Cryoablation with sintilimab and lenvatinib for intrahepatic cholangiocarcinoma: Does the addition of locoregional therapy augment survival outcomes?

Intrahepatic cholangiocarcinoma (iCCA) presents a significant and persistent challenge for oncologists. It is an aggressive malignancy with increasing incidence, and only about 30% of patients are candidates for curative intent surgery (1). For most patients, treatment is palliative despite notable incremental advances in systemic therapy. The current standard of care for advanced biliary tract cancers was established by the TOPAZ-1 trial, which combined durvalumab with gemcitabine and cisplatin chemotherapy. While this regimen represents progress, the magnitude of benefit is modest: median overall survival (OS) of 12.9 months, an objective response rate (ORR) of 26.7%, and fewer than 15% of patients alive at 3 years (2). Nevertheless, TOPAZ-1 demonstrated that iCCA is a disease amenable to immunomodulation, thereby paving the way for further strategies to augment anti-tumor immunity and improve patient survival.

One strategy of particular interest to surgical oncologists and interventional radiologists is the integration of locoregional therapies (LRTs) with systemic immunomodulatory regimens. LRTs not only offer cytoreduction but the potential to induce systemic immune responses through immunogenic cell death (3). The single arm phase II CASTLE-01 trial by Gu et al., which evaluated the anti-PD1 checkpoint inhibitor sintilimab with the multi-targeted tyrosine kinase inhibitor lenvatinib following cryoablation, represents an application of this approach. Twenty-eight patients with advanced or metastatic iCCA who had progressed after standard chemotherapy underwent cryoablation of an intrahepatic lesion followed two weeks later by sintilimab and lenvatinib (4). Non-ablated target lesions were assessed for response to treatment. The reported outcomes far exceed historical benchmarks: ORR 75.0% compared to 26.7% in TOPAZ-1; median progression-free survival (PFS) 16.8 versus 7.2 months; and median OS 25.4 versus 12.9 months. Of note, the median OS is approximately 6.2 months for second-line FOLFOX (5). The results of CASTLE-01 are encouraging for patients following first-line therapy.

For context, retrospective studies combining locoregional therapies with lenvatinib plus anti-programmed cell death protein 1 (PD1)/programmed death-ligand protein 1 (PDL1) systemic therapy have reported ORRs of 51–59% and median OS of 18.5–27.3 months (6-9). For example, a multicenter observational study of sintilimab plus lenvatinib including either transarterial chemoembolization or radiotherapy in some but not all patients (without cryoablation) reported an ORR of 46.3% and median time to progression of 6.6 months. In contrast, systemic approaches including both lenvatinib with anti-PD1/PDL1 therapy have demonstrated ORRs of 18–20% and median OS of 14.3–15.5 months (7,10). It is important to note however that inclusion/exclusion criteria in these retrospective analyses vary widely, complicating comparisons. Nevertheless, taken together these data suggest that the magnitude of response observed in CASTLE-01 may reflect a meaningful effect of cryoablation (11).

Cryoablation is thought to induce immunogenic cell death by the release of damage-associated molecular patterns (DAMPs) such as adenosine triphosphate (ATP), calreticulin, and High Mobility Group Box 1 (HMGB1), which promote dendritic cell recruitment and maturation (3). This process facilitates tumor antigen presentation and priming of tumor-specific T-cell responses, which may be further modulated by checkpoint inhibition. With tissue biopsies taken after 6 weeks of lenvatinib plus sintilimab (8 weeks post cryo-ablation), Gu et al. demonstrated activation of immune pathways in lesions previously lacking this activity. CD8 T cell effectors were consistently increased within tumor biopsies by both single-cell RNA seq and imaging mass cytometry both 2 weeks post-cryoablation (pre-lenvatinib and sintilimab) and 6 weeks following initiation of lenvatinib and sintilimab. T cell clonotypic analysis demonstrated shifts in transcriptional signatures within clonal populations over the course of treatment, and novel clonotypes were identified in T effector subsets within tumor biopsies after initiating lenvatinib plus sintilimab, some of which were clonally expanded. Dendritic cell activation increased as measured by costimulatory and human leukocyte antigen (HLA) gene expression following cryoablation, prior to administration of lenvatinib and sintilimab. The elegant study design with extensive tissue sampling and multiomics analyses demonstrates clear systemic immunomodulation following cryoablation preceding systemic therapy.

Despite these compelling data, the interpretation of CASTLE-01 remains limited by the absence of a control arm, precluding determination of the independent contribution of cryoablation towards clinical response. It should be noted that cryoablation is an older, more time-consuming technique that is not currently included in the National Comprehensive Cancer Network (NCCN) guidelines as a recommended locoregional modality for iCCA, which instead prioritize radiofrequency ablation, microwave ablation, and irreversible electroporation. While comparisons to historical controls are provocative, they may be inherently confounded. The combination of PD-1 inhibition and lenvatinib alone has demonstrated meaningful activity, and the observed benefit may reflect additive or synergistic effects—or simply the cumulative impact of multiple active therapies. For example, a phase II trial evaluating the anti-PD1 antibody toripalimab in combination with lenvatinib reported ORR of 32.3% and median OS of 20.3 months as a first-line therapy (6). Furthermore, it remains to be seen if the impact of anti-PD1/PDL1 is replicated when applied to patients who have progressed on the updated first-line regimen now incorporating anti-PD1/PDL1 together with GemCis. As with all single arm studies, patient selection introduces additional bias; eligibility requiring disease amenable to cryoablation and patients naïve to checkpoint inhibition could have enriched for a population with more favorable biology.

The broader therapeutic landscape further complicates interpretation. Combination strategies incorporating chemotherapy, immunotherapy, and targeted agents are achieving increasingly high response rates. For example, a phase II study of GemOx combined with toripalimab and lenvatinib reported an ORR of 80% and median OS of 22.5 months (12). In this context, the CASTLE-01 regimen may be viewed as a chemotherapy-sparing alternative achieving comparable outcomes through locoregional immune priming rather than systemic cytotoxicity, which would be hugely significant for patients if it meant an improved quality of life. Along these lines, it should be noted that prior studies combining regional therapy with tyrosine kinase inhibitors and checkpoint blockade have demonstrated substantial toxicity, including grade 5 immune-mediated events in up to 5.7% of patients (13).

Ultimately, CASTLE-01 should be viewed as hypothesis-generating rather than practice-defining. It reinforces an important and increasingly accepted principle: locoregional therapies may function not only as cytoreductive tools but as modulators of systemic immunity (14). Gu et al. are to be congratulated for an important milestone study which should be replicated in larger cohorts, inclusive of western populations and patients who have progressed following checkpoint inhibition before it can be broadly endorsed. In the authors’ opinion, progress in iCCA will come in the form of potent combinations as was utilized in the present study. Gu et al. initiated the study following the observation of partial responses in four patients treated with anti-PD1 and lenvatinib following progression with anti-PD1 monotherapy and cryoablation (15). These observations highlight the need for LRT combinations with more than single agent checkpoint inhibition but also underscore the importance of understanding how each component contributes to outcome so that future studies can be appropriately planned. The role of components (including LRTs) must be defined through controlled investigation, which should follow novel exploratory phase II studies documenting improvements in outcomes.

Supplementary

The article’s supplementary files as

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.