The tyranny of non-inferiority trials.

Opportunities to decrease the toxicity and cost of approved treatment regimens with lower dose, less frequent, or shorter duration alternative regimens have been limited by the perception that alternatives must be non-inferior to approved regimens. Non-inferiority trials are large and expensive to do, because they must show statistically that the alternative and approved therapies differ in a single outcome, by a margin far smaller than that required to demonstrate superiority. Non-inferiority's flaws are manifest: it ignores variability expected to occur with repeated evaluation of the approved therapy, fails to recognise that a trial of similar design will be labelled as superiority or non-inferiority depending on whether it is done prior to or after initial registration of the approved treatment, and relegates endpoints such as toxicity and cost. For example, while a less toxic and less costly regimen of 3 months duration would typically be required to demonstrate efficacy that is non-inferior to that of a standard regimen of 6 months to displace it, the longer duration therapy has no such obligation to prove its superiority. This situation is the tyranny of the non-inferiority trial: its statistics perpetuate less cost-effective regimens, which are not patient-centred, even when less intensive therapies confer survival benefits nearly identical to those of the standard, by placing a disproportionately large burden of proof on the alternative. This approach is illogical. We propose that the designation of trials as superiority or non-inferiority be abandoned, and that randomised, controlled trials should henceforth be described simply as "comparative".

Cell proliferation, drug distribution and therapeutic effects in relation to the vascular system of solid tumours.

In this perspective, the authors summarise some properties of the solid tumour micro-environment that have been explored during the last 55 years. It is well established that the concentrations of nutrients, including oxygen, decrease with increasing distance from tumour blood vessels, and that low extracellular pH is found in nutrient-poor regions. Cell proliferation is dependent on nutrient metabolites and decreases in regions distal from patent blood vessels. Proliferating cells cause migration of neighbouring cells further from blood vessels where they may die, and their breakdown products pass into regions of necrosis. Anticancer drugs reach solid tumours via the vascular system and establish concentration gradients such that drug concentration within tumours may be quite variable. Treatment with chemotherapy such as doxorubicin or docetaxel can kill well-nourished proliferating cells close to blood vessels, thereby interrupting migration toward necrotic regions and lead to re-oxygenation and renewed proliferation of distal cells, as can occur with radiotherapy. This effect leads to the paradox that cancer treatment can rescue cells that were destined to die in the untreated tumour. Renewed and sometimes accelerated repopulation of surviving tumour cells can counter the effects of cell killing from repeated treatments, leading to tumour shrinkage and regrowth without changes in the intrinsic sensitivity of cells to the administered treatment. Strategies to prevent these effects include the combined use of chemotherapy with agents that selectively kill hypoxic tumour cells, including inhibitors of autophagy, since this is a process that may allow recycling of cellular macromolecules from dying cells and improve their survival.

Lack of cognitive impairment in long-term survivors of colorectal cancer.

BACKGROUND: Our longitudinal study reported cognitive impairment in 43% of people following diagnosis of localised colorectal cancer (CRC) versus 15% in healthy controls (p < 0.001) and 50% versus 13% 1-2 years later (p < 0.001). Here we evaluate cognitive function and neuroimaging in a subgroup at long-term follow-up. PATIENTS AND METHODS: Cancer-free Australian participants in the study, and controls, completed cognitive and functional assessments. Neuroimaging was optional. Blood tests included inflammatory markers, clotting factors, sex hormones and apolipoprotein E genotype. The primary endpoint was demographically and practice effect-corrected cognitive scores comparing CRC survivors with controls over time examined using a linear mixed model, adjusted for baseline performance. Secondary endpoints included cognitive impairment rate using the Global Deficit Score [GDS > 0.5], Functional Deficit Score, blood results and neuroimaging. RESULTS: The study included 25 CRC survivors (60% men, median age 72) at mean 9 years after baseline (9 received adjuvant chemotherapy) and 25 controls (44% men, median age 68) at mean 6 years after baseline. There were no significant differences in cognitive scores or proportion with cognitive impairment (16 vs. 8%) between survivors and controls and no evidence of accelerated ageing in CRC survivors. Baseline cognitive performance predicted for subsequent cognitive function. There were no differences in functional tests or blood tests between groups. In 18 participants undergoing neuroimaging, 10 CRC survivors had higher myoinositol levels than 8 controls, and lower volume in the right amygdala and caudate and left hippocampal regions. CONCLUSIONS: There was no difference in cognitive capacity and function between CRC survivors and controls 6-12 years after diagnosis. Differences in neuroimaging require confirmation in a larger sample. HIGHLIGHTS: • No evidence of long term cognitive impairment in colorectal cancer survivors compared to controls 6-12 years after diagnosis • No evidence of accelerated cognitive ageing in colorectal cancer survivors • No evidence of long-term functional impairment in colorectal cancer survivors.

