Prolonged Time to Adjuvant Chemotherapy Initiation Was Associated with Worse Disease Outcome in Triple Negative Breast Cancer Patients

The optimal time to adjuvant chemotherapy (TTC) for breast cancer (BC) patients remains uncertain. Herein, we aim to evaluate the association between TTC and prognosis among different subtypes in modern era of adjuvant chemotherapy. BC patients receiving operation and adjuvant chemotherapy between January 2009 and December 2015 were included. Enrolled patients were categorized into TTC ≤4 weeks and >4 weeks groups. Relapse-free survival (RFS) and overall survival (OS) were compared according to TTC and analyzed among different BC molecular subtypes. A total of 2611 patients were included. Elder age (P = 0.005), more comorbidities (P <0.001), breast-conserving surgery (P = 0.001), non-invasive ductal carcinoma (P = 0.012), and HER2-positivity (P <0.001) were associated with prolonged TTC. Among whole BC population, no significant difference was observed between two TTC groups in terms of RFS (P = 0.225) or OS (P = 0.355). However, for triple negative (TNBC) patients, TTC >4 weeks was independently related with worse RFS (5-year RFS 81.9% vs 89.3%; HR, 1.89; 95% CI, 1.09 to 3.27; P = 0.024) and OS (5-year OS 84.0% vs 94.0%; HR, 2.49; 95% CI, 1.30 to 4.76; P = 0.006) compared with those TTC ≤4 weeks. Prolonged TTC >4 weeks after BC surgery was not associated with worse survival outcomes in the whole BC patients. However, TTC >4 weeks may increase risk of relapse or death in TNBC patients, which deserves further clinical evaluation.

treatment decision. A multidisciplinary team (MDT) meeting, which included surgical oncologists, medical oncologists, radiation oncologists, pathologists, BC specialized nurses, and other related specialists was held to decide adjuvant treatment regimens based on patient's clinico-pathological characteristics. Frequently proposed chemotherapy regimens included A (anthracyclines-containing), T (taxanes-containing), A + T (anthracyclines and taxanes-containing), and others (such as platinum-based regimen, capecitabine, etc.). Patient compliance, actual chemotherapy usage, and regimen were confirmed by BC specialized nurses during follow-up.
Follow-up. Patient follow-up was conducted by BC specialized nurses in each center. Relapse-free survival (RFS) was measured from the date of surgery to the date of first documented local or distant recurrence or death due to any cause. OS was measured from the date of surgery to the date of death of any cause. The diagnosis of recurrence was generally based on radiographic images and/or pathologic biopsies when accessible. Last follow-up was completed in June 2019. Statistical analysis. TTC was defined from the definitive surgery to the initiation of adjuvant chemotherapy. Patients were then categorized according to TTC length into two groups: ≤4 weeks and >4 weeks. Patient characteristics were compared according to TTC categories by using Chi-square test or Fisher's exact test. Charlson Comorbidity Index (CCI) was applied to evaluate patient comorbidity. Odds ratio (OR) with 95% confidence interval (CI) were calculated from multivariate logistic regression analysis to identify impact factors for TTC. Kaplan-Meier curve was adopted to estimate the 5-year RFS and 5-year OS according to TTC and patient characteristics. Subgroup analysis was conducted using stratified Mantel-Haenszel test to estimate hazard ratio (HR) with 95% CI. Cox proportional hazards regression models was applied to determine the association between TTC and survival outcomes after adjustment for potential confounders. Two-sided P values < 0.05 were considered statistically significant. Statistical analyses were carried out by using the IBM SPSS statistics software version 25.0 (SPSS, Inc., IL, USA) and GraphPad Prism version 7.0 (GraphPad Software, CA, USA).  Supplementary Table S1.

Results
Adjuvant treatment recommendations were listed in Supplementary Table S2. The most frequently used regimens were EC-T (epirubicin 90 mg/m 2 and cyclophosphamide 600 mg/m 2 every 21 days followed by docetaxcel 80-100 mg/m 2 every 21 days or weekly paclitaxel 80 mg/m 2 ), TC (docetaxel 75 mg/m 2 plus cyclophosphamide 600 mg/m 2 every 21 days), and EC (epirubicin 90 mg/m 2 and cyclophosphamide 600 mg/m 2 every 21 days) in the www.nature.com/scientificreports www.nature.com/scientificreports/ whole population. Chemotherapy regimens were differently distributed among different molecular subtypes (P < 0.001, Supplementary Figure S1). Patients receiving EC-T regimen tended to start adjuvant chemotherapy sooner after surgery (P < 0.001, Supplementary Table S2). However, chemotherapy regimen was not an independent  www.nature.com/scientificreports www.nature.com/scientificreports/ impact factor related to TTC in the multivariate analysis (P = 0.187). Patients treated with targeted therapy was more likely to start chemotherapy later (P < 0.001, Supplementary Table S2). Radiation therapy and endocrine therapy recommendation were comparable between two TTC groups.
factors associated with ttc. The distribution of TTC in the study population was demonstrated in Fig. 1.

