Pancreatic cancer remains one of the most formidable challenges in oncology. It is the seventh leading cause of cancer-related death worldwide, and in the United States it is already the third leading cause of cancer mortality, with projections suggesting it will climb to second place by 2040. Most patients are diagnosed at an advanced, unresectable stage, when treatment is largely palliative and even the best chemotherapy regimens deliver a median overall survival of less than a year. Immune checkpoint inhibitors, which have transformed the treatment of melanoma, lung cancer and many other malignancies, have so far produced disappointing results in pancreatic cancer. Against this grim backdrop, a new post hoc analysis published in Immunity, Inflammation and Disease has examined a question with major implications for how patients are stratified and treated: does the presence of liver metastases shape how long patients live, and does it matter differently depending on whether they receive chemotherapy or immunotherapy?
The liver occupies a unique position in systemic immunity. It is continuously exposed to antigens and toxins arriving from the gut and must tolerate them to avoid destructive autoimmunity, a tolerogenic property that tumors exploit. In pancreatic cancer, liver metastases are present in nearly all patients with advanced disease, and evidence from several cancer types suggests that liver metastases blunt the effectiveness of immune checkpoint inhibitors. Preclinical work points to a plausible mechanism: an increased fraction of FasL-positive, monocyte-derived macrophages in the liver promotes apoptosis of Fas-positive CD8-positive T cells, depleting the cytotoxic T cells that immunotherapy depends on from systemic circulation. Tumor cells, immune cells and inflammatory cytokines in the peripheral blood and portal veins add further layers of systemic immunosuppression. The researchers behind the new study hypothesized that patients with liver metastases would therefore derive less benefit from checkpoint inhibitor-based regimens.
To test this hypothesis, the investigators pooled data from three Danish clinical trials of immune checkpoint inhibitor-based treatment in refractory advanced pancreatic cancer: Triple-R, CheckPAC and Influence. All three combined a single 15 Gy fraction of stereotactic body radiotherapy aimed at a feasible metastasis or the primary tumor with checkpoint blockade, most commonly nivolumab and ipilimumab; the Triple-R and Influence studies also incorporated tocilizumab, an interleukin-6 receptor antibody, and Influence added a single dose of seasonal influenza vaccine as an immune adjuvant. As a chemotherapy comparator, the team used the PACTO study, a randomized phase 2 trial in which 147 treatment-naive patients received gemcitabine and nab-paclitaxel with or without tocilizumab. In total, 276 patients with advanced pancreatic cancer were included in the analysis, with liver metastasis status determined at baseline.
The composition of the two treatment groups reflected the natural history of the disease. In the chemotherapy cohort, 107 patients had liver metastases and 40 did not, while in the checkpoint inhibitor cohort the imbalance was even more pronounced, with 111 patients carrying liver metastases and only 18 free of them. Baseline characteristics were broadly comparable between patients with and without liver metastases in both groups, including sex, age, performance status and, in the immunotherapy group, the number of prior lines of treatment. What distinguished patients with liver metastases was a strikingly more inflamed and immunosuppressed systemic profile: C-reactive protein, the tumor marker CA 19-9, interleukin-6 and interleukin-8 were all significantly elevated in the liver metastasis groups of both cohorts, and the neutrophil-to-lymphocyte ratio was higher in the chemotherapy group. Absolute lymphocyte counts and YKL-40, also known as chitinase-3-like protein 1, showed no significant differences.
The survival findings were unambiguous on the chemotherapy side. Patients with liver metastases who received first-line chemotherapy had a median overall survival of 7.0 months, compared with 12.0 months for those without liver metastases, a difference that was highly statistically significant with a log-rank p value of 0.002. The unadjusted hazard ratio for death was 1.77, meaning liver metastases were associated with a roughly 77 percent higher risk of dying during the observation period. Progression-free survival told the same story: 5.0 months versus 7.5 months, with a hazard ratio of 1.58 and a p value of 0.017. Interestingly, the proportion of patients achieving a partial response was numerically higher among those with liver metastases, 39 percent versus 25 percent, but this radiographic benefit did not translate into longer survival, underscoring that response rates alone can be misleading in this disease.
The immunotherapy results defied expectations. In the pooled checkpoint inhibitor group, median overall survival was 3.7 months for patients with liver metastases and 5.0 months for those without, a difference that did not reach statistical significance, with a hazard ratio of 1.38 and a p value of 0.21. Progression-free survival was essentially identical between the two groups, at 1.7 versus 1.8 months. Best overall response rates were also comparable, with a small number of complete responses occurring in both groups. In multivariate analyses adjusting for age, sex, performance status and a panel of circulating biomarkers, liver metastases did not emerge as an independent predictor of survival in the immunotherapy cohort, whereas in the chemotherapy cohort the negative prognostic signal persisted. Notably, tocilizumab treatment was associated with a significantly reduced hazard of death in the chemotherapy group, and high levels of YKL-40 and interleukin-6 were associated with increased mortality, consistent with earlier work linking these inflammatory markers to poor outcomes.
