INTRODUCTION

The definition of elderly women varies according to the context, but it is recognized that the incidence of many diseases, including cancer, increases with age [1-3]. In 2016, it remains a major public health problem with one of the highest incidences worldwide, and approximately 11,800 deaths per year [4]. In Western countries, about 20% of the population is aged 70 years and older. In 2000, there were 35 million women aged 65 years and over, and given the gradual increase in life expectancy, this figure will double in 2030 [5]. In Cameroon, breast cancer (BC) is the most common cancer in women with 48.12% of cases [6]. In Senegal, BC is the second most common gynecological cancer in women over 60 years [7]. In Guinea, BC in 2018 was ranked at 4th with 605 new cancer cases [8]. In 2030, approximately 70% of cancers will be found in the elderly population [5]; however, the phenomenon of population aging, whatever its origin, is growing steadily.

Currently, nearly one-third of cancers occur after 75 years, and this figure is expected to reach 50% in 2050 [10]. In addition to this incidence, BC in elderly women is diagnosed at a more advanced stage, three times more than the metastatic forms in young women [9].

In most western series, BC in elderly patients has a good prognosis, characterized by an increase of hormone receptor expression, a decreased of HER2 overexpression, and low proliferation index [7]. Geriatric evaluation is essential to avoid "over- or under-medication" of patients [10]. The aim of this retrospective study was to describe the clinicopathological characteristics and prognosis of BC in elderly women at the surgical oncology unit of Donka National Hospital.

MATERIAL AND METHODS

We retrospectively reviewed the database of all patients who have newly diagnosed BC at the Surgical Oncology Unit at Donka National Hospital, between April 2007 and December 2017. A population of 607 patients was identified, and their medical records and treatment charts were retrospectively reviewed to obtain information on the patient and tumor characteristics. From this group of patients, 83 elderly women aged 60 years and older with BC were identified, and these patients were evaluated in this study. 

The clinical information included the socio-demographic characteristics (age, socio-professional categories, marital status, risk factors, cancer history), malignancy-related features (clinical data, time to consultation, mode of discovery), and preclinical and diagnostic investigations results (histology, mammography, ultrasound, radiography, computed tomography). Balducci's geriatric score was based on patient autonomy and co-morbidities [11]. The clinical-stage was classified according to the TNM staging system proposed by the International Union for Cancer Control (UICC, 8th edition) [12]. 

The treatment modalities have been listed. In the case of surgery, the curative or palliative goal and the type of surgery were specified. The postoperative status was described. The associated treatments, chemotherapy, and radiotherapy were reported. Patient follow-up and survival time were listed. 

All these data analyses were performed on SPSS 21.0 software. Categorical variables were calculated in terms of percentage, and quantitative variables were analyzed as mean (± standard deviation) or median with interquartile range (IQR). The overall survival was calculated according to the Kaplan Meier method. Survival data were compared with the Log Rank test. The test was significant if the p-values were less than 0.05.

RESULTS

Of the 607 patients with histologically confirmed BC, 83 (13.6%) elderly women (≥60 years) were diagnosed between 2007 and 2017. The mean age was 67.6 ± 6.6 years (range, 60 to 90 years). They were married in 83 (100%) cases and 72 of them (86.7%) were housewives. A history of first-degree cancer was reported in 2 cases, a second degree in 2 cases, and third-degree in 1 case. The mean age at the first menstruation was 14 (±1.4) years, with a range of 12 and 18 years. Only 2 (3.4%) of them had their first pregnancy over 30 years and in 46 (79.3%) cases they had their first pregnancy before 20 years. The mean age at menopause was 47.7 (± 4.8) years. Oral contraception was reported in 5 (6.02%) cases.

The body mass index (BMI) was recorded in 78 cases, with a range of 18.5 and 24.9, in 35 (44.8%) cases and between 25 and 29.9, in 24 (30.7%) cases, with a mean of 24.5 (±5.4) kg/m2. Comorbidities were found in 38 (45.8%) cases. The most common comorbidities were hypertension in 35 (42.1%) cases and diabetes in 6 (7.2%) cases. Association hypertension and diabetes were found in 3 (3.6%) cases. The geriatric score of Balducci was classified as Balducci II in 49 (59.1%) cases. Patients consulted within 1 to 180 months after the initial clinical signs, with a mean time to consultation of 20 (± 33.7) months. The tumor was in the right breast in 42 (50.7%) cases and to the left breast in 41 (49.4%) cases. The common clinical signs were represented by breast mass or lump in 71 (85.5%) cases, breast pain in 44 (53%) cases, ulceration in 19 (22.9%) cases, and peau d’orange in 18 (21.7%) cases (Figure 1). Axillary lymph nodes metastasis was found in 62 (74.7%) cases and supraclavicular lymph nodes in 13 (15.7%) cases.

