Cancer Occurrence and Associated Factors to Malignancy in BIRADS-3 Lesions in Yaoundé: The Need to Be More Proactive for Patients above 40, Non-Compliant with Imaging Follow-Up and Presenting with Nipple Discharge ()
1. Introduction
Breast cancer is the leading cause of cancer-related death in women in both developed and developing countries [1] . In sub-Saharan Africa, it comes in second place after cervical cancer [2] . In Cameroon, the WHO classified breast cancer as the first cancer in women with a prevalence of 35.1% followed by cervical cancer (14.9%) [3] [4] . The diagnosis of breast cancer involves several paraclinical investigations including medical imaging and histopathology [5] . The combination of mammography and ultrasound has a high sensitivity in the detection of malignant breast pathologies [6] . The results of these explorations are usually presented in a standardised classification to ensure uniformity of interpretation and imaging reports: the Breast Imaging-Reporting and Data System (BIRADS) of the American College of Radiology (ACR). This classification, in category 3 (BIRADS-3) includes probably benign lesions which need a short-term imaging follow-up. According to the guidelines, the predictive value of malignancy in this category is between 0% and 2% [7] .
In Africa, some studies conducted in Burkina Faso, Gabon and Cameroon have shown a higher frequency of malignancies in this category than that described in the guidelines with values ranging from 3.5% to 13% [5] [8] [9] [10] . A south-Korean study reported in 2014 showed that 40% of occult cancers in this category were detected at the second check-up [11] . Assessing cancer occurrence in BIRADS-3 lesions could be important to a better management approach, especially in our specific setting where medical imaging programs are not subsidised and where patients have to support themselves the financial costs induced by their follow-up. Indeed, the lesions graded BIRADS-3 remain insufficiently evaluated during imaging check-ups in our context. We therefore carried out this study to assess the cancer occurrence and associated factors in BIRADS-3 lesions during the follow-up in order to propose an adaptation of the management for this category in our setting.
2. Patients and Methods
2.1. Study Design and Patients
We conducted a retrospective longitudinal study with 395 female patients whose lesions were initially classified in mammography and/or ultrasonography as BIRADS-3 between January 2014 and December 2022 in five hospitals which are representative of the senological activity of the Yaoundé city: the Gynaeco-Obstetric and Paediatric Hospital, the Central Hospital, the University Teaching Hospital, the General Hospital and the Cathedral Medical Centre. Patients’ files with lesions classified as BIRADS-3 by their radiologists in mammography and/or in breast ultrasound during the study period and who realized each at least one additional imaging check-up were included.
2.2. BIRADS Assessment and Data Collection
The BIRADS grades were assessed by the radiologists on duty in the different centres involved, using Siemensâ and General Electricâ’s devices for mammography and breast ultrasound according to ACR classification [7] . The studied variables included age, clinical history, breast symptoms, imaging aspects, BIRADS grades during the follow-up, histopathological founding concerning the patients upgraded to BIRADS-4 and the factors correlated with the presence of malignancies.
2.3. Statistical Analyses
Using an encoded datasheet, all collected clinical, imaging and histopathological data were entered into a Microsoft Excelâ 2019 worksheet and analysed using the “Statistical Package for Social Sciences (IBMâ SPSSâ 21.0)” software for MacBookâ. The comparison of variables and associations were tested by Chi-square and Mann-Whitney U tests. For all statistical analyses, a p-value less than 0.05 was considered to indicate a statistically significant difference.
2.4. Ethics Statement
This study was approved by the institutional competent ethics committee by Clearance Number 226/UY1/FMSB/VDRC/CSD and conducted in accordance with the declaration of Helsinki and the guidelines of good clinical practice issued by the International Conference on Harmonization (ICH).
3. Results
3.1. General Characteristics of Sample at Baseline
As shown in Table 1, a total of 395 files of women patients who had undergone breast imaging with lesions classified as BIRADS-3 were compiled. The patients were aged 13 to 73 (33.0 ± 13.4) years, 68.1% (269 patients) were under 40 years. The main clinical manifestations were the breast mass (315 cases, 79.7%), pains (25 patients; 6.3%) and nipple discharge (20 cases; 5.1%). A family past history of breast cancer encountered for 25 patients (6.33%) and hormonal contraception was found in 26.3%. At baseline, the main imaging abnormalities included mammogram’s opacities (256 cases; 64.8%) as shown in Figure 1A, and microcalcifications (192 cases; 48.6%; Figure 1B). Initial breast ultrasound showed solid masses with posterior echo enhancement (Figure 1C) in 304 patients (77.0%) and cystic lesions (44 cases; 11.1%; Figure 1D). The isolated right-sided involvement was the most common (197 cases; 49.9%) versus 167 cases (42.3%) of single left-sided involvement and 31 cases (7.8%) of bilateral involvement. In overall breasts, the superior-external quadrant was the most affected (40.8%).
![]()
Table 1. Baseline characteristics of patients.
