Predictors of pathologic complete response after standard neoadjuvant chemotherapy in triple-negative breast carcinoma.

The objective of this study was to identify predictors of pathologic complete response and tumor volume reduction in triple-negative breast carcinomas. Consecutive cases of 101 triple-negative carcinomas within the last 3 years treated with standard neoadjuvant chemotherapy were identified. However, 56 cases with sufficient material available (for tissue microarray construction) in the pretherapy core biopsy tissue blocks formed the basis of this study. The pretherapy tumor core biopsy slides were examined for various morphologic features including tumor grade. The tumors were immunohistochemically examined for basal phenotype markers (CK5, CK14, CK17, epidermal growth factor receptor), cell adhesion marker E-cadherin, and proliferation marker Ki-67. The overall rate of pathologic complete response was 34% (19 of 56). Neither any morphologic feature nor any basal marker reactivity predicted for pathologic complete response or >50% tumor volume reduction. Ki-67 proliferation index also failed as a predictive marker. Reduced E-cadherin expression (defined as H score ≤200) was initially seen in 47% of cases with pathologic complete response and in only 6% of cases that failed to achieve pathologic complete response (P=0.001); however, in additional 20 cases from a separate validation set, no such difference was identified. Basal marker reactivity in triple-negative breast carcinomas does not predict pathologic complete response after neoadjuvant chemotherapy. As vast majority of triple-negative tumors are highly proliferative, Ki-67 proliferation index appears to have negligible clinical value in predicting pathologic complete response. E-cadherin expression as a predictor of pathologic complete response in triple-negative tumors should be further assessed on larger number of cases.

Semi-quantitative immunohistochemical assay versus oncotype DX(®) qRT-PCR assay for estrogen and progesterone receptors: an independent quality assurance study.

Estrogen receptor (ER) status is a strong predictor of response to hormonal therapy in breast cancer patients. Presence of ER and level of expression have been shown to correlate with time to recurrence in patients undergoing therapy with tamoxifen or aromatase inhibitors. Risk reduction is also known to occur in ER-negative, progesterone receptor (PR)-positive patients treated with hormonal therapy. Since the 1990s, immunohistochemistry has been the primary method for assessing hormone receptor status. Recently, as a component of its oncotype DX(®) assay, Genomic Health began reporting quantitative estrogen and PR results determined by quantitative reverse transcription polymerase chain reaction (qRT-PCR). As part of an ongoing quality assurance program at our institution, we reviewed 464 breast cancer cases evaluated by both immunohistochemistry and oncotype DX(®) assay for estrogen and PR. We found good correlation for ER status between both assays (98.9% concordance), with immunohistochemistry being slightly more sensitive. Concordance for PR was 94.2% between immunohistochemistry and qRT-PCR with immunohistochemistry again more sensitive than RT-PCR. The results also showed linear correlation between immunohistochemistry H-scores and qRT-PCR expression values for ER (correlation coefficient of 0.579), and PR (correlation coefficient of 0.685). Due to the higher sensitivity of hormone receptor immunohistochemistry and additional advantages (ie preservation of morphology, less expensive, faster, more convenient), we conclude immunohistochemistry is preferable to qRT-PCR for determination of estrogen and PR expression.

Polyradiculopathies from schwannomatosis.

We describe a case of schwannomatosis presenting as radicular pain and numbness in multiple radicular nerve distributions. There were multiple peripheral nerve tumors detected by magnetic resonance imaging (MRI) at the left vestibular nerve, cauda equina, right radial nerve, thoracic paraspinal nerve, and brachial plexi. Several resected tumors have features of schwannomas, including hypercellular Antoni A areas, hypocellular Antoni B areas, Verocay bodies, and hyalinized blood vessels. The specimens are also positive for immunohistochemical staining for INI1 with diffuse nuclear staining. The findings are consistent with sporadic form of schwannomatosis. This case highlights the importance of using MRI and INI1 immunohistochemistry to differentiate familial schwannomatosis, neurofibromatosis 2 (NF2)-associated schwannomatosis, and sporadic schwannomatosis.

The relationship between papillary infarction and microinvasion in ovarian atypical proliferative ("borderline") serous and seromucinous tumors.

Papillary infarction is commonly observed in ovarian atypical proliferative serous and seromucinous tumors (APST/APSMT), but there are no published data on its significance. This study characterizes the features associated with papillary infarcts and microinvasion to further understand these phenomena. From consecutive hospital-based cases, 32 APST/APSMT in 26 patients (6 bilateral) were reviewed and evaluated for papillary infarcts, microinvasion (<5 mm), and other histologic features. Among the tumors, 69% were APSTs and 31% APSMTs. Infarcts were identified in 46% of patients, and microinvasion in 27%. Microinvasion was significantly more common in tumors with infarcts (50%) than in those without (7%; P=0.0261). Papillary infarcts were significantly more common in APSTs (61%) than in APSMTs (13%; P=0.0357). The microinvasive tumors were significantly more likely to be bilateral (57% vs. 11%, P=0.0278). The mean infarct size in the presence of microinvasion was 5.9 mm, and in the absence of microinvasion, 2.2 mm (not significant). The infarcts were topographically separate from the foci of microinvasion. Other features evaluated showed no meaningful correlations with microinvasion or infarction. Proliferative noninvasive serous tumors with papillary infarcts are significantly more likely to have microinvasion, and papillary infarcts are more common in APSTs than in APSMTs. APSTs with microinvasion are more common than earlier appreciated. Whether papillary infarction is pathogenetically related to microinvasion is unknown and warrants further investigation.