Secretory carcinoma is a rare form of breast cancer, generally with a favourable prognosis, since literature reports few cases with locoregional or distant metastases. Although this disease represents 84% of all the histological variants of infant breast carcinoma , secretory carcinoma can also occur in adults, both in men and women. The tumour has been found in patients ranging from three to eighty-seven years of age [1, 8, 9]. Most of the cases reported in literature involve female patients of the average age of 40 years and a median of 33 . Only 15 cases have been observed in males (average age 17) with a male:female ratio of 1:6, which is relatively high compared with the distribution of other histological types of breast carcinoma in males [11, 12].
The correlation between known risk factors for breast carcinoma and secretory carcinoma is not yet fully understood. Since the presence of oestrogen and progesterone receptors is extremely variable in patients with secretory breast carcinoma, the etiopathogenesis of such tumours is probably not linked to female sex hormones.
Many authors have shown the presence of several chromosomic anomalies in secretory carcinomas, although the significance of such genetic alterations has still to be clarified [7, 10, 13]. Our own case showed the two more distinctive pathological characteristics of such tumours: intracellular and extracellular secretion and granular eosinophilic cytoplasm of the neoplastic cells.
Immunohistochemically the negative stain for oestrogen and progesterone receptors and the positive stain for S100 are in agreement with previous reports [9, 14, 15].
The primary tumour showed a low risk factor (diameter < 2 cm, low histological grade, absence of vascular invasion, low proliferation rate assessed by Ki67, Her-2 negative immunostain), with the exception of the hormone receptors, which proved to be negative.
In our case we found strong membranous immunostain for e-cadherin. Previous studies reported that e-cadherin is present in infiltranting ductal carcinoma and its loss has been reported in most of infiltranting lobular carcinoma [16, 17]. This supports the hypothesis that the secretory carcinoma might originate from the ductal component of the mammary gland and might therefore be considered as a variant of ductal carcinoma.
To our knowledge, this is the first report of secretory carcinoma in which the e-cadherin expression has been investigated. On the other hand, positivity for S100 and, focally, for CD10 and negativity for oestrogens and progesterone suggests that the tumour might originate in the basal cells of the duct, whose immunophenotype has recently been described by Kesse-Adu R and Shousha S . If this view is correct, oestrogen and progesterone negativity should not be considered as the indication of an unfavourable outcome, but merely as a feature linked to the histogenetic origin of the tumour. The analysis of the DNA cell content of the secretory carcinoma proves to be diploid with a low proliferative activity [19–23]. In our case, flow cytometry performed on the cell DNA (65,520 cells) also showed a diploid-type histogram (DI = 1.0) with a low proliferative activity (S-phase = 4.5%).
Whatever the age or sex of the patient, the secretory carcinoma can usually be felt at palpation. Reported dimensions of the tumour range from a minimum of 1 cm to a maximum of 16 cm [2, 21].
In agreement with reports published by several other authors , the tumour observed by us also presented clinical and ultrasound features similar to those of a fibroadenoma – clear margins, smooth surface and mobile on both the upper and the lower levels. Partly as a result of the clinical and echographic features of the nodule, and also following the specific request of the patient, who preferred to undergo nodulectomy only, we did not perform a core biopsy of the nodule, which would most certainly have led to a more accurate therapeutic approach and made it possible to avoid the second surgical intervention.
Breast X-rays have rarely been performed in patients with secretory carcinomas; in a multicentre study, Richard et al. described a few cases where there was radio-opacity with defined margins and the presence of microcalcifications .
Since this type of tumour is extremely rare, there is still no unanimous consensus about treatment choice. Many authors maintain that the frequency of local disease relapse after the simple surgical removal of the neoplasia only would suggest that mastectomy should always be performed [8, 14].
According to Richard et al  about 33% of adult women treated with extensive surgery of the tumour site presented local disease relapse.
Whenever possible, prepubertal girls should be treated initially by wide local excision. Preservation of the breast bud should be attempted, but as this is not always possible, breast development may sometimes be impaired .
When the neoplasia is large, mastectomy should be performed in order to avoid the probability of local relapse .
Post-operative radiotherapy should be proposed after conservative surgery in adult patients, but is not advised for children because of the possible secondary effects such as fibrosis of the lung, rib damage and the consequent asymmetry of the rib cage [15, 25]. In our own case, since the patient was an adult woman with a tumour of less than 2 cm, and since the histological examination of one of the nodule resection margins was not sufficiently clear, we decided on quadrantectomy followed by subsequent radiotherapy.
Although locoregional and distant metastases from the secretory carcinoma are extremely rare, they are nevertheless possible. For the correct staging of the disease, therefore, it is necessary to verify the possible presence of sytemic and/or axillary lymph node metastases.
There is still considerable discussion about the indications for a complete axillary lymph node dissection. Since the secretory carcinoma has a reduced metastatic capacity, it is obvious that a total axillary lymphoadenectomy performed on principle would be in fact a form of over-treatment for this particular disease stage. The frequently-reported post-operative complications of axillary dissection, such as pain, parasthesia, seroma, difficulty of shoulder movements and lymphedema of the upper arm are all further valid reasons for a cautious approach towards this therapeutic choice.
In the last few years, in order to avoid where possible the complete dissection of the axillary fossa in ductal and lobular breast cancer, a biopsy of the sentinel lymph node is usually performed. In our own case we also tried to use this technique, which had never been reported in literature at the time of surgery (1999). We used both intravital stain and a radiotracer with a portable radioisotope detector for the intraoperative identification and subsequent removal of the lymph node without any difficulty, even though these substances were injected directly into the surgical scar.
The histological examination of the sentinel lymph node was performed on formalin-fixed, paraffin-embedded material and showed the presence of metastases; this made it imperative to perform further surgery for total axillary lymphoadectomy. Reported data show that in 40% of cases, the other lymph nodes removed after sentinel lymph node positivity have not undergone metastatic colonisation  and this proved to be true in our patient.
Although several cases of secretory carcinoma have been treated with adjuvant chemotherapy, there are no published data about the real effectiveness of this therapeutic choice. Herz et al have reported non-responsiveness of the tumour to chemotherapy . In our own case, since we had found axillary metastases, we administered adjuvant chemotherapy consisting of six cycles of CMF.