In the present study, performed with prospectively collected data from the Netherlands Cancer Registry, statistically significant differences in ER and PR expression levels were observed between patients with ILC and IDC. However, the absolute differences were very small and are unlikely to be of clinical relevance. In ER-positive breast cancer patients, more than 80% showed an expression level of more than 90%. Especially in postmenopausal patients aged 50-69 years with ER-positive breast cancer, no significant differences could be observed in ER expression levels between ILC and IDC patients. Furthermore, the differences in PR expression levels between ILC and IDC were even smaller, compared to ER levels. We are not aware of previous studies reporting on highly detailed ER/PR expression levels of ILC and IDC in a large dataset of more than 25,000 patients.
The primary reason to perform the present study was to determine if differences in quantitative ER/PR expression between ILC and IDC could be an important limitation for the interpretation of the results of our previously published study [9]. In that study, we found that adding chemotherapy to hormonal treatment did not improve the overall survival of patients with ILC, in contrast to those with IDC, who clearly benefited from the use of chemotherapy. According to criticasters, this result could partially be explained by the better response of ILC to hormonal therapy due to higher ER expression levels, compared to patients with IDC [10]. In the present study, we proved that among patients with ER-positive breast cancer, there were no clinically relevant differences with respect to ER/PR expression levels. This conclusion suggests that it is the lobular histology itself that relates to the lower chemo-sensitivity compared to the ductal counterpart.
The discussion about the reduced chemo-sensitivity of ILC is ongoing for several years now. The added value of (neo)adjuvant chemotherapy in ER-positive ILC was already questioned in reviews by Katz et al. and Purushotham et al. [11, 12]. In several randomized neoadjuvant trials and population-based case series, patients with ILC were shown to have a significantly lower pathological complete response percentages, compared to patients with IDC [4–6]. Also in a recent, pooled analysis of nine neoadjuvant trials including 1052 patients with ILC, a low percentage of pathologic complete response was observed. However, the lower response rate did not translate into a poorer long-term outcome. Based on the lower response rates, it was recommended in this study that the use of neoadjuvant chemotherapy in patients with ILC should be restricted to the small proportion with ER-negative disease [7]. Altogether, these data contribute to the assumption that ER-positive ILC and IDC respond differently to chemotherapy.
Concerning the response of ILC and IDC to adjuvant hormonal treatment of breast cancer, a study by Rakha et al. showed that the response was better in patients with ILC and that they had a better survival as compared to matched patients with IDC [13]. A more recent study by van de Water et al. showed a similar effect of endocrine therapy regimens in IDC and ILC. On the other hand, they also reported that patients with ER-rich tumors experienced a larger benefit of upfront Exemestane, while patients with ER-poor tumors had better outcomes with sequential therapy, irrespective of histological subtype [14]. This finding emphasizes the relevance of quantification of ER expression levels in hormonal treatment strategies. In concordance with our study, van de Water et al. observed no significant differences between ILC and IDC when looking at semi-quantitative ER expression levels. Furthermore, studies investigating progesterone as a predictive marker for response to endocrine therapy show that loss of PR expression predicts relative resistance to tamoxifen, whereas maintenance of response to aromatase inhibitors can be observed, suggesting a selective role of this treatment in this subgroup [15–17].
In the current era of molecular characterization of breast cancer, most patients with ILC should be classified as luminal A, since ILC is high in ER and PR expression, often low grade and most often HER2-negative [16]. In general, luminal A type tumors do show a good responsiveness to hormonal therapy and because of this, tamoxifen and aromatase inhibitors serve as keystone therapies in ER/PR-positive ILC. This makes it difficult to prove if the good prognosis of patients with lobular breast cancer treated with endocrine therapy, and the apparent lack of an additional effect of chemotherapy, is the result of an excellent response to endocrine treatment or of a reduced chemo-sensitivity. In this light, the recent St Gallen guideline also stated that it is among these patients with the ‘luminal’ type of breast cancer, of which ILC is typical example, that uncertainty exists whether to use adjuvant chemotherapy [3].
Some limitations in this study should be considered when interpreting its results. Despite the fact that these data are derived from a large prospectively collected dataset, some missing values were observed with respect to tumor size, axillary nodal status, grade, HER2 status, and ER expression. However, the number of missing values is too small to have a real and relevant impact on the results. Moreover, information on other tumor characteristics, such as lymphovascular invasion was not available in this database.
In conclusion, our study provides strong evidence that, when looking at patients with ER and PR-positive breast cancer, ILC and IDC do not differ with respect to quantitative ER and PR expression levels. This finding provides additional proof for the lower chemo-sensitivity of ILC and the opinion that histological subtype should play an important role in the decision-making process regarding the use of chemotherapy in this patient subgroup. Future (neo)adjuvant randomized studies or analyses of existing trial data are warranted to provide further evidence on this subject.