RESUMO
True invasive tubular breast carcinoma (TBC) is unlikely to metastasize to axillary nodes, yet it is routinely subjected to sentinel lymph node biopsy (SLNB), even if the diagnosis was suspected preoperatively. The positive predictive value (PPV) of core biopsy for TBC and the incidence and predictors of axillary metastasis in invasive breast carcinomas identified as tubular-rich on core biopsy are unknown. Prospective patient and tumor data regarding postoperatively confirmed TBCs, and tubular-rich carcinoma identified on preoperative core biopsy between January 2005 and May 2016 was analyzed retrospectively. Axillary metastasis occurred in only 4.2% (4/95) of TBCs, all of which measured >15 mm pathologically. In 11.1% (11/99) of TBCs, the initial core biopsy was either indeterminate/suspicious or ductal carcinoma in situ (DCIS); therefore, their true tubular histotype and size were ascertained following operative excision and before SLNB. Nine were ≤15 mm, and all were node-negative. Only 63.9% (46/72) of tubular-rich core biopsies were confirmed as TBCs; the remaining 36.1% (26/72) were well-differentiated invasive nontubular carcinomas. None of the preoperative patient or tumor features were predictive of true TBC on multivariable analysis; 10.1% (7/69) of carcinomas identified as tubular-rich on core biopsy (regardless of their true histotypes) were node-positive; 23.1% (6/26) in nontubular and 2.3% (1/43) in true tubular carcinomas. Preoperative ultrasound size >15 mm was associated with axillary metastasis in 40.0% (4/10) compared to 5.7% (3/53) in those ≤15 mm (OR = 11.11, 95% CI = 1.99-62.04; multivariable P = .010). Axillary metastasis in TBC is dependent on pathological size; therefore, a case is made for omitting SLNB in small true TBCs confirmed following excision. Preoperative tubular-rich core biopsy is not adequately diagnostic of TBC; however, it selects carcinomas that are well-differentiated, small, and unlikely to metastasize to the axilla, thus allowing for the selective omission of SLNB.