In our study, we collected information of nearly 1200 OBC patients across the US for 15 years and the data is the most up-to-date one. According to the mean age, we found that OBC tended to be diagnosed in elderly population. Majority of OBC patients were diagnosed at relative early stage (Stage IIA) and Luminal A type of OBC took up nearly 30% of all subjects. Radiotherapy and chemotherapy were common and important treatment for OBC patients, both of which were offered to more than half of the complete cohort. In this data, the number of patients receiving mastectomy were 4 times higher than that of patients receiving BCS. Both clinical and pathological traits were similarly distributed no matter in BCS or mastectomy group, except tumor subtypes. For this variable, Luminal-B OBC was the least one to be treated by BCS whereas HER-2-positive OBC was the least one to be treated by mastectomy, which resulted in a marginal difference between OBC and mastectomy group (P = 0.07). However, when specifically discussing every immunohistochemistry marker (ER/PR/HER-2), no significant was identified between them and different types of surgery. Different from all the factors above, rate of patients treated by radiotherapy had a significant difference between two groups, in which 70% subjects in BCS group were accepted this therapy but only half of women took radiotherapy in mastectomy groups (P < 0.001). Speaking of survival analysis, the whole OBC cohort generally had optimistic prognosis. 5-year OS and BCSS exceeding 80% and both of the survival analysis did not reach the timepoint of median survival. When the whole cohort was grouped according to types of surgery, both of the 5 and 10-year BCSS of BCS group were higher than 5 and 10-year OS of BCS, respectively. The same result was also seen in 5 and 10-year BCSS of mastectomy. However, no significant difference was found no matter in OS or BCSS between BCS and mastectomy group (P = 0.058 for OS and P = 0.14 for BCSS). For Cox regression analysis, we found that age was an independent predictor for OS and higher age meant a poorer survival. Race served as an independent predictor for BCSS and Asian/Pacific natives has a better clinical outcome while native Americans had a relatively poor survival. Radiotherapy was clinically beneficial to OBC patients, no matter calculated according to OS or BCSS. But no significant benefit was brought by neither chemotherapy or other systemic regimen. This finding was also supported by multivariate analysis of tumor subtypes and breast cancer markers: conditions of ER, PR, HER-2 and tumor subtypes were not independent predictors for neither OS or BCSS.
OBC is a special form of breast malignancy characterized as persistent lymphadenopathy[18]. The definition of OBC is time and technique-dependent. At early time, OBC only meant that malignant axillary lymph nodes without positive findings in ipsilateral breast merely with clinical examination[19]. As the time goes on and with the quantum leak of development of technique in medical imaging, the definition of OBC has been refined as breast cancer without evidence of primary malignancy by clinical examination, mammography and ultrasonography[19]. It is predictable that future definition of OBC will be much stricter because creation of brand-new medical technique with higher sensitivity, such as magnetic resonance imaging and Fluorodeoxyglucose-positron emission tomography. Literature reported that MRI could detect primary malignancy in breast of nearly 72% patients previously diagnosed as OBC[20]. One case report suggested that FDG PET/CT was also helpful for detection of OBC despite that no other similar supportive evidence for this method to be used in diagnosis of OBC[21]. Hence, present criteria of OBC is still remain on negative findings of breast lesion through clinical examination, mammogram and ultrasound.
Although NCCN guidelines stated that ALND was recommended for OBC patients, the optimal surgical therapy for the ipsilateral breast in OBC has been controversial[19]. A retrospective study performed by Wang et al. included 51 OBC cases from a single center, in which 38 had ALND plus Mastectomy and the other 13 patients had ALND only[22]. 28 of 38 patients having mastectomy had been found the primary tumor in the breast by pathology. Recurrence rate in mastectomy group was 26%, much lower than that in ALND-only group (77%). Patients with mastectomy had a significantly promising disease-free survival and overall survival compared with patients merely received ALND (P < 0.001). Hence, this study drew a conclusion that mastectomy based on ALND is necessary for OBC patients. This is different from many other studies including our results, which have shown that surgical types did not impact prognosis of OBC patients. A clinical study based on National Cancer Database collected 190 patients diagnosed from 2004 to 2014[23]. Treatment strategies included mastectomy alone, radiation alone and mastectomy combined with radiation. No significant difference in OS was found between each two of these three strategies (mastectomy vs. radiation, P = 0.650; mastectomy + radiation vs. mastectomy, P = 0.393; mastectomy vs. radiation, P = 0.872). In another research also based on National Cancer Database, 1231 OBC patients were grouped into MRM ± radiotherapy (N = 592), radiotherapy + ALND (N = 342), ALND alone (N = 106) and no breast surgery (N = 191). They found that patients treated by ALND and radiotherapy have significantly better OS compared with patients received MRM and/without radiotherapy (HR = 0.475, 95% CI 0.306–0.736, P = 0.001). Multivariate analysis proved that ALND with radiation was an independent protective predictor of OS (HR 0.509, 95% CI 0.321–0.808, P = 0.004). A retrospective study from Korea included 142 pathologically diagnosed OBC cases over 20 years. Surgical strategies consisted of ALND alone, BCS + ALND and mastectomy + ALND. No significant discrepancy was found among these three types of surgical treatments (P = 0.061)[10].
