Clinicohistologic Characteristics of Breast Cancer in Ghanaian Patients

Background: Breast cancer is the leading cause of cancer morbidity and mortality worldwide. The management and prognosis of breast cancer depend heavily on the different histologic and molecular biologic features of the tumour. The different histologic types describe the distinct growth patterns and cytological features of the tumour. Methodology: This is a retrospective study of archival breast cancer excision and mastectomy specimen at KBTH from 2012-2016. 729 cases were retrieved over the period and examined by two pathologists independently. Demography of the cases, tumour size, grade, histologic type, stage, mitosis, site of lesion etc. were entered into SPSS and analysed with chi-square done with P-value set at P < 0.05. Results and Conclusion: The mean age of presentation is 52.45 ± 12.75 years. The commonest histologic type of breast cancer is invasive carcinoma (NOS) forming (87%). Only 1.2% of male presented with breast cancer with the other percentage in females. Most (88%) of the tumours were greater than 5cm at the time of surgery. The tumours are of high grade (II and III) forming 88%. Seventy-eight percent of cases presented with late stage of the disease (≥ stage IIB). There was association between histologic type and vascular invasion (P < 0.000) and lymph node involvement (P = 0.010). Moreover, tumour size showed an association with tumour grade (P < 0.05). This study has shown that breast cancer among Ghanaian patients has a worse histologic type with poor tumour characteristics giving it poor prognosis.


Introduction
Breast cancer is the leading cause of cancer morbidity and mortality worldwide. The management and prognosis of breast cancer depend heavily on the different histologic type and molecular biologic features of the tumour [1] . The different histologic type of breast cancer and its variant has already been established [2] . The histological types describe the distinct growth patterns and cytological features of the tumour. They comprise different epidemiology, clinical and prognosis. Breast cancer can broadly be classified as in-situ carcinoma and invasive carcinoma. The in-situ carcinoma is the preinvasive carcinomas they grow from the terminal ductal-lobular normal cells and are therefore referred to as lobular carcinoma in-situ and ductal carcinoma in-situ [3] . Like the premalignant class, the invasive carcinomas fall under the invasive ductal carcinoma and lobular carcinoma broadly. Within these broad categorizations are different variants. Breast cancer can be categorized under 21 different histological categories based on architecture patterns, cell morphology and growth as per WHO classification [2,4,5] . Invasive carcinoma no special type (NST) has over the period been known to be the commonest breast cancer.
The features are varying to fit in any of the special types. The special type of breast cancer which comprises of classical lobular and invasive carcinoma forms about 25% of all breast cancer [4,6] . These varying classifications have prognostic value [4,6] .
Systemic therapy has been determined to be effective in patients with early-stage breast cancer, the three main determining factors are lymph nodes status, tumour size and histologic grade as having been used in routine practice [1] . The Nottingham grading system which is a modification if Bloom Richardson grading system is recognised by most international professional bodies in grading of breast cancer is prognostically relevance was demonstrated in the early part of 1991 [1,5,7] and has also been validated by multiple independent studies. NGS has been combined with lymph node stage and tumour size to form the prognostic index due to the fact that NGS has independent and equally powerful prognostic value. These prognostic indices are referred to as Nottingham Prognostic Index, NPI [1,8] .
The histological grading is the means of identifying the extent of differentiation of a tumour. Although its well known that the histological type of a tumour has prognostic value, most tumours are invasive carcinoma no special type [1,8] . The semi-quantitative approach of determining morphological differentiation of a tumour is quite simple and depends on an examination of haematoxylin and eosin stained slide by a well-trained pathologist. This is based on three morphological features; a) degree of tubule or glandular formation, b) nuclear pleomorphism, c) mitotic count [1,9] .
Several independent studies have shown that the prognostic value of histological grading is equivalent to lymph node involvement and higher than tumour size

Materials and Methods
Study area: The source of the data was from the Department of Pathology Korle-Bu Teaching Hospital which is the premier and largest teaching hospital in Ghana. It is located in Ghana's capital city, Accra and receives cases from all over the country especially the southern half of Ghana.

Data Collection:
This was a retrospective study of archival tissue blocks and slides of breast cancer cases from the Department of Pathology, KBTH. All excision specimen including wide local excision and mastectomy specimen blocks and haematoxylin and eosin stained slides between the periods 2012 to 2016 were retrieved. The request forms were also retrieved for these cases. Demography such as age, sex, site of lesion, the location of the tumour in the breast and size of the tumour were retrieved from the request forms.
New slides were prepared for broken and faded slides from the formalin fixed paraffin embedded blocks. New slides were stained with haematoxylin and eosin stain.
Each case was examined by two independent pathologists. Discordant cases were reviewed by a third pathologist and in some case a fourth pathologist. The pathologists were given strict guidelines as to the histological type, grading, staging and margins of tumours.
Histological Type: Each pathologist was given strict guidelines as pertain to WHO breast cancer classification guidelines. The cases were classified using the WHO breast cancer typing.
Histological Grading: The tumour was graded using the Nottingham grading system which is a modification of the Bloom-Richardson grading system. This grading system is based on three parameters; Tubule or glandular formation, Pleomorphism of the tumour cells, Mitotic figures present the tumour. The field diameter of the microscope used in assessing the mitotic figure was 0.59.

Staging :
The staging was done using revised American joint committee on Cancer (AJCC) 8th edition recommended TNM staging criteria with T representing tumour size, N lymph node involvement and M as distant metastasis.

Data Analysis:
The data were entered into SPSS version 22.0 and analysis were done. Descriptive statistics were done for the parameters with tables, graphs and figures drawn. Univariate and multivariate analysis was done for some of the parameters using a Pearson correlation at 95% confidence intervals.

