1 INTRODUCTION

In patients with early- and mid-stage NSCLC, complete surgical resection is the most effective treatment.¹ However, 30%–70% of patients relapse after resection.^{2, 3} Therefore, perioperative chemotherapy may improve the surgical outcome and thus patient survival.^4-6 Several randomized control studies (RCT) and a pooled analysis of trials indicate that cisplatin-based adjuvant chemotherapy improves the 5-year survival by 5%,^7-9 although the results are controversial for patients with stage IB, as adjuvant chemotherapy, one of the perioperative chemotherapy options, is recommended as a standard treatment for early-stage NSCLC patients by international guidelines.^{10, 11} Neoadjuvant chemotherapy, the other part of perioperative treatment, was not evaluated as extensively but still demonstrated an absolute benefit of 5% at 5-years, which was equal to that of adjuvant chemotherapy.^{12, 13} Overall, studies addressing the role of perioperative chemotherapy showed that it confers a survival benefit.

Differences in the efficacy of pemetrexed between squamous cell carcinoma and nonsquamous cell carcinoma led to the assessment of precision chemotherapy to improve the efficacy of treatment. Several factors, such as stage, chemotherapy regimens, and histologic type, need to be considered in the selection of perioperative chemotherapy to maximize the survival benefits of NSCLC patients. However, tumor location, an important factor affecting the prognosis of NSCLC, is not included in decisions regarding chemotherapy. Because tumors in different lobes are associated with different prognoses after surgery, radiotherapy, and immunotherapy,^14-17 we examined the survival benefits of perioperative chemotherapy according to the lobes affected to provide precise perioperative chemotherapy options that increase the survival benefit to patients.

2 PATIENTS AND METHODS

3 RESULTS

3.2 Survival outcome

Patients who received perioperative chemotherapy had a significant better OS than those who received surgery alone (p < 0.001; Figure S2A). The absolute improvement in 5-year survival was 1.9%. After PSM, the results indicated that perioperative chemotherapy could significantly benefit patients with IB–III disease (p < 0.001; Figure 1A), and the absolute improvement in 5-year survival was 6.3%, which greater than the result before PSM.

To further explore the benefit of perioperative chemotherapy, lobar subgroup analyses were performed. As shown in Figure 1B–F, perioperative chemotherapy was associated with a significantly better OS than surgery alone in most lobar subgroups except the RML subgroup (p = 0.690). The absolute improvement was 7.8% for RUL (p < 0.001), 6.1% for RLL (p = 0.001), 7.7% for LUL (p < 0.001), and 3.9% for LLL (p = 0.013).

3.3 Survival outcome according to stage

Next, we performed survival analysis according to stage subclassification (stage IB, stage II, or stage III). A total of 11,174 patients with stage IB were enrolled in the study. Before PSM, there was an absolute improvement in 5-year survival of 6% after perioperative chemotherapy compared with surgery alone (p = 0.001; Figure S2B). However, the benefit disappeared after PSM (p = 0.880; Figure 2A). Lobar subgroup analyses did not identify a potential benefit of a lobar subgroup (Figure 2B–F).

Consistently significant survival benefits were observed in stage II patients before PSM (p < 0.001; Figure S2C) and after PSM (p < 0.001; Figure 3A), and the absolute improvement in 5-year survival was 10.5% and 9.5%, respectively. Lobar subgroup analyses demonstrated that perioperative chemotherapy significantly benefited patients with RUL, RLL, LUL, and LLL tumors (p < 0.001, p < 0.001, p < 0.001, and p = 0.015, respectively; Figure 3B,D–F), whereas no benefit was observed for patients in the RML subgroup (p = 0.180; Figure 3C).

A statistically significant OS benefit was observed for perioperative chemotherapy in stage III patients, with an absolute improvement in 5-year OS of 5.8% before PSM (p < 0.001; Figure S2D) and 6.1% after PSM (p < 0.001; Figure 4A). Stage III patients in the RML and RLL subgroups did not significantly benefit from perioperative chemotherapy (p = 0.830 and p = 0.360, respectively; Figure 4C,D), whereas patients in the RUL, LUL, and LLL subgroups did (p = 0.002, p < 0.001, and p = 0.028; Figure 4B,E–F).

3.8 Cox regression analysis

To explore any potentially confounding factors related to OS, Cox regression analysis was performed. The HRs, p values, and 95% CIs are summarized in Table S1. Univariate analysis demonstrated that older age, white people, male sex, LUSC, advanced stage, LN metastases, sublobar resection, no LN dissection, and no perioperative chemotherapy were associated with worse OS (Figure 5). Similarly, multivariate Cox regression analysis revealed that older age, male sex, LUSC, advanced stage, LN metastases, no LN dissection, and no perioperative chemotherapy were associated with poor OS (Figure 5).

