Systemic therapies are necessary for patients with early breast cancer, and adjuvant chemotherapy has been shown to decrease mortality due to breast cancer (1). However, patients with breast cancer who have undergone adjuvant chemotherapy have often suffered from adverse effects (AEs), including nausea, diarrhea, peripheral neuropathy, stomatitis, taste disorder, and insomnia, which significantly impact the patients’ quality of life (QOL) and their treatment adherence. Assessments of AEs are conducted using the Common Terminology Criteria for Adverse Events (CTCAE), but these assessments are often based on the clinician’s evaluation, leading to a tendency to underestimate subjective symptoms such as nausea, dysgeusia, insomnia, and neuropathy (2-5).

Patient-reported outcomes (PROs), collected through structured questionnaire surveys, have emerged as a complementary tool that captures patients’ direct experience of symptoms, thereby providing more accurate and patient-centered toxicity data (2, 3). Discrepancies between physician-reported and patient-reported AE profiles in breast cancer chemotherapy have been identified (2, 3). Large-scale surveys in Japan have also highlighted unmet patient needs regarding AE management and the importance of pretreatment information on the expected course and impact of AEs (3). However, few studies have prospectively monitored the course and persistence of chemotherapy-induced toxicity by using PROs after the end of treatment. Our research group has conducted a prospective observational study on chemotherapy-induced alopecia (which significantly impairs patients’ QOL), and we used PROs to evaluate the course of alopecia according to the treatment regimen and patients’ age (6-9). In the present study, we evaluated chemotherapy-induced AEs by obtaining PROs via a structured questionnaire, focusing on outcomes after the completion of chemotherapy and 6 months later in order to examine the frequency and progress of the AEs.

Patients. This single-center prospective observational study included 115 patients with primary breast cancer who underwent perioperative chemotherapy at Gunma University (Maebashi, Japan) between January 2016 and March 2023, based on the approved study protocol. This study was approved by the ethics committee of Gunma University (approval no. 1405, January 28, 2016). The oncologic management for all of the patients included breast surgery with the appropriate axillary lymph node procedures, including the lymph node sentinel procedure and lymph node dissection, radiotherapy, chemotherapy, hormonal therapy according to the patient’s age and the tumor’s hormonal status, and a trastuzumab ± pertuzumab regimen for human epidermal growth factor receptor 2 (HER2)-positive breast cancer.

Treatment. The patients received a standard anthracycline- and taxane (A+T)-based chemotherapy regimen, e.g., FEC (5-fluorouracil, epirubicin, and cyclophosphamide) or AC (adriamycin and cyclophosphamide) followed by docetaxel or paclitaxel, or docetaxel plus cyclophosphamide (TC). The HER2-positive patients received additional trastuzumab ± pertuzumab therapy. The chemotherapy protocols were as follows. The FEC regimen was a combination of epirubicin 100 mg/m², cyclophosphamide 500 mg/m², and 5-fluorouracil 500 mg/m² for four cycles every 3 weeks. The AC regimen was a combination of doxorubicin 60 mg/m² and cyclophosphamide 600 mg/m² every 2-3 weeks for a total of four cycles and docetaxel 75 mg/m² every 3 weeks for four cycles or weekly paclitaxel 80 mg/m² for 12 cycles. The TC regimen was a combination of docetaxel 75 mg/m² and cyclophosphamide 600 mg/m² for four cycles every 3 weeks. Only one patient (scheduled for the A+T regimen) could not receive taxane due to AEs, resulting in the administration of AC alone.

Outcomes (patient-reported adverse events). PROs were collected via a structured self-administered questionnaire designed specifically for this study. The questionnaire was provided to each patient at the completion of her chemotherapy regimen and then again 6 months later. The end-of-treatment questionnaire assessed the presence of nausea, vomiting, oral mucositis, constipation, diarrhea, dysgeusia, insomnia, neuropathy, and nail loss. PROs were collected via a structured self-administered questionnaire developed for this study. The questionnaire focused on symptomatic AEs that commonly occur during perioperative breast cancer chemotherapy and that can be self-reported (nausea, vomiting, oral mucositis, constipation, diarrhea, dysgeusia, insomnia, peripheral neuropathy, and nail loss). For each AE, response options were written in patient-friendly language and anchored to the descriptive criteria of the Common Terminology Criteria for Adverse Events (CTCAE; version 4.0), such as frequency (e.g., vomiting episodes/day) and functional impact (e.g., interference with oral intake or activities of daily living). Because CTCAE was originally developed for clinician-reported grading, we used a CTCAE-informed grading scheme for patient self-reporting rather than clinician-assigned CTCAE. The questionnaire was reviewed by breast oncologists and oncology nurses for content/face validity and feasibility before study initiation. The questionnaire was administered at completion of the chemotherapy regimen and again 6 months later using the same grading scheme.

