Pancreatic cancer is a highly aggressive malignant tumor of the digestive tract with a five-year survival rate of only 10%, which has increased by only 5% over the past 20 years (1). Surgery and postoperative adjuvant chemotherapy (ACT) have always been the standard treatments for pancreatic cancer, and past research mainly focused on ACT (2–4). In order to improve the survival of patients with pancreatic cancer, it is necessary to conduct multimodal research. Neoadjuvant chemotherapy (NACT) with or without radiotherapy is increasingly used in patients with pancreatic cancer. This may be related to the fact that patients who plan to undergo surgery often have local unresectable lesions or micro-metastases that cannot be predicted by imaging during surgical exploration. In the end, only 20% of patients underwent surgical resection. The second reason may be that less than 60% of patients can tolerate systemic therapy after surgery (5), and about 50% of patients who do not receive ACT will relapse within half a year (2). Neoadjuvant therapy can increase the opportunity for patients with pancreatic cancer to receive systemic therapy, increase the R0 resection rate, and improve the overall survival (OS) of patients (6, 7). A Meta analysis that included six randomized controlled trials showed that neoadjuvant therapy significantly improved OS (HR= 0.73, 95%CI: 0.61-0.86) in patients with pancreatic cancer (8). However, it is unclear which neoadjuvant therapy modality is optimal.

The Italian PACT-5 study found that neoadjuvant chemotherapy can reduce the 1-year event-free survival (EFS) rate of resectable pancreatic cancer (RPC) compared with primary surgery (9). The Prep-02/JSAP05 study also found that the combination of gemcitabine and S1 neoadjuvant therapy improved survival by about 10 months compared to direct postoperative S1 adjuvant chemotherapy (10). Two randomized controlled studies have shown that neoadjuvant chemoradiotherapy (NARCT) can improve the OS of patients with pancreatic cancer compared with direct surgery (11–13). However, the results of the phase II A021501 randomized controlled trial showed (14) that compared with 8 cycles of neoadjuvant mF0LFIRINOX chemotherapy regimen, 7 cycles of neoadjuvant mF0LFIRINOX chemotherapy plus SBRT in patients with borderline resectable pancreatic cancer (BRPC) had shorter median EFS (10.2 months VS 15.0 months) and mOS (17.1 months VS 29.8months), and R0 resection rate (33% VS 57%) is lower. This finding raises the question of whether additional radiation therapy may reduce the effectiveness of ACT.

Currently, there are few studies comparing the impact of different neoadjuvant therapy modalities on the survival of patients with pancreatic cancer. Therefore, in this retrospective analysis based on the Surveillance, Epidemiology, and End Results (SEER) database, this study evaluated the difference between NARCT and NACT in improving survival in patients with pancreatic cancer who underwent surgical treatment. Furthermore, we used propensity score matching (PSM) analysis to reduce the impact of confounding factors in non-randomized controlled studies. At the same time, we performed subgroup analysis of patients to determine the most appropriate neoadjuvant therapy mode for each subgroup.

Pancreatic cancer is a malignant tumor with high mortality. It is estimated that by 2030, it will become the second leading cause of cancer death in the United States (15). Due to the fact that most patients with pancreatic cancer have lost the opportunity for surgery at the time of diagnosis, and the completion rate of postoperative chemotherapy is also low, neoadjuvant therapy has become an important means to improve patient survival of pancreatic cancer.

Our study found that before PSM, patients with pancreatic cancer who received NACT had a trend of OS and CSS benefit compared with NARCT, while after PSM, patients with non-metastatic pancreatic cancer who received NACT had significant OS and CSS benefit. Subgroup analysis showed that compared with NARCT, OS and CSS of pancreatic cancer patients receiving NACT had significant benefits in male, white, pancreatic head cancer, stage II, T3, tumor size ≤ 2cm, and non-ACT subgroups. At the same time, in other subgroups, patients receiving NACT have a tendency to have better OS and CSS than pancreatic cancer patients receiving NARCT. Multivariate COX regression analysis confirmed that NACT was an independent prognostic risk factor for OS and CSS in patients with non-metastatic pancreatic cancer. Our findings are consistent with the A021501 study (14) and support NACT as the preferred neoadjuvant therapy strategy for patients with non-metastatic pancreatic cancer.

An increasing number of studies support the application of neoadjuvant therapy, which can convert patients with originally inoperable non-metastatic pancreatic cancer into resectable patients, thereby providing the possibility of curative surgery and improved survival. At the same time, neoadjuvant therapy can also screen out those patients who can really benefit from surgery and avoid ineffective surgery for patients whose disease progresses during neoadjuvant therapy. This may be due to the fact that the tumors in these patients are too aggressive and surgery cannot bring benefits (16–19). Although radiotherapy can further improve the surgical resection rate of patients on the basis of chemotherapy (18, 20, 21), whether this can improve patient survival is controversial. NARCT is associated with a higher surgical resection rate for pancreatic cancer (21). More than half of inoperable locally advanced pancreatic cancer (LAPC) can be converted to surgical resection. Compared with patients without surgery, patients after surgical resection have a significant survival benefit (17) (15.3 months VS 8.5 months, P<0.001). The single-arm LAP1 study (22) showed that compared with historical controls, NARCT improved the survival rate of patients with LAPC. 18% (7/39) of patients were converted to operable and all underwent R0 resection. The median OS of patients in the resection group was better than Unresected group (24.0 months VS 15 months, P= 0.03).

