A total of six RCTs were included in our meta-analysis (N = 2142), which is the most comprehensive study with a larger sample size compare with previous meta-analysis to the best of our knowledge, aiming to evaluate the efficacy and safety of NACT plus ICIs vs. NACT alone in patients with early TNBC [17–22].The results showed that chemo-immunotherapy could improve pCR rates significantly compared with chemotherapy alone in patients with TNBC. Moreover, there was no correlation between the improvement in pCR rates and PD-L1 expression status; nevertheless, the pCR rates were higher in PD-L1 positive patients. Subgroup analysis showed that nodal positive and ECOG PS 0 subgroups benefited from PD-1/PD-L1 inhibitors plus NACT. Furthermore, a statistical difference was observed for EFS in the two RCTs through preliminary analysis. Finally, the addition of PD-1/PD-L1 inhibitors increased the incidence of grade 3–4 diarrhea, vomiting, hypothyroidism, and hyperthyroidism of any grade, did not increase the incidence of other adverse events.
Anthracyclines and taxanes are the most classic chemotherapy regimens for TNBC patients [25]. Platinum-based neoadjuvant chemotherapy can significantly improve pCR rates in patients with TNBC, but with large hematological toxicities [11, 26]. The effective rate of patients with TNBC still needs to be improved, and it is necessary to explore new antitumor drugs to increase efficacy. As immunotherapy is becoming increasingly effective in the treatment of cancer, it has become a promising treatment option for patients with TNBC [27–29]. Immunotherapy mainly uses the human immune system to kill tumor cells, thus achieving its anti-tumor effects, and aims to improve immunity [30, 31]. PD-L1 expressed by tumor cells binds to PD-1 expressed by T cells, thus inactivating T cells and forming immune escape of tumor cells, which can be blocked by PD-1 or PD-L1 inhibitors in the tumor microenvironment [31]. Due to the heterogeneity of the tumors, immunotherapy alone is not outstanding, and immunotherapy plus chemotherapy, target drugs, and local ablation therapy are expected to improve efficacy at present [32].
Based on the results of IMpassion 130 and Keynote-355 trials, atezolizumab and pembrolizumab in combination with chemotherapy have been shown to improve PFS in patients with advanced TNBC (aTNBC), especially PD-L1 positive patients [15, 16]. However, IMpassion131 assessed atezolizumab plus paclitaxel in aTNBC and the result showed did not improve progression-free survival (PFS) and overall survival (OS). The results of both studies were discussed and voted, in favor of continuing to approve atezolizumab in combination with nab-paclitaxel in patients with aTNBC of PD-L1 positive patients. However, recently the FDA stopped supporting this approval due to changes in the treatment outlook. In this situation, it is predicted whether patients with early TNBC will obtain significant benefits from immune checkpoint inhibitors. Some studies exploring the question, such as the Keynote-173 phase 1b study, included 60 high-risk patients with early stage TNBC in six cohorts to assess the safety and curative effect of adding pembrolizumab to six different regimens of chemotherapy [23]. All cohorts of the overall pCR were defined as ypT0/Tis ypN0, and the pCR rates was 60% (90% CI: 0.49–0.71). The EFS and OS at 12 months ranged from 80 to 100%.
In a previous meta-analysis [33], a total of three RCTs were included (N = 883) and the results showed that patients with TNBC had a pCR rate improvement after using immunotherapy plus chemotherapy (OR:1.78; 95% CI:1.34–2.30, P<0.0001) [16–18]. However, there may be high heterogeneity due to the lack of subgroup analysis. In addition, no survival results were reported in the previous meta-analysis. There could be some bias due to the small number of included studies and the small sample size. Another meta-analysis [34] included five RCTs; their sample size was smaller than ours (1496 vs 1557), and the result of the pCR rate analysis was consistent with ours (OR: 1.72; 95% CI: 1.22–2.42, P < 0.0001) [17–21]. Nevertheless, subgroup analysis only analyzed PD-L1 expression status and showed that PD-L1 negative populations did not benefit from NACT plus ICIs (OR: 1.56; 95% CI: 0.80–3.03), whereas we had a different result. Furthermore, the author did not analyze EFS by forest map combined analysis. In the analysis of AEs, their results showed that the addition of PD-1/PD-L1 inhibitors did not increase the incidence of AEs, while our AEs analysis showed different results. Therefore, it is necessary to perform a comprehensive meta-analysis to clarify and define the efficacy of immunotherapy and chemotherapy.
