This study employed a historical cohort design to investigate the differences between two patient groups based on their response to treatment. This study was carried out at two oncology treatment centres in Cameroon: the Yaounde General Hospital (YGH) and the Yaounde Central Hospital (YCH). Both are referral hospitals in Cameroon, located in the administrative capital and serve as teaching hospitals.

This cohort study included patients who were followed up at both study sites from the 1st of January 2019 to the 31st of December 2023; spanning 5 years. Data were collected retrospectively for a duration of 3 months from the 1st of June to the 31st of August 2024.

The target population of the present study was all patients with breast cancer in Yaounde. This study evaluated breast cancer patients with initially non-metastatic breast cancer and who received NACT at the YGH and YCH between the 1st of January 2019 and the 31st of December 2023. Inclusion criteria for good responders were all patients with histologically-confirmed breast cancer, who had staging information, had indication for NACT, and with available data on therapeutic response showing complete, near total or >50% pathological response. For poor responders, all patients with histologically-confirmed breast cancer, who had staging information, had indication for NACT, and with available data on therapeutic response showing <50% pathological response or no response were included. Exclusion criteria included patients with relative or absolute contraindications to standard chemotherapy including pregnant patients and those with heart conditions contraindicating the use of anthracyclines.

Using the Fleiss formula [13] for comparing two proportions, assuming an expected proportion of good responders of 80% and poor responders of 20%, with 80% power and 95% confidence, the calculated minimum sample size was 110 patients.

The Ethical Committee of the Faculty of Medicine and Biomedical Sciences of the University of Yaounde I approved this study. Questionnaires were coded leaving no link between the patient’s record and the questionnaire. The identity and personal details of participants of the study were kept strictly confidential. Moreover, all patient files were examined within the archive of this institution without any tampering or modification of their contents. This study was conducted in accordance with the principles of the Declaration of Helsinki (as revised in 2013) [14].

Patient demographics including age, sex, marital status, profession, and gynecological history were collected. The diagnosis of cancer was based on the pathological confirmation from the files either written and signed by the consultant oncologist or from the histopathological report. Histopathological characteristics including tumour grade, hormone receptor status, Ki67, and human epidermal growth factor receptor 2 (HER2) were noted where available. Staging was noted based on imaging including chest and abdominopelvic CT scan or on chest radiography and abdominopelvic ultrasound; these were done using the Tumour Node Metastasis (TNM) staging method. The chemotherapy regimen used prior to surgery was also recorded, with details of the number of cycles received. The post-operative pathological reports were examined for pathological response using the Sataloff and margin status. The occurrence of an event (progression, relapse, or death) was noted and the dates of these events were recorded. The date of last contact was noted from the files. Data from validated questionnaires was entered into Microsoft excel 2013 spread sheets.

Clinical response was assessed using the response evaluation criteria in solid tumors (RECIST) criteria [15] by measuring tumour size and node size after neoadjuvant chemotherapy. The Sataloff Classification was used to evaluate the pathologic response based on the response of the primary carcinoma and the lymph nodes. The Sataloff classification was used for pathologic response assessment as it is the routine reporting system adopted in our pathology services. Residual Cancer Burden (RCB) and Miller-Payne grading are not systematically implemented in our setting due to limited availability of standardized quantitative tumor bed measurement and dedicated pathology infrastructure. Thus, the Sataloff system was selected for consistency with existing institutional protocols. Those classified as pCR and pPR1 were placed into the good responders’ group and those with pPR2 and pNR were placed in the poor responders’ group (Table 1).

Table 1. Pathologic response category according to Sataloff criteria.

AJCC, American Joint Committee on Cancer; pCR, pathologic complete response; pPR1, pathologic partial response > 50%, pPR2, pathologic response < 50%, pNR, pathologic no response.

