Premature termination of early-phase cancer clinical trials: “It’s the economy, stupid!”

In the 1992 US presidential campaign, strategist James Carville famously reminded campaign staff, “It’s the economy, stupid!” to underscore that a single underlying force was shaping many problems. In contemporary oncology, the analysis by Iannantuono and colleagues in this issue suggests that a similar logic applies to early-phase cancer drug development.1 Although it may seem unsurprising to many clinicians and investigators, when phase 1 cancer clinical trials are stopped before completion, the decisive forces are more often economic and strategic than strictly scientific or ethical. Iannantuono et al evaluated 1740 phase 1 clinical trials in solid tumors initiated between 2018 and 2023, alongside 561 phase 2 to 3 trials from the same period, using an aggregate analysis of the ClinicalTrials.gov database.1 They classified reasons for premature discontinuation into prespecified categories and examined temporal patterns and study-level factors associated with stopping. The central observation is striking: in phase 1 studies, “sponsor/strategic decisions” account for most premature terminations (51.1%), whereas “poor accrual” predominates in phase 2 to 3 trials (35.5%). Rather than being curtailed mainly by unacceptable toxicity, clear futility, or underwhelming efficacy, phase 1 programs are most commonly halted because sponsors realign portfolios, redirect capital, or reconsider strategic priorities.1 First-in-human oncology studies are high-risk, capital-intensive endeavors, and many biotechnology companies reach clinical trial stages with only one or two lead assets.2,3 For such sponsors, a phase 1 trial is a scientific experiment and a financing-sensitive milestone. In the era of targeted therapies and immunotherapies for cancer, early-phase clinical trials are no longer merely safety or dose-finding exercises but are often designed to probe early signals of efficacy through specific expansion cohorts. Ambiguous early pharmacodynamic signals, a rapidly changing competitive landscape, or a tightening funding environment can all trigger “strategic” discontinuation, particularly when portfolio-management frameworks favor reallocating resources to programs perceived as more competitive.4 The novelty of the work by Iannantuono et al. is not that economics matter but that, when carefully counted, economic and portfolio considerations emerge as the dominant proximate reason for losing phase 1 clinical trials in solid tumors. The timing of these decisions is equally revealing. By estimating the cumulative probabilities of trial completion and premature termination over time, the authors show that the risk of stopping rises sharply between 10 and 30 months after trial initiation and then gradually approaches a plateau at approximately 60 months. At about 35 months, the curves for termination and completion intersect; beyond this point, phase 1 trials are more likely to complete as planned than to stop early. This empirically defined 10 to 30-month “critical window” is precisely when sponsors have accumulated enough information—safety, pharmacokinetic, preliminary activity, competitive, and capital-markets signals—to make pivotal go/no-go decisions, but before sunk costs lock development trajectories in place.1 The structural context of a study strongly modulates its vulnerability to this economic triage. In multivariable models, trials funded by academic institutions have a substantially lower probability of premature termination than those funded by industry (odds ratio 0.50 vs industry). This difference likely reflects distinct incentive structures: academic sponsors are relatively insulated from quarterly earnings cycles and formal portfolio realignment and may tolerate slower accrual or incremental progress in pursuit of longer-term scientific aims, even within the constraints of limited budgets and personnel. Industry-sponsored studies, by contrast, are explicitly embedded in product-portfolio decision making, in which phase 1 programs can be curtailed when they are deemed redundant, insufficiently differentiated, or misaligned with evolving strategic priorities, regardless of whether they have “failed” on strictly scientific grounds. Geographical setting is also informative. Cancer clinical trials conducted in Europe, Asia, or Oceania show a lower likelihood of premature discontinuation than those conducted in the United States, whereas globally distributed phase 1 programs—arguably the most complex and resource-intensive—are more likely to terminate early.1 Large, multi-country trials must navigate heterogeneous regulatory requirements and logistical challenges, amplifying operational risk when sponsor enthusiasm wanes, whereas cohesive regional networks with more aligned processes may confer greater resilience once a study is underway. The nature of the enrolled patient population appears to influence vulnerability as well. Phase 1 trials focused on thoracic or gastrointestinal malignancies exhibit a higher probability of premature termination than those enrolling patients across multiple solid tumor types (odds ratios 1.77 and 1.52, respectively, compared with “multiple solid tumors”). These are clinical areas in which standards of care and available agents have evolved rapidly, often with serial approvals of targeted therapies and immunotherapies. In such settings, an investigational agent can quickly lose strategic priority as new competitors emerge or practice changes, even in the absence of unequivocal safety or futility signals. The analysis by Iannantuono et al. therefore suggests that some phase 1 programs in these diseases may be abandoned more for strategic reasons than for translational failure.1 By contrast, the class of investigational agent does not emerge as a robust independent determinant of premature termination. Aside from a heterogeneous “other” category, which the authors appropriately treat cautiously, drug class does not clearly distinguish studies that complete from those that stop early. Together, these observations reinforce the impression that structural, organizational, and economic factors outweigh molecular mechanism in determining whether a phase 1 trial survives long enough to reach its planned conclusion.1 At the same time, it is important to recognize that industry-sponsored phase 1 programs may also be disproportionately responsible for agents that ultimately reach regulatory approval or change clinical practice. In this sense, industry-led phase 1 trials may represent a high-risk, high-reward segment of the portfolio: more exposed to early discontinuation, but also more likely to yield transformative therapies when they succeed. Thus, higher termination risks should not ultimately