TTC and OS by different molecular subtypes.
For patients with TNBC, the 5-year estimated OS was 94.0% and 84.0% for patients who started chemotherapy ≤4 weeks and >4 weeks after surgery (P = 0.010; Fig. 2D). Meanwhile, HER2-positive patients receiving trastuzumab showed similar OS between TTC groups (5-year OS 95.3% and 98.9%, P = 0.065). No significant differences in OS according to TTC were seen among patients with the other BC subtypes (Supplementary Figure S3B, S3D, S3F, S3H).

Discussion
In this current study, we included 2611 continuous BC patients and managed to demonstrate that age, comorbidities, breast surgery type, tumor pathologic type, and molecular subtype were independently associated with prolonged TTC. TTC wasn't an independent impact factor for RFS or OS in the whole population, whereas TTC > 4 weeks was significantly associated with worse RFS (HR = 1.89) and OS (HR = 2.49) in TNBC patients. Our findings encouraged us to start adjuvant chemotherapy within 4 weeks after surgery in TNBC patients.
The optimal TTC remains uncertain for BC patients 8 . The median TTC of our study cohort was 20 (IQR 13-29) days, which was relatively shorter than other studies [9][10][11][12]14,15 . Till now, a 4-week interval from surgery to chemotherapy was most frequently adopted when assessing the impact of TTC 19 . Based on our clinical practice, we also chose 4 weeks as the cutoff value since it was the necessary duration to obtain a complete pathological report especially in those who needed additional FISH or multigene assay testing. Other TTC intervals evaluated in previous studies ranged from 2 weeks to 12 weeks [8][9][10][11][12] . For example, Gagliato et al. adopted TTC intervals of ≤30 days, 31-60 days, and ≥61 days 10 . Downing et al. divided TTC by 3 weeks, 3-6 weeks, 6-10 weeks, and >10 weeks 20 . Further explorations were warranted to decide the optimal TTC cutoff for BC patients.
Diverse impact factors have been identified for TTC, mainly consisted of patient factors, surgical/ tumor factors and socioeconomic factors. For example, it was reported that African American 21 , non-Hispanic black 12 , and Hispanic ethnicity 22 were more likely to undergo longer treatment intervals compared with white patients. Elder age 23 , and more comorbidities 24 also led to increased odds of longer TTC. In addition, surgical factors including postoperative complications and immediate autologous breast reconstruction significantly augmented the risk of prolonged TTC as well 25 . On the other hand, low socioeconomic status or underinsured patients were more likely to experience delayed treatment 12,22 . However, there was still limited data about the impact factors for prolonged TTC in the Chinese population. Overall, our findings were generally consistent with previous studies. As identified in our study, patient characteristics including elder age and more comorbidities, surgical factor like BCS procedure, as well as tumor factors like non-IDC histologic type, and HER2-positive subtype were significantly associated with chemotherapy initiation delay. One explanation would be that it would take longer time to assess the condition in elder patients with comorbidities before the initiation of chemotherapy. Meanwhile, for patients receiving BCS, they usually had relatively small tumor size and negative node, which needs more pathological-histological evaluation, including 21-gene RS testing, to make the final chemotherapy decision making. For example, 194/622 (31.19%) patients receiving BCS had a 21-gene RS test, which was higher than patients treated with mastectomy (255/1989, 12.82%). In addition, patients receiving EC-T regimen tended to start adjuvant chemotherapy sooner after surgery, since they were more likely to have node-positive, TN or HER2-positive diseases, and were often quicker to determine if chemotherapy was needed. Other chemotherapy regimens, like EC or TC, were more likely to be applied in node-negative, Luminal-like patients, and additional FISH or 21-gene RS testing were more often required before the initiation of chemotherapy. To note, some of the impact factors could be altered to improve the treatment efficacy, for example, to shorten the waiting duration www.nature.com/scientificreports www.nature.com/scientificreports/ of tumor histo-pathological evaluation including FISH test and 21-gene recurrence score testing. Moreover, for elderly patients, it would be better to collect a complete medical history with all comorbidities at the first time at consultation, to perform necessary clinical evaluations as soon as possible to rule out chemotherapy contraindications. Furthermore, TN patients with stage II-III disease are recommended to receive neoadjuvant therapy, which would shorten the TTC. Moreover, escalation therapy might be given if TNBC patients have residual disease after neoadjuvant therapy to further improve disease outcome.
The impact of TTC on patient survival has been widely evaluated with controversial results. Yu et al. demonstrated a reduced OS (HR, 1.15, 95% CI 1.03 to 1.28) and disease-free survival (HR, 1.16, 95% CI 1.01 to 1.33) with each 4-week delay to the initiation of adjuvant chemotherapy 13 . Similarly, a meta-analysis which included 15410 patients reported that a 4-week increase in TTC was associated with a significant decrease in both OS (HR, 1.14; 95% CI, 1.10 to 1.17) and disease-free survival (HR, 1.14; 95% CI, 1.10 to 1.18) 26 . However, other individual studies found no survival impact with TTC up to 3 months 27-29 . Hershman et al. showed in a cohort of 5003 patients that those who started chemotherapy within 1, 2, or 3 months after surgery had similar survival 27 . Lohrisch C et al. conducted a retrospective analysis of 2594 patients and drew a similar conclusion 28 . In current study, we did not observe significant differences in terms of RFS or OS between two TTC groups in the whole population. One possible reason for such inconsistent findings was the different distribution of patient baseline characteristics between studies. For example, in the cohort of Yu et al., only 27.2% patients had T1 tumors, 52.4% patients had node-positive diseases, and 47.4% were Luminal A 13 . In comparison, 55.4% of our enrolled population had T1 tumors, 42.5% reported node involvement, and only 13.4% were Luminal A.