Ductal invasive carcinoma in 60 cases was the most common histologic subtype, followed by lobular invasive carcinoma in 3 cases. Regarding the histological grade, the Scarff Bloom Richardson (SBR) II was the most common and has been found in 28 cases of the 42 reported.  Of the eight patients who have had immunohistochemistry, hormone receptors were positive in 6 patients and the Her2 oncogene was overexpressed in 2 patients. The biomarker CA15.3 was assayed in 36 (43.3%) cases among which, 25 were range within normal values, and in 11 cases the values were higher than the normal value 30 IU/ml. The metastasis was reported in 16 cases, including 11 pulmonary, 2 hepatic, and 2 hepatic/pulmonary. Cardiac Doppler ultrasound was performed in 52 (62.6%) cases, among which cardiomyopathy was reported in 11 cases.

At the time of diagnosis, 25.3% of these women had cT3 and 68.7% had cT4, and 19.2% had metastatic cancer. In the current study, more than half of the women were with advanced cancer (cT4 or metastatic). Inflammatory tumors represented 6 (7.2%) cases. Sixty- two patients (51.8%) had initially axillary lymph node metastasis. The clinical-stage was assigned according to the 8th edition of the UICC staging system, 55 (63.3%) patients had stage III, and 17 (20.5%) patients had stage IV.

Surgical treatment was performed in 43 (51.8%) cases including, radical breast surgery (RBS) in 39 cases and conservative breast surgery (CBC) in 4 cases. Postoperative complications occurred in 12 (27.9%) cases, including lymphedema in 7(16.2 %) cases, seroma 3 (6.9 %) cases, hemorrhage 1(2.3%) case, and release of threads 1(2.3%). The chemotherapy was administered in 68 (81.9%) cases including, 41(60.3%) in neoadjuvant and 27(39.7%) cases in adjuvant. In Nine (10.8%) cases radiotherapy was performed to the countries where this therapy is available. Hormonotherapy was administered to 20 (24.0%) patients after chemotherapy and/or radiotherapy, including tamoxifen in 16 cases and letrozole in 4 cases.

The mean duration of follow up was 28.8 (± 28.5) months. At the time of analysis, 11 patients experienced recurrence. The site of recurrence was locoregional in 4 patients, metastatic in 5 patients, and multi-visceral in 2 patients. At last count, 40 (48.1%) patients were alive and 43 (51.8%) patients died. All the deaths were related to cancer. The 2, 3 and 5-year overall survival rates were 47%, 34% and 15%, respectively (Figure1, 2). The survival rates at 24 months were 37% after recurrence and 47% without recurrence. According to the clinical TNM stage, the overall 5-year survival was 34% for the local stage, 16% for the locally advanced stage, and 8% for the metastatic stage.

Figure 1

The overall survival curve

Figure 2

The survival curve according to the recurrence

DISCUSSION

The limitation of this study was the lack of immunochemistry for most of the patients. However, the study allowed to trends relatively high frequency of BC in Guinean elderly women through to hospital data. BC is seen more frequently in the postmenopausal period. In this study, elderly women represented 13.6% of all patients with BC. A similar finding has been reported by some authors in sub-Saharan Africa [6, 13]. This proportion was very lower in Cameroonian study’s that reported 3.1% of elderly women [14]. The low incidence of BC among elderly women in the current study would be related to the relatively younger age (which is 48 years old) of the female BC population in Guinea [15]. Lack of early detection and screening programs could also decrease the number of cases in this older female population. Family history of cancer increases the risk by 80% when there is a first-degree history, by three times fold if two first-degree histories coexist, and by four times if there are three or more [16]. We identified 5 cases of family history of cancer (first-degree cancer was reported in 2 cases, a second degree in 2 cases, and third-degree in 1 case). In our study, we were unable to access genetic tests to identify mutations in BRCA1, BRCA2, and other genes abnormalities. A family history of BC in a family should lead  to genetic counseling for the identification of these mutations.Precocious, or earlier than unexpected menarche is an important risk factor since the first menstruation before 12 years will increase the risk of developing BC in adulthood. The main reason is due to a long duration of breast tissue exposure to estrogen activity in particular. Conversely, the onset of menstruation over 14 years is considered to have a protective role [16]. Compared to our results, the study by Haddad et al [17] indicated that the mean age at puberty was 13 (±1.4) years. This finding is similar to that in our results. The mean time between the first sign and the first oncological visit in our study was 20 (± 33,7) months (range, 1-180 months), this time was longer than that reported in two (2) Maghrebian studies, which were 11.2 months and 12 months [18, 19], respectively. The main reasons for this late visit are multiple and differ from those encountered in developed countries. The lack of interest of physicians and their ignorance of the therapies available in the country is more important than the neglect of patients. In the current study, the long period between the first symptoms and the first visit was frequently found in women with modest incomes. It seems that it is always the patients themselves who have discovered a breast abnormality. In this study, the most common clinical presentation was a breast mass, with 71 (85.5%) cases. This is similar to almost the published data reported in the literature [17,20].

