Introduction

For patients diagnosed with triple-negative or human epidermal growth factor receptor 2 positive (HER2 +) breast cancer, anthracycline-based neoadjuvant chemotherapy is considered one of the standard treatment options. Despite being effective, the use of anthracycline is limited due to its noticeable cardiotoxicity [1, 2]. As a result, researchers have been exploring anthracycline-sparing regimens. Previous studies have shown that adding platinum to anthracycline-based regimens or using platinum-based regimens can increase these patients’ pathological complete response (pCR) rate compared to anthracycline-based regimens [3, 4].

In our center, consistent results were obtained from two prospective randomized controlled trials, neoCARH and neoCART [5, 6]. In the neoCARH trial, patients with HER2 + early-stage breast cancer (EBC) achieved a higher pCR rate with the neoadjuvant TCbH regimen compared to the EC-TH regimen (55.9% versus (vs) 37.3%, P = 0.032). Similarly, the neoCART trial found that the neoadjuvant TCb regimen resulted in a higher pCR rate than the EC-T regimen in patients with triple-negative EBC (61.4% vs 38.6%, P = 0.004). The management of treatment-related adverse events (AEs) in patients receiving platinum-based regimens was also found to be acceptable compared to anthracycline-based regimens. Considering these positive findings, the National Comprehensive Cancer Network (NCCN) guidelines currently still endorse the results of the neoCART trial [7].

However, treatment-related AEs are objective measures that are evaluated by clinicians or researchers, and there is a lack of subjective feedback directly from patients. Patient-reported outcomes (PROs) provide direct reports from patients about their health status and treatment-related experiences without modification by clinicians or others [8, 9]. The Food and Drug Administration (FDA) recognizes the potential value of core PROs data in the benefit-risk assessment of anticancer drugs and has incorporated PROs as evidence for drug approval [10,11,12]. Moreover, numerous trials, such as Destiny-Breast03, SOLTI CORALLEEN, ASCENT, KEYNOTE-407, and so on, have integrated PROs as secondary or exploratory endpoints to enhance patients' quality of life (QoL) while ensuring the efficacy of anticancer drugs [13,14,15,16,17,18,19,20,21].

The number of breast cancer survivors is increasing, with a 5-year survival rate of over 90% for early-stage patients [22]. Both the American Society of Clinical Oncology (ASCO) Breast Cancer Survivorship Care Guidelines and the European Society for Medical Oncology (ESMO) Cancer Survivorship Expert Consensus emphasize the importance of assessing the long-term impact of treatment on patients’ physical, psychosocial, and social functioning, to provide high-quality care and improve their long-term QoL [23, 24]. Hence, it is crucial to evaluate PROs following treatment.

In this study, we present the health-related quality of life (HRQoL) data collected from the neoCARH and neoCART trials. Our primary objective was to compare the HRQoL differences between neoadjuvant chemotherapy regimens of platinum-based (TCb or TCbH) and anthracycline-based (EC-T or EC-TH) among patients diagnosed with triple-negative or HER2 + EBC during (post-neoadjuvant chemotherapy) and after (1-year, 2-year post-diagnosis) primary treatment.

Materials and methods

Study design

The neoCARH (ClinicalTrials.gov NCT03140553) and neoCART (ClinicalTrials.gov NCT03154749) are randomized control, multicenter, phase II trials conducted in our center. In the neoCARH trial, patients with clinical stage II–IIIC HER2 + breast cancer were randomized in a 1:1 ratio to either the EC-TH group, which involved four cycles of intravenous epirubicin (90 mg/m2) and cyclophosphamide followed by four cycles of docetaxel and trastuzumab every 3 weeks, or the TCbH group, which consisted of six cycles of docetaxel plus carboplatin (area under the curve, 6 mg/ml per min) administered every 3 weeks concurrently with trastuzumab. Similarly, in the neoCART trial, patients with clinical stage II–III triple-negative breast cancer were randomly assigned in a 1:1 ratio to either the experimental TCb (carboplatin, area under the curve, 6 mg/ml per min) group or the EC-T (epirubicin, 90 mg/m2) group. The primary endpoint for both trials is the pCR rate. Previous publications have provided detailed information about these trials [5, 6]. In this study, patients were categorized into the anthracycline-based group (EC-T or EC-TH) or the platinum-based group (TCb or TCbH) based on the regimens they received.

