Pancreatic cancer (PC) is a highly aggressive tumor. Its five-year survival rate is 5%–10%, and life expectancy at diagnosis is less than 5 months (1). PC is the fourth leading cause of cancer death in the United States and the sixth in China (2, 3). Since PC occurs deep in the abdomen behind the stomach and in front of the spine, it does not cause obvious symptoms in its early stages. About 50% of patients develop metastases at initial diagnosis, which is a major factor in poor outcomes (4). Among all patients receiving first-line chemotherapy for PC, 57% went on to receive second-line therapy and 22% received third-line therapy (5).

Systemic therapy for locally advanced or metastatic disease has been documented in the National Comprehensive Cancer Network guidelines (6). FOLFIRINOX (category 1) and AG (category 1) are listed as the preferred recommended first-line chemotherapy treatments for patients in good physical condition, while gemcitabine monotherapy is recommended for patients in poor physical condition with metastatic PC (mPC) (7, 8). Almost all PC patients progress within a few months during or after first-line chemotherapy (9). Fluoropyrimidine-based regimen is the recommended subsequent treatment option for patients with a good performance status and those previously treated with gemcitabine-based therapy. Gemcitabine-based regimen is advised for patients with a good performance status and those previously treated with fluoropyrimidine-based therapy. Gemcitabine (category 1), capecitabine, and 5-fluoropyrimidine are suggested for patients with a poor performance status (10, 11). Pembrolizumab is used in an advanced disease setting as the first-line and subsequent treatment for PC patients with high microsatellite instability and mismatch repair-deficiency (12). Larotrectinib or entrectinib can be considered for NTRK gene fusion-positive diseases (13, 14).

Unfortunately, most patients face the challenges of tumor progression, chemotherapy resistance, and toxic effects after receiving second-line chemotherapy. Chemotherapy remains the standard of care for advanced disease. Research into novel therapies is ongoing and includes immunotherapy, targeted therapy, vaccines, and oncolytic viruses. Although most PC patients have gene mutations, there are few approved targeted therapies. New antitumor drugs for various targets are currently being developed and tested (15). PC is considered to be a ‘cold tumor’ in immunotherapy due to its typical bone marrow cell infiltration, lack of CD8+ T cells, and low activation markers. Except for 1% of patients with high microsatellite instability, PC is almost completely unsuitable for immunotherapy (16). According to the national guidelines for diagnosis and treatment of PC in China in 2022, continuing chemotherapy for PC patients who failed to respond to second-line chemotherapy is controversial, and there are no clear chemotherapy regimens to recommend (17). However, chemotherapy is still the most common choice for the third-line treatment in PC patients. There are only a handful of third-line chemotherapy drugs available, and many doctors choose to implement chemotherapy re-challenge programs for these patients (18). The efficacy and safety of various third-line treatments in PC patients are still awaiting confirmation, and clinical predictors for third-line treatment option selection are still lacking.

A considerable number of patients still have sufficient physical strength to receive antitumor therapy when the disease progresses to the third-line stage. The present study aimed to compare the efficacy and safety of different third-line therapies for mPC. The efficacy, safety, and relevance of various combinations of third-line antitumor therapies, including chemotherapy, targeted therapy, and immunotherapy, were explored in order to investigate the status of third-line therapy in mPC.

