Neoadjuvant immunochemotherapy and pathological response in locally advanced cervical cancer: a real-world bicenter cohort study

Neoadjuvant chemotherapy (NACT) followed by radical surgery remains an option for selected patients with locally advanced cervical cancer (LACC) in settings where access to definitive chemoradiotherapy is constrained. Whether adding an immune checkpoint inhibitor (ICI) -neoadjuvant immunochemotherapy (NAIC) - improves pathologic outcomes without increasing surgical risk is unknown outside clinical trials. We retrospectively analyzed consecutive patients with FIGO 2018 stage IB3, IIA2, or IIIC1 (r/p) cervical cancer who received neoadjuvant therapy followed by radical hysterectomy at two tertiary centers. Of 490 patients screened, 246 were eligible (NACT, n = 151; NAIC, n = 95). Confounding was addressed by 1:1 propensity score matching (PSM) and stabilized inverse probability of treatment weighting (IPTW); outcome models in the matched cohort were further adjusted for covariates that remained imbalanced after matching (doubly robust estimation). The primary endpoint was pathologic complete response (pCR), major pathologic response (MPR), and optimal pathologic response (OPR = pCR+ MPR). Other endpoints include adverse pathologic features, perioperative outcomes, toxicity, and survival. Because ICIs became available only in 2021, calendar-time confounding was interrogated by contemporaneous-period, temporal-falsification, restriction, and quantitative bias (E-value) analyses. pCR occurred in 22 of 95 patients (23.2%) receiving NAIC versus 10 of 151 (6.6%) receiving NACT (OR 4.25, 95% CI 1.91–9.44; P < 0.001); the association persisted after PSM (23.5% vs. 8.8%; OR 3.18, 95% CI 1.19–8.48; P = 0.021) and IPTW (20.6% vs. 6.7%; OR 3.60, 95% CI 1.44–8.99; P = 0.007). OPR was higher with NAIC in all three models (crude 47.4% vs. 13.9%, OR 5.57, 95% CI 3.02–10.27; PSM OR 4.61, 95% CI 2.11–10.07; IPTW OR 3.00, 95% CI 1.33–6.78; all P ≤ 0.009). MPR favored NAIC in the crude (OR 4.07) and matched (OR 3.88) cohorts but was attenuated after IPTW (OR 1.86, 95% CI 0.64–5.43; P = 0.257). Nodal metastasis in the surgical specimen was less frequent with NAIC after adjustment (PSM 14.7% vs. 38.2%, OR 0.28, 95% CI 0.12–0.66; IPTW OR 0.33, 95% CI 0.16–0.68), whereas the use of adjuvant radiotherapy did not differ. Operating time was longer with NAIC in the crude cohort (median 220 vs. 190 min; P < 0.001) but not after matching ( P = 0.173); estimated blood loss, transfusion, any complication (Clavien–Dindo ≥ I, 48.4% vs. 47.7%; P = 0.91), and severe complications (6.3% vs. 4.0%; P = 0.41) were comparable. Any-grade treatment-related adverse events were similar (95.8% vs. 94.0%; P = 0.55), with no significant between-group difference in any individual hematologic or non-hematologic toxicity; thrombocytopenia was numerically less frequent with NAIC (16.8% vs. 27.8%; P = 0.05). Immune-related adverse events were confined to the NAIC group and were uncommon (hypothyroidism 13.7%, pneumonitis 5.3%, enteritis 4.2%, adrenal insufficiency 3.2%). Survival data were immature (28 progression events, 16 deaths overall): 24-month progression-free survival was 93.4% with NAIC versus 91.2% with NACT, and no death occurred in the NAIC group; all survival comparisons were underpowered and are reported descriptively. For the primary PSM estimate of pCR, the E-value was 2.96 (1.41 for the confidence limit closest to the null). In this real-world, two-center cohort, NAIC was associated with substantially higher pathologic response rates than conventional NACT among patients with LACC who proceeded to radical surgery, with no detectable increase in perioperative morbidity. Because treatment assignment was almost perfectly confounded with calendar time, these findings represent associations only; they are hypothesis-generating and require confirmation in prospective randomized trials.