The Research Ethics Committee of the Guangdong Provincial People’s Hospital, Guangdong Academy of Medical Sciences approved both of these trials, and the protocols underwent review by the respective ethics committees at each center. Both trials were performed in accordance with the ethical standards as laid down in the 1964 Declaration of Helsinki and its later amendments or comparable ethical standards, and all participants provided written informed consent.

Prespecified exploratory endpoint

HRQoL is a prespecified exploratory endpoint in the neoCARH and neoCART trials. The assessment time points for HRQoL include T0 (baseline), T1 (1 week after the last dose of neoadjuvant chemotherapy), T2 (1 year after diagnosis), and T3 (2 years after diagnosis). HRQoL assessments utilize the validated and widely used European Organization for Research and Treatment of Cancer Quality of Life-Core30 (EORTC QLQ-C30) questionnaire (oncology-specific), along with the EORTC QLQ- Breast23 (BR23) questionnaire (breast cancer-specific). The C30 questionnaire consists of 30 items arranged into 15 domains: a 2-item global health status (GHS)/QoL domain, 5 multi-item functioning domains, 3 multi-item symptom domains, and 6 single-item symptom domains [25]. Similarly, the BR23 questionnaire consists of 23 items arranged into 8 domains: 4 multi-item functioning domains and 4 multi-item symptom domains [26]. The questionnaires were scored according to the scoring manuals of EORTC QLQ-C30 and BR23 [27, 28]. The C30 summary score (C30-SumSc) was calculated by taking the mean of the scores from the 13 domains, except for the GHS/QoL and financial difficulties domains (the scores from the symptom domains were reversed before calculation) [29]. For the C30-SumSc, GHS/QoL, and functioning domains, scores ranged from 0 to 100; higher scores indicate better HRQoL. Scores for the symptom domains also ranged from 0 to 100, but higher scores indicate worse symptomatology. The primary variable of interest is the C30-SumSc [29]. Secondary variables of interest include GHS/QoL, physical functioning, role functioning, emotional functioning, social functioning, fatigue, pain, diarrhea, constipation, appetite loss, and systemic therapy side effects of the C30 and BR23 questionnaires. These variables were chosen based on their relationship with typical drug-related AEs, their inclusion as core PROs data that the FDA advises collecting, or their identification as important PRO domains [5, 6, 10, 30].

The left ventricular ejection fraction (LVEF) evaluation

LVEF of patients was evaluated through color Doppler echocardiography at baseline and after neoadjuvant chemotherapy.