For radiotherapy, we found that radiotherapy is clinically beneficial for OBC. This finding is in line with majority of previous clinical research. A study from the UK finally collected 29 OBC patients across more than 25 years. Median follow-up time was 44 months. Amongst, 16 patients got local radiotherapy and 13 patients did not receive local radiation. Local recurrence rate was 12.5% in radiation group while 69% of patients in non-radiotherapy group got local recurrence (P = 0.02). Moreover, radiotherapy could significantly improve relapse-free survival (HR = 0.31, P = 0.04) and local relapse-free survival (HR = 0.09, P = 0.004). Overall survival, unfortunately, was not positively impacted by radiation therapy[24]. In 2010, another research conducted by team from UK led by Masinghe reported 53 OBC patients diagnosed from 1974 to 2003. Similar to the above studies, this study was also a mixture of clinical and pathological OBC. All the patients received axillary surgery but only 25 of them (47%) had ALND. 5-year recurrence rate of ipsilateral breast was 16.4% (95% CI 4.3–28.5%) in population receiving radiotherapy, lower than that in non-radiation population (35.8%, 95% confidence interval 7.6–64.1%). Similar results was presented in 10-year ipsilateral relapse rate, where radiotherapy group was 23.2% (95% CI 8.9–37.6%) and non-radiotherapy group was 51.9% (95% CI 17.4–86.4%). Furthermore, patients receiving radiotherapy had a significantly higher 5-year and 10-year BCSS rate compared with patients without radiation [P = 0.0073; 5-year: 72.8% (95% CI 59.1–86.5%) vs. 58.3% (95% CI 30.4–86.2%); 10-year: 66.2% (95% CI 51.0–81.5%) vs. 14.6% (0.0–43.3%), respectively). Radiotherapy was the only significant predictor of survival[25]. In a Korean study, 3 of 66 OBC patients did not receive breast radiation. In 15 patients received BCS therapy (blind local excision of the breast), no positive finding of tumor was identified in all these patients. The local recurrence rate was significantly lower in patients without breast radiation than that in patients with breast radiation (6.3% vs. 66.7% ; P = 0.02). Breast radiation could significantly improve 8-year disease-free survival in patients with breast radiation compared with patients without ipsilateral breast radiation (89.5% vs. 50.0%, P = 0.02). The shortcoming of this study is that number of patients without breast radiation was much lower than that of radiation group so that the conclusion should be further examined [26]. Different from studies above, research team from Memorial Sloan Kettering Cancer Center gave an opposite answer for this topic. In this study, all the 38 OBC patients were evaluated by MRI before recruited, stricter examination than the above studies. Among the subjects, 25 patients received ALND + whole breast radiotherapy while 13 patients got MRM (6 patients got chest-wall radiation)[27]. This totally met the criteria of NCCN guideline mentioned before. To make the population in different groups more homogenous, all the patients got chemotherapy. Surprisingly, no recurrence occurred in lymph node during a median follow-up of 7 years. Even two breast recurrence occurred in patients receiving ALND combined with whole breast radiotherapy. No local recurrence occurred in patients with MRM. Similar rates of distal failure were between the two groups (7.7% in MRM cohort vs. 8.0% in ALND + radiotherapy group). Nevertheless, the conclusion of this study should be further discussed because of a small sample size.
For systemic therapy, especially chemotherapy for OBC patients, no specific guidelines, recommendation and consensus have been laid out. Conventionally, neoadjuvant chemotherapy was commonly used for locally advanced/ inoperable/inflammatory breast cancers[28]. This kind of regimen can degrade the tumor stage and serve to evaluate treatment response[29]. Several studies reported results of effectiveness of neoadjuvant chemotherapy in treating of OBC. A case series from China including 5 OBC patients suggested an effective neoadjuvant chemotherapy with pathological complete response (PCR) rate reached 80% (4/5). Among the 4 patients, they were all received mastectomy followed by radiotherapy, two of whom also underwent endocrine therapy. Another study led by Rueth et.al collected 36 OBC patients. Nearly 95% of subjects (34/36) received chemotherapy and the rate of neoadjuvant therapy was up to almost 70% (25/36). In 33 patients with ALND, 15 of them got neoadjuvant therapy. Thus, the PCR rate of neoadjuvant therapy was about 60%, which supported the effectiveness of neoadjuvant therapy in the case series above[13]. In the most recent study with the largest sample size, Cohen et.al collected 684 OBC patients and 214 (31.3%) of them received neoadjuvant chemotherapy. AND PCR was recognized in 55 of 214 (26%) patients took neoadjuvant therapy. In our study, counterintuitively, we found that patients without chemotherapy tended to have a better prognosis. Partly because patients administered with chemo-drug had a relatively higher tumor stage. In our study, 21% cases with stage-IIIA OBC and 19% cases with stage-IIIC cases in chemotherapy group patients, whereas both of the stage-IIIA and stage-IIIC OBC patients accounted for 13% in the other group (P = 0.002, χ2test). As the number of studies focusing on neoadjuvant chemotherapy was small and the PCR rate varied among these studies by virtue of sample size in huge disparity, it is still debatable whether systemic therapy is beneficial for OBC patients.
Despite our analysis have collected the latest data and further expanded the sample size as well as the time span, several limitations are necessary to be noticed: 1) For systemic therapy, current database is shortage of specific data such as the chemotherapy formula and dosage of each drug. Thus, we could not get more information about chemotherapy and its relationship with survival. Moreover, data of endocrine therapy is not available in current version of SEER database; 2) SEER database predominately describes clinical characteristic of American citizens including white and black citizens. Data of Asian OBC patients were insufficient so that it is unclear whether the results of this analysis were also generalized to Asian population; 3) Because the data of HER-2 status originates from 2010, it is not available in the cases diagnosed before 2010, which impacted the descriptive results and following analysis about HER-2 status, tumor subtypes as well as their associations with OS and BCSS.