Our results revealed that tumor progression during NST has a negative impact on survival outcomes. To the best of our knowledge, this is the first report assessing patients divided into three groups (PD group, non-pCR group, and pCR group). The association of tumor progression during NST with poor outcomes was also supported in recurrence settings. Many studies have reported the clinical and molecular predictors of pCR [(4, 12–19)], and pooled analyses and meta-analyses have supported the same findings [(8, 9)]. However, few reports have focused on the predictors of tumor progression during NST. In the literature, cancer progression while receiving NST was reported by the MD Anderson Cancer Center in America [(10)] and the Sunnybrook Odette Cancer Center in Canada [(20)].
Predictive factors of tumor progression during neoadjuvant chemotherapy were reported by Abigail Set al [(10)]. They identified advanced tumor stage, high nuclear grade, high Ki-67 score, and ER/PgR negativity as predictive factors. However, these statuses were also reported as predictive factors for pCR [(21–23)]. The mechanisms or molecular markers that differentiate chemotherapy-sensitive tumors from chemotherapy-resistant tumors are not well known. Furthermore, there are no novel therapeutic targets for patients with tumor progression during NST.
Recently, circulating tumor DNA (ctDNA) has shown an important role in monitoring the early response and predicting survival outcomes after chemotherapy treatment [(24–26)]. The presence of ctDNA has a marked prognostic value when detected after surgical resection [(27)]. Several studies have reported that ctDNA detection at the end of neoadjuvant chemotherapy indicates worse survival outcomes [(28)]. Shunying Li et al demonstrated that ctDNA can be used to predict the tumor response during NST in breast cancer, which may provide personalized treatment with escalating the treatment for the patients with poor outcomes [(29)]. This study showed that the presence of ctDNA before NST was correlated with a worse prognosis, particularly in ER-negative patients. These results suggest that monitoring and detecting ctDNA may help identify the patients with tumor progression during NST.
Enhancing the efficacy of neoadjuvant chemotherapy with immune checkpoint inhibitors might be one of the methods for overcoming primary resistance, which results in the decrease of tumor progression during NST. Two studies in neoadjuvant settings have reported a high response rate to treatment with immune checkpoint inhibitors in a phase 3 trial [(30)-(31)]. These results suggest that there are very large benefits, in terms of pCR, from adding immune checkpoint inhibitors to chemotherapy, particularly in high-risk triple-negative patients, who have a similar status to the patients who experienced tumor progression during NST in our study.
Response-guided therapy with neoadjuvant chemotherapy is the standard strategy for early breast cancer [(32)]. Triple-negative and HER2-positive subtypes are particularly good candidate populations for this strategy, as was shown in the CREATE-X trial [(33)] and KATHERINE trial [(34)]. However, there is no effective therapy for patients with PD during NST. In this study, almost all patients experienced early recurrence and extremely shorter survival regardless of whether they received standard or optional treatment after recurrence in the PD group.
A single institution retrospective study demonstrated the efficacy of salvage therapy for patients with early breast cancer who experience disease progression during NST [(20)]. The salvage treatment includes surgery, concurrent chemoradiation with cisplatin, and chemotherapy alone. They concluded that patients progressing on NST responded well to salvage therapy; however, our study reported poor outcomes in patients with tumor progression during NST compared with those without tumor progression. Interestingly, it was suggested that the efficacy of a platin-based regimen was the same in as our study.
Moreover, one of the differences between the KEYNOTE-522 and IMpassion051 trials is the combination of chemotherapy, specifically whether carboplatin is included. The KEYNOTE-522 trial evaluated treatment including carboplatin and reported a PD rate during neoadjuvant chemotherapy that was lower than in the IMpassion031 trial (1.3% in the KEYNOTE-522 trial vs. 3.0% in IMpassion051 trial) [(30, 31)].
Our study has several limitations. First, this study was a retrospective study conducted based on medical record; therefore, it was difficult to perform multivariable regression analyses to account for potential confounders. Second, this study was performed in a single institution. However, 595 patients were enrolled in this study, which was thought to be a sufficient sample size.