Retrospective analysis of real-world data to evaluate actionability of a comprehensive molecular profiling panel in solid tumor tissue samples (REALM study).

INTRODUCTION: Considering the growing interest in matched cancer treatment, our aim was to evaluate the ability of a comprehensive genomic profiling (CGP) assay to propose at least one targeted therapy given an identified genomic alteration or signature (actionability), and to collect the treatment modifications based on the CGP test results in clinical practise for solid tumors. METHODS: This retrospective, multicentre French study was conducted among 25 centres that participated in a free of charge program between 2017 and 2019 for a tissue CGP test. Data were collected on the patient, disease, tumor genomic profile, treatment suggested in the report (related to the genomic profile results) and subsequent therapeutic decisions according to the physician's declaration. RESULTS: Among the 416 patients, most had lung cancer (35.6%), followed by biliary tract cancer (11.5%) or rare cancers (11.1%); 75% had a metastatic disease. The actionability was 75.0% (95% CI [70.6%-78.9%]) for all patients, 85.1% and 78.4%, respectively in lung cancer and metastatic patients. After exclusion of clinical trial suggestions, the actionability decreased to 62.3% (95% CI [57.5%-66.8%]). Treatment modification based on the test results was observed in 17.3% of the patients and was more frequent in metastatic disease (OR = 2.73, 95% CI [1.31-5.71], p = 0.007). The main reasons for no treatment modification were poor general condition (33.2%) and stable disease or remission (30.2%). The genomic-directed treatment changes were performed mostly during the first six months after the CGP test, and interestingly a substantial part was observed from six to 24 months after the genomic profiling. CONCLUSION: This French study provides information on the real-life actionability of a CGP test based on tissue samples, and trends to confirm its utility in clinical practice across the course of the disease, in particularly for patients with lung cancer and/or advanced disease.

Interactive statistical monitoring to optimize review of potential clinical trial issues during study conduct.

Background: Statistical monitoring involves the review of prospective study data collected in participating sites to detect intra/inter patients and sites inconsistencies. We report methods and results of statistical monitoring in a phase IV clinical trial. Method: PRO-MSACTIVE is a study evaluating ocrelizumab in active relapsing multiple sclerosis (RMS) patients in France. Specific statistical methods (volcano plots, mahalanobis distance, funnel plot …) have been applied to a SDTM database to detect potential issues. R-Shiny application was developed to generate an interactive web application in order to ease site and/or patients identification during statistical data review meetings. Results: The PRO-MSACTIVE study enrolled 422 patients in 46 centers between July 2018 and August 2019. Three data review meetings were held between April and October 2019 and 14 standard and planned tests were run on study data, with a total of 15 (32.6%) sites identified as needing review or investigation. Overall 36 findings were identified during the meetings: duplicate records, outliers, inconsistent delays between dates. Conclusion: Statistical monitoring is useful to identify unusual or clustered data patterns that might be revealing issues that could impact the data integrity and/or may potentially impact patients' safety. With anticipated and appropriate interactive data visualization, early signals can easily be identified or reviewed by the study team and appropriate actions be set up and assigned to the most appropriate function for a close follow-up and resolution. Interactive statistical monitoring is time consuming to initiate using R-Shiny, but is time saving after the 1st data review meeting (DRV).(ClinicalTrials.gov identifier: NCT03589105; EudraCT identifier: 2018-000780-91).

Identification of predictive factors of diabetic ketoacidosis in type 1 diabetes using a subgroup discovery algorithm.

AIM: To identify predictive factors for diabetic ketoacidosis (DKA) by retrospective analysis of registry data and the use of a subgroup discovery algorithm. MATERIALS AND METHODS: Data from adults and children with type 1 diabetes and more than two diabetes-related visits were analysed from the Diabetes Prospective Follow-up Registry. Q-Finder, a supervised non-parametric proprietary subgroup discovery algorithm, was used to identify subgroups with clinical characteristics associated with increased DKA risk. DKA was defined as pH less than 7.3 during a hospitalization event. RESULTS: Data for 108 223 adults and children, of whom 5609 (5.2%) had DKA, were studied. Q-Finder analysis identified 11 profiles associated with an increased risk of DKA: low body mass index standard deviation score; DKA at diagnosis; age 6-10 years; age 11-15 years; an HbA1c of 8.87% or higher (≥ 73 mmol/mol); no fast-acting insulin intake; age younger than 15 years and not using a continuous glucose monitoring system; physician diagnosis of nephrotic kidney disease; severe hypoglycaemia; hypoglycaemic coma; and autoimmune thyroiditis. Risk of DKA increased with the number of risk profiles matching patients' characteristics. CONCLUSIONS: Q-Finder confirmed common risk profiles identified by conventional statistical methods and allowed the generation of new profiles that may help predict patients with type 1 diabetes who are at a greater risk of experiencing DKA.