![]()
Figure 1. Mammography (A, B) and ultrasound (C, D) images classified BIRADS-3. (A) Mammogram’s nodular opacity with polycyclic shape (black star); (B) Mammogram’s microcalcifications (black arrow) of different shapes including rounded and polymorphic; (C) Breast ultrasound showing an oval hypoechoic solid nodule with posterior echo enhancement and long axis not parallel to the skin; (D) Breast ultrasound of a palpable tender mass showing a cystic lesion with more echogenic parts (white arrows).
3.2. Compliance with Imaging Follow-Up and Cancer Occurrence
In overall sample (Table 2), 636 imaging controls were planned to follow-up and 510 check-ups were performed as follows: 395/395 patients at the first check-up, 97/162 patients for the second control and 18/79 patients to the last check-up. Within these controls, only 152 were carried out in the recommended timeframe (23.9%). As shown in Table 2, during the follow-up, the number of patients upgraded to BIRADS-4 reached 115 (29.1% of the sample). Among these patients, 107 (93.0%) underwent biopsy and histopathology revealed the presence of 43 cancers representing 10.9% of overall initial BIRADS-3 sample. At the time of study termination, 35 patients (8.9%) had undergone breast surgery (mastectomies) for malignancies and height (2.0%) patients were undergoing neo-adjuvant chemotherapy.
3.3. Factors Associated to Malignancies
The univariate analysis of epidemiological and clinical characteristics with cancer occurrence during the follow-up (Table 3) showed that the factors associated to malignancies were age above 40 years (p = 0.0001) and a presence of nipple discharge (p = 0.0375).
![]()
Table 2. Imaging follow-up and occurrence of malignancies at the three controls.
![]()
Table 3. Univariate analysis of factors associated to malignancies.
4. Discussion
In this retrospective longitudinal study, we aimed to assess the cancer occurrence and associated factors in BIRADS-3 lesions during the follow-up in order to propose a justified adaptation of the management for this category in our setting. Our results demonstrated that the global frequency of cancers among initial BIRADS-3 lesions reached 10.9%, imaging follow-up compliance is 23.9% and malignancies were more associated to age above 40 and nipple discharge.
The mean age of our study population, was 33.0 ± 13.4 years with extremes of 13 and 73 years, 68.1% of patients being under 40 years, making our sample a predominantly young population. These results are similar to that found in Nigeria in 2018 [12] on the management of BIRADS-3 breast lesions where the mean age was 29.7 years (13 - 68 years) and the most represented age group was 20 - 29 years. Ndé-Ouedraogo et al in Burkina-Faso in 2018 [5] reported a mean age of 38.2 years. But some European and Asian series [7] [13] [14] found higher average ages ranging from 41 to 59.6 years. The method of recruitment in our series from symptomatic patients could explain why our patients are younger. Indeed, at least 91.1% of our patients were symptomatic, with palpable breast mass or nodule, breast pain or nipple discharge whereas in the above referenced studies, inclusion was done from voluntary screenings which are generally the prerogative of women above 40 years. Concerning risk factors of breast cancer, hormonal contraception was found in 26.3% and family history of cancer in 12.4% but without any correlation with malignancy, contrary to what is described in the literature [15] . This could be explained by the youth of our study population.
According to the recommendations, BIRADS-3 lesions must be monitored clinically and radiologically at 4 to 6 months, and then at 12 and 24 months [12] [13] [14] . During this monitoring, if the lesion progresses to BIRADS-4, a histopathology evidence is necessary. As stated in the ancillary analysis paper early reported [16] , the participation to imaging controls was low and decreasing over time. A more aggressive strategy was then advocated such as a telephone reminder system or an iterative biopsy as credible alternative to the low compliance with imaging follow-up. Our global compliance was 23.9% and the number of lesions upgraded to BIRADS-4 reached 29.1% with 10.9% of overall initial BIRADS-3 being malignant, 61.0% of lesions being stable. This occurrence of cancer was mainly correlated with the age above 40 and to the presence of nipple discharge. Studies done in the United States [17] [18] found 3% of cancers during the follow-up of BIRADS-3 lesions. In South-Korea in 2016, 80.6% of initial BIRADS-3 lesions were stable and only 1.0% were found to be malignant at 1 year or more [14] . In Spain in 2001, only 1% of lesions were reclassified as BIRADS-4 [19] . All these studies found a positive predictive value of malignancy in concordance with the literature. The high incidence of malignancy in our series may be explained by a poor imaging follow-up due to low financial income and then lower compliance with controls than those in the studies above. In our setting, would it not be preferable to consider an upgrade to BIRADS-4 at the term of the first control if the lesion is not reclassified as BIRADS-2 so that histopathological samples can be collected? A similar recommendation was proposed in others african studies [5] [10] .
5. Conclusion
The global frequency of malignancies among initial BIRADS-3 lesions in this series, of 10.9%, is higher than that described in the international guidelines. This study highlights the need to be more proactive in the management of BIRADS-3 lesions in our setting as the compliance with follow-up is low. So, upgrading to BIRADS-4 in order to perform biopsy should be considered as an alternative to long-term follow-up for patients above 40, non reclassified as BIRADS-2, non-compliant with the first imaging check-up and presenting with nipple discharge. The proposed adaptation of the management approach might ensure an earlier diagnosis of breast malignancy for better patient outcomes.