The role of clinical breast examination and fine needle aspiration cytology in early detection of breast cancer: A cross-sectional study nested in a cohort in a low-resource setting

Abstract

Background:

Breast cancer is prevalent worldwide, with disparities in screening, diagnosis, treatment outcomes, and survival. In Africa, the majority of women are diagnosed at advanced stages, affecting treatment outcomes. Screening is one of the best strategies to reduce mortality rates caused by this cancer. Yet in a resource-constrained setting, there is limited access to screening and early detection services, which are available only at a few referral hospitals.

Objectives:

We aimed to evaluate the prevalence and screening results of breast cancer using clinical breast examination coupled with fine needle aspiration cytology in a resource-constraint setting.

Design:

A combined cross-sectional and cohort study.

Methods:

Women at risk of developing breast cancer in the Kilimanjaro region of Tanzania were invited, through public announcements, to their primary healthcare facilities. A questionnaire was used to assess the participants’ characteristics. The women received a clinical breast examination, and detectable lesions were subjected to a confirmatory fine needle aspiration cytology or an excisional biopsy. Preliminary data from this ongoing breast cancer control program were extracted and analyzed for this study.

Results:

A total of 3577 women were screened for breast cancer; their mean age was 47 ± 7.53 years. About a third of them (1145, 32%) were practicing self-breast examination at least once a month. Of 200 (5.6%) with abnormal clinical breast examination, 18 (9%) were confirmed to be breast cancer, making the prevalence to be 0.5%. The vast majority of participants with breast cancer (13, 72.2%) had early disease stages, and infiltrating ductal carcinoma, no special type, was the most common (15, 83.3%) histopathology subtype. Hormonal receptor status determination results indicated that 11 (61.1%), 7 (38.9%), and 5 (27.8%) of the tumors overexpressed estrogen receptor, progesterone receptor, and human epidermal receptor-2, respectively.

Conclusion:

Our study demonstrates 5.6% of Tanzanian women have abnormal clinical breast examination findings, with 9% having breast cancer. Nearly three-quarters (72.2%) of breast cancer screened for early disease were detected in the early disease stages. This finding suggests that organized screening with clinical breast examination coupled with fine needle aspiration cytology, which is a simple and cost-effective screening method, has the potential to improve early detection and outcomes for breast cancer patients in a resource-constraint setting.

Introduction

Breast cancer (BC) is the most common malignancy worldwide, and among four women with cancer, one is likely to be BC. ¹ In sub-Saharan Africa, BC is the second most common cancer in females and the second leading cause of cancer-related deaths.^1,2 This high incidence of BC has been attributed to the improved lifespan of people coupled with a sedentary and improper lifestyle that has increased the risk of non-communicable diseases, including cancer. ² In Tanzania, BC is the second leading cause of cancer morbidity and mortality in women, with an incidence of 23.4 per 100,000 women per year.^3,4 The majority of women with BC present with advanced-stage disease with a poor prognosis.^4,5 The burden is expected to rise in the next 20 years.^1,2

There is a great discrepancy in the overall survival rate between developed and developing countries, such as Tanzania, where the 5-year survival rate of BC is less than 45% as compared to more than 90% in developed countries. ⁶ Studies have also shown that women at risk of developing BC had poor knowledge of BC and related warning signs and were unaware of self-breast examination (SBE), risk factors, and screening methods. ⁷ A study conducted in the Lake Zone of Tanzania found that only a small proportion of the participants received health education about BC. In addition, more than half did not know any risk factors for BC. ³ The Lake Zone study also showed that more than 90% of participants would go to seek health care only if they experienced severe symptoms because they were afraid of either losing the breast or receiving bad results.