Clinical variables. Clinical and pathological data were collected from medical records: the patient’s age, tumor TNM stage, estrogen receptor (ER) expression status, progesterone receptor (PgR) expression status, and the HER2 score of the primary tumor. The ER and PgR statuses were assessed by using the Allred score ≥3 points, which indicates ER and PgR positivity (10, 11). HER2 overexpression was determined by an immunohistochemistry (IHC) analysis and a fluorescence in situ hybridization (FISH) analysis with IHC 3+ or IHC 2+/FISH+ scores indicating HER2 positivity (10). Breast cancer subtypes were defined based on the patient’s ER, PgR, and HER2 status as luminal-type breast cancer (ER-positive, HER2-negative), luminal-HER2 type (ER-positive and HER2-positive), HER2 type (ER-negative and HER2-negative), and triple-negative (ER-negative and HER2-negative) breast cancer.

Statistical analysis. We conducted a univariate statistical analysis using Fisher’s exact test or the χ²-test with Yates’ correction. Continuous variables such as age were evaluated for normality using the Shapiro-Wilk test. Since the distribution of age in both groups did not follow a normal distribution, intergroup differences were assessed using the Mann–Whitney U-test. A p-value <0.05 was considered significant. Longitudinal changes between the end of the chemotherapy regimen and the 6-month timepoint were evaluated.

Informed consent. Informed consent was obtained from all individual participants included in the study.

Patient characteristics. Table I provides the details of the 115 female breast cancer patients who received perioperative chemotherapy during the study period. The median age was 50 years (range=24-76 years), and 48 patients (41.7%) underwent neoadjuvant chemotherapy (NAC). The breast cancer subtypes were luminal in 56 patients (48.7%), luminal-HER2 in 30 patients (26.2%), pure HER2 in 12 patients (10.4%), and triple-negative in 17 patients (14.8%). The treatment regimens were A+T in 53 patients (46.1%), A+T plus anti-HER2 therapy in 39 patients (33.9%), TC therapy in 19 patients (16.5%), TC therapy plus anti-HER2 therapy in 3 patients (2.6%), and AC therapy alone in a single patient (0.9%).

Incidence and severity of adverse events. Table II shows the incidence of AEs in all 115 cases. The frequency of AEs at the end of the chemotherapy regimens, across all grades, was as follows: nausea 68.7%, vomiting 17.4%, oral mucositis 53.0%, constipation 73.0%, diarrhea 51.3%, dysgeusia 87.0%, insomnia 77.4%, peripheral neuropathy 86.1%, and nail loss 39.1%. The AEs with particularly high frequencies were dysgeusia, peripheral neuropathy, insomnia, and constipation. At 6 months post-chemotherapy, the incidence rates for all grades were: nausea 17.4%, vomiting 1.7%, oral mucositis 19.1%, constipation 40.9%, diarrhea 20.9%, dysgeusia 47.8%, insomnia 45.2%, peripheral neuropathy 64.3%, and nail loss 52.2%. Although gastrointestinal symptoms such as nausea and oral mucositis had decreased, constipation remained relatively common. Neuropathy, dysgeusia, insomnia, and nail loss were still frequently observed at the 6-month timepoint. Grade 3 symptoms persisted for neuropathy, and nail loss was confirmed to have increased at 6 months, emphasizing the importance of long-term PRO monitoring.

Comparison between regimens (TC vs. A+T). Table III summarizes the background characteristics of the TC group and the A+T group. TC chemotherapy is administered as postoperative adjuvant therapy, and no cases of NAC were observed in this patient series. The A+T group included more cases with higher stages and a higher proportion of HER2-positive and TN cases. In the comparison of the groups’ AEs (Table IV), peripheral sensory neuropathy also occurred more frequently and persisted longer in the A+T group (71.0% vs. 36.4% at 6 months). The incidence of nail loss in the A+T group was relatively higher at 6 months than at the end of chemotherapy (55.9% vs. 34.4%). The incidence of AEs by grade is detailed in Table V. Most of the symptoms were Grade 1-2, but persistent Grade 3 neuropathy and dysgeusia were observed in some patients who had received A+T therapy.

The results of our analyses of 115 patients demonstrated that PRO-based monitoring, specifically through a structured patient questionnaire aligned with the CTCAE, provides important insights into chemotherapy-induced symptoms after perioperative treatment for breast cancer. The design of the questionnaire allowed for systematic collection of patient-reported symptom severity and recovery timing across multiple AEs, including nausea, vomiting, oral mucositis, constipation, diarrhea, dysgeusia, insomnia, neuropathy, and nail loss. The persistence of AEs beyond 6 months underscores the necessity of continuing toxicity monitoring beyond the treatment period. This study revealed that patients continue to suffer long-term symptoms such as peripheral neuropathy, nail loss, dysgeusia and insomnia even after their chemotherapy ends, which can significantly impair patients’ QOL and daily functioning (12-15).