Results
From table 2.1. A total number of 726 age of cases were retrieved over the five year period (2012-2016) with a mean and SD of 52.45±12.75. Of the age of cases, the number of lymph nodes retrieved and involved in the study was 520 for each with a mean and standard deviation of 10.79±5.36 and 3.83±4.74 With regards to the age group of cases (table 2.2) with breast cancer, participants from 50-59years recorded the highest frequency (218) representing 30% of the participants whereas participants ˂20 had the least frequency (4) representing 0.6% From table 2.3, males represented 1.2% while females represented 98% of the participants. Patients who have undergone mastectomy were 68% while those with excision were 32%. Tumours located at the right recorded the highest (49.7%) of laterality of tumour location and tumours with solitary were 94%. However, tumour with size ˃5 had 88.6% with grade II tumour recording 54.9%. With regards to the margin of excision, the complete excision was 80%.the mitotic figures of tumour ˂10 were 47.6% and tumours located within the outer quadrant of the breast was 57%. There was a significant association between age group and histological subtype Malignant phyllodes with a p-value = 0.004 while there was no significant association between the other histological subtypes and age.

Discussion
The mean age of this study is 52.45±12.75 years which is contrary to some of the researches done within the subregion. In a study done by Ngwogu, Offiah [10] , the mean age was reported as 46.5±5.4 years. In that same study, 79.1% of the cases were within 21-50years. Other studies show similar findings which are different from this study with 42.4% less than 50years [10,11].
The mean age is consistent with similar studies done in Ghana [12,13] . The peak incidence age of breast cancer by Ohene-Yeboah and Adjei [13] was 40-49year while in this study it was noted to be 50-59years. The percentage of male to female was consistent with other researchers [10,13] .
Histologic subtypes of breast cancer have over the years been shown as one of the established factors upon which treatment is based on. The histologic diversity of breast adenocarcinoma has different morphological and cytological patterns that have been associated with distinctive clinical presentation and/or outcomes.
In this current study the commonest histologic type of breast cancer is invasive ductal carcinoma no special type (NST) as described by literature [5,10,13,14] .
This was followed by DCIS, mucinous carcinoma and invasive lobular as seen in the study by Edmund et al [14] . There is no appreciable difference between tumours in the left versus right breast as seen by Ngwogu, Offiah [10] . Most of the tumours were located in the upper outer quadrants.
In this study, malignant phyllodes were the commonest sarcoma with 7cases forming 1% of all the malignant cases with an average age of 35±11.04 years. This value is consistent with the literature that put the incidence of malignant phyllodes at less than 1% with peak age incidence of 35 and 55 years. Few cases have been reported in men [15][16][17] . There was no case reported in men. There was a significant case of malignant phyllodes before age 50years (p=0.004).
Annals of Pathology and Laboratory Medicine, Vol. 7, Issue 8, August, 2020 The commonest grade of the tumour was identified as grade 2 followed by grade 3 and grade 1 was the least identified. This phenomenon was similar to most of the studies done in Ghana and other parts of Africa [10,13,14,18,19] . This study has shown that there is a relationship between tumour grade and tumour size (p=0.029), tumour stage (p=0.005) and number of lymph nodes involvement (p=0.010) but no association between the grade and histologic type. This is so because of the aggressive nature of the malignant tumours.
Certain studies have shown that the 5year survival rate of high-grade tumours [14,20,21] and this may be so because of the association that exists between tumour grade and these factors.
The number of lymph nodes retrieved from our study ranges from 1 -42 lymph nodes with a mean of 10.79±5.364 with mean lymph nodes involvement of 3.86±4.744. These findings support that of Edmund, Naaeder [14] . We also identified an association between histologic type of tumour (p=0.010), and tumour grade (p=0.010) and tumour stage (p˂0.001). These associations indicate treatment failure, high recurrence and survival rate of patients.
Staging is one of the most important factors in determining the survival of patients with the disease. This study reports a higher stage disease of breast cancer of stage III and above accounting for 57.2% of the cases consistent with the findings of other researchers in Ghana, Nigeria and other parts of Africa [10,11,13,14,19,22] . We also report of an increasing percentage of early disease probably due to the effectiveness of breast cancer education, self-breast examination and increasing numbers of patients reporting early to the hospital.
The high stage of the disease was also identified to be associated with tumour size (p=0.001). Most of the tumours measure more than 5cm. Tumour size has been shown to be a prognostic indicator; the larger the size the poorer the survival rate [23,24] . It can, therefore, be said that our patients with larger tumour size have a low survival rate. There is no association between tumour size and histologic grade.
Vascular invasion in breast cancer has shown to be an important prognostic factor. We had 70.6% of our cases with vascular invasion as opposed to 5.1% by Edmund, Naaeder [14] this figure stresses the fact that breast cancer in Ghanaian patients has a poor prognosis. Our study also demonstrated an association between vascular invasion and nodal involvement (p˂0.001), histologic type (p˂0.001) and tumour stage (p˂0.001); none of the previous studies done in Ghana has ever shown any associations between vascular invasion and any of these factors.
A multivariate analysis of tumour stage against tumour size, grade, vascular invasion and histologic type has shown positive correlation indicating that breast cancer in Ghanaian woman is generally poor using staging as the prognostic indicator.

Conclusion
It can be deduced from the study that, the frequent histological subtype of breast cancer is invasive ductal carcinoma(NOS) and breast cancer in Ghanaian woman is generally poor using staging as the prognostic indicator.