Screening is critical for early detection of BC, improves treatment outcomes, and thus reduces mortality from BC. Despite an increasing BC burden, screening, early detection, diagnosis, and management services remain inadequate in low- and middle-income countries (LMICs). ⁸ In these settings, cancer services are largely based at a few zonal or national hospitals that are always in urban areas or cities and are often overwhelmed by patients.^9,10 Moreover, the referral system is long and leads to late detection of cancer and delays in treatment. Although the use of mammography screening has expanded rapidly in many industrialized nations in recent years, this method is not popular in LMICs, partly due to resource constraints. Potential solutions in these settings include the use of cost-effective techniques, such as clinical breast examination (CBE). ¹¹ Some studies have documented that CBE may detect tumors that mammography misses and is an effective screening tool that can complement the performance of mammography in lowering BC mortality. ¹² A pilot study demonstrated the feasibility of combining CBE, fine needle aspiration biopsy (FNAB), and rapid on-site evaluation (ROSE) to expedite patient triage for breast cancer care. ⁶ However, some clinicians wonder if screening with CBE is still useful. ¹³ We aimed to evaluate the prevalence and screening results of BC using CBE coupled with fine needle aspiration cytology (FNAC) in a low-resource setting.

Materials and methods

The study design, eligibility criteria, and setting

This is an ongoing combined cross-sectional and cohort study carried out in rural areas within the Kilimanjaro region of Tanzania, involving women aged ⩾ 30 years who were permanent residents in the identified areas and willing to participate in the BC screening program. The younger inclusion age was selected based on the fact that African BC presents at an earlier age and has more aggressive biological behavior.^9,13,14 Moreover, the index study was also intended to provide health education regarding BC for women potentially at high risk. Women with disease conditions that precluded the standard breast examination, those unable to provide informed consent, or those unwilling to participate in the study, were excluded from this study.

The local government and religious leaders of the selected communities were informed about the study through stakeholders’ meetings. Community members were informed about the study using different media, such as public announcements through vehicles with loudspeakers, announcements in health centers, churches, mosques, local markets, local radios, brochures, and social media. The public announcements were also used to invite women to the primary health facilities for BC screening. The Kilimanjaro region is located in the northern zone of Tanzania and is one of the 31 administrative regions. According to the 2022 census, Kilimanjaro region had an estimated population of 1,861,934, of which females aged ⩾ 30 years were 386,421. ¹⁵ The region has six districts, and the study was undertaken in selected primary health facilities in all districts within the region. Based on previous experiences,^{16 –21} the sample size required for this study was estimated using a single population proportion formula with a 95% confidence interval, a precision of 4.5%, and an assumed screening practice of BC of 20% (0.2) to get a maximum sample size. The calculated sample size was 304, but by adding a 15% non-response rate and incomplete data, the minimum sample size was 350 women. However, to address secondary objectives that are beyond the scope of this article, a larger-than-required minimal sample size was obtained. A non-probability convenience sampling technique was used to fulfill the inclusion criteria. The study had 3577 participants.

Study procedures

Participants who agreed to join the study signed an informed consent form first, followed by a face-to-face interview using a pre-tested, semi-structured questionnaire (Appendix 1). The questionnaire was pre-tested on about 5% of the study population. The questionnaire assessed the sociodemographic and clinicopathological characteristics of respondents. In addition, the questionnaire assessed participants’ BC screening practices. Research assistants provided overall supervision of data collection in the field and quality checks for the completeness of the questionnaires. All completed questionnaires were numbered serially and stored for data entry and analysis. The study team, which included medical doctors in various clinical disciplines with the help of experienced reproductive health study nurses, provided theoretical and practical instructions on BC control, including self-CBE, to the participants. The emphasis was placed on recognition of early signs of BC and the importance of early diagnosis. Visual demonstrations (Figure 1) were used to facilitate the understanding of the message. This was done both on an individual and a small group basis. Participants were encouraged to ask questions in case they had any.

Figure 1.

SBE demonstration, adopted from thirdeyemalta.com.

Participants who were willing to undergo BC screening after the interview underwent CBE, which was performed by the study team. During CBE, if pathological lesions (detectable masses, ulcers, discharges, lymph nodes, etc.) were encountered, samples were obtained for cytological analysis. Two to three glass slides of smears were made from each case, and the slides were preserved in 95% alcohol. The slides were sent to the department of pathology at the Kilimanjaro Christian Medical Center (KCMC) for routine staining and analysis. Participants found with suspicious lesions, including atypical smears or malignant lesions on FNAC, were immediately scheduled for a follow-up where a breast tru-cut needle core biopsy was performed for confirmation. Similarly, other abnormal breast findings on CBE underwent excisional biopsies for histopathologic confirmation. The biopsies were done as outpatient procedures under local anesthesia, whereby at least six cores of biopsy tissue were collected and tissue samples were stored in a container with 10% buffered formalin for transportation to the pathology department of the KCMC, where routine processing and histopathological evaluation were performed. The samples were processed and analyzed according to the KCMC standard operating procedures. The histopathology reporting and sign-out of the cases were done by the study pathologists. Within seven working days of sample collection, pathology reports were communicated back to the participants. This work has been reported in line with the STROBE Guidelines. ²²