Persistent chemotherapy-induced peripheral neuropathy (CIPN) is a significant problem for patients. CIPN has been shown to significantly affect patients’ QOL (12, 13). Rivera et al. emphasized the impact of CIPN on treatment discontinuation (16), and other researchers reported that CIPN is associated with psychological distress and falls (17). Our group recently demonstrated the potential efficacy of mirogabalin in managing taxane-related CIPN during perioperative breast cancer chemotherapy (18). These findings confirm that effective interventions are both necessary and feasible through symptom monitoring.

In addition to neuropathy, our findings revealed the long-term persistence of nail disorders among the patients. Nail changes induced by taxane-based agents are highly prevalent, occurring in up to 89% of patients, with functional impairment reported in approx. 32% (14). It has been reported that nearly half of paclitaxel-treated patients developed Grade 2 nail toxicity, including painful hemorrhagic onycholysis, which significantly impacted the patients’ daily lives (19). Our findings corroborate these observations, underscoring the importance of (i) providing accurate information to patients regarding nail toxicity and (ii) systematic PRO-based assessment.

Insomnia emerged as another important symptom persisting 6 months post-treatment in our patient population. Ancoli-Israel et al. reported that 30%-75% of cancer patients experience sleep disturbances and that chemotherapy frequently exacerbates insomnia (20). A resent prospective study of breast cancer patients undergoing chemotherapy confirmed that the severity of insomnia is strongly associated with impaired QOL (15). Our findings align with these reports, highlighting the importance of monitoring sleep disturbances as part of supportive care strategies.

We also observed that gastrointestinal (GI) symptoms such as nausea, diarrhea, and constipation, though prevalent during active chemotherapy, generally resolved within months after the chemotherapy’s completion. The Chemo-Gut Study conducted by Deleemans et al. demonstrated that while some survivors report persistent GI complaints, the majority experience a resolution of acute GI symptoms over time (21). Similarly, gastrointestinal symptoms were reported to be only rarely persistent in long-term breast cancer survivors (22). These findings are consistent with our present results, suggesting that although the management of patients’ GI symptoms is crucial during active therapy, long-term follow-ups should prioritize persistent symptoms such as neuropathy, nail disorders, and insomnia. In this study, among gastrointestinal symptoms, dysgeusia was found to be relatively common and potentially persistent. Evaluating dysgeusia may be important for nutritional management in early-stage breast cancer patients undergoing chemotherapy (23, 24).

Our study is among the few prospective investigations using PROs to capture chemotherapy symptoms longitudinally. The utility of PRO-CTCAE to characterize long-term AEs in breast cancer survivors has been described (25), supporting the validity of our methodology. By explicitly using a questionnaire as a PRO tool, our study ensured that the patients’ voices were central in the evaluation of treatment-related AEs. Electronic PRO (ePRO) systems represent the next step in AE monitoring. They allow real-time symptom reporting, integration into clinical workflows, and automated alerts for clinicians. A recent systematic review confirmed the value of ePRO systems in cancer care, describing improved monitoring capabilities and patient engagement (26). In addition, digital follow-up methods may be cost-effective alternatives to traditional surveillance for breast cancer (27), enhancing both efficiency and patient-centered care.

This study had several limitations. The number of cases was relatively small (n=115); however, the data were based on self-reported questionnaires, and while sampling bias may have occurred, there was no researcher bias at the time of sampling. Moreover, the study results were derived from patients’ subjective outcome reports, not from objective findings by researchers. It should be noted that CTCAE was designed as a clinician-reported grading framework. For patient self-reporting, the National Cancer Institute developed and validated the Patient-Reported Outcomes version of the CTCAE (PRO-CTCAE) (28), which captures symptomatic toxicities using patient-reported frequency, severity, and interference items. In addition, the questionnaire was developed for this study and used a CTCAE-informed grading scheme for patient self-reporting; it was not psychometrically validated against PRO-CTCAE or clinician-assigned CTCAE, which may have introduced measurement error. Future studies involving a larger number of cases and using an ePRO system are necessary.

In conclusion, the PRO-based monitoring of patients treated with perioperative chemotherapy revealed persistent symptoms, especially neuropathy, nail changes, and insomnia, while gastrointestinal symptoms generally resolved after treatment. Importantly, adverse events persisted even at 6 months after the completion of the patients’ chemotherapy regimens, indicating that monitoring should continue beyond the completion of treatment. Integrating PROs, or potentially ePRO systems, into clinical practice could enhance patient-centered care and improve supportive care in breast cancer chemotherapy.

The Authors have no conflicts of interest to declare.

TF analyzed data and wrote the initial draft of the manuscript. KI, MA, KT, MO and SO collected data and were involved in the initial study conception and design. TF, KI, MA, KT, MO and HT were involved in drafting and revising the manuscript. All Authors have read and approved the final manuscript.

We thank the Gunma University School of Nursing for their cooperation.

During the preparation of this manuscript, a large language model (DeepL) was used solely for language editing and stylistic improvements in select paragraphs. No sections involving the generation, analysis, or interpretation of research data were produced by generative AI. All scientific content was created and verified by the authors. Furthermore, no figures or visual data were generated or modified using generative AI or machine learning–based image enhancement tools.