Abstract OT3-02-01: Development of cell-free nucleic acid-based tests for early detection of breast cancer: The STRIVE study

Background:

Mammography (digital 2D or digital 3D/tomosynthesis) is the cornerstone of current screening strategies for breast cancer, but new approaches are needed to further reduce the proportion of cancers diagnosed at advanced stages and more effectively identify those women in need of additional testing and biopsies. Circulating cell-free nucleic acids (cfNAs) shed from tumors, isolated from peripheral blood, and analyzed with ultra-deep and broad sequencing of cancer-associated genes, have great potential for early cancer detection. The ultimate goal is to develop blood cfNA cancer screening tests for use in conjunction with established risk factors and/or radiographic features for improved cancer detection. Development of these tests requires large, well-annotated cohorts of asymptomatic participants with adequate volumes of prediagnostic blood. The STRIVE Study cohort was recently established to train and validate cfNA-based tests for early detection of breast cancer.

Eligibility criteria and trial design: The STRIVE Study is a new prospective, multi-ethnic mammography cohort that will recruit 120,000 subjects from 15+ US breast cancer screening centers (including Mayo Clinic and Sutter Health sites). Eligibility criteria require only that a participant has a scheduled routine screening mammogram at a participating center and has not received a biopsy prior to the research blood draw. Participants are recruited within 28 days of screening mammography (digital or tomosynthesis), consent electronically, provide blood samples, and complete an on-line risk factor questionnaire. Participants will be followed for all cancer diagnoses, cancer recurrences, and death for at least 5 years. Pertinent medical record information, imaging findings (including breast density), and follow-up data will be transferred electronically to a central database throughout the study period. Additional blood samples will be collected from participants with abnormal mammogram results, or who are diagnosed with cancer, to document and better understand the evolution of cfNA signals. Recruitment began in February 2017.

Primary Aims: To train and validate a cfNA blood-based test to identify breast cancer overall in a cohort of women undergoing screening mammography.

Statistical Methods: The study will be divided into a training phase (1/3 of participants) and an independent clinical validation phase (remaining 2/3 of participants). In the training phase, statistical machine learning techniques will be used to develop algorithms incorporating cfNA signals, clinical characteristics, or radiological features. In the validation phase, the prespecified locked algorithm developed from the training phase will be clinically validated in an independent group of women.

Contact information for people with a specific interest in the trial: Additional details regarding the STRIVE Study are available on the ClinicalTrials.gov website (NCT03085888). For site-specific questions, please call 844-366-9738 for the Mayo Clinic and 1-855-578-7483 for Sutter Health.

Citation Format: Liu MC, Cummings S, Vachon CM, Kerlikowske K, Couch FJ, Morris EA, Olson JE, Polley EC, Conners AL, Ellis RL, Patel B, Maimone IV S, Zhang N, Hamilton S, Clarke CA, Allen BA, Maddala T, Hartman A-R. Development of cell-free nucleic acid-based tests for early detection of breast cancer: The STRIVE study [abstract]. In: Proceedings of the 2017 San Antonio Breast Cancer Symposium; 2017 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2018;78(4 Suppl):Abstract nr OT3-02-01.

Prevalence of germ-line mutations in cancer genes among pancreatic cancer patients with a positive family history

Background

Panel-based genetic testing has identified increasing numbers of patients with pancreatic ductal adenocarcinoma (PDAC) who carry germline mutations. However, small sample sizes or number of genes evaluated limit prevalence estimates of these mutations. We estimated prevalence of mutations in PDAC patients with positive family history.

Methods

We sequenced 25 cancer susceptibility genes in lymphocyte DNA from 302 PDAC patients in the Mayo Clinic Biospecimen Resource for Pancreatic Research Registry. Kindreds containing at least two first-degree relatives with PDAC met criteria for Familial Pancreatic Cancer (FPC), while the remaining were familial, but not FPC.

Results

Thirty-six patients (12%) carried at least one deleterious mutation in one of 11 genes. Of FPC patients, 25/185 (14%) were carriers, while 11/117 (9%) non-FPC patients with family history were carriers. Deleterious mutations (n) identified in PDAC patients were BRCA2 (11), ATM (8), CDKN2A (4), CHEK2 (4), MUTYH/MYH (3 heterozygotes, not biallelic), BRCA1 (2), and 1 each in BARD1, MSH2, NBN, PALB2, and PMS2. Novel mutations were found in ATM, BARD1, and PMS2.

