KEYNOTE-671 Extends the View Beyond Pathological Complete Response in Early-Stage NSCLC
An exploratory analysis from KEYNOTE-671 turns attention to the patients who did not reach pathological complete response after perioperative pembrolizumab for early-stage non-small cell lung cancer. The focus is clinically important because pCR often serves as an early marker of treatment effect, yet many patients seen in routine practice and in trials will not achieve it. For those patients, the more relevant question is not whether the tumor disappeared entirely in the surgical specimen, but whether the perioperative strategy still altered recurrence risk, delayed progression, or changed the postoperative course in a durable way.
Five-year follow-up gives this analysis added weight. In curative-intent lung cancer trials, time reveals the difference between a transient pathologic response and a treatment effect that survives surgery, adjuvant therapy, and the biology of micrometastatic disease. Early-stage NSCLC remains a disease in which relapse after apparently successful local therapy is still common enough to justify systemic treatment before and after resection. Perioperative immunotherapy is designed precisely for that gap: it aims to prime antitumor immunity before the tumor is removed, then maintain immune pressure during the postoperative period when residual disease may be most vulnerable.
The title alone makes clear what this analysis is trying to answer. If a patient does not achieve pCR, does that automatically mean the perioperative regimen failed? In a disease driven by both local tumor burden and occult systemic spread, the answer is rarely so simple. Pathologic response is one layer of evidence, not the entire clinical story. Immune checkpoint blockade can produce benefits that extend beyond complete eradication of tumor cells in the resection specimen, including control of microscopic disease that is not captured by surgical pathology alone.
That distinction matters most in the clinic, where postoperative decision-making often hinges on how much confidence to place in pathologic findings. A patient without pCR may still have derived meaningful benefit from pembrolizumab if the regimen lowered the chance of recurrence or delayed the appearance of metastatic disease. Conversely, a lack of pCR can still identify a group with higher residual risk, which may be useful when considering surveillance intensity and the need for continued systemic therapy. The practical value of this exploratory analysis lies in separating those possibilities rather than treating all non-pCR patients as a single prognostic block.
Why pCR Is Useful but Not Sufficient
Pathological complete response has become a familiar surrogate in neoadjuvant and perioperative oncology because it is immediate, measurable, and biologically meaningful. In resected lung cancer, a complete disappearance of viable tumor suggests robust sensitivity to treatment and a lower residual burden entering the postoperative phase. Yet tumor immunology does not always conform to binary pathology. A specimen with residual viable cells may still sit within a broader immune-responsive state, especially when treatment has already altered the tumor microenvironment and reduced the likelihood that scattered malignant clones can re-establish themselves after surgery.
That is why exploratory subgroup analyses deserve careful reading. They help determine whether the treatment effect extends across response categories or is concentrated among patients who achieve the deepest pathologic response. For surgeons, oncologists, and thoracic multidisciplinary teams, that information affects how perioperative immunotherapy is discussed before treatment starts. Patients frequently want to know whether a less-than-perfect pathologic result means the regimen failed. A study centered on non-pCR outcomes can answer that question more precisely than headline pCR rates alone.
Perioperative Immunotherapy and the Biology of Residual Disease
Pembrolizumab targets the PD-1 pathway, removing a brake on activated T cells. In early-stage lung cancer, that immune activation is most compelling when delivered before resection, because the intact tumor can serve as an antigen source and sustain immune priming. Once surgery removes the primary lesion, the challenge shifts to eradicating cells that escaped detection—cells that may already have seeded lymph nodes or distant sites long before the operation.
For patients without pCR, the biological question becomes whether the immune response was still strong enough to suppress those surviving clones. That is a very different endpoint from complete histologic sterilization of the tumor bed. It also explains why long follow-up is essential. Recurrence risk in NSCLC can remain hidden for years, and a treatment that fails to deliver pCR may still succeed in altering the pace or pattern of relapse. Five-year follow-up is long enough to begin judging whether the perioperative strategy changed the natural history of the disease rather than merely reshaped the surgical pathology report.
Residual disease after neoadjuvant treatment is not a single entity. Some tumors retain a small amount of viable cancer with a microenvironment still amenable to immune attack. Others may be intrinsically less responsive to checkpoint inhibition and more likely to recur despite treatment. An exploratory analysis of non-pCR participants can help distinguish these scenarios, even when the abstract does not yet provide the full numerical detail.
What This Analysis Signals for Practice
For now, the most defensible reading is conceptual rather than quantitative. The trial asks whether the benefit of perioperative pembrolizumab reaches beyond the subgroup with complete pathologic eradication and remains meaningful in patients who do not fully clear viable tumor at surgery. That question is central to real-world thoracic oncology, where pathologic response is helpful but never the sole determinant of prognosis.
This also has implications for how future perioperative studies are designed. Response-adapted strategies, biomarker selection, and postoperative escalation all depend on understanding the fate of non-pCR patients. If the benefit of pembrolizumab persists in that group, then pCR becomes one marker among several rather than a gatekeeper for treatment success. If benefit narrows sharply, then residual viable tumor may identify a subgroup in need of different postoperative intensification.
For readers looking for a simple endpoint summary, this abstract does not offer one in the source material provided. What it does offer is a clinically meaningful frame: five-year follow-up, early-stage non-small cell lung cancer, perioperative pembrolizumab, and a deliberate focus on patients who did not achieve pathological complete response. The value of that lens is straightforward. It forces the field to judge perioperative immunotherapy by recurrence biology and long-term disease control, not by pCR alone.