Patient characteristics
Totally, 1596 LARC patients receiving direct surgery followed by adjuvant chemotherapy from the SEER database were eligible for this study, including 288 patients (18.0%) were present with PNI. As was depicted in Table 1, significant differences were observed between patients with and without PNI stratified by tumor characteristics, including tumor size, tumor differentiation, TD, T stage, N stage, and T/N stage (P < 0.05).
Effect Of Pni On Prognosis
Univariate regression analysis showed that PNI was associated with decreased OS and CSS, as well as other established factors including age, marital status, CEA, tumor size, tumor differentiation, TD, and T/N stage (all P < 0.05, Table 2). And then, PNI was also found to be an independent risk factors of OS and CSS using multivariate Cox regression analysis, as well as age, marital status, CEA, tumor size, TD, and T/N stage (all P < 0.05, Table 2).
Taken into consideration of prognostic interaction between PNI and other factors, we further calculated the adjusted HR of PNI for OS and CSS. As was depicted in Table 3, PNI remained to be an independent risk factor of OS with slightly decreased HR after adjusting age, sex, race, martial status (model 1), mode l + T stage, N stage (model 2), model 2 + tumor differentiation, tumor size, CEA, TD (model 3), and model 3 + number of LND, RT (model 4), respectively (all P < 0.05). Importantly, similar findings were observed in terms of CSS (all P < 0.05, Table 3).
Effect Of Adjuvant Rt On Pni Prognosis
Among 288 patients with PNI, 185 patients (64.2%) receiving adjuvant RT. Not as expected, PNI patients with stage III, and N + were less likely to receive adjuavnt RT (all P < 0.05, Table 4). Nonetheless, the baseline characteristics between patients receiving adjuvant RT or not became comparable after 1:1 PSM (all P > 0.05, Table 4).
K-M survival curves showed that the pooled HR for the OS was in favor of S + CRT group compared to S + CT group, although the difference lacked statistical significance (P > 0.05, Fig. 1A). Likewise, corresponding advantage of adjuvant RT was also observed in terms of CSS without statistical significance (P > 0.05, Fig. 1B). Of note, the survival advantage of adjuvant RT remained after 1:1 PSM regardless of both P > 0.05 (Fig. 1C, D).
Univariate analysis showed that age, marital status, CEA levels, tumor size, TD, T stage and N stage were associated with OS of patients with PNI (all P < 0.05, Table 5), and then age, CEA levels, tumor size and N stage were identified to be independent risk factors of OS (all P < 0.05, Table 5). Likewise, CEA, tumor size, T stage and N stage were identified to be independent risk factors of CSS (all P < 0.05, Table 5).
Constructed A Nomogram To Predict The Prognosis Of Pni(+) Patients
Based on the results of multivariable analysis, we constructed a nomogram to predict the OS of patients with PNI (Fig. 2). The C-index of the current nomogram and the 8th American Joint Committee on Cancer (AJCC) staging system were 0.750 and 0.552, respectively. Good calibration was observed between the predicted outcome of the current nomogram and the observed outcome in terms of the 3- and 5-year survival rates (Fig. 3A-B).
Subgroup Analysis
Each patient with PNI was scored according to the current nomogram, and the median total value was 180 (25–233). We supposed that only a small number of people could be exempt from AC, and the patients were then divided into high- and low-risk subgroups with the ratio of 2:1. KM survival curves revealed significant difference between subgroups of high- and low-risk (P < 0.0001, Fig. 4A), and the corresponding 1-, 3-, and 5-year survival rates comparing the high-risk subgroup versus the low-risk subgroup were 93.0% vs. 98.4%, 63.3% vs. 95.3%, 52.2% vs. 89.5%, respectively. Further analysis showed that adjuvant RT could improve the OS of high-risk subgroup (P < 0.05, Fig. 4B) but not the low-risk subgroup (P > 0.05, Fig. 4C).
Validation By Fujian Cancer Hospital
98 LARC patients receiving direct surgery followed adjuvant chemotherapy and present with PNI from FJCH were taken as an externation validation cohort, including 15 patients (15.3%) receiving adjuvant RT. The baseline characteristics were summarized in Supplementary Table 1. Good calibration was also observed between the predicted outcome of the current nomogram and the observed outcome in terms of the 3- and 5-year survival rates (Fig. 3C-D). The C-index of the current nomogram model and the 8th AJCC staging system were 0.699 and 0.540, respectively. Patients were also divided into high- and low-risk subgroups with divergent survival curves (P < 0.005, Fig. 5A). Similarly, the survival advantage of adjuvant RT was observed in subgroup of high-risk regardless of P > 0.05 (Fig. 5B). Of note, only three patients in the low-risk subgroup received adjuvant RT without recurrence and death during the follow-up, and consequently the corresponding survival results deserved further validation (Fig. 5C).