• Users Online: 328
  • Print this page
  • Email this page

Table of Contents
Year : 2022  |  Volume : 27  |  Issue : 3  |  Page : 251-260

Accuracy of Mammography in the Diagnosis of Breast Cancer

1 Department of Radiology, Nnamdi Azikiwe University Teaching Hospital, Nnewi and Nnamdi Azikiwe University, Nnewi Campus, Nnewi, Nigeria
2 Department of Histopathology, Nnamdi Azikiwe University Teaching Hospital, Nnewi and Nnamdi Azikiwe University, Nnewi Campus, Nnewi, Nigeria
3 Department of Surgery, Nnamdi Azikiwe University Teaching Hospital, Nnewi and Nnamdi Azikiwe University, Nnewi Campus, Nnewi, Nigeria

Date of Submission14-Dec-2021
Date of Decision01-Mar-2022
Date of Acceptance14-Mar-2022
Date of Web Publication2-Jun-2022

Correspondence Address:
Brenda C Nwammuo
Department of Radiology, Nnamdi Azikiwe University Teaching Hospital, Nnewi and Nnamdi Azikiwe University, Nnewi Campus, Nnewi
Login to access the Email id

Source of Support: None, Conflict of Interest: None

DOI: 10.4103/ijmh.IJMH_48_21

Rights and Permissions

Background: Breast cancer is one of the leading causes of cancer-related deaths among women worldwide. Mammography was recently introduced at the Nnamdi Azikiwe University Teaching Hospital (NAUTH), Nnewi, Nigeria for breast cancer screening and diagnosis at the institution. Objective: The objective of this study is to determine the diagnostic accuracy of mammography in the assessment of palpable malignant breast lump among adult females referred from the NAUTH outpatient clinics using histopathology as gold standard. Materials and Methods: This prospective cross sectional study involved recruitment of consenting adult females with palpable breast lump referred for mammography between October 2017 and November 2019 at NAUTH, Nnewi, Nigeria. Two standard views—cranio-caudal (CC) and medio-lateral oblique (MLO)—were done for each patient. Spot magnification or other supplementary views were only done when indicated. Data analysis was done using Statistical Package for the Social Sciences (SPSS) version 21.0 for Windows software. Results: The sensitivity, specificity, positive predictive value, negative predictive value, and overall accuracy of malignant breast lesions were 41.4%, 98%, 96.7%, 54.4%, and 65%, respectively. The most common features of malignancy were mammographic opacity with an irregular outline (86.7%) and architectural distortion (80%), whereas asymmetric density (10%) was the least common feature. Conclusion: Mammography has a fairly high overall diagnostic accuracy in breast cancer diagnosis. However, an alternative imaging modality is required for dense breasts. This study’s reported specificity (98%) suggests that the test has a very high ability to correctly identify non-cancerous breast lumps. This is particularly relevant in breast cancer screening with an objective to rule out the presence of the disease among adult females in the population.

Keywords: Breast cancer, diagnosis, mammography, screening

How to cite this article:
Nwammuo BC, Umeh EO, Ebubedike UR, Nwosu SC, Elendu KC, Umeokafor CC, Ukah CO, Egwuonwu OA. Accuracy of Mammography in the Diagnosis of Breast Cancer. Int J Med Health Dev 2022;27:251-60

How to cite this URL:
Nwammuo BC, Umeh EO, Ebubedike UR, Nwosu SC, Elendu KC, Umeokafor CC, Ukah CO, Egwuonwu OA. Accuracy of Mammography in the Diagnosis of Breast Cancer. Int J Med Health Dev [serial online] 2022 [cited 2022 Aug 12];27:251-60. Available from: https://www.ijmhdev.com/text.asp?2022/27/3/251/346438

  Introduction Top

Globally, in females, breast cancer is a major public health issue, with annual diagnosis of new cases being over a million and about 4.4 million women living with the disease. Resultant annual deaths from this disease are over 400,000.[1],[2],[3] Recent reports have shown a rise in the global indices for both newly diagnosed cases and annual deaths from breast cancer, being about 2.1 million new cases, accounting for almost 1 in 4 cancer cases among women and 627,000 deaths, respectively.[4]

As of 2018, the incidence of breast cancer has risen to 37.3 age-standardized rate per 100,000 females in western Africa,[4] which is slightly higher than what was obtained in the previous years.[5] This rise in incidence, especially in areas that historically had low incidence, is largely attributed to rising life expectancy and a combination of demographic factors allied to socioeconomic development, including nulliparity, late age at first birth and having fewer children, intake of exogenous hormones, greater levels of obesity and physical inactivity, as well as increases in breast cancer screening and awareness, among others.[4],[6],[7]