be equated with lower overall clinical impact. How does this fit with prior evidence? Earlier work across oncology and other fields has documented high rates of trial discontinuation for reasons unrelated to prespecified stopping rules and has emphasized poor accrual, operational problems, and sponsor decisions as recurring themes.2,3 Concerns that premature discontinuation for financial or strategic reasons undermines the ethical basis of clinical research were notably articulated by Lièvre and colleagues, who showed that nonscientific termination can mislead participants and violate the spirit of the Helsinki Declaration.5 Most of these prior analyses, however, focused on mixed-phase cohorts or specific diseases and tended to highlight low enrollment as the dominant driver, particularly in phase 2 and 3 oncology trials.2,3 Commentaries and case reports have pointed to financial constraints and corporate strategy as forces that can terminate trials, but these have often been qualitative and anecdotal.5 The delay in seeing an analysis like that of Iannantuono et al. reflects that only recently have trial registries become sufficiently complete and structured, the phase 1 landscape sufficiently crowded and capital-intensive, and the motivation strong enough to justify the labor of classifying thousands of unstructured termination reports.6 The present study’s main contribution is to quantify, in a large and contemporary cohort of phase 1 solid-tumor trials, that sponsor/strategic decisions are the single most common reason for premature discontinuation and that this pattern contrasts sharply with the accrual-driven terminations that predominate in later-phase studies.1,2,3 There are important limitations. The broad category of “sponsor/strategic decisions” inevitably conflates ethically and scientifically justified actions (for example, stopping in the face of emerging safety concerns that do not meet formal stopping thresholds, or clear inferiority to a new standard of care) with decisions driven primarily by changing business strategy or capital constraints.1 Because termination reasons in registries are provided in unstructured free text and are not externally adjudicated, it is difficult to disentangle these subtypes, and registry data may be incomplete or delayed.1,6 Furthermore, the study window spans the global COVID-19 pandemic, which severely disrupted trial activation, patient mobility, and site operations worldwide; pandemic-related effects on both trial conduct and the coding of termination reasons are challenging to separate from underlying secular trends.1,6 Although the period overlaps the pandemic—with its well-documented disruptions to oncology trials—these effects alone are unlikely to account for the predominance of sponsor and strategic decisions among phase 1 terminations.1,6 Finally, the analysis is restricted to phase 1 trials in solid tumors; patterns of premature termination may differ in hematological malignancies, pediatric oncology, or non-oncology fields, and these will require dedicated study.1 Fragile patients who altruistically volunteer for phase 1 trials do so with the understanding that they are contributing to knowledge that may shape the future of cancer therapy, accepting uncertain personal benefit.7 Investigators and institutions invest considerable effort to design, activate, and run these complex studies. When a trial stops because of compelling safety concerns or evident futility, that shared investment has served its purpose. When a trial is discontinued primarily because a sponsor has shifted strategic direction or capital has evaporated, the balance between participant burden and collective scientific gain becomes even more troubling.5 The present data do not inform us exactly how often that occurs within the broad “sponsor/strategic” category, but they make clear that the economic and strategic context is not an occasional side story; it is central to the fate of many phase 1 programs. What, if anything, can be done? First, improving the granularity and standardization of termination-reason categories in trial registries would help distinguish scientific and ethical justifications from operational failures and explicitly financial or strategic choices.6 Moreover, recognizing the 10 to 30-month period after trial initiation as a critical window, sponsors and partners could formalize “sustainability reviews” that integrate accrual, safety, early activity, competitive landscape, and funding, with the goal of identifying scientifically valuable trials at high risk of purely economic termination and exploring options for rescue or redesign. In selected cases, the transition of a declining industry-sponsored study to an academically led continuation might help preserve important scientific opportunities at lower incremental cost. Finally, funders, regulators, and research institutions should recognize that safeguarding patients’ contributions in phase 1 requires attention not only to dose-escalation design and biomarker strategy but also to the economic and organizational frameworks that determine whether a trial can realistically be carried to completion.1-4 Phase 1 oncology trials—especially those launched by small biotechnology companies with narrow pipelines—are acutely sensitive to portfolio economics and capital markets.3,4 But the fact that these pressures are well recognized does not mean they have been adequately addressed in policy and practice. By demonstrating that sponsor and strategic decisions account for most premature discontinuations in a phase 1 cohort, Iannantuono et al.1 provide empirical support for what many have long suspected.2,4,5 If Carville’s admonition in this context—that for early-phase cancer clinical trials, it really is the economy that matters—is to be accepted, then improving oncology research will require structural changes in how early-phase studies are prioritized, funded, and stewarded. This editorial reflects the viewpoint of the authors and is not the official position of Rutgers, the State University of New Jersey, or the New York University. The funders had no role in the writing of this editorial or the decision to submit it for publication. Wadih Arap (Conceptualization, Project administration, Writing—original draft, Writing—review & editing), Shridar Ganesan (Conceptualization, Writing—original draft, Writing—review & editing), and Renata Pasqualini (Conceptualization, Writing—original draft, Writing—review & editing) The authors were supported by National Institutes of Health (NIH) funding as follows: WA (P30CA072720); SG (P30CA016087); RP (P30CA072720). This work was also supported by research awards from the Levy-Longenbaugh Donor-Advised Fund (to WA and RP). No new data were generated or analyzed for this editorial.