IMPORTANCE Limited information about the relationship between specific mutations in BRCA1 or BRCA2 (BRCA1/2) and cancer risk exists. OBJECTIVE To identify mutation-specific cancer risks for carriers of BRCA1/2. DESIGN, SETTING, AND PARTICIPANTS Observational study of women who were ascertained between 1937 and 2011 (median, 1999) and found to carry disease-associated BRCA1 or BRCA2 mutations. The international sample comprised 19 581 carriers of BRCA1 mutations and 11 900 carriers of BRCA2 mutations from 55 centers in 33 countries on 6 continents. We estimated hazard ratios for breast and ovarian cancer based on mutation type, function, and nucleotide position. We also estimated RHR, the ratio of breast vs ovarian cancer hazard ratios. A value of RHR greater than 1 indicated elevated breast cancer risk; a value of RHR less than 1 indicated elevated ovarian cancer risk. EXPOSURES Mutations of BRCA1 or BRCA2. MAIN OUTCOMES AND MEASURES Breast and ovarian cancer risks. RESULTS Among BRCA1 mutation carriers, 9052 women (46%) were diagnosed with breast cancer, 2317 (12%) with ovarian cancer, 1041 (5%) with breast and ovarian cancer, and 7171 (37%) without cancer. Among BRCA2 mutation carriers, 6180 women (52%) were diagnosed with breast cancer, 682 (6%) with ovarian cancer, 272 (2%) with breast and ovarian cancer, and 4766 (40%) without cancer. In BRCA1, we identified 3 breast cancer cluster regions (BCCRs) located at c.179 to c.505 (BCCR1; RHR = 1.46; 95% CI, 1.22–1.74; P = 2 × 10−6), c.4328 to c.4945 (BCCR2; RHR = 1.34; 95% CI, 1.01–1.78; P = .04), and c. 5261 to c.5563 (BCCR23, RHR = 1.38; 95% CI, 1.22–1.55; P = 6 × 10−9). We also identified an ovarian cancer cluster region (OCCR) from c.1380 to c.4062 (approximately exon 11) with RHR = 0.62 (95% CI, 0.56–0.70; P = 9 × 10−17). In BRCA2, we observed multiple BCCRs spanning c.1 to c.596 (BCCR1; RHR = 1.71; 95% CI, 1.06–2.78; P = .03), c.772 to c.1806 (BCCR13; RHR = 1.63; 95% CI, 1.10–2.40; P = .01), and c.7394 to c.8904 (BCCR2; RHR = 2.31; 95% CI, 1.69–3.16; P = .00002). We also identified 3 OCCRs: the first (OCCR1) spanned c.3249 to c.5681 that was adjacent to c.5946delT (6174delT; RHR = 0.51; 95% CI, 0.44–0.60; P = 6 × 10−17). The second OCCR spanned c.6645 to c.7471 (OCCR2; RHR = 0.57; 95% CI, 0.41–0.80; P = .001). Mutations conferring nonsense-mediated decay were associated with differential breast or ovarian cancer risks and an earlier age of breast cancer diagnosis for both BRCA1 and BRCA2 mutation carriers. CONCLUSIONS AND RELEVANCE Breast and ovarian cancer risks varied by type and location of BRCA1/2 mutations. With appropriate validation, these data may have implications for risk assessment and cancer prevention decision making for carriers of BRCA1 and BRCA2 mutations.

Stromelysin is a collagenase-related connective-tissue-degrading metalloproteinase. We have detected RNAs capable of hybridizing to a rat stromelysin cDNA in 11 of 69 human tumours tested. Molecular cloning of cDNAs to these RNAs has identified them as a mixture of stromelysin RNA and a transcript of a hitherto-undescribed related human gene, the stromelysin-2 gene. We have also isolated cDNAs corresponding to a more distantly related new human gene, the pump-1 gene. A comparison of the cDNA-derived amino acid sequences of stromelysin-2 and pump-1 with the known sequences of stromelysin and collagenase reveals significant similarities, with conservation of sequence motifs believed to have functional importance in metalloproteinase action. We conclude that the collagenase gene family in humans consists of at least four members, and speculate that expression of these genes plays a role in cancer.

Head and neck squamous cell carcinoma (HNSCC) is the sixth most common cancer among men in the developed world. There is a need, for both clinical and scientific reasons, to find markers to identify patients with aggressive disease as early as possible, and to understand the events leading to malignant transformation and susceptibility to metastasis. We report the first largescale gene expression analysis of a unique HNSCC location, the hypopharynx. Four normal and 34 tumour samples were analysed with 12 600 gene microarrays. Clusters of differentially expressed genes were identified in the chromosomal regions 3q27.3, 17q21.2-q21.31, 7q11.22-q22.1 and 11q13.1-q13.3, which, interestingly, have already been identified by comparative genomic hybridization (CGH) as major regions of gene amplification. We showed that six overexpressed genes (EIF4G1, DVL3, EPHB4, MCM7, BRMS1 and SART1) located in these regions are indeed amplified. We report 119 genes that are highly differentially expressed between 'early' tumours and normal samples. Of these, we validated by quantitative PCR six novel poorly characterized genes. These genes are potential new markers of HNSCC. Comparing patients with relatively nonaggressive and aggressive tumours (without or with clinical evidence of metastasis 3 years after surgery), we identified 164 differentially expressed genes potentially involved in the acquisition of metastatic potential. This study contributes to the understanding of HNSCC, staging patients into prognostic groups and identifying high-risk patients who may benefit from more aggressive treatment.

Propensity for subsequent distant metastasis in head and neck squamous-cell carcinoma (HNSCC) was analysed using 186 primary tumours from patients initially treated by surgery that developed (M) or did not develop (NM) metastases as the first recurrent event. Transcriptome (Affymetrix HGU133_Plus2, QRT-PCR) and arraycomparative genomic hybridization data were collected. Non-supervised hierarchical clustering based on Affymetrix data distinguished tumours differing in pathological differentiation, and identified associated functional changes. Propensity for metastasis was not associated with these subgroups. Using QRT-PCR data we identified a four-gene model (PSMD10, HSD17B12, FLOT2 and KRT17) that predicts M/NM status with 77% success in a separate 79-sample validation group of HNSCC samples. This prediction is independent of clinical criteria (age, lymph node status, stage, differentiation and localization). The most significantly altered transcripts in M versus NM were significantly associated to metastasis-related functions, including adhesion, mobility and cell survival. Several genomic modifications were significantly associated with M/NM status (most notably gains at 4q11-22 and Xq12-28; losses at 11q14-24 and 17q11 losses) and partly linked to transcription modifications. This work yields a basis for the development of prognostic molecular signatures, markers and therapeutic targets for HNSCC metastasis.

We address the question as to whether increased metalloproteinase production might be related to the high regional recurrence rate of some carcinomas, and particularly head and neck squamous-cell carcinomas (SCC). Northern blot of total RNA prepared from 26 lung carcinomas, 107 head and neck carcinoma samples and corresponding normal tissue samples demonstrates the frequent and sometimes concomitant over-expression of the 2 stromelysin genes, the type-I collagenase gene and the pump-I gene in the head and neck tumour tissue samples. In these SCC, over-expression of the 2 stromelysin genes and the type-I collagenase gene (but not the pump-I gene) is associated with a high degree of tumour differentiation. Moreover, a tumour with high levels of the stromelysin mRNAs is more likely to show high local invasiveness, suggesting that the stromelysins may be implicated in the clinical course of head and neck tumours. Evaluation of the corresponding mRNA levels may prove a useful indicator for predicting the clinical aggressiveness of these tumours.