Despite the significant international and geographical variations with regard to incidence, the variability with regard to mortality rate is less.[4] Africa has been noted to be significantly contributing to the mortality rate worldwide, with mortality rates of 18.4, 17.8, 15.8, 15.6, and 15.4 age-standardized rate per 100,000 females in northern, western, middle/central, southern, and eastern Africa in the descending order, respectively.[4]

In Nigeria, today, breast cancer still remains one of the commonest cancers ravaging a great percentage of our female population[8],[9] with about 100,000 new cases of cancer occurring every year and an associated high case fatality ratio.[10] Based on the WHO report in 2014, Nigeria was ranked fourth in the world in breast cancer mortality rate, with age-adjusted death rate being 28.11 per 100,000 of the population.[8],[9] Like every other developing/low income countries, this high mortality rate from breast cancer has been greatly attributed to late presentation; when the disease is in the advanced stage, only palliative care can be given.[11]

Breast cancer occurs in females across all races; however, in the black race, it occurs a decade earlier than their white counterparts.[11],[12] In blacks, it has been widely reported to have a peak age incidence of 40–49 years.[13] Nevertheless, some studies have noted a double age peak incidence, whereas others have reported a peak age incidence below 40 years.[14],[15],[16] A study done in southeastern Nigeria noted a peak age incidence of 35–39 years with the condition constituting 30% of all patients with breast disease.[17] Notably, the disease in those aged below 40 years is usually more aggressive, associated with high mortality, diagnosed at the advanced stage, and likely to reoccur than in older women.[18]

The gold standard for the diagnosis of breast cancer is histopathology, and it depends mainly on appropriate biopsy (tissue sampling) techniques, optimal tissue processing, and competent interpretation of gross and microscopic pathology findings. Successful pathology service requires timely and accurate comprehensive reporting, as well as archiving of slides, tissue blocks, and reports with accurate patient and specimen identification.[19] These still remain as pertinent challenging issues in developing countries. Given these numerous challenges experienced by countries in Sub-Saharan Africa with regard to histopathology, other modalities such as mammography could be explored.

Mammography itself is the standard of reference for the clinical detection of breast cancer in older females (40 years and above), but there are limited studies evaluating its use for diagnostic purposes, especially in our environment.[11],[20] Mammography is available to an extent, easy to operate, well-tolerated, non-invasive, and inexpensive.[21] The major limitation of mammography is the dense breast in young females.

Various studies have recorded a fairly high sensitivity and specificity for mammography in the diagnosis of breast cancer.[22],[23] Some other studies documented low sensitivity and specificity and proposed the use of other imaging modalities as adjunct.[21],[24]

However, this study is aimed at determining the accuracy of mammography diagnosis of the malignant breast disease, among adult females with palpable breast lumps, using histopathology as gold standard at Nnewi, Nigeria. With the inception of mammography in the hospital, there is a need to determine the accuracy of mammography as a diagnostic tool and also the proficiency of the operatives within the radiology department of the institute, in evaluation of mammographic images specifically with regard to detection of features indicative of malignant breast disease, ahead of the commencement of routine screening mammography across the state. This study will also help identify our weaknesses and work on them in order to improve accuracy.

  Materials and Methods Top

This study was a prospective study done at the Radiology Department, Nnamdi Azikiwe University Teaching Hospital (NAUTH), Nnewi and was carried out over a period of 2 years, from October 2017 to November 2019. The convenience sampling technique was used. The sample size was calculated using sample size determination by Isreal.[25] The prevalence rate of breast cancer is 9.12%.[18] The absolute precision was 5%, so 93 patients were the minimum required sample size. However, we recruited 120 patients in the study.

The study included all female patients above 35 years of age with palpable breast lumps referred from the outpatient clinics in the hospital as well as those below 35 years with a clinically suspicious palpable breast lump and women below 35 years who presented with a palpable breast lump and also have a positive family history of breast cancer. Excluded from the study were those with biopsy-proven malignancy, pregnant women, lactating mothers, and those with fungating breast lesions. Similarly, patients whose histopathology results were not available or those who did not undergo biopsy/surgery were also excluded.

The subjects were informed about the nature and purpose of the study, and a written consent was obtained. A detailed history was taken regarding patient’s signs and symptoms and duration. Also age at first conception, age at menarche, age at menopause, last menstrual period, history of breastfeeding, number of children, history of hormone therapy, family history of breast or gynecological cancer in mother or sisters, a personal history of breast cancer or benign proliferative breast disease, radiation, chemical exposure, or smoking data were obtained.

Mammogram was performed using a General Electric (GE) Performa Mammography Machine MGF-110 (manufactured by GE Hungary zrl. Healthcare, in 2009). Mammography was done for both breasts, in the two standard views cranio-caudal (CC) and medio-lateral oblique (MLO), with additional views (spot magnification and supplementary views) done when required. The mammograms acquired were evaluated by the radiologists and findings were documented using the ACR-BIRADS lexicon. Patients with mammograms which showed a combination of highly suspicious findings, such as irregular, ill-defined, microlobulated or spiculated high-density masses, in association with other secondary suspicious findings such as suspicious calcifications (pleomorphic, fine linear, or fine linear branching calcifications) in segmental distribution, nipple retraction, architectural distortion, skin thickening, nipple or skin retraction, trabecular thickening, and replaced or eccentric enlarged lymph nodes, were categorized as BIRADS 5 in this study. Histopathology results were provided by the pathologist in the same institute, which served as gold standard. The mammography and histopathology findings were in the proforma by the researcher.