Can Oncologists Predict the Efficacy of Treatments in Randomized Trials?

BACKGROUND: Decisions about trial funding, ethical approval, or clinical practice guideline recommendations require expert judgments about the potential efficacy of new treatments. We tested whether individual and aggregated expert opinion of oncologists could predict reliably the efficacy of cancer treatments tested in randomized controlled trials. MATERIALS AND METHODS: An international sample of 137 oncologists specializing in genitourinary, lung, and colorectal cancer provided forecasts on primary outcome attainment for five active randomized cancer trials within their subspecialty; skill was assessed using Brier scores (BS), which measure the average squared deviation between forecasts and outcomes. RESULTS: A total of 40% of trials in our sample reported positive primary outcomes. Experts generally anticipated this overall frequency (mean forecast, 34%). Individual experts on average outperformed random predictions (mean BS = 0.29 [95% confidence interval (CI), 0.28-0.33] vs. 0.33) but underperformed prediction algorithms that always guessed 50% (BS = 0.25) or that were trained on base rates (BS = 0.19). Aggregating forecasts improved accuracy (BS = 0.25; 95% CI, 0.16-0.36]). Neither individual experts nor aggregated predictions showed appreciable discrimination between positive and nonpositive trials (area under the curve of a receiver operating characteristic curve, 0.52 and 0.43, respectively). CONCLUSION: These findings are based on a limited sample of trials. However, they reinforce the importance of basing research and policy decisions on the results of randomized trials rather than expert opinion or low-level evidence. IMPLICATIONS FOR PRACTICE: Predictions of oncologists, either individually or in the aggregate, did not anticipate reliably outcomes for randomized trials in cancer. These findings suggest that pooled expert opinion about treatment efficacy is no substitute for randomized trials. They also underscore the challenges of using expert opinion to prioritize interventions for clinical trials or to make recommendations in clinical practice guidelines.

Design of placebo-controlled randomized trials of anticancer agents: Ethical considerations based on a review of published trials.

BACKGROUND: Limited information exists about the design of placebo-controlled cancer trials. Through a systematic review of trials published in 2013, we describe placebo use in randomized trials testing anticancer agents and analyze strategies that increase exposure to the experimental regimen. METHODS: Trials were classified as add-on (placebo in combination with standard treatment) or placebo-only. Strategies to allow more than half of the participants to receive the experimental regimen were reviewed. The risk-benefit ratio of receiving the experimental agent was considered favorable if the difference in primary outcome was significant (p ≤ 0.05), neutral if there was no significant difference in the primary outcome and the experimental agent did not add substantial toxicity, and unfavorable otherwise. RESULTS: Eighty trials were included (32,694 participants). Most trials were add-on (69%). The risk-benefit outcome was favorable, neutral, and unfavorable to the experimental agent in 52%, 32%, and 16% of placebo-only trials and 25%, 53%, and 22%, respectively, of add-on trials. Four strategies increased exposure to the experimental regimen: one-way crossover (23%), uneven randomization (21%), three-arms (13%), and randomized discontinuation design (4%); these strategies were used more often in placebo-only trials. CONCLUSION: A minority of participants received placebo alone and strategies to increase experimental exposure were used commonly. Fewer than half of the studies had favorable outcomes, thus defending the use of placebo controls, when there is no established treatment. Strategies that increase patient exposure to experimental agents rather than placebo may expose them to non-beneficial, sometimes toxic, experimental agents.

The fragility of phase 3 trials supporting FDA-approved anticancer medicines: a retrospective analysis.