Some related studies have shown that pCR as a biomarker after neoadjuvant chemotherapy has a prognostic value in TNBC patients [35, 36]. It has been proven to be a substitute indicator for disease-free survival (DFS), EFS, and OS among patients with TNBC [37, 38]. There were six RCTs included in our meta-analysis (N = 2142) [17–22], and the results showed that immunotherapy combined with chemotherapy significantly improved pCR rates (OR: 1.91; 95% CI: 1.32–2.78, P < 0.001). The pCR rates were increased by 16.5% in IMpassion031 [20], 13.6% in KEYNOTE-52217, and 36.8% in Nci 10,013 [22]. However, the negative result in NeoTRIPaPDL1 [21] may be associated with the enrolled population being at high risk, the disease stage being late, and the immune induction effect of the chemotherapy regimens in the study design being insufficient.
However, our meta-analysis has some heterogeneity (I2 = 62%), and subgroup analysis is required. Some studies, such as IMpassion 130 and Keynote-355, showed that PD-L1 expression status is associated with better efficacy in patients with metastatic TNBC, but not in patients with early TNBC [15, 16]. The IMpassion 031, Keynote-522, and GeparNuevo studies showed benefits regardless of PD-L1 positivity or negativity in patients with early TNBC [17, 19, 20]. Moreover, the measurement of PD-L1 expression status was inconsistent between IMpassion 130 and Keynote-522. According to the PD-L1 staining of tumor-infiltrating immune cells as a percentage of the tumor area (PD-L1 negative < 1 vs. PD-L1 positive ≥1) in IMpassion 130, while calculating the combined positive score (CPS) to evaluate the number of PD-L1 positive cells (PD-L1 positive CPS ≥1) in Keynote-522 [20, 39]. In our meta-analysis, according to a subgroup analysis of PD-1/PD-L1 expression status, the results showed that PD-1/PD-L1 inhibitors plus chemotherapy had an improvement in pCR rates compared with chemotherapy alone, especially higher in patients with PD-L1 positive TNBC. The disease stage may indicate that different stages have different efficacy. The explanation is that the tumor immune microenvironment is stronger in early stage TNBC than in advanced-stage TNBC for PD-L1 patient populations; as a result, immunotherapy may further enhance the anti-tumor response. In addition, chemotherapy could stimulate some immune responses through the immune system [20]. Furthermore, two subgroup analyses showed that NACT plus ICIs increased pCR rates in nodal positive, ECOG PS 0 subgroups.
EFS was used to evaluate the efficacy of NACT. Three studies reported survival outcomes, and the results showed a statistical significance (HR: 0.66; 95% CI: 0.48–0.92, P = 0.01), although the data were immature. After 15.5 months of follow-up, KEYNOTE-522 showed that EFS was not achieved in the two groups at the time of the second interim analysis, however, after 39.1 months of follow-up, the result was beneficial. Among them, the I-SPY2 Trial showed that the addition of PD-1/PD-L1 inhibitors could improve EFS in patients with TNBC, but the IMpassion031 trial showed that EFS was not beneficial, possibly because of the short follow-up time [17, 18, 20].
According to our meta-analysis, NACT plus ICIs increased the incidence of diarrhea, vomiting, hypothyroidism, and hyperthyroidism and did not increase the incidence of other AEs compared with chemotherapy alone in patients with early TNBC. The intestinal tract as an immune organ, when attacked by the immunotherapy drugs, will cause some immune-related reactions, such as diarrhea [40, 41]. The highest incidence of AEs was neutropenia (349/1106 vs. 201/819), followed by anemia (165/1244 vs. 82/819), which was consistent with chemotherapy alone.