There is limited knowledge on the influence of breast cancer subtypes, based on hormone receptor status and HER2-status, on clinical outcome in stage IV IBC. This large study demonstrates that breast cancer subtypes in stage IV IBC are associated with unique patterns of distant metastatic spread and differences in OS.
To our knowledge, this is the first extensive analysis of the influence of HR/HER2-based breast cancer subtypes on the preferential site of metastases and OS in stage IV IBC. Data of previous studies were derived from patients treated in single institutions [
11,
12], whereas our study contains data based on a nationwide population-based cancer registry including unselected and unbiased data of all hospitals (both academic and non-academic) in the Netherlands. Therefore, our study, which includes the largest patient number so far, represents valuable data on current clinical presentation and practice.
Inflammatory breast cancer subtypes and preferential site of metastasis
At first presentation, patients with IBC display significantly higher rates of distant metastases compared to non-inflammatory locally advanced breast cancer (39.7% versus 34.1%) [
6].
In our current cohort, 744 patients with stage IV IBC at diagnosis were evaluated, which accounted for 33.3% of all patients diagnosed with IBC in the studied period. Over 50% of patients presented with multiple metastatic sites with different frequencies of the site of distant metastases among the various breast cancer subtypes. Multivariable analysis revealed that metastatic involvement of multiple sites was independently associated with a worse survival.
Bone metastases were most commonly diagnosed in all stage IV IBC subtypes with a significant predominance in the HR+/HER2− group. Liver metastases were more frequently observed in the HER2-enriched group and lung metastases in the HR−/HER2− group. This is not in agreement with a previous SEER analysis which did not show a significant association between IBC subtypes and site of metastases, which may well be attributed to the small sample with only 83 patients with stage IV IBC analyzed in that study [
13]. We used data from the NCR to demonstrate that the metastatic patterns of stage IV IBC seem rather comparable to stage IV breast cancer in general [
14]. A SEER analysis demonstrated that HR+/HER2+ and HER2-enriched subtypes are prone to abdominal/pelvic metastases and HR+/HER2− and HR+/HER2+ subtypes to bone metastases, while the HR−/HER2− subtype was prone to lung/mediastinal metastases [
14]. In the present study, the occurrence of brain metastases was evidently lower than previously reported by Warren et al., who advised to incorporate surveillance brain MRI after diagnosis of extracranial metastatic disease in IBC [
15]. However, in their study, no statistically significant association was found between primary tumor subtype and increased risk of development of brain metastases. A US-based single-center analysis in 203 IBC patients showed that the median time to development of brain metastases was 19 months [
16]. One of the reasons of the low incidence of brain metastases in our study might be the fact that in the NCR only synchronous metastases are recorded, and as such, we were not able to analyze subsequent brain metastases that occurred more than 3 months after diagnosis.
Treatment of stage IV inflammatory breast cancer
Management of synchronous stage IV IBC includes primary systemic cytotoxic therapies and targeted HER2 therapy in case of HER2 positivity [
17]. One area of debate is whether patients with stage IV IBC also should undergo local resection of the breast tumor. In the absence of prospective data, a potential survival benefit from removal of the breast tumor is suggested by retrospective evidence [
11,
12,
18].
To our knowledge, there are three prospective trials conducted evaluating the effect of removal of the primary tumor in stage IV breast cancer, in which conflicting results were presented: two studies could not demonstrate a survival benefit [
19,
20] and one showed an improved survival after 40 months follow-up (initially not showing any survival benefit of surgery after 36 months of follow-up) [
21].
Just over 20% of all patients in our analysis underwent surgical resection of the primary tumor and just over 8% received trimodality treatment. This combination of neoadjuvant chemotherapy, surgery, and adjuvant radiation therapy is considered to be the most effective treatment regimen in stage III IBC [
17]. Our numbers indicate that, also in the metastatic setting, surgery is used for locoregional management.
Inflammatory breast cancer subtypes and survival
An important finding of the present study is the highly variable prognosis among the different breast cancer subtypes in stage IV IBC. Patients with HR+/HER2+ IBC had the best survival among the four subgroups, whereas HR−/HER2− IBC is an independent prognostic factor for decreased survival, compared to the other subtypes. These results are consistent with previous studies in stage III IBC as well as stage IV non-IBC, both showing that HR−/HER2− tumors have the worst prognosis [
5,
22,
23]. The 5-year OS of HR+/HER2+ IBC patients is 3.5-fold higher than that of patients with HR−/HER2− IBC, whereas HR+/HER2− and HR−/HER2+ subtypes display similar survival rates to each other but evidently lower compared to the HR+/HER2+ subtype. These ratios are comparable to a recent SEER analysis of metastatic breast cancer. Improved survival of the HR+/HER2+ subtype most likely reflects the use of HER2-targeted therapies. After the development of HER2-targeted therapies for the treatment of metastatic breast cancer in general, the survival of patients with HER2-positive tumors was greatly improved. This effect was irrespective of the HR status of the tumor [
24]. Data were collected from patients who presented with stage IV IBC in the period 2005–2016. The last years, trastuzumab emtansine and pertuzumab enlarged the therapeutic arsenal for HER-2 positive patients, and as a consequence, the prognosis for survival will be even better nowadays than in the study period [
25].