Statistical analyses

The statistical analysis was conducted using Statistical Product and Service Solutions version 26.0 (SPSS 26.0). The GraphPad Prism 8 was used to create the figures. The mean least squares (MLS) changes from baseline and between-group differences in MLS change from baseline for the scores of the C30 and BR23 questionnaires were calculated using linear mixed-effect models for repeated measures (MMRM). The model considered groups (the platinum-based group and the anthracycline-based group), time points (modeled as a categorical variable, including T0, T1, T2, and T3), baseline score, the interaction between baseline score and time points, and the interaction between groups and time points as fixed effects. Only patients who had HRQoL data at baseline and at least one subsequent time point were included in the analysis. The definition of completion rate is the percentage of patients who complete more than half of the items in each domain of the EPRTC QLQ-C30 questionnaire and ≥ 1 item of the EPRTC QLQ-BR23 questionnaire at each time point out of the number of patients randomized. The compliance rate was defined as the percentage of patients who complete more than half of the items in each domain of the EPRTC QLQ-C30 questionnaire and ≥ 1 item of the EPRTC QLQ-BR23 questionnaire at each time point over the number of patients who were expected to complete the HRQoL assessment, excluding patients who were missing by design (i.e., death, discontinuation). In the C30 and BR23 domains, the absolute value of the MLS changes from baseline of ≥ 10 points was deemed clinically relevant [31]. Furthermore, referring to a previous study [19], the absolute value of the difference in MLS changes from baseline (MLSCFB) between groups was considered clinically significant if it exceeded the minimal important difference (MID) derived from evidence-based guidelines for the C30 questionnaire [32], or calculated as 0.3 × standard deviation for the mean change score between baseline and a subsequent evaluation time point [33]. Missing data for C30 and BR23 questionnaires’ scores were analyzed using multiple imputations. In parallel, the chi-squared test or Fisher’s exact test was employed for categorical variables, while the two-sample t test or Mann–Whitney U test was utilized for continuous variables. To deal with the clustering effect of the neoCARH and neoCART tests, we carried out the following sensitivity analysis. First, the two trials were taken as stratified factors and incorporated into the linear mixed-effect model as random effects. Second, the two trials were analyzed as subgroups respectively. In the sensitivity analysis, we only focus on the primary variable of interest, namely the C30-SumSc. Since HRQoL was not included in the hierarchical testing plan for the neoCARH or neoCART trials, the reported P values are considered nominal, and there was not adjusted for multiple comparisons. All tests were conducted using a two-sided approach with 95% confidence interval (CI). A significance level of < 0.05 was deemed statistically significant.

Results

Baseline characteristics

The process by which patients were enrolled in this study is depicted in Fig. 1. From September 1, 2016, to December 31, 2019, a total of 223 patients from 9 centers were randomly assigned to either the anthracycline-based group (n = 111) or the platinum-based group (n = 112). Table 1 demonstrates that the baseline characteristics of patients were well-balanced between the two groups.

Fig. 1
figure 1

Flow diagram for patients enrolled in this HRQoL study. Abbreviation. HRQoL, health-related quality of life; HER2, human epidermal growth factor receptor 2; EC-TH, four cycles of epirubicin and cyclophosphamide, followed by four cycles of docetaxel and trastuzumab every 3 weeks; EC-T, four cycles of epirubicin and cyclophosphamide, followed by four cycles of docetaxel; TCbH, docetaxel, carboplatin, plus trastuzumab administered every 3 weeks for six cycles; TCb, docetaxel plus carboplatin administered every 3 weeks for six cycles

Full size image
Table 1 Baseline characteristics of all patients and patients in the anthracycline-based or platinum-based group
Full size table

Completion rate, compliance rate, and baseline score of questionnaires

The average completion rate of the EORTC QLQ-C30 and EORTC QLQ-BR23 questionnaires across all follow-up periods was 90.9%. Specifically, the completion rates were 97.3% at baseline (217/223), 94.2% after neoadjuvant chemotherapy (210/223), 89.2% 1 year after diagnosis (199 /223), and 83.0% 2 years after diagnosis (185/223). The average compliance rate across all follow-up periods was 95%, including 97.3% at baseline (217/223), 98.6% after neoadjuvant chemotherapy (210/213), 94.3% (199/211) 1 year after diagnosis, and 89.8% (185/206) 2 years after diagnosis. There were no statistical differences in the C30-SumSc and the majority of domain scores in the C30 and BR23 questionnaires at baseline between the two groups (Table S1).

Primary variable of interest

The calculated MLSCFB for post-neoadjuvant chemotherapy in the C30-SumSc for the platinum-based group was − 15.997 (95%CI: − 17.877 to − 14.117) and for the anthracycline-based group was − 20.156 (95%CI: − 22.053 to − 18.258). The difference between the two groups was found to be 4.159 (95%CI: 1.462 to 6.855, P = 0.003, MID = 3), indicating both statistically and clinically better HRQoL related to the platinum-based group (Fig. 2 and Table S2). During the 1-year and 2-year follow-ups post-diagnosis, the C30-SumSc gradually returned to the baseline level, and no statistical differences were observed between the two groups.