Conclusions

Multiple susceptibility gene testing in PDAC patients with family history of pancreatic cancer is warranted regardless of FPC status, and will inform genetic risk counseling for families.

Inherited Mutations in Men Undergoing Multigene Panel Testing for Prostate Cancer: Emerging Implications for Personalized Prostate Cancer Genetic Evaluation

PURPOSE

Multigene panels are commercially available for the evaluation of prostate cancer (PCA) predisposition, which necessitates tailored genetic counseling (GC) for men. Here we describe emerging results of Genetic Evaluation of Men, prospective multigene testing study in PCA to inform personalized genetic counseling, with emerging implications for referrals, cancer screening, and precision therapy.

PATIENTS AND METHODS

Eligibility criteria for men affected by or at high risk for PCA encompass age, race, family history (FH), and PCA stage/grade. Detailed demographic, clinical, and FH data were obtained from participants and medical records. Multigene testing was conducted after GC. Mutation rates were summarized by eligibility criteria and compared across FH data. The 95% CI of mutation prevalence was constructed by using Poisson distribution.

RESULTS

Of 200 men enrolled, 62.5% had PCA. Eleven (5.5%; 95% CI, 3.0% to 9.9%) had mutations; 63.6% of mutations were in DNA repair genes. FH of breast cancer was significantly associated with mutation status (P = .004), and FH that met criteria for hereditary breast and ovarian cancer syndrome was significantly associated with PCA (odds ratio, 2.33; 95% CI, 1.05 to 5.18). Variants of uncertain significance were reported in 70 men (35.0%). Among mutation carriers, 45.5% had personal/FH concordant with the gene. A tailored GC model was developed based on emerging findings.

CONCLUSION

Multigene testing for PCA identifies mutations mostly in DNA repair genes, with implications for precision therapy. The study highlights the importance of comprehensive genetic evaluation for PCA beyond metastatic disease, including early-stage disease with strong FH. Detailed FH is important for referrals of men for genetic evaluation. The results inform precision GC and cancer screening for men and their male and female blood relatives.

Cancer Susceptibility Gene Mutations in Individuals With Colorectal Cancer

Purpose Hereditary factors play an important role in colorectal cancer (CRC) risk, yet the prevalence of germline cancer susceptibility gene mutations in patients with CRC unselected for high-risk features (eg, early age at diagnosis, personal/family history of cancer or polyps, tumor microsatellite instability [MSI], mismatch repair [MMR] deficiency) is unknown. Patients and Methods We recruited 1,058 participants who received CRC care in a clinic-based setting without preselection for age at diagnosis, personal/family history, or MSI/MMR results. All participants underwent germline testing for mutations in 25 genes associated with inherited cancer risk. Each gene was categorized as high penetrance or moderate penetrance on the basis of published estimates of the lifetime cancer risks conferred by pathogenic germline mutations in that gene. Results One hundred five (9.9%; 95% CI, 8.2% to 11.9%) of 1,058 participants carried one or more pathogenic mutations, including 33 (3.1%) with Lynch syndrome (LS). Twenty-eight (96.6%) of 29 available LS CRCs demonstrated abnormal MSI/MMR results. Seventy-four (7.0%) of 1,058 participants carried non-LS gene mutations, including 23 (2.2%) with mutations in high-penetrance genes (five APC, three biallelic MUTYH, 11 BRCA1/2, two PALB2, one CDKN2A, and one TP53), 15 of whom lacked clinical histories suggestive of their underlying mutation. Thirty-eight (3.6%) participants had moderate-penetrance CRC risk gene mutations (19 monoallelic MUTYH, 17 APC*I1307K, two CHEK2). Neither proband age at CRC diagnosis, family history of CRC, nor personal history of other cancers significantly predicted the presence of pathogenic mutations in non-LS genes. Conclusion Germline cancer susceptibility gene mutations are carried by 9.9% of patients with CRC. MSI/MMR testing reliably identifies LS probands, although 7.0% of patients with CRC carry non-LS mutations, including 1.0% with BRCA1/2 mutations.

