Pancreatic ductal adenocarcinoma (PDAC) is highly malignant with a dismal prognosis. Although surgery is the only potentially curative method, up to 80% of patients who undergo curative resection experience recurrence within two years, and the 5-year survival rate is 10-25% (1, 2). Lymph node metastasis and positive margins indicate a poor prognosis. Lymph node metastasis is associated with worse disease-free survival (DFS) and overall survival (OS) (3, 4). R1 resection is also associated with decreased OS (5). Adjuvant chemotherapy (CT) can improve long-term outcomes compared with observation. Based on the PRODIGE-24 and ESPAC-4 trials (6, 7), six months of mFOLFIRINOX or gemcitabine with capecitabine is the recommended adjuvant CT regimen. Gemcitabine monotherapy is an alternative option for patients who cannot tolerate combination therapy. In the Asian population, S-1 was shown to markedly prolong OS with lower toxicity than gemcitabine (8). However, after receiving surgical treatment and standard adjuvant CT, patients are still at high risk of local relapse (6). Therefore, adjuvant radiotherapy has been investigated for many years in the treatment of resectable pancreatic adenocarcinoma.

Due to the lack of high-quality clinical trials, the role of radiotherapy is still controversial. Adjuvant chemoradiation (CRT) alone did not show any OS benefit over observation due to the toxicity of radiotherapy and the delay of the administration of CT (9–11). Therefore, some researchers have investigated the role of adjuvant chemotherapy and chemoradiation (CT-CRT). However, for non-selected patients, no differences in DFS and OS were observed between adding CRT to adjuvant CT and administering CT alone according to several phase 2 and phase 3 trials (12, 13). Patients with certain pathological characteristics may benefit from additional CRT. Lymph node metastasis favors improved OS from CT-CRT when compared to CT or CRT alone (0 positive nodes: hazard ratio (HR) 0.96, P = 0.67; 1‐3 positive nodes: HR 0.74, P < 0.001; 4+ positive nodes: HR 0.75, P < 0.001) (14). Positive resection margin is also an indicator of survival benefit from CRT (15). Therefore, the American Society of Clinical Oncology (ASCO) guidelines (16) recommend that adjuvant CRT may be offered to patients with R1 resection or node-positive disease after 4 to 6 months of systemic adjuvant CT. However, the view of the role of CT-CRT is not consistent among different international academic societies, and the European Society for Medical Oncology (ESMO) guidelines do not recommend adjuvant CT-CRT (17).

Our study enrolled 266 patients with R1 resection or node-positive pancreatic cancer after surgery. Survival analysis based on baseline characteristics and treatment was performed. Propensity score matching (PSM) was used to investigate the role of radiotherapy in 122 patients who had medical records of adjuvant therapy.

Patients with lymph node metastasis or R1 resection after surgery had a poor prognosis (3–5). These two factors were also reported as possible indicators for selecting patients who may benefit from adjuvant radiotherapy (14, 15). Therefore, in our study, we focused on these patients to analyze possible prognostic factors and explored the role of postoperative CT-CRT.

In our analyses, pancreatic tumors located in the pancreatic body or tail indicated a better prognosis than tumors located in the pancreatic head for patients undergoing curative resection. Several studies reported similar results (20–22). Pathologically unfavorable is a possible explanation for the worse OS associated with tumors located in the pancreatic head, which tend to have lymph node metastases, advanced pathologic stages and worse tumor grades (21). We also found that the pathological T stage was a significant prognostic factor. Indeed, the significance of tumor size in clinical outcomes has received more attention since the reversion of the T stage in the 8th edition of the AJCC TNM staging system (18). Many studies have validated the more detailed T staging’s superiority in stratifying patients by survival (23, 24). Larger tumors correlated with higher CA19-9 levels, higher grades, perineural invasion, R1 resection and more positive lymph nodes (25). Our study indicated that tumor location and size should be given more attention in patients with R1 resection or positive lymph nodes. The role of tumor location and size in treatment stratification for these high recurrence risk patients could be further explored. In this study, the N and TNM stage were significantly associated with survival in univariate Cox regression analyses but not in multivariate analyses. The few N0 and TNM stage I patients led to the patients’ uneven stage distribution, contributing to the bias. An association between marginal status (R0 vs R1) after resection and OS was not found. The criteria we set for patient selection resulted in all patients with R0 resection included in this study having lymph node metastasis, which also indicated a poor prognosis. Thus, the survival difference between different marginal status groups was not revealed. The CT regimen was also recorded and analyzed in our study. For patients after resection of pancreatic adenocarcinoma, mFOLFIRINOX or gemcitabine with capecitabine is the recommended adjuvant CT regimen, although the increased toxicities of mFOLFIRINOX limit its widespread use (7). Gemcitabine monotherapy is an alternative option. In the Asian population, the S-1 regimen was indicated to markedly prolong OS with lower toxicity than gemcitabine (8). In our study, only four patients received the mFOLFIRINOX regimen, and most of the patients included received the GEM, GS or S-1 regimen. CT showed a significant improvement in OS compared with observation, while there were no significant differences among the different CT regimens.