In this study, the third-line antitumor treatment was demonstrated to benefit patients and prolong their survival time compared to palliative care. Baseline characteristics were analyzed in all patients to identify efficacy predictors. Results showed that female patients, those with ECOG PS 0–1, and patients with family history of cancer were independent prognostic factors for longer OS in a group of mPC patients who received the third-line treatment. In particular, patients with a normal body mass index and family history of cancer were independent prognostic factors for longer OS in a group of mPC patients who received chemotherapy-based treatment. This indicates that not all patients are suitable for third-line antitumor therapy and some screening is still needed. Many retrospective studies have concluded that ECOG PS is an independent prognostic factor associated with treatment efficacy. Most notably, a family history of cancer was an independent factor for longer survival time among different treatment regimes in our study. Existing research studies have reported that family history of BRCA-related tumors may correlate with the response to chemotherapy and OS in PC, which is similar to the results of our study (19). This may be related to genetic differences and lifestyle changes. Patients with a family history of tumor disease may have some genetic mutations and are more likely to have malignant changes when affected by the external environment compared to those without a family history (20). In addition, it has been suggested that young patients who are aware of their family history may adopt healthy behaviors, such as opportunistic screening, and/or make healthy lifestyle changes, thereby improving their prognosis (21). Surprisingly, female patients were more likely to benefit from third-line treatments for mPC. Patient characteristics were not balanced between each group, including ECOG PS, first-line treatment, and second-line treatment. In order to determine the cause of this imbalance at baseline, a review of the case data revealed that patients with ECOG PS of ≥ 2 had a poor physical performance and were more inclined to choose chemotherapy-free regimens before the third-line treatment, while patients with a better physical performance were more suitable for chemotherapy. Patients receiving palliative treatment with ECOG PS of ≥ 2 at the beginning of the second-line treatment only received fluorouracil single-agent in the second-line treatment. A significant proportion of patients (24, 33.3%) received gemcitabine-based regimen as the first-line treatment and fluorouracil-based regimen as the second-line treatment.

Since our research data and previous studies have shown that third-line antitumor therapy can bring survival benefits to patients, it was necessary to determine whether chemotherapy combined with other treatments can further improve treatment efficacy. With the recent development of novel therapies, such as immunotherapy and targeted therapy, some clinicians have chosen to combine these therapies with chemotherapy or to directly use chemotherapy-free therapies when selecting third-line treatments. This is the first real-world study to compare the efficacy and safety of various third-line treatments for advanced PC. Although there was no statistical difference in P value, Kaplan-Meier analysis showed that the chemotherapy alone (mOS, 6.9 months; 95% CI, 0.5–13.9 months) and chemotherapy combined with targeted therapy or immunotherapy (mOS, 5.9 months; 95% CI, 1.6–10.2 months) groups had a better OS compared to the chemotherapy-free group (mOS, 3.3 months; 95% CI, 0.2–5.0 months). This may be due to the insufficient sample size in the study. This investigation demonstrated for the first time that the combined targeting/immunotherapy based on chemotherapy cannot improve third-line mOS (6.9 months vs. 5.9 months, P = 0.588) compared to chemotherapy alone. This notion has previously been introduced in other studies investigating the first-line treatment. Previous research has revealed that PC promotes an immunosuppressive microenvironment through formation of dense stromal desmoplasia, and concurrent administration of gemcitabine plus nab-paclitaxel was poised to improve immunotherapy drug access to tumor cells via structural disruption/remodeling of the PC tumor microenvironment (22). Negative results in the CCTG PA.7 trial indicated that it was not sufficient to increase immunotherapy efficacy in the overall patient population. The CCTG PA.7 trial demonstrated no survival benefits from adding durvalumab and tremelimumab to gemcitabine and nab-paclitaxel as the first-line therapy in an unselected population of patients with PC (mOS, 9.8 months vs. 8.8 months, P = 0.72). Moreover, the combination of epidermal growth factor receptor inhibitors and chemotherapy failed to achieve the preclinical model estimates. In preclinical models of PC, ibrutinib combined with gemcitabine increased the levels of effector CD8+ T cells and mast cell inhibition, decreased angiogenesis, and reduced desmoplasia in multiple mouse models, resulting in reduced tumor size and increased survival rate (10, 23). In the phase III RESOLVE study, the combination of ibrutinib plus nab-paclitaxel and gemcitabine did not improve survival in patients without any previous cytotoxic chemotherapy for primary PC compared to chemotherapy alone (mOS, 10.8 months vs. 9.7 months, P = 0.323) (24). The phase II ACCEPT study demonstrated an mOS of 7.3 months in the afatinib combined with gemcitabine group compared to 7.4 months in the gemcitabine group (P = 0.80) (25). This was likely because the addition of targeted therapy to chemotherapy may have mitigated the ability to deliver the complete chemotherapy regimen and the tumor received a lower cumulative dose of all agents compared to patients in the chemotherapy alone group. In the phase II ACCEPT study, AEs were more frequent in the combination therapy group, which was consistent with the present study results showing a significantly higher toxicity burden in the combination group. In our study, we found that chemotherapy combined with targeted therapy was more prone to result in grade 3/4 leukopenia compared to chemotherapy alone. Family history of cancer was an independent mOS predictor for PC patients who received the chemotherapy alone treatment. In addition, the present research also showed that there was no statistical difference in mOS between single- and multi-agent chemotherapy (5.1 months vs. 7.9 months, P = 0.967) groups, which was consistent with the randomized phase III NAPOLI-1 trial results (26). The response in the NAPOLI-1 trial was less prominent in patients treated with nanoliposomal irinotecan with fluorouracil/leucovorin compared to fluorouracil/leucovorin as the third-line treatment (mOS, 5.4 months vs. 4.3 months, P = 0.178). Therefore, the appropriate use of low-intensity regimen in third-line antitumor therapy can also prolong survival. To our knowledge, no clinical studies have been carried out on third-line antitumor therapy to support this conclusion, which still needs to be confirmed by studies with a larger sample size.