Data entry and analysis were done using the statistical analysis of the obtained data from mammography, and histopathology reports were conducted using the Statistical Package for the Social Sciences (SPSS) version 21.0 for Windows (IBM Corporation). Statistical analysis was performed for all variables with Pearson’s χ2 test. Variables were deemed significant if P-value is less than 0.05.

The sensitivity, specificity, positive predictive value, negative predictive value, and diagnostic accuracy of digital mammography in the detection of breast cancer were calculated against histopathology as the gold standard, as defined in the operational definition.

  Results Top

The data used for this study were collected from a total of 120 women who presented with palpable breast lumps, age ranging between 25 and 74 years. The mean age of 47.37 ± 10.67 years is shown in [Figure 1]. The highest proportion of females were aged between 41 and 45 years (20.8%), followed by those aged between 46 and 50 years (17.5%), whereas the least proportion of females were aged between 21 and 25 years (1.7%).
Figure 1: Histogram of distribution of age, with normal distribution curve superimposed. Note that the age distribution is skewed to the left

Click here to view

Of the 120 females recruited for this study, 7 presented with bilateral palpable breast lumps, whereas 113 had unilateral palpable breast lumps.

Mammography done on 120 patients revealed bilateral or unilateral lump on 112 cases, whereas 8 cases were inconclusive due to dense breast or marked edema obscuring possible mass within the breast. On mammography, 30 (25%) cases were classified under BIRADS 5 and 8 (6.67%) were classified under BIRADS 0. The others were grouped under BIRADS 2 (34), BIRADS 3 (14), and BIRADS 4 (34), depending on the mammographic features seen, as shown in [Table 2]. All 120 patients underwent Tru-cut or surgical excision biopsy, including the eight patients who had inconclusive reports. On histopathology, 70 (58.33%) lumps were malignant, whereas 50 (41.67%) were benign lumps, as also shown in [Table 2].
Table 1: Overall mammographic diagnostic statistics of breast cancer in adult females using BIRADS 5, showing very high specificity and positive predictive value and low sensitivity

Click here to view
Table 2: Distribution of histological diagnosis, mammographic diagnosis, and number of breast mass on mammography among the patients, showing higher number of malignant lesions than benign lesions on histopathology and also showing BIRADS 2 and BIRADS 4 having the highest frequency

Click here to view

Each mammographic final BIRADS assessment was confirmed on histopathology to be either malignant or benign, as shown in [Table 3], with BIRADS 5 having 25 cases confirmed to be malignant (true positives) and 1 benign (false positive), as seen in [Table 3].
Table 3: Relationship between mammographic diagnosis and histological diagnosis among the patients, showing approximately 97% corroboration of cases categorized as BIRADS 5 with histopathology

Click here to view

Accuracy of mammography in the assessment of palpable malignant breast lump is shown in [Table 1], as follows: sensitivity (41.4%), specificity (98%), NPV (54.4%), PPV (96.7%), and overall accuracy (65%).

[Figure 2][Figure 3][Figure 4][Figure 5] are images showing some of the appearances of some of these features suggestive of malignancy on mammography.
Figure 2: MLO view of the right breast showing irregular spiculated high density opacity (thick black arrow), with associated thickened trabecular (black arrow head), architectural distortion (thin black arrows), and subtle focal skin thickening (white arrow)

Click here to view
Figure 3: A: CC view of the left breast showing an irregular high density spiculated mass (black arrow head), with associated skin thickening (white arrow head), nipple retraction (white arrow), and trabecular thickening (black arrows). B: MLO view of same breast showing skin retraction (white arrow), architectural distortion (black arrow head), and axillary lymphadenopathy (black arrow)

Click here to view
Figure 4: CC view of the right breast showing a fairly round spiculated high-density mass (white arrow head) with associated skin thickening (black arrow)

Click here to view
Figure 5: MLO view showing a fairly oval-shaped opacity with obscured margin (black arrows), a huge axillary mass (black star) with associated skin thickening (white arrows)

Click here to view

Mammographic breast compositions showed that category “c,” heterogeneously dense breasts which may obscure small masses, had the highest frequency (52.50%), followed by category “b”; breasts with scattered areas of fibroglandular density (23.33%). The extremely dense breasts which lower the sensitivity of mammography had the least frequency (7.50%).

  Discussion Top

To the best of our knowledge, this study has not been carried out before within the southeastern region, hence there is a need to compare with studies done elsewhere.