Hypertension and diabetes were the most frequent comorbidities that were associated with poor prognostic factors. Our data were lower than those of Jedidi et al [18] who reported hypertension in 74.4% of patients and diabetes in 52.9%, and Haddad et al [16] who found hypertension and diabetes in 79.2% of patients.  These differences observed in our study could probably be explained by the lifestyle, including sedentary and eating habits. At the time of diagnosis, more than half of the women were with advanced cancer (cT3/cT4 or metastatic). This is different from most data in the literature [17;18]. Also, some studies have shown that women beyond 80 years are more likely to have advanced disease at diagnosis [21]. Generally, it has been found that the multiplicity of clinical consultations before the first oncological visit delayed the diagnosis. The most histological subtype frequently encountered in female BC is ductal invasive carcinoma (85–90%) [22; 23]. The result in our study was similar with a ratio of 95.1%, and this was significantly higher than the other histology. While ductal invasive carcinoma was found to be 94.9% in our study and confirms by other studies, we had no patients with ductal carcinoma in- situ. In our study, receptor analysis was performed in 8 (9.6 %) cases. In these patients analyzed for receptors, while the frequency of hormone receptor-positivity was similar to literature, overexpression of the oncogene Her2 ratio was lower, with 2.25% of cases [24]. This low frequency in our study could be explained by the absence of immunochemistry in our country.  Balducci's score II was the most common 49 (59%) cases. Besides, Flipo et al [25] in Nice found 56% of cases for the score I and 44% of cases for scores II and III. This difference could be explained by the unbalanced lifestyle and the irregular follow-up of co-morbidities in our context. BC in elderly women should be treated with the same strategy at a young age. Although no consensus is reached regarding locoregional therapy in elderly patients, mastectomy with axillary node dissection remained the cornerstone of the therapy. In previous studies, RBS was preferred, since the rationale for this was the localization of the lesion near to the pectoralis major muscle and the tumor being in a more advanced stage in elderly women compared to young women at the time of diagnosis [17]. Recent studies are in favor of modified radical or simple mastectomy combined with radiation therapy [18]. There seems to be no prominent difference in survival between radical mastectomy and other methods [17, 18]. In the present study, RBS was performed in 39 patients which is lower than that reported by Haddad et al [17] and Jedidi et al [18]. In these two studies, RBS was performed in 70.1% and 73.9% of elderly women, respectively [17, 18].  The mains reasons of this difference could be explained on the one hand, by the treatment cost, patients' refusal despite physicians’ recommendation in 56% of cases [26], on the other hand, the medically unfit patients’ condition to undergo surgery justified by the WHO performance index and Co-morbidities. The surgical complications were dominated in our study by lymphedema with 16.3%, which was lower than that of Ben et al [27] in Tunisia who reported  23% of cases in their study. Similarly, the data on chemotherapy (81.9%) were different from those reported by Haddad et al [17] who found 71.6% of cases. This difference could be explained by the late stage of cancer at diagnosis in our country. The hormone therapy effect is more related to the positivity of hormone receptors. Our results 24.0% were lower than those reported by Haddad et al [17] where 35.8% of patients received hormone therapy. This difference could be explained by the lack of immunohistochemistry in our context. Radiotherapy reduces the local relapse risk by more than 60% both after conservative surgery and after mastectomy [5]. Our results 10.8% were inferior to those of Colombo et al [28] in Montpellier who reported 88% of radiotherapy. The absence of a radiotherapy unit in Guinea would explain this discrepancy. With a mean follow-up time of 60 months, our survival data were lower than those reported by Jedidi et al [18]. At the last count, they found 50 (73.5%)patients alive and 18 (26.5%) dead. The overall survival rate was 88% at 2 years, 80% at 3 years, and 65% at 5 years. In our study, the survival rate was 47% at 2 years, 34% at 3 years, and 15% at 5 years. This difference can be explained by the delay in diagnosis, the cost related to the BC treatment the irregular follow-up, and the non- observance in the treatment in our context.

CONCLUSION

Breast cancer remains a public health problem. In elderly women, it accounted for about one-seventh of all histologically confirmed breast cancers. Effort must be directed towards early detection, access to genetic tests to identify families at risk, and regular monitoring of comorbidities; all necessary to I, prove the survival of elderly women with BC.