KRAS mutations, occurring in 20–30% of non-small cell lung cancer (NSCLC) cases in Western populations and approximately 13% in Chinese patients, define a distinct molecular subset historically associated with limited therapeutic options and poor prognosis (1). Unlike EGFR or ALK-driven tumors, KRAS-mutant NSCLCs have historically been less amenable to targeted therapies. While immunotherapy (with or without chemotherapy) is now standard in the first-line setting, and KRAS G12C inhibitors (e.g., sotorasib, adagrasib) are approved in the second-line and beyond, access to these agents remains limited in many regions, including China, and efficacy in non-G12C subtypes is modest (2, 3).

Combining ICIs with anti-angiogenic agents represents a rational strategy to remodel the immunosuppressive tumor microenvironment (TME). Preclinically, anti-angiogenic drugs can normalize aberrant vasculature, enhance immune cell infiltration, and potentiate ICI activity (4). Clinically, regimens such as atezolizumab plus bevacizumab and chemotherapy have shown benefit in KRAS-mutant NSCLC, with a median PFS improvement of approximately 2 months in exploratory subgroup analyses (5). Notably, KRAS-mutant tumors are often characterized by an immunosuppressive tumor microenvironment, which may render them particularly susceptible to combined PD-1/PD-L1 and VEGF pathway blockade (6, 7). This provides a strong biological rationale for evaluating such combinations specifically in this molecular subset. However, chemotherapy-free combinations, particularly using oral multi-target tyrosine kinase inhibitors (TKIs) with ICIs, are underexplored in this molecular context within real-world Chinese practice.

Sintilimab, a fully human anti-PD-1 antibody, and anlotinib, a multi-target TKI against VEGFR, FGFR, and PDGFR, have demonstrated synergistic activity in advanced NSCLC (8). While a phase II trial reported encouraging first-line activity of this combination in driver-negative NSCLC (9), data specifically focusing on its efficacy and safety in pretreated KRAS-mutant patients are lacking.

Therefore, this multicenter, retrospective case series aimed to evaluate the real-world performance of sintilimab plus anlotinib in Chinese patients with advanced KRAS-mutant NSCLC who had progressed on prior therapies. Additionally, we sought to explore potential clinical predictors of benefit, with a specific focus on the impact of prior anti-angiogenic therapy, to inform optimal treatment sequencing.

This multicenter retrospective study provides preliminary evidence that the combination of sintilimab and anlotinib may offer clinically relevant activity and an acceptable safety profile as a later-line therapy for patients with advanced KRAS-mutant NSCLC, warranting further prospective investigation. With a median PFS of 8.0 months and median OS of 12.4 months, this chemotherapy-free regimen represents a strategy worthy of future exploration, particularly in regions with limited access to KRAS G12C inhibitors. A key finding of our analysis is that patients without prior exposure to anti-angiogenic therapy had numerically longer median PFS (9.1 vs. 3.0 months) and OS (13.5 vs. 5.7 months) compared to those with prior exposure, suggesting a potential association between treatment history and clinical outcomes. This observation underscores the critical influence of prior therapy on treatment sequencing and clinical outcomes.

Our results compare favorably with existing later-line options for KRAS-mutant NSCLC. In the context of second-line therapy, the observed PFS of 8.0 months compares favorably with docetaxel (median PFS 3.0–4.2 months) and is within range of the PFS reported for sotorasib (5.6 months) and adagrasib (6.5 months) in KRAS G12C-mutant populations (10, 11). Moreover, the disease control rate of 92.6% is notable, particularly when considering that our cohort included both G12C (48.1%) and non-G12C (51.9%) subtypes (3). This suggests that the combination may provide clinical benefit across KRAS variants, addressing an important unmet need for non-G12C patients who are not eligible for G12C-specific inhibitors. The efficacy also aligns with that of the atezolizumab-bevacizumab-chemotherapy triplet (mPFS 8.1 months) from the IMpower150 cohort (12). The observed grade 3–4 toxicity rate of 25.9% in our cohort appears lower than the 60–70% reported in some prospective trials of chemo-immunotherapy combinations (13, 14); however, direct comparisons are limited by differences in data collection (retrospective vs. prospective) and potential under-ascertainment of adverse events in our study.