Germline multi-gene hereditary cancer panel testing in an unselected endometrial cancer cohort

Hereditary endometrial carcinoma is associated with germline mutations in Lynch syndrome genes. The role of other cancer predisposition genes is unclear. We aimed to determine the prevalence of cancer predisposition gene mutations in an unselected endometrial carcinoma patient cohort. Mutations in 25 genes were identified using a next-generation sequencing-based panel applied in 381 endometrial carcinoma patients who had undergone tumor testing to screen for Lynch syndrome. Thirty-five patients (9.2%) had a deleterious mutation: 22 (5.8%) in Lynch syndrome genes (three MLH1, five MSH2, two EPCAM-MSH2, six MSH6, and six PMS2) and 13 (3.4%) in 10 non-Lynch syndrome genes (four CHEK2, one each in APC, ATM, BARD1, BRCA1, BRCA2, BRIP1, NBN, PTEN, and RAD51C). Of 21 patients with deleterious mutations in Lynch syndrome genes with tumor testing, 2 (9.5%) had tumor testing results suggestive of sporadic cancer. Of 12 patients with deleterious mutations in MSH6 and PMS2, 10 were diagnosed at age >50 and 8 did not have a family history of Lynch syndrome-associated cancers. Patients with deleterious mutations in non-Lynch syndrome genes were more likely to have serous tumor histology (23.1 vs 6.4%, P=0.02). The three patients with non-Lynch syndrome deleterious mutations and serous histology had mutations in BRCA2, BRIP1, and RAD51C. Current clinical criteria fail to identify a portion of actionable mutations in Lynch syndrome and other hereditary cancer syndromes. Performance characteristics of tumor testing are sufficiently robust to implement universal tumor testing to identify patients with Lynch syndrome. Germline multi-gene panel testing is feasible and informative, leading to the identification of additional actionable mutations.

Frequency of Germline Mutations in 25 Cancer Susceptibility Genes in a Sequential Series of Patients With Breast Cancer

PURPOSE

Testing for germline mutations in BRCA1/2 is standard for select patients with breast cancer to guide clinical management. Next-generation sequencing (NGS) allows testing for mutations in additional breast cancer predisposition genes. The frequency of germline mutations detected by using NGS has been reported in patients with breast cancer who were referred for BRCA1/2 testing or with triple-negative breast cancer. We assessed the frequency and predictors of mutations in 25 cancer predisposition genes, including BRCA1/2, in a sequential series of patients with breast cancer at an academic institution to examine the utility of genetic testing in this population.

METHODS

Patients with stages I to III breast cancer who were seen at a single cancer center between 2010 and 2012, and who agreed to participate in research DNA banking, were included (N = 488). Personal and family cancer histories were collected and germline DNA was sequenced with NGS to identify mutations.

RESULTS

Deleterious mutations were identified in 10.7% of women, including 6.1% in BRCA1/2 (5.1% in non-Ashkenazi Jewish patients) and 4.6% in other breast/ovarian cancer predisposition genes including CHEK2 (n = 10), ATM (n = 4), BRIP1 (n = 4), and one each in PALB2, PTEN, NBN, RAD51C, RAD51D, MSH6, and PMS2. Whereas young age (P < .01), Ashkenazi Jewish ancestry (P < .01), triple-negative breast cancer (P = .01), and family history of breast/ovarian cancer (P = .01) predicted for BRCA1/2 mutations, no factors predicted for mutations in other breast cancer predisposition genes.

CONCLUSION

Among sequential patients with breast cancer, 10.7% were found to have a germline mutation in a gene that predisposes women to breast or ovarian cancer, using a panel of 25 predisposition genes. Factors that predict for BRCA1/2 mutations do not predict for mutations in other breast/ovarian cancer susceptibility genes when these genes are analyzed as a single group. Additional cohorts will be helpful to define individuals at higher risk of carrying mutations in genes other than BRCA1/2.

