In the current study, we enrolled 682 pre-OP DCIS patients and compared ALN metastasis pattern with another cohort of 2268 pre-OP diagnosed invasive cancer. We found 34.2% of these pre-OP DCIS patients upgraded to DCIS-IC in final pathology, and this upgrade rate was consistent with literate reported range (3.5–56%)[1–17]. The risk of ALN metastasis rate varied widely depends on the pre-operative pathology, and in our current study, the ALN metastasis rate is 7.6% in pre-OP DCIS patients, and up to 39% in pre-OP invasive cancer group (Table 2). In patients with post-OP pure DCIS, the LN metastasis rate is lowest (0.8%), and ALN metastasis was found in 20.6% of patients with DCIS-IC. We also demonstrated that the ALN disease burden (N1-3) varied among different categories of patients (Table 2). These data reminded us the differences of ALN metastatic risk in each category of patients, and more tailored or individualized ALN evaluation strategy should be offered.
Indication and threshold for surgical lymph biopsy in patients with pre-OP biopsy diagnosed DCIS remained a controversy issue for decades. When invasive component present, SLNB was viewed as “adequate-treatment” while no axillary surgery seemed likely to “under-treatment”. When pure DCIS found in permanent pathology, SLNB seemed “over-treatment”. Under this concept, in our pre-OP DCIS patients, 32.6% were adequate-treated with SLNB, 58.4% over-treated, and 1.5% under-treated. Another 7.6% patients with post-OP pure DCIS did not receive surgical axillary biopsy, which should view as adequate-observation. If only “invasive component” present in final pathology viewed as an indication for SLNB, then up to 60% of patients diagnosed with pre-OP DCIS received inadequate axillary management (Table 2). This highlights the “unmet medical need” in modern era of personalized breast cancer care, which over-treated the low risk and/or low disease burden axilla.
In post-OP pure DICS patients, 0.8% ALN metastasis rate is not justified for surgical ALN biopsy. In pre-OP invasive cancer patients, ALN metastasis rate could up to 39% and surgical lymph node biopsy with either SLN and/or ALND is indicated (Table 2). The group of patients, who diagnosed initially as DCIS, had a substantial risk of upgrade to invasive cancer (34.2% in current study) and 7.6% overall lymph node metastasis cases detected. In patients of DCIS-IC, the ALN metastasis rate up to 20.6%. One could speculate that patients with biopsy proven pre-OP DCIS present a category of “low” (pure DCIS) to “intermediate” (DCIS-IC) risk of ALN metastasis, which possessed some controversy whether routine lymph node surgery should be performed and it remained unsolved issue even in current modern breast imaging era.
According to NCCN and other practice guideline, patients with pre-OP biopsy diagnosed DCIS, selected for BCS and adjuvant radiotherapy, do not routinely indicate for SLNB. As most patients (77% in current study) with pre-OP biopsy diagnosed DCIS received BCS would remain pure DCIS in final pathology report (Fig. 1, Table 2). Furthermore, according to ACOSZ0011 trial, even 1–2 positive sentinel lymph node present, patients could omit ALND if breast cancer with tumor size T1-T2, received BCS and whole breast radiotherapy would be performed. These reasons supported that SLNB is not routinely needed in pre-OP biopsy diagnosed DCIS patients indicated for BCS (Table 2).
However, some patients with pre-OP biopsy diagnosed DCIS and indicated for BCS would have post-op upgrade to DCIS-IC and with more than 2 lymph nodes metastases. In current study, we observed 7 pre-OP DCIS patients who received BCS presented with ALN metastases in final pathology check-up. Five of them were N1, 1 was N2, and the other 1 was N3 status (Fig. 1). This accounts that about 2.7% (7/252) of pre-OP DCIS and indicated BCS patients would have ALN metastasis. Although only few patients would have > 2 occult metastatic lymph nodes, routinely abandon SLNB is also worrisome.
Controversy did persist about the role of SLNB in patients with pre-OP biopsy diagnosed DCIS and indicated for mastectomy. Parallel with increasing early detection of breast cancer, DCIS rate increased up to 14% -20%, and around half (68% in current study) of them would receive mastectomy. This group of patients constituted about 5% of annual breast cancer cases, which constituted a non-ignorable minority. Current practice guideline had suggested SLNB to be performed in pre-OP biopsy diagnosed DCIS and received mastectomy as secondary lymph node staging surgery seems unreliable when post-OP pathology upgrade to invasive breast cancer[31, 32].
In our 430 patients with pre-OP biopsy diagnosed DCIS received mastectomy, 59.3% remained DCIS-post-OP, and in these patients SLNB seemed unnecessary or “over-treatment”. About 40.7%(175/430) of patients with pre-OP DCIS received mastectomy and upgrade to DCIS-IC post-OP. Among these 175 patients, 171 (98%) received SLNB, and 40 (23.5%) was found to have LN metastasis. This accounts for 9.8% (42/430) of pre-OP biopsy diagnosed DCIS and indicated mastectomy patients would have ALN metastasis. This result is consisted with Price et al.’s study, which showed about 10% patients would have ≧ 1 LN metastasis when SLNB was performed in patients with pre-OP biopsy diagnosed DCIS and indicated for mastectomy.
The reason that mastectomy patients tend to receive SLNB is that the upgrade rate is higher in patients received total mastectomy than partial mastectomy (BCS) (40.7% versus 23%, P < 0.01, Table 2). This is related to more multicentric breast cancer, more larger tumor size of patients would receive mastectomy than BCS, and larger tumor or multicentric lesions would possess higher risk lead to upgrade to DCIS-IC. Among the predictors of ALN metastasis for patients received SLNB, patients with post-OP invasive component was the highest risk of lymph node metastasis (Table 5).
In current study, we tried to evaluate the pre-operative MRI for pre-OP DCIS patients, and we had 386 patients with detail ALN evaluation and post-operative pathologic report were analyzed for concordance. The accuracy of MRI: sensitivity 53.8%, specificity 77.8%, PPV 14.9%, NPV 95.9%, accuracy 76.2%. This high NPV rate is very useful, especial for patients selected for BCS patients. In patients with MRI estimated tumor size ˂ 3cm & lymph node negative cases, the NPV of ALN metastasis is high (up to 96.4%), and all (3.6%, 5/137) the FN cases were limited to N1 (with 1–2 positive nodes only). Our data supported that pre-OP MRI evaluated node negative patients suitable for BCS patients could safety omitted SLNB if whole breast radiotherapy is to be performed. In mastectomy patients, this high NPV could be used as shared decision making, however, could not guarantee metastatic free lymph node to omit SLNB (Table 2).
Our current study is limited in its retrospective study and limited cases, which could not answer whether subgroup of patients could omit axillary surgery in pre-OP DCIS patients indicated for mastectomy. Our current study, however, did collect of 682 pre-OP DCIS patients with detailed pre-op and post-OP lymph node pathologic report. Of special note that we had 386 patients with detailed pre-OP breast MRI evaluation and post-operative pathologic results, which enable us to evaluate the role of breast MRI in decision of lymph node surgery.