The Comparator Arm Is the Trial: Why OptiTROP-Lung05’s Design Debate Exposes a Structural Problem in Oncology Phase 3s

Pull up the OptiTROP-Lung05 efficacy table and the number that stops you is the hazard ratio: 0.35. Sacituzumab tirumotecan plus pembrolizumab against pembrolizumab alone, first-line, PD-L1-positive advanced non-small-cell lung cancer, median PFS not reached versus 5.7 months. That is not a marginal signal. That is a result that, under most oncology frameworks, rewrites treatment sequencing. But read the Lancet correspondence responding to Anwen Xiong and colleagues’ report and a quieter argument surfaces: was pembrolizumab monotherapy the right comparator in the first place?

That question sounds technical. It carries billion-dollar consequences. Between June 7, 2024, and March 27, 2025, 741 patients were enrolled across the trial’s multicenter randomized arms, patients whose treatment assignments were locked to a comparator arm that critics now argue may understate what patients actually receive in community practice. If the benchmark moves, so does everything the hazard ratio proves.

This is the through-line running beneath every major oncology Phase 3 filed in the last four years: the comparator arm is not a neutral scientific choice. It is a regulatory and commercial decision, dressed in the language of clinical equipoise. Getting it wrong does not just compromise scientific validity, it delays approval, invites Complete Response Letters, and hands regulators a justification to demand additional trials. The OptiTROP-Lung05 correspondence makes that tension visible. The question worth asking is how the industry got here, and what it will take to get out.

Where Standard of Care Stops Being Standard

Start with what pembrolizumab monotherapy actually represents in 2026. When KEYNOTE-024 reported a median PFS of 10.3 months for single-agent pembrolizumab in PD-L1 ≥ 50% NSCLC patients, that result established pembrolizumab as the first-line standard against which every subsequent regimen would be measured. That was the right call in 2016. The problem is that oncology has not stood still while sponsors have continued building comparator arms around that benchmark.

In the decade since KEYNOTE-024, combination chemoimmunotherapy regimens entered guidelines, histology-specific approvals proliferated, and real-world prescribing drifted toward doublet and triplet regimens for patients who tolerate them. A sponsor designing a Phase 3 today must answer a question the FDA’s guidance on adequate and well-controlled studies does not cleanly resolve: does “standard of care” mean the approved labeled indication, the NCCN preferred regimen, or what oncologists at academic centers actually prescribe on a Tuesday afternoon? Those three answers produce three different comparator arms, and potentially three different hazard ratios.

The OptiTROP-Lung05 correspondence crystallizes exactly this ambiguity. Pembrolizumab monotherapy remains approved and guideline-listed for PD-L1-positive NSCLC. But the correspondents’ implicit challenge, and it is a fair one, is whether a trial powered against a monotherapy comparator still generates evidence that is clinically actionable when many treating physicians have moved toward combinations. An HR of 0.35 against a weakened benchmark still produces a regulatory win. Whether it produces clinical practice change is a separate question, and regulators are increasingly distinguishing between the two.

The ESMO Magnitude of Clinical Benefit Scale captures this pressure explicitly: it evaluates new treatments against comparators using both absolute and relative gains for OS and PFS, with explicit scoring penalties when the comparator is judged suboptimal. A trial that wins by a wide margin against a stale comparator earns a lower ESMO score than one that wins by a narrower margin against an aggressive current standard. Payers and formulary committees in Europe read those scores. So, increasingly, do FDA reviewers who want to understand whether a new approval will displace an existing regimen or merely add to a crowded shelf.

The Mechanism Problem No One Talks About Loudly

Sacituzumab tirumotecan complicates this debate in a way that is specific to its mechanism. As a Trop-2-targeting antibody-drug conjugate, sac-TMT delivers a cytotoxic payload selectively to tumor cells expressing the trophoblast cell-surface antigen 2. Trop-2 is broadly expressed across NSCLC histologies, which is precisely why this ADC-immunotherapy combination looked attractive as a first-line strategy: you get checkpoint inhibition and targeted cytotoxicity in a single regimen, without requiring a targetable genomic alteration.