The association between prior anti-angiogenic therapy and inferior outcomes with the sintilimab-anlotinib regimen is logically consistent with expectations, as retreatment with agents targeting similar pathways may yield diminished benefit. We hypothesize that this may be driven by two interrelated mechanisms: first, prolonged anti-angiogenic pressure may induce a fibrotic, hypoxic tumor microenvironment characterized by impaired T-cell infiltration and function—a state often termed vascular “abnormalization” (4); second, tumor cells may undergo clonal selection under VEGF/VEGFR inhibition, leading to upregulation of alternative pro-angiogenic pathways (e.g., FGF, ANG-2) that diminish the efficacy of subsequent VEGFR inhibition (5). This biological rationale is consistent with subgroup analyses from the IMpower150 trial, wherein patients previously treated with bevacizumab derived less pronounced benefit from the atezolizumab-based combination (6).

When contextualized within the broader therapeutic landscape, our findings suggest that the sintilimab-anlotinib combination warrants further investigation. While its efficacy in our selected cohort appears numerically higher than historical data for anlotinib monotherapy (ALTER-0303: mPFS 4.8 months), direct cross-trial comparisons are limited by differences in patient selection, sample size, and study design (8). Regarding prior anti-angiogenic exposure, our findings differ from those of the ORIENT-31 study (9)—which evaluated sintilimab combinations in EGFR-mutant NSCLC after TKI failure—a distinct clinical and biological context. These differences highlight the need for prospective studies designed specifically for KRAS-mutant populations.

The safety profile observed was consistent with the known effects of each agent, and no new safety signals were identified. The most common grade 3–4 TRAEs were hypertension and hand-foot syndrome, which were readily managed with supportive care and dose modification. It is noteworthy that no grade ≥3 immune-related pneumonitis or hepatitis occurred in our cohort. While this contrasts with the higher rates (8–12%) sometimes observed with PD-1 inhibitor plus chemotherapy combinations (15), it is possible that the absence of chemotherapy, rather than a specific protective effect of VEGFR-TKI, contributed to this favorable safety profile.

Due to the modest sample size (n = 27) and limited number of events, multivariate analysis to adjust for potential confounders (e.g., age, ECOG performance status) was not feasible. Therefore, unmeasured confounding cannot be excluded, and the findings should be interpreted as hypothesis-generating rather than definitive. The lack of data on PD-L1 expression and tumor mutational burden also limits our ability to assess the contribution of these potential confounding factors. Additionally, the retrospective abstraction of safety data from medical records may have led to under-ascertainment of mild or asymptomatic adverse events. The inclusion criterion requiring completion of at least two treatment cycles may have introduced selection bias toward patients with better initial tolerance and longer intrinsic PFS, potentially overestimating the true treatment effect. These factors should be considered when interpreting the reported survival outcomes.

Future research should seek to prospectively validate “prior anti-angiogenic therapy” as a stratification factor in clinical trials. It will also be crucial to explore optimal therapeutic sequencing, such as prioritizing KRAS G12C inhibitors (e.g., fulzerasib) when available, followed by sintilimab-anlotinib upon progression. For patients with non-G12C mutations, investigating rational combinations such as VS-6766 plus defactinib (16) represents another promising direction.

In summary, this case series provides hypothesis-generating evidence that the combination of sintilimab and anlotinib may have clinical activity in pretreated, advanced KRAS-mutant NSCLC. The observation of numerically longer survival in patients without prior anti-angiogenic therapy highlights the potential importance of treatment sequencing and warrants further investigation in prospective studies. In the context of constrained access to KRAS G12C inhibitors in some regions, this regimen represents a strategy deserving of future exploration.