Pathologic Response, Immune Learning, and the Nonlinear Nature of Treatment Duration

Dear Author The incorporation of immune checkpoint inhibitors into neoadjuvant and adjuvant treatment strategies for early-stage solid tumors has prompted a critical reassessment of how treatment response should be interpreted and translated into long-term therapeutic decisions. Data from contemporary neoadjuvant immunotherapy trials across multiple tumor types have revealed a recurring and biologically intriguing pattern: patients achieving deep pathologic responses may experience durable clinical benefit despite limited or even absent adjuvant immune checkpoint inhibition, whereas patients with residual disease appear to derive disproportionate benefit from continued postoperative therapy. This divergence—observed in KEYNOTE-522 in early-stage triple-negative breast cancer, CheckMate 816 in resectable non–small-cell lung cancer, response-adapted melanoma studies such as PRADO and NADINA, and short-course neoadjuvant immunotherapy in mismatch repair–deficient colon cancer (NICHE-2)—suggests that mechanisms beyond treatment intensity alone are operative and challenges the implicit assumption that immunotherapy benefit follows a linear “dose–duration–benefit” paradigm (1–6). The immune system is inherently adaptive; however, immune learning does not progress in a strictly linear or indefinitely amplifiable manner. Once critical thresholds of antigen recognition, effector T-cell activation, and immune memory formation are achieved, additional immune stimulation may yield diminishing marginal returns. This biologic ceiling effect provides a plausible explanation for the limited incremental benefit of continued adjuvant immunotherapy in patients who have already achieved a pathologic complete response (pCR). In such cases, treatment continuation may primarily serve to preserve an already established immune equilibrium rather than generate further clinically meaningful immune education (7). Importantly, the efficiency and durability of this immune learning process likely vary across tumor types and individual patients, influenced by baseline tumor immunogenicity, tumor mutational burden, and immune contexture, including PD-L1 expression. In contrast, patients with residual disease but without clear evidence of immune escape may represent a distinct and potentially modifiable biologic state. In these individuals, neoadjuvant therapy may have successfully primed antitumor immunity without surpassing the threshold required for complete tumor eradication. Following surgical debulking, the persistence of minimal residual disease and continued antigen exposure may allow adjuvant immunotherapy to stabilize this partially trained immune response and suppress micrometastatic progression. Accordingly, the observed benefit in patients with residual disease may reflect not greater therapeutic potency, but rather the continued presence of a biologically relevant immune target (8). These considerations raise important questions regarding the adequacy of pCR as a standalone biomarker for predicting immunotherapy benefit. While surgical pathology effectively captures local tumor eradication, the principal contribution of immunotherapy likely lies in the establishment of durable, systemic immune surveillance. Consequently, pCR—although a robust surrogate for chemotherapy sensitivity—may incompletely capture the long-term clinical impact of immune-based therapies (9). The integration of additional biomarkers reflecting systemic immune activation, such as circulating tumor DNA clearance or markers of immune memory, may therefore be necessary to more accurately define the threshold of effective immune learning. In summary, although the efficacy of immunotherapy in early-stage disease is well established, increasing attention should be directed toward defining treatment duration according to biological necessity rather than maximal exposure, as schematically illustrated in Figure 1. Recognizing the nonlinear nature of immune learning supports the development of more refined treatment strategies that integrate pathologic response, residual disease biology, and markers of immune memory. Future clinical trials should be explicitly designed to test response-adapted and biomarker-driven de-escalation strategies, including randomized studies in which patients achieving pCR or deep molecular response are assigned to abbreviated or omitted adjuvant immunotherapy. Such approaches may represent a critical step toward more rational, personalized, and toxicity-conscious treatment paradigms (10).