Fig. 2
figure 2

MLS change from baseline for EORTC QLQ-C30 summary score. EORTC QLQ-C30 summary score range from 0 to 100, a higher score represents better quality of life. Data are from a mixed-effect model for repeated measures analysis. *, P < 0.05 (platinum-based group versus anthracycline-based group). Abbreviation: MLS, mean least square; EORTC QLQ-C30, European Organization for Research and Treatment of Cancer Quality of Life Questionnaire-Core30; CI, confidence interval; T0, baseline; T1, 1 week after the last dose of neoadjuvant chemotherapy; T2, 1 year after diagnosis; T3, 2 years after diagnosis

Full size image

Secondary variables of interest

In terms of functional domains, the platinum-based group had statistically and clinically better physical functioning (difference in MLSCFB: after neoadjuvant chemotherapy: 7.093, 95%CI: 1.664 to 12.522, P = 0.011, MID = 5; 1 year after diagnosis: 5.385, 95%CI: 0.030 to 10.739, P = 0.049, MID = 5) and role functioning (difference in MLSCFB: after neoadjuvant chemotherapy: 12.405, 95% CI: 6.672 to 18.138, P < 0.001, MID = 6; 1 year after diagnosis: 7.912, 95% CI: 2.015 to 13.808, P = 0.009, MID = 6) compared to the anthracycline-based group. Statistically and clinically significant better social functioning was observed after neoadjuvant chemotherapy for the platinum-based group compared to the anthracycline-based group (difference in MLSCFB: 14.483, 95% CI: 8.853 to 20.113, P < 0.001, MID = 5). Additionally, statistical improvement in emotional functioning in the platinum-based group compared to the anthracycline-based group could be seen 1 year (difference in MLSCFB: 6.719, 95% CI: 2.235 to 11.202, P = 0.003, MID = 7) and 2 years after diagnosis (difference in MLSCFB: 5.419, 95% CI: 0.729 to 10.109, P = 0.024, MID = 8) (Fig. 3 and Table S2).

Fig. 3
figure 3

Forest plots of MLS difference between groups for focused EORTC QLQ-C30 domains’ scores. aRange from 0 to 100, a higher score represents better quality of life or functioning. bFor the EORTC QLQ-C30 summary score and emotional functioning, the MID was derived as 0.3 × SD for the mean change score between two visits. Abbreviation: MLS, mean least square; EORTC QLQ-C30, European Organization for Research and Treatment of Cancer Quality of Life Questionnaire-Core30; GHS, global health status; QoL, quality of life; MID, minimal important difference; CI, confidence interval; T1, 1 week after the last dose of neoadjuvant chemotherapy; T2, 1 year after diagnosis; T3, 2 years after diagnosis; SD, standard deviation

Full size image

In terms of symptom domains, patients in the platinum-based group exhibited significantly less fatigue both statistically and clinically after neoadjuvant chemotherapy, in comparison to the anthracycline-based group (difference in MLSCFB: − 17.846, 95% CI: − 25.445 to − 10.247, P < 0.001, MID = 5). Furthermore, it has been consistently observed that the platinum-based regimens result in lower levels of pain at all follow-up time points in comparison to the anthracycline-based group. These differences between the two groups are statistically significant and also hold clinical significance. Overall, patients in the platinum-based group reported statistically fewer side effects from systemic therapy 1-year post-diagnosis compared to the anthracycline-based group (difference in MLSCFB: − 6.084, 95% CI: − 11.734 to − 0.434, P = 0.035, MID = 7). Nevertheless, patients in the platinum-based group experienced statistically and clinically more severe constipation 1-year post-diagnosis when compared to the anthracycline-based group (difference in MLSCFB: 8.457, 95% CI: 2.753 to 14.160, P = 0.004, MID = 5). Additionally, the platinum-based group exhibited more statistically significant appetite loss at 2 years post-diagnosis compared to the anthracycline-based group (difference in MLSCFB: 4.055, 95% CI: 0.556 to 7.554, P = 0.023, MID = 5) (Fig. 4 and Table S2).