The mean age of 47.37 ± 10.67 years, recorded in this study among women with palpable breast lumps referred for imaging, is similar to the mean age of 48.3 ± 9.5 years, as reported by Ebubedike et al.,[26] in a study of women in South-East Nigeria aged between 20 and 79 years who presented with breast lumps. It also noted a predominant age range of 45–49 years. Akinola et al.[27] in South-West Nigeria in a study of 300 patients, aged between 35 and 82 years with breast symptoms, reported a mean age of 46.8 ± 10.5 years, which is similar to that reported in this study.

This study shows that mammography has a sensitivity of 41.4% in the detection of malignant breast disease, which is indicative of the proportion of histologically confirmed cases reported earlier as malignant on mammography. However, this apparently low sensitivity can be attributed to carefulness in categorizing lesions as BIRADS 5, as suspicious cases were rather categorized as BIRADS 4; hence, only 30 cases were reported malignant (BIRADS 5), with 29 confirmed on histopathology, giving a 96.6% corroboration with histopathology diagnosis and a high PPV (94.7%). A greater number of cases (34) were categorized as BIRADS 4, as the mammographic features for malignancy seen arguably did not qualify for BIRADS category 5. However, while this study noted a low sensitivity, Majeed et al.[28] in Islamabad, Pakistan noted a 99.9% much higher sensitivity for mammography in the diagnosis of breast malignancy. This report may be explained by a consensus from more than one radiologist reviewing images acquired using a higher quality mammography facility, with better image resolution.

It is also possible that the fairly conservative sensitivity value in this study may be reflective of the limited number of referrals for mammography within this region of Nigeria, which in turn impacts on the experience and exposure gathered by radiologists in practice and training. In line with this, Esserman et al.[29] proposed that reader volume is an important determinant of mammography sensitivity and specificity and concluded that high sensitivity and high specificity can be achieved in high volume centers.

The reported sensitivity of mammography in the detection of malignant breast lumps in this study may have been affected by the predominant breast density among the participants. More than half of the participants were aged below 50 years and approximately half of the participants (53%) had the ACR_BIRADS category c (heterogeneously dense breasts which may obscure small masses), whereas another 9.8% of the participants had the ACR_BIRADS category d (extremely dense breasts which lower the sensitivity of mammography).

As was established in a study by Devolli-Disha et al.,[30] sensitivity in a study population will improve with images showing predominantly fatty tissue, i.e., ACR_BIRADS “a” breast composition (almost entirely fatty breasts) and “b” (there are scattered areas of fibroglandular density) breast composition, which was not the case in this study in which 63% of the participants had a breast composition that could obscure underlying lesions and affect negatively the sensitivity as well as the overall accuracy of mammography. As stated by Barlow et al.[20] and authors from some other studies,[30],[31] a dense breast is a major limitation in the sensitivity of mammography; hence, breast composition can affect positively or negatively the sensitivity of mammography.

In this study, eight participants had inconclusive results (BIRADS 0) either due to extremely dense breast or extensive edema which obscured lesions and characteristics which may suggest malignancy in the breast. However, seven cases out of the aforementioned eight participants turned out malignant on histopathology. This is in keeping with what Majid et al.[32] proposed in their study that breast cancers can be missed due to dense parenchyma that obscures a lesion, among other causes such as edema from inflammation.

As noted in this study, a total of 10 cases which were grouped under BIRADS 2 and BIRADS 3 (five cases each) due to the benign features they portrayed such as fairly round or oval shapes, well-circumscribed margins some of which show lobulated outlines, without any associated suspicious finding, were reported malignant on histopathology. This is consistent with Burrell et al.,[33] who reported that breast cancers are easily missed when they appear as focal areas of asymmetry or distortion (as seen in invasive lobular carcinoma) or when their appearance suggests a benign cause as seen in medullary and mucinous (colloid) invasive ductal carcinomas, which usually manifest as mostly circumscribed masses.

This study reports a specificity of 98% for mammography in the diagnosis of breast malignancy, which is indicative of proportion of cases deemed non-malignant on histopathology and earlier corroborated on mammography. This suggests that there is greater confidence on the exclusion of malignancy in the absence of characteristic features on mammography. This finding is similar to that noted by Majeed et al.[28] in Pakistan, who reported a mammography specificity of 83.4% in the diagnosis of malignant breast disease. However, a similar study by Mustapha et al.[34] reported much lower specificity (50%) among their study groups in Maiduguri, Nigeria, with a significant proportion of false positives.

The PPV in this study was high (94.74%), which means that lesions reported as malignant on mammography have a 94.7% corroboration on histopathology. Similar findings were seen in a study by Taif et al.[35] (94.1%) in Oman and Orel et al.[36] (97%) in Philadelphia, USA.