Descriptive statistics were employed to summarize demographic, clinical, and pathological characteristics of the study participants. Chi-square tests and Fisher's exact tests were conducted to compare categorical variables between patients with OPR and those without. The independent samples t-test was used to compare continuous variables between these groups. Kaplan-Meier curves were used to estimate OS and EFS for the entire cohort. The log-rank tests were employed to compare survival curves between different response groups. Multivariate Cox regression analysis was performed to assess whether pathologic response to NACT significantly was independently associated with survival outcomes (EFS and OS), adjusting for potential confounding variables. The multivariate Cox regression models included age at diagnosis, civil Status, professional status, menopausal status, histological type, histological grade, molecular subtype, AJCC Stage, surgical technique, surgical margin status, and chemotherapy regimen. For molecular subtype analysis, patients with missing immunohistochemical data were excluded from subtype-specific analyses but retained in overall survival analyses. No imputation method was applied due to the retrospective nature of data collection. Statistical significance was set at p < 0.05.

Our study included 119 female participants: 30 (25.21%) in the good responders’ group (pCR: 10, 8.40% and pPR1: 20, 16.81%) and 89 (74.79%) in the poor responders’ group (pPR2: 82, 68.91% and NR 7, 5.88%). The median age of the overall participants was 47.00 ± 11.44 years (range 28 to 80) years (Table 2). When the mean age was compared between subgroups, good responders were averagely one year older than poor responders, albeit without statistical significance (48 ± 12 versus 47 ± 11.26, p = 0.75). The average age at menarche was 13.34 ± 2.17 (range 9 to 18) years, while that at menopause was 50.41 ± 3.72 (range 42 to 59) years.

Table 2. Sociodemographic characteristics of the participants per subgroup.

Percentages show distribution of characteristics within good and poor responders.

Triple-negative breast cancer was the most common subtype (33, 42.31%), while non-luminal HER2+ was the least represented (3, 3.85%). The majority (60, 61.22%) had grade 2 tumours. Poor responders had a higher proportion of histological grade III than good responders (26.58% versus 15.79%, p = 0.01, Table 3).

Table3. Clinical and pathobiologic characteristics of the study participants.

IHC, Immunohistochemistry; HER2+, human epidermal growth factor receptor 2 positive.

All participants received at least one line of chemotherapy, while 76 (63.87%) received an additional cycle (NACT cycle 2). The AC plus taxane combination was the most common NACT protocol used (64, 53.78%), and most (114, 95.80%) underwent a radical surgical procedure (mastectomy) (Table 4).

Table4.Treatment modalities observed in the study population.

AC, Doxorubicin plus Cyclophosphamide; NACT, neoadjuvant chemotherapy.

The clinical response after NACT showed that a majority of the participants had a partial clinical response (69, 57.98%) followed by complete response (32, 26.89%). Notably, a higher proportion of good responders had complete clinical response albeit without statistical significance (33.33 % versus 24.72 %, p = 0.61). Good responders had a significantly higher proportion of negative surgical margins than poor responders (83.33% versus 56.18%, p = 0.01).

The median follow-up period was 27 months. At the end of the study, 31 (26.05%) participants had died and up to 47 (39.5%) participants had disease progression.

The comparison of EFS between the two groups revealed a significantly better median EFS in the good responders; 41.91 ± 2.43 months (95% CI: 35.64 - 45.15) versus 19.86 ± 5.26 months (95% CI: 9.49 - 30.11), p = 0.00 (Figure 1).

Figure1. Event-free survival with respect to treatment-response subgroups.

When overall survival was compared between the two groups, the good responders had a better median overall survival; 6.20 ± 0.27 years (95% CI: 4.10 - 8.15) versus 3.26 ± 0.47 years (95% CI: 2.34 - 4.18), p = 0.12, albeit without statistical significance, as shown in Figure 2.

Figure2. Overall survival with respect to treatment response groups.

In the multivariate Cox regression analysis investigating the association of pathologic response to NACT and survival duration, pathological response to NACT was not significantly associated to EFS (p = 0.79) or OS (p = 0.37) (Table 5). However, the use of AC plus taxane was independently associated with a longer EFS (HR: 0.23, p = 0.01) and OS (HR: 0.10, p = 0.00). A positive surgical margin was associated with shorter OS (HR: 1.81, p = 0.03).

Table 5. Cox regression model of pathological response and survival.

EFS, Event-free survival; OS, Overall survival.