Fig. 4
figure 4

Forest plots of MLS difference between groups for focused EORTC QLQ-C30 and BR23 domains’ scores. aRange from 0 to 100, a higher score represents worse symptomatology. bFor the EORTC QLQ-BR23 systemic therapy side effects, the MID was derived as 0.3 × SD for the mean change score between two visits. Abbreviation: MLS, mean least square; EORTC QLQ-C30, European Organization for Research and Treatment of Cancer Quality of Life Questionnaire-Core30; BR23, Breast23; MID, minimal important difference; CI, confidence interval; T1, 1 week after the last dose of neoadjuvant chemotherapy; T2, 1 year after diagnosis; T3, 2 years after diagnosis; SD, standard deviation

Full size image

LVEF

Up until the completion of the neoadjuvant chemotherapy, LVEF data was available for a total of 92.8% (207/223) of the patients. Both groups exhibited high levels of LVEF both at the baseline and after completing the neoadjuvant chemotherapy, with an average LVEF greater than 66%. Additionally, there were no statistical differences in LVEF observed between the two groups (Table 2).

Table 2 LVEF measurements at T0 and T1 for anthracycline-based or platinum-based group
Full size table

Sensitivity analyses for the clustering effects of the neoCARH and neoCART trials

The results of the sensitivity analyses, presented in Table S3, were generally consistent with those of the primary analysis. Specifically, HRQoL was significantly impaired in both the platinum-based and anthracycline-based treatment groups after neoadjuvant chemotherapy. However, the HRQoL of the platinum-based group was relatively better than the anthracycline-based group. Over time, the HRQoL of both groups gradually returned to baseline levels at 1 and 2 years after diagnosis, with no significant difference between the groups.

Discussion

Analysis of the HRQoL data from the neoCARH and neoCART trials [5, 6] revealed that after neoadjuvant chemotherapy, the platinum-based regimens outperformed the anthracycline-based regimens in various domains of the EORTC QLQ-C30 questionnaire including the C30-SumSc, physical functioning, role functioning, social functioning, fatigue, and pain, showing that the platinum-based regimens are related to a better HRQoL compared to the anthracycline-based regimens.

Although AEs related to the neoCARH and neoCART trials have been reported, these are objective indicators assessed by clinicians or researchers. It is important to note that toxicities associated with anticancer therapies, especially subjective symptoms like fatigue and pain, are often overlooked or underestimated by clinicians. However, this information is crucial in clinical practice [34, 35]. At this moment, there is an urgent need for PROs. Including PROs in clinical practice promotes effective communication between patients and clinicians, facilitating early detection and intervention for treatment-related symptoms [34, 36]. Studies have shown that the use of PROs can lead to a reduction in emergency visits and hospitalizations, as well as improvements in symptoms, physical functioning, quality of life, treatment adherence, and overall survival [37,38,39,40]. Furthermore, this information has already been incorporated into the labeling claims of drugs by the FDA, providing supportive evidence for drug approval [11, 12, 35]. Additionally, it is worth noting that breast cancer is the most common malignant tumor in women [41, 42], further highlighting the significance of PROs data in this patient population. Various questionnaires are commonly used to assess PROs, with the EORTC QLQ-C30 and BR23/45 frequently employed in clinical trials on breast cancer. Trials such as Destiny-Breast03, CANTO, ASCENT, and UK TACT have utilized these two questionnaires to assess PROs [13, 14, 17, 19, 43]. In our study, we are also utilizing these two questionnaires.