The influence of chemotherapy delay on survival was subtype-dependent 9,10 . For example, Colleoni et al. found that early TTC could improve the survival of premenopausal ER-negative BC patients rather than those with ER-positive tumors 30 . Notably, several recent studies linked late TTC to poor prognosis in TNBC patients 11,12,31 . Li et al. reported that a prolonged TTC >60 days significantly decreased the RFS in TNBC patients 11 . Another recent study showed that longer TTC was associated with worse OS in HR-negative patients, not in HR-positive ones 31 . In addition, TTC ≥91 days was shown to be associated with an increased risk in BC-specific death only in TNBC patients (HR, 1.53, 95% CI, 1.17-2.07), but not in patients with HR-positive or HER2-positive tumors 12 . Consistent with these findings, our data showed a substantially detrimental effect of adjuvant chemotherapy delay on the RFS (HR, 1.89, 95% CI, 1.09-3.27, P = 0.024) and OS (HR, 2.49, 95% CI, 1.30-4.76, P = 0.006) of TNBC patients. Since TNBC tumor had relatively aggressive biology, rapid proliferation, and relevant sensibility to chemotherapy 32,33 , TNBC patients were expected to derive most benefit from chemotherapy [34][35][36] . The results of our study strongly suggested an early initiation of adjuvant chemotherapy in TNBC patients. However, we didn't find a substantial survival difference between TTC groups in patients with HER2-positive diseases, while it was previously reported that a TTC ≥61 days led to a worse survival (HR, 3.09; 95% CI, 1.49 to 6.39) compared to a TTC <30 days in HER2-positive subtype 10 . This could also be explained by the improved clinical outcomes with the application of combined anti-HER2 targeted therapy [37][38][39] . Moreover, delayed TTC was not associated with poor survival outcomes in patients with HR-positive tumors, which adhered to previous studies 9,10,12 . Potential explanations included efficient endocrine therapy and relatively superior survival in HR-positive patients 3,40-43 .
To our knowledge, this is the largest study in Chinese population to provide insights into the influence of TTC on the clinical outcomes of BC patients, and to evaluate the optimal treatment interval for TNBC subtype. Our highlights lied in the fact that we reported a large series of consecutive patients, and we performed subgroup analysis for all patients according to clinico-pathological characteristics and molecular type. We managed to reveal that TN patients had a poorer prognosis with a TTC of more than 4 weeks. Second, the database we applied, SJTU-BCDB, had a follow-up rate of 99.98% and an error rate of 0.01-0.02%, guaranteeing that our results were creditable and consistent with the real-world practice. Thirdly, all the patients enrolled had gone through multidisciplinary discussion and had received standardized adjuvant treatment, so as to avoid the bias caused by irregular treatment. As a result, our finding was important to shed light on a better treatment schedule for Chinese patients, especially for TNBC patients, who would better initiate chemotherapy within 4 weeks after surgery, to maximize treatment efficacy. Nevertheless, several limitations should be considered when interpreting our results. First, given the non-randomized and retrospective nature of the study design, the current study had potential bias and needed to be further validated in other cohorts and larger population. Second, the baseline characteristics of patients treated in the two study centers weren't well balanced, since Jiaxing Center was a specific breast health center which treated younger patients with fewer comorbidities and earlier diseases, while Shanghai Center was a comprehensive center which treated elder patients with more comorbidities and more locally advanced diseases. We here presented a joint analysis of data because our two study centers had a long-term cooperation in clinical practice, had similar multidisciplinary management procedure and shared the same breast cancer database. As a result, the treatment that patients received was comparable between these two centers. Furthermore, no data regarding patient germline variant was currently available in this study and further study was warranted to give more information on genetic risk of the patients. Meanwhile, thanks to the application of systemic treatment, patients had relatively superior survival, and the recurrent events were relatively insufficient, which warrants a continuous long-term follow-up. On the other hand, further analysis should be conducted to identify the most optimal cutoff of TTC for TNBC patients as well as other molecular subtypes.

conclusions
In conclusion, prolonged TTC >4 weeks after BC surgery was not associated with worse survival outcomes in the whole BC patients. However, TTC >4 weeks may increase the risk of relapse or death in TNBC patients. Early initiation of adjuvant chemotherapy should be encouraged for TNBC patients, which deserves further clinical evaluation.