Statistical analysis

The data used in this study were electronically captured using REDCAP between May 2022 and July 2023. The data were stored using participants’ unique identification numbers to preserve confidentiality. Descriptive statistics, such as mean and standard deviation, were used to summarize continuous variables. A chi-square test was used to investigate associations between breast pathologies and sociodemographic variables. All analyses were done at the 5% level of significance, and a value of p-value less than 0.05 was considered to be statistically significant. Data analysis was done using STATA version 17 software (Stata Corp., LLC).

Ethical statement

The research ethics committee of Kilimanjaro Christian Medical University College approved this study protocol on 10 May 2022 (Certificate No. 2568). All participants provided informed consent to participate in the study before recruitment. The participants with various pathological breast lesions on CBE were followed up for further investigations, including lumpectomy and histopathology. The management was performed accordingly, either at their respective primary healthcare facilities or at the KCMC and Ocean Road Cancer Institute (ORCI), in line with the Tanzanian national cancer treatment guidelines. ²³

Results

Of the 4000 women targeted in this project and thus offered BC screening services, only 3697 women responded to the invitation and were assessed for eligibility, making the acceptability of BC screening services in the population being studied to be 92.4%. However, 120 participants did not meet the eligibility criteria and were excluded. The reasons for exclusion were age < 30 years (n = 110), history of BC (n = 5), or did not accept BC examination screening (n = 5). A total of 3577 women participated in this BC screening program. Their mean age was 47 ± 7.53 years, ranging from 30 to 93 years. Over half (56%) of them were between 30 and 41 years old. The majority were residents of Moshi district (66%), married (65.6%), had primary education (53.1%), had one to three children (55.1%), and were self-employed (75.2%), respectively. In addition, 59.9% had a normal body mass index (BMI), while the vast majority (84.8%) reported that it takes them 1 h to get to their nearest primary healthcare facility. The vast majority of study participants (95%) were able to perform SBE appropriately after receiving instructions. About two-thirds (67.6%) of the respondents have never been screened for BC before. All sociodemographic variables except age at menarche did not show a significant relationship with breast pathology status (malignant, benign, and normal) (p > 0.05), Table 1. A considerable proportion of the study participants (42%) has never practiced SBE, and only about a third (32%) of them were practicing SBE at least once a month (Figure 2).

Table 1.

Baseline characteristics of study participants, N = 3577.

Variable	Total, n (%)	Normal	Benign	Malignant	p
Age (years)					0.42
30–40	2003 (56)	1941	58	4
41–50	1166 (32.6)	1055	101	10
>51	408 (11.4)	381	23	4
Residence					0.59
Moshi	2262 (66)	2218	37	7
Hai	251 (7)	210	39	2
Siha	358 (10)	313	42	3
Rombo	447 (12.5)	407	37	3
Mwanga	84 (2.3)	68	15	1
Same	175 (4.9)	161	12	2
Age at menarche (years)					0.01
<17	3277 (91.6)	3102	164	11
⩾17	300 (8.4)	275	18	7
Marital status					0.87
Single	869 (24.3)	821	44	4
Married	2346 (65.6)	2216	119	11
Widow	135 (3.8)	127	7	1
Divorced	96 (2.7)	90	5	1
Other	131 (3.7)	123	7	1
Education					0.47
No formal education	132 (3.7)	124	7	1
Primary education	1899 (53.1)	1794	96	9
Secondary education	987 (27.6)	931	51	5
Tertiary education	559 (15.6)	528	28	3
Children					0.68
0	304 (8.5)	286	15	3
1–3	1970 (55.1)	1881	100	9
>4	1303 (36.4)	1230	67	6
Employment					0.25
Self-employed	2689 (75.2)	2541	136	12
Employed	550 (15.4)	519	28	3
Retired/unemployed	338 (9.4)	317	18	3
BMI					0.76
<18.5	75 (2.1)	72	3	0
18.5–24.9	2143 (59.9)	2022	112	9
25.0–29.9	1119 (31.3)	1060	53	6
⩾30	240 (6.7)	223	14	3
Time to health center (h)					0.58
<1	3033 (84.8)	2869	150	14
1–3	490 (13.7)	459	29	2
>3	54 (1.5)	49	3	2
Ever screened for BC					0.37
Yes	1159 (32.4)	1096	58	5
No	2418 (67.6)	2281	124	13
Able to perform SBE
Yes	3395 (95)	3209	170	16	0.73
No	182 (5)	168	12	2

