workforce to meet the future healthcare needs of Australia and New Zealand: report from the first summit meeting

The delivery of healthcare that meets the requirements for quality, safety and cost-effectiveness relies on a well-trained medical workforce, including clinical academics whose career includes a specific commitment to research, education and/or leadership. In 2011, the Medical Deans of Australia and New Zealand published a review on the clinical academic workforce and recommended the development of an integrated training pathway for clinical academics. A bi-national Summit on Clinical Academic Training was recently convened to bring together all relevant stakeholders to determine how best to do this. An important part understood the lessons learnt from the UK experience after 10 years since the introduction of an integrated training pathway. The outcome of the summit was to endorse strongly the recommendations of the medical deans. A steering committee has been established to identify further stakeholders, solicit more information from stakeholder organisations, convene a follow-up summit meeting in late 2015, recruit pilot host institutions and engage the government and future funders.

Intra-parotid dermoid cyst: excision through a face lift incision

Intra-parotid dermoid cysts are extremely rare. Clinically, they present like any other parotid lump and pre-operative diagnosis is rarely possible. A 62 years old Caucasian man presented with a painless lump in his right parotid region. The Magnetic Resonance Imaging (MRI) scan revealed a 3 x 3 cm mass in the parotid tail. Cytological examination was unhelpful. The patient underwent successful excision of the cyst through a small face lift incision. No drain was inserted and the patient was discharged home the same day. The patient made an uneventful recovery and was highly satisfied with the aesthetic outcome. The definite histopathological diagnosis was an intra-parotid dermoid cyst.

Preservation of speech and swallowing after excision of a pharyngeal pouch tumour

A pharyngeal pouch is not an uncommon cause of dysphagia in older population. However, finding a tumour in a pharyngeal pouch is very rare. A 79-year-old gentleman presented with dysphagia and a neck lump. Videofluoroscopy showed a pharyngeal pouch. Rigid endoscopy confirmed a tumour arising from the pharyngeal pouch. The histology showed it to be squamous cell carcinoma. The MRI scan based staging was T3N1M0. The patient underwent modified right radical neck dissection, partial pharyngectomy with primary repair and tracheostomy. He completed postoperative radiotherapy and remained disease-free at 4 years follow-up.