In general, the present study confirmed that compared with other primary tumor characteristics, breast cancer subtypes based on the HR and HER2 status of the primary tumor in stage IV IBC are important predictors of OS.
Several limitations of the present study have to be discussed. Firstly, the NCR does not register cause of death, and therefore, breast cancer-specific survival could not be determined. However, since all included patients already were diagnosed with stage IV disease at diagnosis, and since metastatic disease is the major cause of cancer-related deaths among breast cancer patients, the cause of death in our population is most likely to be breast cancer specific [
26]. Secondly, it should be noted that 98 patients who were excluded with an unknown HR/HER2 less often underwent any form of treatment. These patients were significantly older and represent a specific subgroup of stage IV IBC patients. Reasons why older patients with cancer accept or decline treatment vary considerably, but the most consistent determinant found in the literature is physician recommendation [
27]. Unfortunately, we are not able to draw firm conclusions on the absence of HR/HER2 data, since reasons for the waiver of treatment modalities could not be investigated in this database. These factors, as well as comorbidity, were not registered in the NCR and could not be accounted for in our study. Moreover, local therapy of the sites of metastatic disease (for example, resection of metastases and/or radiation therapy) could not be analyzed in this study.
We chose to only analyze clinical T4d breast cancers, instead of analyzing both clinical and pathological T4d breast cancers. However, since IBC is typically diagnosed clinically (dermal lymphatic invasion without typical clinical findings is not sufficient for a diagnosis of IBC), analysis of clinical T4d breast cancers seems to be the most accurate approach. As with any information obtained retrospectively from the abstraction of medical records, we acknowledge the dependency on the availability of data and reporting accuracy.
Furthermore, a central pathology review during treatment was not conducted, which might have led to an altered HR/HER2 status in several patients. Therefore, the potential impact of inter-institutional discordance was not evaluated. However, our current analysis reflects daily clinical practice in which local laboratories do not send all samples to a central laboratory and limited discordance was found in previous analyses which addressed the possibility of potential discordance [
28‐
30]. Furthermore, discordances in ER/PR/HER2 test results between tumor core needle biopsy taken at the time of diagnosis and tumor resection material are low, also in patients receiving any form of neoadjuvant therapy [
31].
Finally, information regarding the type of diagnostic modalities is lacking. Given the high rate of metastatic disease at presentation, patients with IBC undergo extensive staging, including whole body bone scintigraphy, ultrasonography of the liver, and a chest X-ray. Some institutions might have used other modalities such as 18-fluorodeoxyglucose positron emission tomography/computed tomography (PET/CT), which might influence the detection of distant metastases compared to traditional modalities [
32]. More than 20% of patients appear to have distant metastases after FDG-PET scanning in comparison with conventional staging in locally advanced breast cancer [
33]. However, with regard to subtypes, no difference is expected, since there is no guideline of which breast cancer subtype should receive a specific type of diagnostic modality and a potential diagnostic bias would be present for all breast cancer subtypes and this will not affect the differences we report.
Clinical relevance
IBC is diagnosed at a younger age with survival rates which are clearly inferior to the average rates for patients with non-IBC [
6]. Similar to IBC in general, the incidence of stage IV IBC seems to be increasing (data not shown) [
6]. This might, among others, be due to increased use of improved staging modalities [
32]. Knowledge of the biology of IBC has to increase to achieve improvement on the treatment of IBC. This applies for both stage III and IV IBC. Stratification of breast cancer subtypes in stage IV IBC is of clinical use for estimating prognosis, since OS differed significantly between the subtypes with the worst OS for HR−/HER2− IBC. These data might aid physicians in patient counseling regarding prognosis and underscribe the need of new systemic (targeted) therapies to improve OS in stage IV IBC and HR−/HER2− disease in particular.
Moreover, the differences seen in sites of metastases between breast cancer subtypes can guide a more patient-tailored staging directed to metastatic sites. Since metastatic disease remains the principal cause of cancer-related deaths [
34], this tailored staging could lead to the identification of more effective prognostication and, hopefully in the future, individualized targeted approaches to treat these patients. Some evidence suggests a potential role for metastasis-specific local treatment (e.g., metastasectomy and radiation therapy) in the prolongation of survival, especially in oligometastatic disease, although prospective data are lacking [
35]. Consequently, in patients with multiple organ metastases, locoregional treatment of metastases should be discussed and potentially omitted. This will prevent potential morbidity of non-beneficial treatments [
36].