BMI: body mass index; SBE: self-breast examination.

Figure 2.

Practice of self-CBE.

Of the 3577 women who participated in the screening program, only 200 (5.6%) had abnormal findings on CBE. The vast majority182 (91%) of the women with abnormal findings on CBE had benign lesions on a confirmatory tru-cut or an excisional biopsy. Fibroadenoma (32.5%) and abnormal benign nipple discharge due to duct ectasia (25.5%) were the most common benign pathologies. A total of 18 (9%) participants with abnormal CBE were confirmed to be BC on tru-cut histopathology, making the proportion of BC in the studied population to be 0.5%. Of participants with BC on FNAC, 2 (1.8%) did not attend follow-up at all for further investigation and treatment. Most of the participants with BC (72.2%) had clinical stage I or II and infiltrating ductal carcinoma; no special type (NST) was the commonest (83.3%) histopathological tumor type (Table 2).

Table 2.

Distribution of clinicopathological characteristics of study participants on CBE and histopathology, N = 3577.

Variable	Frequency	%
CBE findings
Normal	3377	94.4
Abnormal	200	5.6
FNAC/histopathology, (n = 200)
Fibroadenoma	65	32.5
Benign nipple discharge, NOS	55	25.5
Inflammatory/infectious	37	18.5
Other benign lesions	25	12.5
Breast cancer	18	9
Clinical stage breast cancer, (n = 18)
Stage I–II	13	72.2
Stage III–IV	5	27.8
Histological tumor type, (n = 18)
Infiltrating ductal carcinoma, NST	15	83.3
Lobular carcinoma	1	5.6
Squamous cell carcinoma	1	5.6
Adenoid cystic carcinoma	1	5.6

FNAC: fine needle aspiration cytology; NOS: not otherwise specified.

Hormonal receptor status determination analysis results indicated that 61.1%, 38.9%, and 27.8% of BC tumors overexpressed estrogen receptors (ER), progesterone receptors (PR), and human epidermal receptor-2 (HER-2), respectively. Seven of the cases were HER-2 negative. Fluorescence in situ hybridization (FISH) was not performed to evaluate the HER-2 + equivocal results of the six cases because it is not available in our setting. Thus, the equivocal cases were considered HER-2 negative (Table 3).

Table 3.

Hormonal receptors status (n = 18).

Hormonal receptor	Positive	Negative	Total
ER	11 (61.1%)	7 (38.9%)	18 (100%)
PR	7 (38.9%)	11 (61.1%)	18 (100%)
HER-2	5 (27.8%)	13 (72.2%)	18 (100%)

Discussion

Our study evaluated the prevalence and screening results of BC using CBE coupled with FNAC among women at risk of BC in Tanzania. Two-thirds (67.6%) of the respondents had never been screened for BC before, and the vast majority (95%) were able to perform SBE confidently after being instructed. We also noted that 5.6% of the participants in this screening program had abnormal findings on CBE, of whom 9% were confirmed to be BC histopathologically, making the prevalence of BC to be 0.5%. Of participants with BC, the majority (72.2%) had an early clinical disease stage.