BACKGROUND: The fragility index of trial results-ie, the minimum number of changes from non-events to events resulting in loss of statistical significance-can provide a measure of confidence that a positive effect reported in a randomised controlled trial is real. We aimed to calculate the fragility index of randomised controlled trials supporting US Food and Drug Administration (FDA)-approved anticancer drugs. METHODS: This is a retrospective analysis of phase 3, randomised, controlled trials supporting anticancer drugs that were approved by the FDA between Jan 1, 2014, and Dec 31, 2018. Two-arm studies with 1:1 randomisation and significant positive results for a time-to-event outcome were eligible for the fragility index calculation, which involves the iterative addition of an event to the experimental group (defined as the group with the smaller number of events in positive trials) and concomitant subtraction of a non-event from that group, until positive significance (defined as p<0·05 by Fisher's exact test) is lost. FINDINGS: We identified 36 phase 3 randomised controlled trials, of which 17 (47%) were included in the fragility index analysis. The median fragility index was 2 (IQR 0-27). The fragility index was 2 or less in nine (53%) of 17 trials; for these trials, the fragility index was 1% or less of the total sample size. In five (29%) of 17 trials, the number lost to follow-up was more than the fragility index. INTERPRETATION: Many phase 3 randomised controlled trials supporting FDA-approved anticancer drugs have a low fragility index, challenging confidence for concluding their superiority over control treatments. Although not a measure of effect, the fragility index might provide an additional means of assessing the robustness of clinical trial data. FUNDING: None.

Pantoprazole Affecting Docetaxel Resistance Pathways via Autophagy (PANDORA): Phase II Trial of High Dose Pantoprazole (Autophagy Inhibitor) with Docetaxel in Metastatic Castration-Resistant Prostate Cancer (mCRPC).

BACKGROUND: Enhancing the effectiveness of docetaxel for men with metastatic castration-resistant prostate cancer (mCRPC) is an unmet clinical need. Preclinical studies demonstrated that high-dose pantoprazole can prevent or delay resistance to docetaxel via the inhibition of autophagy in several solid tumor xenografts. MATERIALS AND METHODS: Men with chemotherapy-naive mCRPC with a prostate-specific antigen (PSA) >10 ng/mL were eligible for enrolment. Men received intravenous pantoprazole (240 mg) prior to docetaxel (75 mg/m2) every 21 days, with continuous prednisone 5 mg twice daily. Primary endpoint was a confirmed ≥50% decline of PSA. The trial used a Simon's two-stage design. RESULTS: Between November 2012 and March 2015, 21 men with a median age of 70 years (range, 58-81) were treated (median, 6 cycles; range, 2-11). Men had received prior systemic therapies (median, 1; range, 0-3), and 14 had received abiraterone and/or enzalutamide. PSA response rate was 52% (11/21), which did not meet the prespecified criterion (≥13/21 responders) to proceed to stage 2 of the study. At interim analysis with a median follow-up of 17 months, 18 (86%) men were deceased (15 castration-resistant prostate cancer, 2 unknown, 1 radiation complication). Of the men with RECIST measurable disease, the radiographic partial response rate was 31% (4/13). The estimated median overall survival was 15.7 months (95% confidence interval [CI], 9.3-19.6) and median PFS was 5.3 months (95% CI, 2.6-12.9). There were no toxic deaths, and all adverse events were attributed to docetaxel. CONCLUSION: The combination of docetaxel and pantoprazole was tolerable, but the resultant clinical activity was not sufficient to meet the ambitious predefined target to warrant further testing. IMPLICATIONS FOR PRACTICE: To date, no docetaxel combination regimen has reported superior efficacy over docetaxel alone in men with metastatic castration-resistant prostate cancer (mCRPC). The PANDORA trial has demonstrated that the combination of high dose pantoprazole with docetaxel is tolerable, but the clinical activity was not sufficient to warrant further testing. The chemotherapy standard of care for men with mCRPC remains docetaxel with prednisone. Future studies of autophagy inhibitors will need to measure autophagy inhibition accurately and determine the degree of autophagy inhibition required to produce a meaningful clinical response.

Delivery of meaningful cancer care: a retrospective cohort study assessing cost and benefit with the ASCO and ESMO frameworks.