4 DISCUSSION

Over the past two decades, perioperative chemotherapy has become a standard treatment for NSCLC, and it has benefited many early-stage patients with resectable tumors.¹⁸ However, whether adjuvant chemotherapy, a type of perioperative chemotherapy, benefits stage IB patients remains controversial because of confounding results. Gary's study, a mature analysis of Cancer and Leukemia Group B (CALGB) 9633, is the only RCT designed specifically for stage IB NSCLC patients. It showed no difference in survival between the adjuvant chemotherapy group and the nonadjuvant chemotherapy group (HR, 0.83; CI, 0.64–1.08; p = 0.120).¹⁹ However, exploratory subgroup analysis in the CALGB 9633 trail showed that patients with stage IB with tumors >4 cm in diameter had a significantly better OS (HR, 0.69; CI, 0.48–0.99; p = 0.043). Several RCTs including IALT,⁷ NCIC-CTG-JBR-10,⁹ and ANITA⁸ showed a significant OS advantage of adjuvant chemotherapy in stage II-IIIA patients, but no significant OS advantage in the stage IB subset. The study by Harubumi et al. was the only RCT that showed a statistically significant benefit of adjuvant chemotherapy in stage IB patients. However, the patients received uracil–tegafur, which is not approved in the USA and Europe. To date, no cisplatin-based RCTs have demonstrated a significant OS advantage of adjuvant chemotherapy in the stage IB subset.

Unlike adjuvant chemotherapy, neoadjuvant chemotherapy, the other part of perioperative chemotherapy, has not been evaluated extensively. There are no studies assessing the effect of neoadjuvant chemotherapy specifically in stage IB patients. In Scagliotti's study,²⁰ stage IB/IIA group (89.8% for IB and 10.2% for IIA), OS (HR, 1.02; 95% CI, 0.58–1.19; log-rank p = 0.940) and PFS (HR, 1.06; 95% CI, 0.65–1.72; p = 0.830) were not significantly different between treatment groups (neoadjuvant chemotherapy vs. surgery alone), whereas a positive effect of neoadjuvant chemotherapy on OS and PFS was revealed in the IIB/IIIA group (HR, 0.42, 95% CI, 0.25–0.71, p < 0.001; HR, 0.51, 95% CI, 0.32–0.80, p = 0.002). In addition, a meta-analysis¹³ of 15 RCTs (2385 patients with stage IB to IIIA) showed a significant benefit of neoadjuvant chemotherapy regarding survival (HR, 0.87, 95% CI 0.78–0.96, p = 0.007) and an absolute survival improvement at 5 years of 5% for all stages, which was similar to the survival benefit with adjuvant chemotherapy.⁶ Overall, perioperative chemotherapy, regardless of whether neoadjuvant chemotherapy or adjuvant chemotherapy, plays an important role in the systemic treatment of NSCLC and significantly benefits patients. However, there is no clear evidence of the positive effect of perioperative chemotherapy in patients with stage IB disease.

This retrospective study showed that patients who received perioperative chemotherapy had a better OS than those who underwent surgery alone including lobectomy and sublobar resection. However, the survival benefit was not significant in patients with IB disease. These results are consistent with those of previous studies. In this study, the absolute improvement in 5-year OS after PSM was 6.3%, which was higher than the improvement reported in other studies, including the LACE Collaborative Group study (5.4% for adjuvant chemotherapy) and NSCLC Meta-analysis Collaborative Group study (5% for neoadjuvant chemotherapy).^{6, 13} The difference between this study and previous reports may be related to differences in the patient population enrolled. Patients with stage I (IA/IB) accounted for 49.4% in the NSCLC Meta-analysis Collaborative Group study and 37.5% in the LACE Collaborative Group study, whereas they accounted for 26.9% in this study after PSM. As reported previously, perioperative chemotherapy showed no meaningful clinical benefit in this population, suggesting that the large proportion of patients with stage I would influence the final absolute improvement in the LACE Collaborative Group study and the NSCLC Meta-analysis Collaborative Group study. Furthermore, both patients who received neoadjuvant chemotherapy and adjuvant chemotherapy were included in this study because of the limitations of the SEER database, which may also have contributed to the discrepancy in the final result.