That mechanistic logic is sound. But it creates a comparator problem that the trial’s critics are right to flag. If sac-TMT is hypothesized to work partly through enhanced tumor antigen release, augmenting the immune response that pembrolizumab is trying to sustain, then measuring it against pembrolizumab alone is not simply a benchmark question. It potentially conflates two different effects: the additive benefit of the ADC payload and the synergistic enhancement of PD-1 blockade. A comparator arm that included chemotherapy would have allowed investigators to disentangle those contributions. Without it, the HR of 0.35 tells you that the combination is dramatically better than monotherapy. It does not tell you why, and that limits what the label will ultimately be able to claim.

Regulators notice this.

When the FDA reviews a combination IND and sees a trial structured to maximize the probability of a positive PFS result rather than to illuminate the mechanism of benefit, reviewers ask about contribution of components analyses. If sponsors cannot provide a credible answer, because the trial was not designed to answer it, the agency defaults to a more conservative label. The win on paper becomes a qualified approval in practice, and the commercial team spends the next three years doing post-marketing studies that should have been built into the Phase 3 design in 2023.

What Trial Architects Get Wrong at the Protocol Stage

The OptiTROP-Lung05 correspondence is, at its core, a protocol design critique arriving after the trial has already enrolled 741 patients and reported primary results. That timing matters, because it illustrates how the industry’s quality-control mechanisms for comparator arm selection operate too late. The Lancet correspondence process is rigorous, but it is a post-hoc peer review of a randomized trial that took two years to execute. The moment to challenge the comparator arm was at the IND-enabling protocol stage, during Type B meeting discussions with the FDA, or at the latest during the Special Protocol Assessment process, not after the data are in.

Sponsors avoid these early challenges for understandable commercial reasons. A more aggressive comparator arm narrows the expected treatment effect, increases the required sample size, extends the enrollment timeline, and raises the risk of a negative result. Against pembrolizumab monotherapy with a historical median PFS of 5.7 months in this population, a power calculation looks tractable. Against a chemoimmunotherapy backbone, the math gets harder and the budget gets bigger. So sponsors pick the comparator that makes the trial feasible rather than the one that makes the evidence definitive.

The FDA’s guidance on “adequate and well-controlled studies” under 21 CFR 314.126 requires that the comparator reflect an established effective regimen, but it does not require sponsors to select the most rigorous one available. That gap is where commercial pressures fill the space that scientific standards leave open. And it is a gap the agency has been slowly closing: the FDA’s 2019 guidance on oncology trial design increasingly signals that single-arm and suboptimal-comparator submissions face heightened scrutiny, particularly in indications with multiple available active therapies.

The practical consequence for sponsors designing NSCLC Phase 3s today is that the comparator arm question cannot be deferred to the statistical analysis plan. It must be resolved before the protocol is finalized, with explicit regulatory alignment, because the FDA reviewer who reads your submission will ask what the NCCN guideline said when your trial opened, and if your comparator lagged behind that guideline on day one of enrollment, the review clock starts ticking backward.

OptiTROP-Lung05’s HR of 0.35 will almost certainly survive regulatory scrutiny. The magnitude is simply too large to argue away. But the Lancet correspondence signals something the next sponsor planning a PD-L1-positive NSCLC combination trial should read carefully: a result that dramatic, against a comparator that contested, will spend years being relitigated in payer submissions, guideline committee meetings, and head-to-head trial designs by competitors who pick a harder benchmark and win on credibility what they cannot win on effect size. The first sponsor to file a Phase 3 combination trial against a chemoimmunotherapy backbone in this space will set the evidentiary standard, and every filing that came before it will be measured against what that trial proves.

References

1. The Lancet, “Optimising standard of care in non-small-cell lung cancer trials” (Correspondence, 2026)
2. Targeted Oncology, “Sacituzumab Tirumotecan Plus Pembrolizumab Improves PFS in NSCLC” (OptiTROP-Lung05 enrollment and results)
3. Patsnap Synapse, “Mechanism of Action of Sacituzumab Tirumotecan”
4. Oncology Nursing News, “Phase III Trial Finds Significant Improvement with Pembrolizumab for Advanced NSCLC” (KEYNOTE-024 median PFS data)
5. ASCO Post, “Assessing the Clinical Utility of ASCO’s and ESMO’s Value Frameworks” (ESMO MCBS comparator scoring)