Tan et al.[37] (84.2%) in Singapore, Hirunpat et al.[38] (80%) in Songklanagarind, Thailand, and Mustapha et al.[34] (82.6%) in Maiduguri, North-Eastern Nigeria got slightly lower values than that in this study, possibly due to lower proportions of breast malignancy in their various studies. It can also be attributed to the number of false positives recorded in their various studies.

The PPV in studies by Zonderland et al.[39] in the Netherlands and Majeed et al.[28] in Islamabad, Pakistan was reportedly 100% each, which could be attributed to multiple reviewers, arriving at a consensus on the final BIRADS score. This could also be attributed to a combination of screening and diagnostic exams with additional ultrasonography done for a significant number of participants, before a BIRADS score was assigned as documented by Zonderland et al.[39]

While reporting mammography images, one is careful at assigning the Final BIRADS 5 (Malignant) score, following the identification of mammographic features on an observed opacity that may be indicative of malignancy. When unconvinced that a lesion is definitely malignant but does show some suspicious features, which may be seen in malignancy, one settles for a conclusion of the Final BIRADS 4 (Suspicious) score. This explains the fairly high PPV in this study and the rather large proportion of cases reported as the Final BIRADS Category 4 (34) with approximately two-thirds of them eventually reported malignant on histopathology. Consequently, the PPV in this study is fairly high, similar to figures reported in the studies mentioned earlier.[35],[36]

The NPV, reported in this study, indicates corroboration of a reportedly benign lesion on mammography in 54.4% of the cases. This is lower than what was reported by Bukhari et al.[40] (92.8%) in Pakistan, and this could be attributed to the fact that only BIRADS 1 and 2 were considered negative for breast cancer (benign) in their study, whereas BIRADS category 3 and BIRADS category 0 were further investigated using ultrasonography and then re-categorized which was not the case in this study.

In this study, the overall accuracy (65%) is slightly higher than that reported by Omidiji et al.[41] (56%) in South-West Nigeria, possibly due to consideration of only suspicious mammographic findings (BIRADS 4) and exclusion of definite malignant cases, BIRADS 5 and 6 in their study, which recorded a significant number of false positives.

Also the study by Omidiji et al.[41] was based on screening mammography, with exclusion of women with symptoms such as mastalgia, nipple discharge or lump, previous breast lump excisions or biopsies, history of trauma, who had never breastfed, and pregnant and lactating mothers, whereas this study was on diagnostic mammography, with palpable breast lump being a compulsory criterion for recruitment.

Barlow et al.[20] inferred in their study that diagnostic mammography may have superior performance over screening mammography, due to noticeable symptoms or clinical findings which may indicate a more advanced tumor that is easier to locate and identify. Dee and Sickles[42] in their study also reported that tumors detected by diagnostic mammography were larger than those detected by screening mammography, which could be a reason for improved accuracy of diagnostic mammography.

The overall accuracy of mammography in the diagnosis of malignant breast disease by Barlow et al.[20] (91.3%) in the USA and Jensen et al.[31] (>90%) in their comparative study of diagnostic mammography performance in the USA and Denmark was higher than that reported in this study. This difference could be due to the inclusion of BIRADS 4 and BIRADS 5 categories as true positives (malignant) in their studies, which was not done in this study. Only mammograms categorized BIRADS 5 were considered in determining the overall accuracy in this study. Furthermore, supplementary ultrasound examination was done for a significant number of participants in their studies before the final BIRADS was assigned,[20],[31] unlike in this study in which ultrasound was not utilized.

Other suggestions for the higher accuracy recorded in their studies were due to more than a single reading by two or more radiologists done, before the final BIRADS score was designated and also due to separate mammography facilities dedicated to either screening or diagnosis alone, leading to improved or better diagnostic mammography expertise in facilities dedicated to symptomatic patients.[31]

Barlow et al.[20] in their study also noted that patient population and age distribution, especially when more than half of the patients population is above 50 years of age and when supposedly the breast is less dense, are important factors that can enhance mammography accuracy. However, in this study, more than half of the patients recruited were below 50 years of age with a significant number having the heterogeneously dense breasts which may obscure small masses (ACR-BIRADS breast composition c) and hence the lower accuracy.

According to the ACR-BIRADS 5th edition,[43] no single mammographic feature on its own is associated with likelihood of malignancy with confidence level above 95%. However, it takes a combination of suspicious findings on mammography to justify a final BIRADS 5 score, which has ≥95% probability for malignancy. As mentioned earlier, suspicious findings on mammography may include an irregular shape and margins (spiculated and indistinct) of opacity,[43] pleomorphic, fine linear, or cluster calcifications,[43] architectural distortion, skin thickening and nipple, and tissue/skin retraction.[43]

Images evaluated for patients classified as the Final BIRADS category 5 showed a combination of some of the aforementioned features. Large proportions of mammographic images in this group showed architectural distortion (80%), skin thickening (90%), and opacity with indistinct margin (50%). These features showed high corroboration on histopathology, i.e., 76.7%, 86.7%, and 46.7% for architectural distortion, skin thickening, and opacity with indistinct margins, respectively, although not statistically significant.