BACKGROUND: The American Society of Clinical Oncology (ASCO) and the European Society for Medical Oncology (ESMO) have developed frameworks that quantify survival gains in light of toxicity and quality of life to assess the benefits of cancer therapies. We applied these frameworks to a cohort of contemporary randomised controlled trials to explore agreement between the two approaches and to assess the relation between treatment benefit and cost. METHODS: We identified all randomised controlled trials of systemic therapies in non-small-cell lung cancer, breast cancer, colorectal cancer, and pancreatic cancer published between Jan 1, 2011, and Dec 31, 2015, and assessed their abstracts and methods. Trials were eligible for inclusion in our cohort if significant differences favouring the experimental group in a prespecified primary or secondary outcome were reported (secondary outcomes were assessed only if primary outcomes were not significant). We assessed trial endpoints with the ASCO and ESMO frameworks at two timepoints 3 months apart to confirm intra-rater reliability. Cohen's κ statistic was calculated to establish agreement between the two frameworks on the basis of the median ASCO score, which was used as an arbitrary threshold of benefit, and the framework-recommended ESMO threshold. Differences in monthly drug cost between the experimental and control groups of each randomised controlled trial (ie, incremental drug cost) were derived from 2016 average wholesale prices. FINDINGS: 109 randomised controlled trials were eligible for inclusion, 42 (39%) in non-small-cell lung cancer, 36 (33%) in breast cancer, 25 (23%) in colorectal cancer, and six (6%) in pancreatic cancer. ASCO scores ranged from 2 to 77; median score was 25 (IQR 16-35). 41 (38%) trials met the benefit thresholds in the ESMO framework. Agreement between the two frameworks was fair (κ=0·326). Among the 100 randomised controlled trials for which drug costing data were available, ASCO benefit score and monthly incremental drug costs were negatively correlated (ρ=-0·207; p=0·039). Treatments that met ESMO benefit thresholds had a lower median incremental drug cost than did those that did not meet benefit thresholds (US$2981 [IQR 320-9059] vs $8621 [1174-13 930]; p=0·018). INTERPRETATION: There is only fair correlation between these two major value care frameworks, and negative correlations between framework outputs and drug costs. Delivery of optimal cancer care in a sustainable health system will necessitate future oncologists, investigators, and policy makers to reconcile the disconnect between drug cost and clinical benefit. FUNDING: None.

Effects of long-term androgen deprivation therapy on cognitive function over 36 months in men with prostate cancer.

BACKGROUND: Many men with prostate cancer (PC) require long-term androgen deprivation therapy (ADT), but to the authors' knowledge, its effects on cognitive function beyond 1 year are not described. METHODS: Three groups of men aged ≥50 years who were matched based on age and education were enrolled: 77 patients with nonmetastatic PC who initiated continuous ADT, 82 patients with PC who were not receiving ADT (PC controls), and 82 healthy controls. A battery of 14 neuropsychological tests, examining 8 cognitive domains, was administered on 5 occasions over 36 months. Changes in cognitive scores over time were analyzed using 3 approaches: linear mixed effects regression, the percentage of participants per group with declines in ≥1/2 cognitive tests, and a global summary of cognitive change. RESULTS: The mean age of the study subjects was 68.9 years, with a median of 16 years of education. In mixed effects models adjusted for age and education, ADT use was not found to be associated with significant changes over time in any cognitive test compared with healthy controls. The percentage of participants declining by ≥1.5 standard deviations in ≥2 tests or ≥2 standard deviations in ≥1 tests was similar across groups. A global summary of cognitive change found no statistically significant worsening of cognitive function among ADT users compared with controls. Sensitivity analyses adjusting for duration of ADT and using multiple imputation for missing data did not materially alter the study findings. CONCLUSIONS: The ongoing use of ADT for up to 36 months does not appear to be associated with cognitive decline. Cancer 2017;123:237-244. © 2016 American Cancer Society.

Under-reporting of harm in clinical trials.

Appropriate safety evaluations of anticancer drugs are crucial to assess their benefit-risk ratio. Substantial evidence shows that clinicians under-report harm in clinical trials, and at least three factors contribute to this problem: assessment of harm by clinicians might not represent the experience of patients; harm might be detected within trials, but is not reported appropriately by investigators or reporting is influenced by sponsors; and short-term follow-up might not detect long-term and potentially serious toxicities. Additionally, because of the selection of patients with good functional status in clinical trials, study results might not apply to patients treated in everyday clinical practice. New approaches for the conduct, oversight, and reporting of clinical trials should include patient-reported assessment of side-effects. Effective pharmacovigilance programmes and large-scale observational studies are needed to improve understanding of the tolerability of anticancer drugs in a real world setting.

Relevance of randomised controlled trials in oncology.

Well-designed randomised controlled trials (RCTs) can prevent bias in the comparison of treatments and provide a sound basis for changes in clinical practice. However, the design and reporting of many RCTs can render their results of little relevance to clinical practice. In this Personal View, we discuss the limitations of RCT data and suggest some ways to improve the clinical relevance of RCTs in the everyday management of patients with cancer. RCTs should ask questions of clinical rather than commercial interest, avoid non-validated surrogate endpoints in registration trials, and have entry criteria that allow inclusion of all patients who are fit to receive treatment. Furthermore, RCTs should be reported with complete accounting of frequency and management of toxicities, and with strict guidelines to ensure freedom from bias. Premature reporting of results should be avoided. The bar for clinical benefit should be raised for drug registration, which should require publication and review of mature data from RCTs, post-marketing health outcome studies, and value-based pricing.