The phase III CONKO-007 study (23) found that compared to NACT, although NARCT can improve the negative circumferential margin rate and pathological complete response (pCR) rate in patients with LAPC, the primary endpoint of both groups was R0 There was no significant difference in resection rates, and this resectability effect was not associated with a progression-free survival (PFS) or OS benefit. Another phase III LAP07 study (24) found that although NARCT could improve the local control rate of LAPC compared with induction chemotherapy, this did not translate into an OS benefit (15.2 months VS 16.5 months, P= 0.83). The PREOPANC long-term follow-up study (12) found that NARCT can significantly improve the 5-year OS rate of patients with RPC, but the results of this study are also somewhat controversial. Firstly, for patients with RPC, there is no significant difference in OS between NARCT and primary surgery (HR= 0.79, 95% CI: 0.54-1.16, P= 0.23). Secondly, NARCT only prolonged mOS by 1.4 months, and at baseline, more patients in the NARCT group had a WHO performance status score of 0 (58% VS 39%), fewer patients were over 65 years old (54% VS 65%), and fewer patients had tumors in the pancreatic head (83% VS 92%). Additionally, a smaller proportion of patients had CA19-9 levels greater than 500 U/ml (29% vs. 35%). These factors may have affected the comparability of the two groups of patients, making interpretation of the results more difficult. The PREOPANC-2 study (25) found that FOLFIRINOX-based NACT and gemcitabine-based NARCT were similar in terms of surgical resection rate and OS for RPC and BRPC, indicating that radiotherapy may make up for the shortcomings of gemcitabine because FOLFIRINOX is more effective than gemcitabine alone (26, 27). In the phase II randomized controlled A021501 study (14), both NARCT and NACT were based on the same mFOLFIRINOX regimen, excluding the impact of differences in the effectiveness of different chemotherapy regimens. The results show that additional radiotherapy is a factor for the poor prognosis of pancreatic cancer, and this is not due to the increased toxicity caused by radiotherapy. The possible explanation for this result is that the mFOLFIRINOX treatment delay rate and dose reduction rate were higher in the chemoradiotherapy group, which may lead to a decrease or disappearance of treatment efficacy for systemic diseases.

Improvements in radiotherapy strategies, such as SBRT, may be the focus of neoadjuvant therapy of pancreatic cancer in the future. The advantages of SBRT are short treatment time, reduced risk of chemotherapy interruption, and higher biologically effective dose (BED) that can increase local tumor control rate and translate into survival benefit (14, 20, 28). Studies have shown that BED greater than 70Gy is associated with better survival benefit in pancreatic cancer (29, 30). In the A021501 study (14), the maximum BED (55Gy) of SBRT was far lower than the ablation dose required for SBRT, which may be the reason for the poor efficacy of the SBRT group. Improved radiotherapy technology, such as Isotoxic High-Dose Stereotactic Body Radiotherapy (iHD-SBRT), can individually adjust the radiotherapy dose and fractionation times to maximize the killing of tumor cells and minimize damage to surrounding normal tissues (28, 31). In addition, combined radiotherapy and immunotherapy may be a potential therapeutic strategy, although the effects of immunotherapy alone are limited in pancreatic cancer due to the characteristics of low tumor mutation load and immunosuppressive microenvironment (32, 33). However, in a phase II study (34) involving 29 patients, the efficacy of tislelizumab combined with AG-based concurrent chemoradiotherapy as neoadjuvant therapy for BRPC and LAPC was evaluated. The objective response rate was 60%, the R0 resection rate was 90% (9/10), and the 1-year OS rate was 72%. This suggests that radiotherapy can synergize with immunotherapy to enhance the immune response of tumors.

Research results based on the SEER database show that NACT may be a better treatment option than NARCT for patients with non-metastatic pancreatic cancer. However, this finding should be treated with caution because the study was based on the TNM staging system rather than clinical staging based on surgical resectability. This difference may bias the study results. For example, those who receive NARCT may include more cases with localized disease that is difficult to remove. therefore, in order to develop individualized radiotherapy regimens, more prospective randomized controlled studies are needed to compare and select different neoadjuvant therapy strategies for pancreatic cancer.

Publicly available datasets were analyzed in this study. This data can be found here: https://seer.cancer.gov/data/access.html, SEER Research Plus Dataset, 17 Registries (2000–2019).

JY: Conceptualization, Methodology, Software, Validation, Visualization, Writing – original draft, Writing – review & editing. XQ: Conceptualization, Data curation, Software, Validation, Visualization, Writing – review & editing. FJ: Methodology, Supervision, Validation, Writing – review & editing. H-MQ: Data curation, Formal analysis, Investigation, Writing – review & editing. JZ: Data curation, Formal analysis, Investigation, Writing – review & editing. J-RZ: Conceptualization, Data curation, Writing – review & editing. L-JY: Conceptualization, Visualization, Writing – review & editing. A-JZ: Investigation, Project administration, Supervision, Writing – review & editing. PN: Software, Supervision, Validation, Visualization, Writing – review & editing.