Spiculated masses, on their own, in the absence of any other feature or secondary characteristic for malignancy, have a >2% to <95% likelihood of malignancy and hence do not qualify as BIRADS 5, but qualify for BIRADS 4.[43] In this study, masses with spiculated margins were seen in 36.7% of the cases (11 out of 30 cases) categorized as BIRADS 5; however, this feature when present in this study was associated with the report of malignancy in histopathology in 100% of the cases, which is similar to what Sambasivarao et al.[21] and Gurung et al.[44] observed in their studies.

However, Gurung et al.[44] in their study, of 100 women with palpable breast lumps, of which 21 out of 36 mammographically malignant cases had spiculated margins, all of which turned out malignant on histopathology, reported a higher frequency of spiculated masses (60%); this could be attributed to the study population. Akinola et al.[27] reported a lower frequency of spiculated masses (11%), whereby out of the nine spiculated masses seen on mammography, three were proven to be malignant histologically.

This lower frequency reported by Akinola et al.[27] could be attributed to the combination of screening and diagnostic mammography, as well as a larger study population (300), whereas this study was purely diagnostic with emphasis on women with palpable breast lump only and had a total of 120 participants. It can also be attributed to late presentation. It is however noted that spiculated breast opacity may also be due to radial and surgical scars, which can be differentiated from carcinoma by histology.[45]

Axillary opacities indicative of enlarged lymph nodes with replaced hilum were seen in about 83% of images that were classified as the Final BIRADS category 5, and 24 out of 30 were reported malignant on histopathology. This is higher than the frequency reported by Sambasivarao et al.[21] (36%) among patients with palpable breast malignancies in their study, who stated that with increase in size of carcinoma there is increased involvement of the axillary nodes; this could be attributed to what was reported in this study, due to the peculiarity of late presentation, with attendant advanced disease, common in our environment.

Suspicious calcification within a mass, which is a strong indicator for malignancy, was seen in 26.7% of case categorized as BIRADS 5 in this study, and it also had a 100% association with malignancy on histopathology. This is similar to the frequency reported by Sambasivarao et al.,[21] in which 6 cases (21.4%) of malignant masses showed suspicious calcification on mammography.

In this study, nipple retraction was seen in 13 (43%) out of 30 patients categorized as BIRADS 5, with 100% association with malignancy on histopathology. This is higher than the 1% (3 cases) reported by Akinola et al.[27] in their study and could be attributed to the fact that theirs was a combined screening and diagnostic study with a significant proportion (41.7%) being asymptomatic and a lesser proportion (23.3%) presenting with lumps. Nipple retraction, a secondary sign of malignancy, generally is associated with cancers that are large enough to be visible on mammograms and palpable at physical examination.[46]

This study shows that there is a fairly high proportion of mammograms showing features suggestive of breast malignancy that were indeed confirmed to be malignant on histopathology.

Few literatures seen on pattern of findings discussed features seen on mammography in relation with PPV, which was not done in this study. Some literatures seen also considered the effect of age on the accuracy of mammography, which was also not done in this study.

  Conclusion Top

The recent introduction of the facility for mammography at the NAUTH, Nnewi inspired this study to determine the accuracy of the modality in the detection of breast cancer at this institution. Results from this work confirm the overall accuracy and high PPV of the imaging modality as provided by the operators at the NAUTH Radiology Department. Mammography should therefore be deployed for breast cancer screening among the adult population within the southeast region of Nigeria, with referral for follow-up imaging or biopsy when indicated.

Financial support and sponsorship


Conflicts of interest

There are no conflicts of interest.

