News in Brief

Abstract

According to the National Cancer Institute, less than 3% of people who have cancer are considered eligible for enrolment in clinical research studies. Conventional measures for setting eligibility have been criticised as being overly restrictive, arbitrarily reusing criteria from previous trials or prioritising the inclusion of patients deemed to have low-risk profiles (eg, participants who are young, do not have severe disease, or do not have other comorbidities besides the condition of interest). However, despite over half of surveyed clinicians agreeing that eligibility criteria are too stringent and harmful to cancer trials, there is a lack of consensus on which criteria are not needed. To analyse the effects of overly restrictive trial eligibility criteria on low patient enrolment and treatment efficacy in cancer trials, Ruishan Liu (Stanford University, Stanford, CA, USA) and colleagues designed Trial Pathfinder, an open source artificial intelligence tool linked to the electronic health records of 61 094 real-world patients with advanced non-small-cell lung cancer (NSCLC). By using the inclusion criteria and treatment information from ten clinical trials in advanced NSCLC, Trial Pathfinder, first, emulated the enrolment of these trials using real-world patients from its dataset (adjusting for baseline confounding factors and to emulate randomisation) and, second, systematically relaxed different combinations of eligibility criteria in their model to quantify how the hazard ratio of overall survival changed in each scenario. They found that, on average, relaxing eligibility criteria (eg, whether patients had previously taken antineoplastic therapies) specified in the trial protocols doubled the number of eligible patients. Moreover, analysing patient safety, the hazard ratio for overall survival of the Trial Pathfinder dataset was either comparable to or smaller than that reported in the original trials (decreasing by an average of 0·05), indicating that many of the patients who were excluded from the original trials could have also benefitted from treatment in the trial. Crucially, their results also showed that the ten trials analysed in their study tended to unnecessarily exclude women and people older than 75 years, suggesting a systematic bias towards excluding these groups during enrolment. Taken together, Liu and colleagues' study provides important experimental evidence to show that broadening eligibility criteria to align with successful trials that used more relaxed laboratory thresholds could enhance trial inclusivity, diversity, quality, completion, and cost-effectiveness in NSCLC trials, with far-reaching implications for the reform of clinical trial design across multiple disease types. Current treatments for painful diabetic neuropathy (PDN), a chronic neuropathic pain condition linked to increasing age that affects approximately 20% of patients with diabetes, are associated with low efficacy, high incidence of adverse events, and significantly decreased quality of life. Consequently, gabapentin and pregabalin, the two most commonly prescribed drugs for PDN, are associated with discontinuation rates of more than 60% by 6 months, leading to a large patient population with significant unmet needs. Investigating the non-pharmacological approach of high frequency (10 kHz) spinal cord stimulation (SCS) on 216 patients (mean age 60·8 years) with PDN who were refractory to gabapentinoids and at least one other analgesic, Erika A Petersen (University of Arkansas for Medical Sciences, AR, USA) and colleagues randomly assigned patients to either 10 kHz SCS via an implanted device plus conventional medical management (CMM; n=113) or CMM alone (n=103). In the SCS group, 74 (85%) out of 87 participants achieved at least 50% pain relief on a visual analogue scale, compared with only 5% in the control group. Moreover, improved health-related quality of life was reported only in the SCS group, suggesting that SCS could represent a more tolerable long-term treatment option than analgesics for patients with PDN. As an independent risk factor for many age-related diseases, including cardiovascular disease, dementia, osteoporosis, and diabetes, major depressive disorder (MDD) has been long associated with premature morbidity and mortality. Using blood samples from participants with moderate-to-severe MDD who were somatically healthy and who had not taken medication at least 6 weeks before enrolment (n=49) and from a group of medically and psychiatrically healthy controls (n=60), Ekaterina Protsenko (University of California San Francisco, CA, USA) and colleagues analysed participants' methylation patterns using the so-called GrimAge clock, a composite mathematical algorithm of epigenetic ageing. Their results showed that participants with MDD had a significantly greater GrimAge (with a median difference of 2 years of excess cellular aging) relative to their chronological age compared with healthy controls, controlling for sex, current smoking status, and body-mass index. These early findings suggest accelerated epigenetic aging in patients with MDD and should be followed with long-term longitudinal studies to better clarify a possible association between MDD-associated GrimAge and mortality.