C30-SumSc provides a comprehensive summary of the C30 questionnaire. Using it as the primary variable of interest can reduce the risk of Type I errors when comparing repeated measurements using the other 15 domains [29]. Additionally, the C30-SumSc has a stronger predictive value for overall survival compared to the other C30 domains [44]. The PRO study of the CANTO trial used the C30-SumSc as its primary endpoint [13, 14]. Similarly, in this study, the C30-SumSc serves as the primary variable of interest. It was found that, after neoadjuvant chemotherapy, patients in the platinum-based group had higher C30-SumSc, which was statistically and clinically significant, in comparison to the anthracycline-based group. However, at 1 year and 2 years after diagnosis, there was no statistical difference in the C30-SumSc between the two groups, and it gradually returned to the baseline level. This finding is consistent with previous studies, which demonstrate that most physical and psychosocial symptoms of treatment typically resolve within the first year after diagnosis, and most breast cancer survivors can regain a higher level of quality of life [45,46,47,48].

For the secondary variables of interest, such as fatigue, pain, and systemic therapy side effects, the platinum-based group exhibited better outcomes compared to the anthracycline-based group at various follow-up time points. However, it is important to note that patients in the platinum-based group also experienced more severe constipation or appetite loss at 1 or 2 years after diagnosis. These HRQoL findings align with previous reports of AEs in the neoCARH and neoCART trials [5, 6]. For example, the incidence rates of fatigue, arthralgia, and bone pain were higher in the anthracycline-based regimens (30.6%, 23.9%, and 43.2%, respectively) compared to the platinum-based regimens (22.3%, 11.8%, and 27.3%, respectively). Similarly, the platinum-based regimens had higher incidence rates of constipation (17%), anorexia (13.2%), and dysgeusia (4.4%) compared to the anthracycline-based regimens (11.7%, 6.8%, and 1.5%, respectively).

The BCIRG 006 trial evaluated the HRQoL in HER2 + breast cancer patients receiving adjuvant chemotherapy using three different treatment regimens: TCbH, AC-T, and AC-TH. The PROs questionnaires utilized in the trial were the EORTC QLQ-C30 and BR23. The findings of the trial indicated that the TCbH group experienced fewer systemic therapy side effects after adjuvant chemotherapy. Overall, there were no statistically significant differences observed in HRQoL among the three regimens [49]. According to our investigation, we also observed that the group receiving platinum-based therapy encountered numerically fewer adverse effects from systemic therapy in comparison to the group receiving anthracycline-based therapy after neoadjuvant chemotherapy.

One of the main AEs associated with anthracyclines is their cardiotoxicity. A study conducted over a median follow-up of 5.2 years observed an incidence rate of 9% for anthracycline-induced cardiotoxicity. The median time to occurrence of this cardiotoxicity was 3.5 months after completion of chemotherapy, with almost all patients (98%) experiencing it within the first year. The incidence rate is positively correlated with the cumulative drug dose, with a rate as high as 36% when the cumulative dose exceeds 601 mg/m2 [1, 2]. Our study only collected LVEF data at baseline and after neoadjuvant chemotherapy. We found that both the group receiving platinum-based and the group receiving anthracycline-based regimens showed high LVEF (average LVEF all > 66%), with no statistical differences between the two groups. However, a longer follow-up period is required to perform a more precise comparison of cardiotoxicity between anthracyclines and platinum.

This study demonstrates a notable strength in its high questionnaire compliance rate, which is consistently maintained at an average of 95.0%. Moreover, even after a 2-year follow-up period, the compliance rate remains at a commendable 89.8%. Nonetheless, the study does have certain limitations that should be acknowledged. Firstly, both the neoCART and neoCART trials utilize a non-blinded design, and patients’ understanding of the treatment assignment may influence their responses to questions on PRO assessments. However, one study has shown that this does not affect patients’ responses to the PRO questionnaires [50]. Secondly, the analyses were not adjusted for multiple comparisons.

Conclusions

Our study focused on patients with triple-negative or HER2 + EBC who received neoadjuvant chemotherapy in the neoCARH and neoCART trials. We found that both patients receiving platinum-based or platinum-based regimens experienced worsened HRQoL after neoadjuvant chemotherapy; however, the former provided relatively better HRQoL compared with the latter, which also showed clinically relevant. These results, along with their higher pCR rate and manageable safety profiles as observed in the neoCARH and neoCART trials, provide supporting evidence for the application of platinum-based regimens in clinical practice.