The optimal therapy sequencing remains unknown and is largely defined by physician preference in practice. In our results, the order of gemcitabine- and fluorouracil-based regimens as the first- and second-line treatment pairs was not an independent predictor of third-line treatment OS, which is consistent with what has been reported so far. In the study by Jung et al., first-line palliative chemotherapy regimens and the order of subsequent chemotherapy regimens were not associated with survival outcomes in third-line treatment patients (27). Advances in systemic chemotherapy over the past decade have been limited, and the mechanism of chemotherapy resistance is still unclear. More trials will be carried out to explore the mechanism of chemotherapy resistance and provide credible data to identify the prognostic factors for chemotherapy rechallenge. According to our results, the order of chemotherapy drug treatment in the process of PC management can be selected according to the patient’s physical condition in the first- and second-line treatments. Patients can also receive personalized treatment.

The current study had limitations, as the research was performed at a single institution using retrospective analysis, and the number of patients included in each group was not balanced. First, the retrospective nature of the analysis may result in a potential selection bias such as the increase of survival would be due to the ECOG, and lack of medical records including molecular pathological information and immunohistochemistry may affect independent prognostic factor results. Second, therapeutic drug subgroups could not be analyzed further due to insufficient sample size. Third, the impact of different treatments after disease progression was not estimated in the study, which may affect the OS analysis. Nevertheless, our results are encouraging and support continued use and study of chemotherapy rechallenge to treat patients who fail to respond to the second-line treatment, as well as further optimization of selection of patients who are most likely to benefit.

Treatment with chemotherapy-based therapy as the third-line treatment combined with targeted therapy or immunotherapy failed to improve survival benefits and demonstrated higher safety risks. In particular, blood test results should be monitored more closely in patients undergoing multi-agent chemotherapy combined with targeted therapy or immunotherapy to prevent the occurrence of grade 3/4 leukopenia. The treatment order of gemcitabine- and fluorouracil-based regimens in the first- and second-line therapy does not affect third-line OS. Thus, in the first- and second-line therapy, the treatment order of chemotherapy drugs can be selected according to the patient’s physical condition. Therefore, patients who tolerate it should be treated mainly with chemotherapy. For patients who cannot tolerate chemotherapy, the use of targeted therapy or immunotherapy represents a survival benefit over supportive therapy. The targeted and immunotherapy drugs included in the present study were mixed, and more clinical trials are needed to explore the feasibility of chemotherapy-free treatment in advanced PC.

The original contributions presented in the study are included in the article/ Supplementary Material . Further inquiries can be directed to the corresponding authors.

The studies involving humans were approved by the Ethics Committee of Zhejiang Provincial People’s Hospital. The studies were conducted in accordance with the local legislation and institutional requirements. The participants provided their written informed consent to participate in this study. Written informed consent was obtained from the individual(s) for the publication of any potentially identifiable images or data included in this article.

H-RL and P-FZ contributed equally to this work. Z-LC, and LY revised the manuscript. All authors wrote and revised the manuscript and issued the final approval for the version to be submitted.