The estimated prevalence of BC in this study is 0.5%; however, previous studies have reported a relatively higher prevalence compared to the index study, which is inconsistent with our findings.^24,25 This disparity could be attributed to the differences in the study population dynamics and settings where the studies were conducted, and the detection methods used. Generally, the relatively low incidence of this malignancy in most LMICs is partly due to underreporting resulting from poor access to screening and diagnostic services in countries with limited resources.^10,13,25,26 The prevalence of BC among Tanzanian women with cancer is estimated to be 15.9%. The age-standardized BC incidence in Tanzania is 23.4/100,000 women, and the age-standardized BC mortality rate is 12/100,000.^1,2 This translates to a mortality-to-incidence ratio (MIR) of 0.5, indicating that half of all women diagnosed with BC will die of the disease. The number of new BC cases is projected to an increase to 4961 cases in 2030, an increase in 82%. Similarly, projections for BC deaths follow the same pattern, with an increase in 80% in BC deaths by 2030.^1,2 Thus, governments and other stakeholders in health should make deliberate efforts to effectively address this growing public health problem.

The primary purpose of screening is to detect cancer at an early stage. Mammography is the most common screening test for BC in a high-resource setting. However, mammography is not an appropriate approach in LMICs because of its cost and complexity. ²⁷ Furthermore, as it was observed in our study, most women in LMICs are younger than 50 years, and mammography is less effective in this age group. A fast, low-cost, efficient, and safe way of identifying malignant and benign breast masses in outpatient settings would considerably benefit patients and healthcare providers while lowering the waiting list for special investigations. The FNAC is a simple, patient-friendly, and cost-effective method. ²⁰ It can rapidly and successfully screen out benign breast lesions at no additional cost or burden to the patient. FNAC also plays a vital role in BC management. In benign disease conditions, FNAC provides a one-stop examination that results in either the patient is discharged or a scheduled follow-up. This immediate reassurance is one of the main benefits of this method for the patients. Being minimally invasive, with minimal pain and discomfort, and with a decreased risk of complications, FNAC is the most patient-friendly option. When performed by expert pathologists, FNAC accuracy for benign lesions is very high. Previous research has shown that sensitivity and specificity of breast FNAC range from 77% to 100%. FNA-material cell blocks are beneficial in determining HER-2 and hormone receptor status. Also, FNAC can be used to investigate axillary lymph nodes in individuals diagnosed with BC, assist with staging the disease, select patients for sentinel node biopsy or axillary lymph node dissection, and plan for neoadjuvant therapy in cases of advanced disease. FNAC can also be used to detect distant metastatic disease or recurrence.^20,21

Despite many advantages, FNAC has limitations in assessing some breast conditions where malignancy and benignity are not always obvious. In these circumstances, a tru-cut biopsy or excisional biopsy is recommended for confirmation, as has been done in this study. Similarly, FNAC cannot give direct information on tissue architecture or detect specific breast diseases. High levels of experience are required to achieve satisfactory results. Because of the lack of complete histological features, FNAC cannot provide firm proof of basement membrane invasion. FNAC cannot reliably distinguish between in situ and invasive carcinoma. For very small lesions (sub-centimeter), FNAC is not the primary choice, and a tru-cut needle biopsy is preferred instead. Ultrasound guidance is needed in non-palpable lesions to boost FNAC efficiency. In addition, FNAC is not ideal for calcified lesions because they are technically difficult to sample, and, as a result, insufficient samples are frequently obtained. Papillary lesions are another area where FNAC is limited. As a result, diagnosis should primarily be based on histologic examination.^20,21

A CBE is a physical examination of the breast by a health professional. CBE is an alternative screening method to mammography in LMICs. CBE is responsible for almost two-thirds of the reduction in BC mortality.^11,12 The use of CBE (alone or in combination with mammography) for screening asymptomatic women at average risk of BC varies greatly.^13,17 To be a reliable screening test, CBE must minimize the risk of dying from BC in an asymptomatic population while being cost-effective and providing more overall benefit than harm. Most guidelines now require that breast lesions be assessed using a triple-assessment approach that includes clinical, radiological, and pathological evaluation. The current international guidelines indicate CBE as the recommended technique in resource-constraint settings. ¹⁸ However, there is some debate over the usefulness of CBE for population-based screening programs. For instance, a Japanese study found a trade-off between sensitivity and specificity associated with CBE added to mammography and that CBE may be omitted from BC screening among women aged 60 and 70 years. ²⁷ However, budgetary and health system hurdles in LMICs prevent efficient implementation of mammography for BC screening, often due to a lack of equipment and a skilled workforce. As a result, there is an opportunity to maximize the use of the available resources and screening techniques in countries with limited resources. Previous research studies have found that CBE with ROSE and FNAC were feasible in a BC screening program in a limited-resource setting. ⁶ In these settings with constrained resources for cancer care, this may be an effective method for triaging patients with breast masses. However, if CBE screening is to be employed as a public health strategy, great care must be taken, particularly when women present without any suspicion of lesions.

