Analysis of our cases of mixed ductal/lobular histology has yielded several facts. Tumor size was larger in our mixed ductal/lobular histology patients, which may account for the higher rate of mastectomies. Tumor grade did not substantially differ between the ductal and mixed ductal/lobular histologies. Mixed ductal/lobular histology tumors were not significantly more likely to be ER negative and PR negative. Treatment did not also significantly differ between the 2 groups, with both groups receiving chemotherapy, radiation, and hormonal therapy at similar rates. Paired sample t-testing revealed no statistically significant difference between the 2 groups of patients (mixed ductal/lobular versus ductal histology) with regard to tumor grade or TNM stage.
Although the literature on this topic is scant, our data is in agreement with the published report of Sastre-Garau et al, who included 249 cases of mixed ductal/lobular histology tumors in their review. They noted a 10-year contralateral disease-free rate of 90% in ductal breast cancers and 90% in lobular cancers while mixed ductal/lobular histology tumors had a significantly diminished rate of 80%. As in our series, overall survival among the two subtypes did not significantly vary. In our more recently diagnosed cohort of patients, for whom breast MRI had become available, none of our patients had a contralateral synchronous invasive breast tumor and only one patient had a breast lesion (in the mixed ductal/lobular tumor group) which was pathologically a case of ductal carcinoma in-situ (DCIS). One patient in the IDC group had a benign fiboadenoma in the contralateral breast. Thus, there was no significant difference between the two groups of patients (mixed ductal/lobular histology and purely ductal carcinoma) with regard to synchronous contralateral breast tumors.
Arpino et al. recently published their institutional experience with ILC. The median age of their patients with ILC was 64.6. The proportion of ER-positive tumors was 92.7%, and PR was expressed in 67.4% of ILCs. The pattern of metastatic dissemination in ILC was also different. ILC was three times more likely to metastasize to the peritoneum, gastrointestinal tract, and ovaries (6.7% versus 1.8%). Information on contralateral breast tumors was also available on the subset of 2,855 patients in whom sites of breast cancer distant from the primary could be assessed. Contralateral breast cancers in this group were more frequent among those with ILC (20.9%) than among those with IDC (11.2%). No difference in overall or disease-free survival was noted between ILC and IDC. Bharat et al. recently compared the outcomes for IDC, ILC, and mixed ductal/lobular histology tumors at their institution. Patients with mixed ductal/lobular histology tumors and ILC were more likely to have low grade and hormone-receptor positive tumors, but also more likely to have stage III disease. The 10-year long-term survival, however, was better in patients with ILC (69%) and mixed ductal/lobular histology tumors (68%) than in patients with IDC (61%).
Clinically, in our series, the behavior of mixed ductal/lobular tumors seemed to demonstrate some important differences from their ductal counterparts. The rate of distant metastatic spread was much lower at 8% compared to a rate of 19% for the ductal tumors. The sites of spread were the lungs and bones, and mimicked the pattern of metastases by IDC. The rates of local/regional relapse were identical between the pure ductal and mixed ductal/lobular histology tumors, with both histologies demonstrating a 6% rate. However, the rate of second primary breast cancers was much higher at 30% compared with a rate of 11% with ductal histology. The rates of hormone-receptor positivity and age of onset were similar between the mixed ductal/lobular histology and purely ductal histology and did not differ from historically cited data in the literature. However, unlike ILC, the rate of synchronous contralateral breast cancer (0%) was much lower, mimicking the ductal histology. This value was highly significant and suggests a different pattern of recurrence and different tumor biology. The literature also indicates that the mixed ductal/lobular histology, like ILC, is more likely to be associated with hormone replacement therapy[27, 28]. Reeves et al. conducted 3.6 million person-years of follow-up on over 1 million post-menopausal women in the UK. They found that the largest relative risks in current users of hormone therapy compared with never users of hormone therapy were seen for lobular (relative risk 2.25, 95% CI 2.00--2.52), mixed ductal/lobular (2.13, 1.68--2.70), and tubular cancers (2.66, 2.16--3.28). The only subtype that did not rise in incidence in their study was medullary carcinoma. Reeves et al. followed up with an analysis of reproductive factors and histologic subtype in their cohort of patients and interestingly found that "for most of the reproductive factors considered, the relative risks for mixed ductal/lobular carcinoma were intermediate between those found for ductal and lobular cancer."
Our immunohistochemistry data suggests that mixed ductal/lobular carcinoma is another pleomorphic variant of IDC as 90% of the cases stained positively for E-cadherin. However, more tissue will be required to statistically confirm this trend. Of the available blocks, only one case did not stain for E-cadherin and may truly represent a lobular carcinoma. Wheeler et al. have characterized tubulocarcinomas of the breast. Although our tumors did mostly stain for E-cadherin as well, we do not think our tumors represent tubulolobular carcinomas as "this histologic pattern is distinct from other mixed ductal/lobular carcinomas in which the invasive components are often separate and the tubular component lacks a lobular growth pattern." Additionally, morphologically, our cases did not meet the diagnostic criteria for tubulolobular carcinoma due to a lack of any tubular elements.