Brem Rachel F, Lechner Mary C, Jackman Roger J, Rapelyea Jocelyn A, Evans W Phil, Philpotts Liane E, Hargreaves Jonathan, Wasden Shane
Department of Radiology, George Washington University, 2150 Pennsylvania Ave., NW, Washington, DC 20037, USA.
AJR Am J Roentgenol. 2008 Mar;190(3):637-41. doi: 10.2214/AJR.07.2768.
The purpose of our study was to better define the rate and variables associated with cancer underestimation when lobular neoplasia is found at minimally invasive breast biopsy.
The records of 32,420 patients who underwent imaging-guided needle biopsy of the breast for mammographic or sonographic abnormalities from 1988 to 2000 were retrospectively reviewed. The 278 cases in which lobular neoplasia was the highest-risk lesion at biopsy were included. Of the 278 cases, 164 proceeded to surgical excision, allowing calculation of rates of underestimation from minimally invasive biopsy.
Of the 32,420 minimally invasive breast biopsies, lobular neoplasia was found in 278 (0.9%). One hundred sixty-four of the 278 (59%) continued to surgical excision, where cancer was pathologically confirmed in 38 (23%). No difference was seen in the underestimation rates for lesions diagnosed as lobular carcinoma in situ (25%, 17 of 67 lesions) versus atypical lobular hyperplasia (22%, 21 of 97 lesions). Statistically significant underestimation of carcinoma was found with biopsy of masses (with or without associated microcalcifications) rather than calcifications only, a higher BI-RADS category (p < 0.0001), use of a core biopsy device rather than a vacuum device (p < 0.01), and obtaining fewer specimens (p < 0.0001).
Significant sampling error occurs regardless of the type of core biopsy device, number of specimens obtained, histologic-radiographic concordance, mammographic appearance, and complete excision of the lesion as determined by imaging. For this reason, all patients with lobular neoplasia at core or vacuum-assisted biopsy should undergo surgical excision until further differentiating criteria can be determined.
我们研究的目的是更准确地界定在微创乳腺活检中发现小叶瘤变时癌症低估的发生率及相关变量。
回顾性分析了1988年至2000年间因乳腺钼靶或超声异常接受影像引导下乳腺针吸活检的32420例患者的记录。纳入了278例活检时小叶瘤变为最高风险病变的病例。在这278例病例中，164例行手术切除，从而能够计算微创活检的低估率。
在32420例微创乳腺活检中，发现小叶瘤变278例（0.9%）。278例中的164例（59%）继续接受手术切除，其中38例（23%）病理确诊为癌症。原位小叶癌（67例病变中的17例，25%）与非典型小叶增生（97例病变中的21例，22%）的低估率未见差异。发现对肿块（有或无相关微钙化）而非仅对钙化进行活检、较高的BI-RADS分类（p < 0.0001）、使用粗针活检装置而非真空装置（p < 0.01）以及获取标本较少（p < 0.0001）时，癌症的低估具有统计学意义。
无论粗针活检装置的类型、获取的标本数量、组织学与影像学的一致性、乳腺钼靶表现以及影像确定的病变完整切除情况如何，都会出现显著的取样误差。因此，所有在粗针或真空辅助活检中发现小叶瘤变的患者均应接受手术切除，直至能确定进一步的鉴别标准。