Due to the high risk of pancreatic cancer recurrence after surgery and CT (6), the effectiveness of adjuvant radiotherapy has been explored for many years. However, compared with CT, the role of radiotherapy in the adjuvant setting is controversial. In our study, PSM was applied and generated a cohort for further analysis (31 patients in the CT group and 20 patients in the CT-CRT group). According to the following Cox regression analysis, adjuvant CT-CRT was a significant prognostic factor for a more favorable OS (p=0.014). Some studies also supported the survival benefit of CT-CRT (26). The prolonged OS of the CT-CRT group might benefit from local control improvement. The local recurrence rate of the CT-CRT group was lower than that of the CT group. The improvement of local control by adding CRT in pancreatic cancer therapy was illustrated in other studies. In a phase II prospective randomized study comparing adjuvant CT alone and adjuvant CT-CRT, the local and simultaneous local and distant progression rates were 11% and 13%, respectively, in the CT-CRT group, which were lower than those (24% and 20%) in the CT group (12). Other retrospective studies also reached similar results (26–28). According to previous studies, the risk of local recurrence correlated with lymph node metastasis (19, 29) and R1 resection (30–32). The better outcome in the CT-CRT group in this study, of which patients have lymph node metastasis or R1 resection, might indicate the importance of patients selecting. Patients who have a higher risk of local recurrence were likely to benefit from CT-CRT. Meanwhile, some pathological characteristics have been reported to indicate benefit from adding adjuvant radiotherapy, such as pT3 (33), lymph node metastasis (14, 29, 33, 34) and positive margin status (35).

In the ESPAC 4 study, for patients with a positive margin, gemcitabine combined with capecitabine did not significantly improve patients’ overall survival compared to gemcitabine alone (23.7 vs. 23.0 months) (6). Although the mFOLFIRINOX regimen showed higher activity in nodal positive or R1 resection pancreatic cancer, few patients could tolerate the toxicities. Therefore, adjuvant radiotherapy could be taken into consideration in these patients. Our study focused on these patients for the first time to analyze and prove the survival benefit of CT-CRT. According to previous studies, the delayed administration of CT, toxicities of radiotherapy (9–11) and radiation dose were potential reasons for the lack of OS benefit observed in CRT or CT-CRT (36). In our study, all patients received a standard regimen, including CRT following 4-6 months of CT, recommended by the ASCO guidelines (16). Intensity-modulated radiation therapy (IMRT) was used in all patients with a sufficient radiotherapy dose (median total dose of 50 Gy and fraction dose of 1.8 Gy). IMRT has been widely used in recent years to achieve more conformal dose delivery with less toxicity. In pancreatic cancer, IMRT had significantly reduced toxicities without changes in the therapeutic outcome compared with conventional 3D-conformal radiation therapy (37). Thus, our study results could be more convincing by adapting reasonable regimens and advanced technology. For the CT regimen given before adjuvant CRT, there were no differences in DFS and OS between gemcitabine and fluoropyrimidine-based regimens (38, 39). However, a trend to extend OS in the gemcitabine group for the patients with pancreatic adenocarcinoma in the pancreatic head was reported (39). In our study, the number of patients who received GS was equal to those who received gemcitabine alone.

There are some limitations of our study. First, it was a retrospective study based on single center. Second, the small sample size of patients who received CT-CRT limited a definitive conclusion. Third, about half of the patients enrolled in our study received S1 chemotherapy, which markedly prolonged OS with low toxicity in Asian patients but not demonstrated in European and North American patients. Hence, our results should be cautiously illustrated when applied in non-Asian populations.

Further studies are in need to illustrate the survival benefit in selected patients. Biomarkers such as genetic alteration, circulating tumor DNA or circulating tumor cells should also be incorporated to investigate the predictive biomarker for chemoradiation. Phase III trial aimed at offering more convincing evidence for CT-CRT is ongoing, in which patients receive adjuvant CRT [50.4 Gy in 28 fractions with concomitant 5-fluorouracil (5-FU)] after five cycles of adjuvant CT (40).

In conclusion, our study focused on a subgroup of patients with resected pancreatic cancer who had a worse prognosis due to lymph node metastasis or positive margin status. Among the baseline and pathological characteristics, tumor location and the T stage were significantly associated with prognosis. There was no significant difference in the effectiveness of different CT regimens. PSM and Cox regression analysis illustrated that adjuvant CT-CRT was associated with prolonged patient OS and lower local recurrence rate compared with CT alone.

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 human participants were reviewed and approved by Peking Union Medical College Hospital. The patients/participants provided their written informed consent to participate in this study.

JX, YC and CB conceived and designed the study. BY and XH collected the clinical data. NJ, XG, XL and NZ finished patients’ follow-up. JX and YC performed the statistical analyses. JX wrote the manuscript. YC and CB reviewed and revised the manuscript. All authors contributed to the article and approved the submitted version.

This study received grants from the Chinese Academy of Medical Sciences (CAMS) Initiative for Innovative Medicine (CAMS- I2M) 2017-I2M-1-001.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.