This study also discovered that SBE practice among the study population was poor. A considerable proportion of the study respondents (42%) reported having never practiced SBE, and the majority (67.6%) of them had never been screened for BC before. The findings are consistent with the findings from a hospital-based study in Nigeria, in which only 23.9% of participants inspected their breasts on a monthly basis. ⁷ Similar reports of low SBE practice have been recorded among women across sub-Saharan Africa. ²⁸ This could be due to a lack of awareness about how to perform SBE, and some participants are unaware that SBE should be done on a monthly basis. As a result, this could explain why patients with BC in LMICs present with advanced disease stages and, thus, a poor prognosis. ²⁹ However, despite the poor SBE practices observed in our study, yet the vast majority of women were able to perform SBE after receiving instructions. The finding suggests that SBE is simple and feasible. If women are given regular health education on SBE, it may increase their knowledge and understanding of BC, warning signals, and risk factors as they practice regular SBE. If performed consistently and appropriately, SBE has the potential to be a cost-effective method for detecting early changes in the breast and presenting to the hospital for further evaluation. Therefore, SBE can be extremely beneficial. Thus, continuous health education opportunities should be made available to the communities so as to improve health-seeking behavior and, therefore, potentially treatment outcomes.

In this study, most of the women found with BC experienced their menarche at a younger age compared to older age counterparts (p < 0.05). The results are consistent with previous studies.^{30 –32} Surprisingly, other previously established BC risk factors, such as nulliparity and BMI, did not show any significant association. The discrepant findings between the previous and the current study could be explained by the small number of participants with BC in the current study. Notably, the vast majority of participants (72.2%) found with BC in this study were detected at early disease stages (I or II). The findings imply that CBE screening coupled with FNAC has the potential to improve the detection of early BC and, thus, the treatment outcomes. BC screening is critical for early detection and staging, and it has the potential to reduce morbidity and death in LMICs. Combining cytology with CBE-based screening programs may be viable and aid in expediting and triage for patients with breast masses. Currently, approximately 80% of women diagnosed with BC are diagnosed at advanced stages of disease and have limited access to early detection, diagnosis, and treatment services.^4,5 Consensus findings from a variety of retrospective studies in the sub-Saharan region point to a late stage of diagnosis, with the majority of patients presenting with stage III or IV disease.^25,26

This study has some limitations. First, the current study’s low BC prevalence could be attributed to the low test performance of CBE in cancer screening, and we did not have a comparison reference standard to assess the CBE test performance. As the primary technique for BC screening, CBE may miss tiny lesions that may need mammography or ultrasound to increase sensitivity. Second, false positives are not uncommon with CBE screening. Therefore, it may lead to unneeded procedures. However, one should know that high-resolution ultrasonography, which is the standard BC screening approach, is both prohibitively expensive and commonly unavailable in many resource-constrained contexts. Furthermore, no trials have shown a difference in mortality benefit between CBE and placebo. Moreover, the data on the cohort of females who were not offered BC screening to compare the prevalence of BC and the stage of detection were not available.

Conclusion

Our study demonstrates that 5.6% of women in Tanzania have abnormal findings on CBE; of these, 9% are BC. Nearly three-quarters (72.2%) of women with BC were detected in early disease stages. The findings from the current study suggest that organized screening with CBE coupled with FNAC may have the potential to be transferrable to other LMICs with a high burden of BC and thus improve early detection and, thus, enhance treatment outcomes for BC patients in limited-resource settings. CBE is a simple and cost-effective screening modality for breast examination in the context of resource availability. However, further studies are recommended before this approach is considered standard BC screening in LMICs. In addition, BC control requires the integration of early detection programs, accurate diagnosis, and timely, accessible, and effective therapies. Addressing any of these elements separately will not enhance BC control. Increasing public awareness through education is a critical first step in increasing BC control because early detection programs cannot be successful if the public is unaware of the value of early detection.