  References Top

Veronesi U, Boyle P, Goldhirsch A, Orecchia R, Viale G Breast cancer. Lancet 2005;365:1727-41.  Back to cited text no. 1
Parkin DM, Bray F, Ferlay J, Pisani P Global Cancer Statistics, 2002. CA Cancer J Clin 2005;55:74-108.  Back to cited text no. 2
The Lancet. Breast cancer in developing countries. Lancet2009;374:1567-2131.  Back to cited text no. 3
Bray F, Ferlay J, Soerjomataram I, Siegel RL, Torre LA, Jemal A Global Cancer Statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin 2018;68:394-424.  Back to cited text no. 4
International Agency for Research on Cancer, IARC GLOBOCAN 2008. Available from: http://globocan.iarc.fr/factsheet.asp.Accessed on January 17, 2020.  Back to cited text no. 5
Ojewusi AA, Obembe T, Arulogun OS, Olugbayela T Breast cancer awareness, attitude and screening practices in Nigeria: A systematic review. Acad J 2016;7:11-25.  Back to cited text no. 6
Anderson BO, Jakesz R Breast cancer issues in developing countries: An overview of the Breast Health Global Initiative. World J Surg 2008;32:2578-85.  Back to cited text no. 7
Adebamowo CA, Ajayi OO Breast cancer in Nigeria. West Afr J Med 2000;19:179-91.  Back to cited text no. 8
Nigeria: Breast Cancer 2014. World Health Rankings: Live Longer Live Better. Available from: http://www.worldlifeexpectancy.com/nigeria-breast-cancer. Accessed on November 23, 2019.  Back to cited text no. 9
Ferlay J, Shin HR, Bray F, Forman D, Mathers C, Parkin DM Estimates of worldwide burden of cancer in 2008: GLOBOCAN 2008. Int J Cancer 2010;127:2893-917.  Back to cited text no. 10
Onyiaorah IV, Ukah CO, Anyiam DC, Onyiaorah AA, Efobi CC Statistical comparison of clinical and histologic diagnoses of breast tumours in public and private hospitals. Am J Cancer Epidemiol Prevent 2013;1:23-29.  Back to cited text no. 11
Okobia MN, Osime U Clinicopathological study of carcinoma of the breast in Benin city. Afr J Reprod Health 2001;5:56-62.  Back to cited text no. 12
Unger-Saldaña K. Challenges to the early diagnosis and treatment of breast cancer in developing countries. World J Clin Oncol 2014;5:465-77.  Back to cited text no. 13
Gakwaya A, Kigula-Mugambe JB, Kavuma A, Luwaga A, Fualal J, Jombwe J, et al. Cancer of the breast: 5-year survival in a tertiary hospital in Uganda. Br J Cancer 2008;9:63-7.  Back to cited text no. 14
Anyanwu SNC Temporary trends in breast cancer presentation in the third world. J Exp Clin Cancer Res 2008;27:17.  Back to cited text no. 15
Danfulani M, Ahmed SS, Mohammed MS, Awwal MM Pattern of mammographic findings in Sokoto, Nigeria. Asian J Med Sci 2014;5:79-83.  Back to cited text no. 16
Anyanwu SN Breast cancer in Eastern Nigeria: A ten year review. West Afr J Med 2000;19:120-5.  Back to cited text no. 17
Ntekim A, Nufu FT, Campbell OB Breast cancer in young women in Ibadan, Nigeria. Afr Health Sci 2009;9:242-6.  Back to cited text no. 18
Shyyan R, Masood S, Badwe RA, Errico KM, Liberman L, Ozmen V, et al; Global Summit Diagnosis and Pathology Panel. Breast cancer in limited-resource countries: Diagnosis and pathology. Breast J 2006;12(Suppl. 1):S27-37.  Back to cited text no. 19
Barlow WE, Lehman CD, Zheng Y, Balllard-Barbash R, Yankaskas BC, Cutter GR, et al. Performance of diagnostic mammography for women with signs and symptoms of breast cancer. J Natl Cancer Inst 2002:94:115.  Back to cited text no. 20
Sambasivarao K, Chand VS, Anuradha B, Sudharani B Correlation between sonomammography and mammography in the evaluation of breast lesions. IOSR J Dent Med Sci 2016;15:13-20.  Back to cited text no. 21
Quershi SA, Rehman K, Muhammad D, Khan MI, Jaffar H Validity of mammography according to BIRADS scoring in relation with histopathology among females presenting with clinically palpable breast lump or nipple discharge. Ann Pak Inst Med Sci 2014;10:150-4.  Back to cited text no. 22
Arsalan F, Subhan A, Rasul S, Jalali U, Yousuf M, Mehmood Z, et al. Sensitivity and specificity of BIRADS scoring system in carcinoma of breast. J Surg Pak (Int) 2010;15:38-43.  Back to cited text no. 23
Tan KP, Mohamad Azlan Z, Rumaisa MP, Siti Aisyah Murni MR, Radhika S, Nurismah MI, et al. The comparative accuracy of ultrasound and mammography in the detection of breast cancer. Med J Malaysia 2014;69:79-85.  Back to cited text no. 24
Isreal GD Determining sample size, University of Florida Cooperative Extension Service, Institute of Food and Agriculture Sciences, EDIS. Fact sheet PEOD-6:1992:1-5.  Back to cited text no. 25