Assay Harmonization and Use of Biological Standards To Improve the Reproducibility of the Hemagglutination Inhibition Assay: a FLUCOP Collaborative Study.

The hemagglutination inhibition (HAI) assay is an established technique for assessing influenza immunity, through measurement of antihemagglutinin antibodies. Improved reproducibility of this assay is required to provide meaningful data across different testing laboratories. This study assessed the impact of harmonizing the HAI assay protocol/reagents and using standards on interlaboratory variability. Human pre- and postvaccination sera from individuals (n = 30) vaccinated against influenza were tested across six laboratories. We used a design of experiment (DOE) method to evaluate the impact of assay parameters on interlaboratory HAI assay variability. Statistical and mathematical approaches were used for data analysis. We developed a consensus protocol and assessed its performance against in-house HAI testing. We additionally tested the performance of several potential biological standards. In-house testing with four reassortant viruses showed considerable interlaboratory variation (geometric coefficient of variation [GCV] range of 50% to 117%). The age, concentration of turkey red blood cells, incubation duration, and temperature were key assay parameters affecting variability. Use of a consensus protocol with common reagents, including viruses, significantly reduced GCV between laboratories to 22% to 54%. Pooled postvaccination human sera from different vaccination campaigns were effective as biological standards. Our results demonstrate that the harmonization of protocols and critical reagents is effective in reducing interlaboratory variability in HAI assay results and that pools of postvaccination human sera have potential as biological standards that can be used over multiple vaccination campaigns. Moreover, the use of standards together with in-house protocols is as potent as the use of common protocols and reagents in reducing interlaboratory variability. IMPORTANCE The hemagglutination inhibition (HAI) assay is the most commonly used serology assay to detect antibodies from influenza vaccination or influenza virus infection. This assay has been used for decades but requires improved standardization of procedures to provide meaningful data. We designed a large study to assess selected parameters for their contribution to assay variability and developed a standard protocol to promote consistent HAI testing methods across laboratories. The use of this protocol and common reagents resulted in lower levels of variability in results between participating laboratories than achieved using in-house HAI testing. Human sera sourced from vaccination campaigns over several years, and thus including antibody to different influenza vaccine strains, served as effective assay standards. Based on our findings, we recommend the use of a common protocol and/or human serum standards, if available, for testing human sera for the presence of antibodies against seasonal influenza using turkey red blood cells.

A predictive model for risk of early grade ≥ 3 infection in patients with multiple myeloma not eligible for transplant: analysis of the FIRST trial.

Infections are a major cause of death in patients with multiple myeloma. A post hoc analysis of the phase 3 FIRST trial was conducted to characterize treatment-emergent (TE) infections and study risk factors for TE grade ≥ 3 infection. The number of TE infections/month was highest during the first 4 months of treatment (defined as early infection). Of 1613 treated patients, 340 (21.1%) experienced TE grade ≥ 3 infections in the first 18 months and 56.2% of these patients experienced their first grade ≥ 3 infection in the first 4 months. Risk of early infection was similar regardless of treatment. Based on the analyses of data in 1378 patients through multivariate logistic regression, a predictive model of first TE grade ≥ 3 infection in the first 4 months retained Eastern Cooperative Oncology Group performance status and serum ß2-microglobulin, lactate dehydrogenase, and hemoglobin levels to define high- and low-risk groups showing significantly different rates of infection (24.0% vs. 7.0%, respectively; P < 0.0001). The predictive model was validated with data from three clinical trials. This predictive model of early TE grade ≥ 3 infection may be applied in the clinical setting to guide infection monitoring and strategies for infection prevention.

MMS19 as a potential predictive marker of adjuvant chemotherapy benefit in resected non-small cell lung cancer.

BACKGROUND: Resectable non-small cell lung cancer (NSCLC) treatment options most often consist of surgical resection along with adjuvant chemotherapy (ACT). The benefit of ACT however is modest and is accompanied by important side effects. OBJECTIVE: One central quest in the field is therefore the identification of a predictive marker of the response to ACT. METHODS: We applied an unbiased approach based on high content analysis of expression data generated from a discovery patient cohort. RESULTS: We identified MMS19, a component of the cytoplasmic Iron-Sulfur Assembly (CIA) machinery important for the Nucleotide Excision Repair (NER) pathway as a pivotal gene for cisplatin toxicity. We then confirmed the association between MMS19 expression and the response to Cisplatin treatment in a panel of NSCLC cell lines. Finally we validated these pre-clinical data in a subgroup of JBR.10 trial patients through a hypothesis-driven analysis, and showed that MMS19 levels associated with ACT benefit. CONCLUSIONS: We therefore propose the expression level of MMS19 as a candidate predictive marker of ACT benefit in resected NSCLC patients.