Primary tumor location may affect the prognosis of NSCLC after surgery, radiotherapy, and immunotherapy.^{15-17, 21, 22} Shaverdian et al.¹⁶ retrospectively analyzed 122 cases and showed that a lower lobe tumor location was associated with poor OS among all patients and those with biopsy-confirmed disease. Similar results were reported by Stran et al. after Cox analysis²²; patients whose primary tumor was located in the lower lobes had a worse OS than those with tumors in upper lobes after surgery (HR, 1.17; 95% CI, 1.03–1.32; p < 0.001). Thus, lobe-specific analysis was performed in this study to explore whether tumor location affects the prognosis of NSCLC after perioperative chemotherapy and to identify the population that could benefit from perioperative chemotherapy. Previous studies classify RML into upper and lower groups to determine the effect on prognosis by Cox analysis. In this study, we consider RML as a subgroup to analyze the survival difference between perioperative chemotherapy and surgery alone, and found that patients with RML tumors did not significantly benefit from perioperative chemotherapy regardless of stage (stage IB–III), which is the first time this result is reported. The outcome of RML was reported to be the worst among all locations of NSCLC, which makes it unique and worthy of more attention.^23-25 In addition, patients with stage III whose primary tumor located in RLL do not have a better OS after perioperative chemotherapy. These results challenge the hypothesis that the prognosis is similar for different lobes treated with the same strategy. However, the reasons that different lobes have different prognoses remain unknown. However, our previous study²⁶ shed light on the different survival outcomes of NSCLC patients with different tumor location after sublobar lung resection, which may explain the reason of different outcomes in this study. Moreover, several studies found postoperative complications and postoperative respiratory failure predominantly develop in lung cancer patients with lower lobes,^{27, 28} which also contributed to the different survival outcomes among patients with different tumor location. In addition, study by Okamoto et al. revealed that tumor mutation burden (TMB) level differed according to tumor location²⁹ in lung cancer. As a biomarker for the efficacy of chemotherapy in lung cancer,³⁰ TMB level may differ in RUL, RML, RLL, LUL, and LLL. We speculate that this may also play a critical role in the different survival outcomes of NSCLC patients with different tumor location in our study.

To the best of our knowledge, this study is the first to discuss the difference between perioperative chemotherapy and surgery alone for NSCLC in different lung lobes. The study had several limitations. Firstly, although PSM was performed, there are still some unavoidable biases. Secondly, the SEER database provided no information on the regime on chemotherapy, such as dose and scheme. These factors may also influence the effect of perioperative chemotherapy. In addition, preoperative and postoperative CT would bring about different outcome in the surgery-oriented treatments owing to their different mechanisms; however, we could not further analyze separately because the SEER database did not provide the information, which is also one of the limitations of our study. Meanwhile, the patient number of certain subgroup is low, especially for RML, which may lead to an inaccurate conclusion, therefore, results in RML needed to be validated in a larger sample size cohort. Lastly, due to inability to distinguish between neoadjuvant and adjuvant populations, we were unable to validate our findings in these two populations separately.

5 CONCLUSION

During the past decades, detecting technology and treatment in lung cancer have been developing rapidly with the rapid developments of precision medicine. For example, targeted therapy can help those with specific gene mutation and gain a better prognosis. Biomarker, such as TMB, can distinguish patient who may benefit from PD-1/PD-L1 inhibitor treatment. And our study, which is still needed to be verified in a larger cohort, may also guide physicians to provide a precise perioperative treatment strategy in early-stage NSCLC. In conclusion, we proposed a precise lobe-specific perioperative chemotherapy. For stage IB NSCLC, right middle lobe NSCLC from stage IB-III and right lower lobe NSCLC from stage III, perioperative chemotherapy might not confer survival benefits.

AUTHOR CONTRIBUTIONS

Xi Lei: Conceptualization (equal); data curation (equal); resources (equal); software (equal). Tong Li: Conceptualization (equal); data curation (equal); formal analysis (equal); funding acquisition (equal); investigation (equal); resources (equal); software (equal); supervision (equal). Fuling Mao: Conceptualization (equal); data curation (equal); resources (equal); software (equal). Fan Ren: Data curation (equal); formal analysis (equal); supervision (equal); writing – original draft (equal). Quanying Tang: Data curation (equal); methodology (equal); software (equal); supervision (equal); validation (equal). Weibo Cao: Resources (equal); software (equal); writing – original draft (equal); writing – review and editing (equal). Lingling Zu: Methodology (equal); project administration (equal); resources (equal); supervision (equal); validation (equal); visualization (equal). Song XU: Conceptualization (equal); data curation (equal); formal analysis (equal); funding acquisition (equal); investigation (equal); methodology (equal); project administration (equal); resources (equal).

FUNDING INFORMATION

The present study was funded by the National Natural Science Foundation of China (82172776), Tianjin Key Medical Discipline (Specialty) Construction Project (TJYXZDXK-061B), and Diversified Input Project of Tianjin National Natural Science Foundation (21JCYBJC01770).

CONFLICT OF INTEREST STATEMENT

The authors declare that they have no conflicts of interest regarding the content or publication of this paper.

ETHICS STATEMENT

SEER data are available to eligible individuals from the National Cancer Institute (NCI) upon completion of a data use agreement. We completed a data use agreement to access the data reported in this manuscript. A data use agreement per NCI/SEER policies was signed in order to obtain approval to access deidentified SEER public-use data through SEER*Stat.