These findings challenge the widely held assumption that liver metastases universally impair checkpoint inhibitor efficacy through immune-suppressive mechanisms. The expectation was grounded in solid evidence from other malignancies: in microsatellite-stable metastatic colorectal cancer, liver metastases are linked to shorter overall survival, shorter progression-free survival and lower disease control rates on immunotherapy, and in melanoma, renal cell carcinoma and lung cancer, their presence correlates with worse long-term survival on nivolumab. Yet a meta-analysis of seven studies in gastroesophageal cancer found no significant impact of liver metastases on checkpoint inhibitor outcomes, and the new pancreatic cancer data align more closely with that pattern. The authors suggest the effect of liver metastases on immunotherapy may be cancer-type dependent, and they caution that their own study had limited statistical power, particularly given the small number of immunotherapy patients without liver metastases.
Several important caveats temper the conclusions. All analyses were post hoc, since none of the original trials was designed to evaluate the impact of liver metastases. The checkpoint inhibitor regimens all included stereotactic body radiotherapy, and preclinical models suggest liver-directed radiotherapy can itself restore immunotherapy efficacy by reducing hepatic tumor burden, increasing T-cell infiltration and depleting immunosuppressive myeloid cells, so radiotherapy may have partially neutralized the liver’s immunosuppressive influence. The immunotherapy patients were also heavily pretreated, with roughly half having received at least two prior lines of chemotherapy, and all had refractory disease, raising the possibility that treatment line or disease stage influenced the results. A small subset of chemotherapy patients later crossed over to immunotherapy, though a sensitivity analysis excluding them left the survival findings intact. The analysis was also limited to ipilimumab and nivolumab, which may restrict generalizability to other checkpoint inhibitors.
Beyond prognosis, the study points toward therapeutic strategies aimed squarely at the liver metastasis microenvironment. Emerging evidence implicates immunosuppressive macrophage subtypes that reprogram the tumor microenvironment through efferocytosis, the engulfment of dying cells, and preclinical work suggests that blocking this process can restore tumor immunity and enhance checkpoint inhibitor activity. The same group has previously shown that an immune response against transforming growth factor beta, a master regulator of fibrosis and immunosuppression in pancreatic cancer, is associated with clinical benefit from checkpoint inhibitor regimens, and this insight has spawned an ongoing trial of a TGF-beta peptide vaccine combined with immunotherapy and radiotherapy specifically in patients with liver-metastatic pancreatic cancer. Pancreatic cancer’s recurrent genetic alterations in KRAS, TP53, CDKN2A and SMAD4 may also modulate TGF-beta signaling in ways that can be exploited therapeutically to remodel the tumor microenvironment.
The broader message is that metastatic pattern matters, and that personalized treatment decisions in pancreatic cancer should increasingly account for where the disease has spread. Promising results from personalized mRNA neoantigen vaccines combined with atezolizumab and modified FOLFIRINOX chemotherapy in the adjuvant setting, which improved recurrence-free survival and generated long-lived CD8-positive T cells, have led to a phase 2 randomized trial that may eventually extend such approaches to advanced disease. For now, the new analysis establishes liver metastases as a strong negative prognostic marker for chemotherapy-treated patients while leaving open, and scientifically tantalizing, the question of whether the immunosuppressive liver microenvironment truly limits immunotherapy in pancreatic cancer. Prospective studies that stratify patients by metastatic site, incorporate detailed immune profiling of the liver and circulation, and test microenvironment-targeted combinations will be needed to convert these observations into longer survival for the patients who need it most.
Subject of Research: The impact of liver metastases on survival outcomes with chemotherapy versus immune checkpoint inhibitors in advanced pancreatic cancer.
Article Title: Impact of Liver Metastases on Outcomes With Chemotherapy or Immune Checkpoint Inhibitors in Pancreatic Cancer
Article References: Kjær, L. K., Theile, S., Madsen, K., Johansen, J. S., Holmström, M. O., Andersen, M. H., & Chen, I. M. (2026). Impact of Liver Metastases on Outcomes With Chemotherapy or Immune Checkpoint Inhibitors in Pancreatic Cancer. Immunity, Inflammation and Disease, 14(10), Article e70545. https://doi.org/10.1002/iid3.70545
Image Credits: AI Generated
DOI: 10.1002/iid3.70545
Keywords: pancreatic cancer, liver metastases, immune checkpoint inhibitors, chemotherapy, overall survival, progression-free survival, tumor microenvironment, biomarkers, SBRT, IL-6, CA 19-9, immunosuppression