Ebubedike UR, Umeh EO, Anyanwu SNC, Ukah CO, Ikegwuonu NC Pattern of mammography findings among symptomatic females referred for diagnostic mammography at a tertiary center in South-East Nigeria. West Afr J Radiol 2016;23:23-27.  Back to cited text no. 26
Akinola RA, Akinola OL, Shittu L, Balogun BO, Tayo AO Appraisal of mammography in Nigeria women in a new teaching Hospital scientific. Res Essay 2007;2:325-9.  Back to cited text no. 27
Majeed AI, Naz N, Arif M, Majeed A Diagnostic accuracy of mammography in detecting breast cancer keeping histopathology as gold standard. Ann Pak Inst Med Sci 2016;12:118-21.  Back to cited text no. 28
Esserman L, Cowley H, Eberle C, Kirkpatrick A, Chang S, Berbaum K, et al. Improving the accuracy of mammography: Volume and outcome relationships. J Natl Cancer Inst 2002;94:369-75.  Back to cited text no. 29
Devolli-Disha E, Manxhuka-Kërliu S, Ymeri H, Kutllovci A Comparative accuracy of mammography and ultrasound in women with breast symptoms according to age and breast density. Bosn J Basic Med Sci 2009;9:131-6.  Back to cited text no. 30
Jensen A, Geller BM, Gard CC, Miglioretti DL, Yankaskas B, Carney PA, et al. Performance of diagnostic mammography differs in the United States and Denmark. Int J Cancer 2010;127:1905-12.  Back to cited text no. 31
Majid AS, de Paredes ES, Doherty RD, Sharma NR, Salvador X Missed breast carcinoma: Pitfalls and pearls. Radiographics 2003;23:881-95.  Back to cited text no. 32
Burrell HC, Sibbering DM, Wilson AR, Pinder SE, Evans AJ, Yeoman LJ, et al. Screening interval breast cancers: Mammographic features and prognosis factors. Radiology 1996;199:811-7.  Back to cited text no. 33
Mustapha Z, Abubakar A, Modu AA, Pindiga UH, Okedayo M, Annongu IT. Histological outcome of BIRADS 5 breast lesions in Maiduguri, north eastern Nigeria. Borno Med J 2014;11:119-22.  Back to cited text no. 34
Taif S, Tufail F, Alnuaim AS Mammography performance in Oman: Review of factors influencing cancer yield and positive predictive value. Asia-Pac J Clin Oncol 2014;1:1-9.  Back to cited text no. 35
Orel SG, Kay N, Reynolds C, Sullivan DC BI-RADS categorization as a predictor of malignancy. Radiology 1999;211:845-50.  Back to cited text no. 36
Tan YY, Wee SB, Tan MP, Chong BK Positive predictive value of BI-RADS categorization in an Asian population. Asian J Surg 2004;27:186-91.  Back to cited text no. 37
Hirunpat S, Tanomkiat W, Khojarern R, Arpakupakul N Accuracy of the mammographic report category according to BIRADS. J Med Assoc Thai 2005;88:62-5.  Back to cited text no. 38
Zonderland HM, Pope TL Jr, Nieborg AJ The positive predictive value of the Breast Imaging Reporting and Data System (BI-RADS) as a method of quality assessment in breast imaging in a hospital population. Eur Radiol 2004;14:1743-50.  Back to cited text no. 39
Bukhari H, Shaukat A, Ahmad N Breast cancer screening; mammography versus dynamic MRI breast. Prof Med J 2017;24:42-6.  Back to cited text no. 40
Omidiji OA, Campbell PC, Irurhe NK, Atalabi OM, Toyobo OO Breast cancer screening in a resource poor country: Ultrasound versus mammography. Ghana Med J 2017;51:6-12.  Back to cited text no. 41
Dee KE, Sickles EA Medical audit of diagnostic mammography examinations: Comparison with screening outcomes obtained concurrently. Am J Roentgenol 2001;176:729-33.  Back to cited text no. 42
D’Orsi CJ, Sickles EA, Mendelson EB, Morris EA ACR BI-RADS® Atlas, Breast Imaging Reporting and Data System. Reston, VA: American College of Radiology; 2013. Available from: https://www.acr.org. Accessed on December 12, 2020.  Back to cited text no. 43
Gurung G, Ghimire RK, Lohani B Mammographic evaluation of palpable breast masses with pathological correlation: A tertiary care centre study in Nepal. J Inst Med 2010;32:21-5.  Back to cited text no. 44
Michell MJ, Lawinski C, Teh W, Vinnicombe S The breast. In: Sutton D, editor. Textbook of Radiology and Imaging. Vol. 2. Edinburg, UK: Churchill Livingstone; 2003. p. 1464-75.  Back to cited text no. 45
Kopans DB Anatomy, histology, physiology, and pathology. In: Breast Imaging. 2nd ed. Philadelphia, PA: Lippincott-Raven; 1998. p. 14-5.  Back to cited text no. 46


  [Figure 1], [Figure 2], [Figure 3], [Figure 4], [Figure 5]

  [Table 1], [Table 2], [Table 3]


    Similar in PUBMED
   Search Pubmed for
   Search in Google Scholar for
 Related articles
    Access Statistics
    Email Alert *
    Add to My List *
* Registration required (free)  

  In this article
Materials and Me...
Article Figures
Article Tables

 Article Access Statistics
    PDF Downloaded39    
    Comments [Add]    

Recommend this journal