Abstract P1-08-07: Predisposing germline mutations in a clinic based breast cancer (BC) population

Background: Evaluation of women with BC for germline mutations associated with hereditary breast and ovarian cancer (HBOC) has become increasingly common due to its impact on management. Guidelines for genetic evaluation indicate testing for cases with early onset, triple negative disease or family cancer history. However, the majority of breast cancer occurs in patients without these high risk characteristics. The prevalence of mutations associated with HBOC has not been well characterized in this population.

Methods: We performed a cross sectional study using DNA from blood samples from consecutive new invasive BC patients seen at the Dana-Farber Cancer Institute (01/01/2010 to 07/31/2102) who consented to research. Subjects were otherwise unselected. Mutations in 25 cancer genes were identified using a next generation sequencing based panel. Germline sequence variations and large rearrangements were classified for pathogenicity.

Results: 456 samples from eligible subjects were included. The mean age of BC diagnosis was 50 years. Mutations were found in 51 women, 49 of which were associated with breast cancer (10.8%, 95% CI 8.1-14.0). BRCA1/2 mutations were found in 6.6% [95% CI 4.5-9.2%] while mutations in other BC-associated genes were found in 4.4% [95% CI 2.7-6.7%], particularly CHEK2 (2.2%, 95% CI 1.1, 4.0). Of the 49 women with BC-related mutations, 21 (43%) had BC diagnosed after age 45. In univariate analyses, age at diagnosis, Ashkenazi Jewish ancestry, triple negative histology and family BC/ovarian cancer (OC) history were associated with BRCA1/2 mutations, but no factors were significantly associated with mutations in other genes. Among 261 women with no FDR/SDR with BC/OC, 26 (10.0%) had a mutation. Nineteen mutations (10 BRCA1/2) were found in the 256 women (7.4%) who had not had previous genetic testing.

Conclusions: In a single academic institution, 11% of new breast cancer patients had a germline mutation in a breast cancer predisposition gene: 6.6% were in BRCA1/2. The elevated prevalence compared to population-based series may reflect the practice composition of academic centers, which often attract women younger at BC diagnosis. In an academic practice with an active cancer genetics program, 10 women with BRCA1/2 and 9 with other mutations had not had genetic testing. Expanded testing identifies additional predisposing mutations, the utility of which are being defined for the care of breast cancer patients and their families.

Citation Format: Garber JE, Tung NM, Elkin EP, Allen BA, Singh NU, Wenstrup R, Hartman A-R, Winer EP, Lin NU. Predisposing germline mutations in a clinic based breast cancer (BC) population. [abstract]. In: Proceedings of the Thirty-Eighth Annual CTRC-AACR San Antonio Breast Cancer Symposium: 2015 Dec 8-12; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2016;76(4 Suppl):Abstract nr P1-08-07.

A comparison of the phenotype and genotype in adenomatous polyposis patients with and without a family history

OBJECTIVES

Adenomatous polyposis of the colon is often secondary to an inherited mutation in adenomatous polyposis coli (APC) gene, however, approximately one third of patients have no family history of the disease. We studied the phenotype and genotype of adenomatous polyposis in patients without a family history.

METHODS

A cohort of 57 unrelated adenomatous polyposis patients were evaluated. Seventeen patients with no family history were compared with 40 patients who had a positive family history of the disease. Family history and medical records were collected and analyzed. Germline APC and Mut Y homologue (MYH) testing was undertaken.

RESULTS

Patients without a family history were diagnosed with polyposis at an older age (41 years vs. 32 years) and presenting more frequently with symptoms (76 vs 20, P < 0.05). The number of colonic polyps and frequency of extracolonic manifestation associated with adenomatous polyposis did not differ between the two groups. APC mutations were detected less frequently among patients without a family history of the disease (4 out of 17 vs 25 out of 40, P=0.007), even among those with greater than 100 colorectal adenomas (4 out of 12 versus 21 out of 29, P=0.03). One homozygous MYH mutation carrier (G382D) was detected among the six patients without a family history and without a germline APC mutation who were tested.

CONCLUSIONS

Adenomatous polyposis patients without a family history are usually diagnosed with symptoms, and at a later age. Phenotypically, they are similar to those with a family history. However, germline APC mutations are detected far less frequently in patients without a family history. A small percentage of these cases may be secondary to biallelic germline MYH mutations.

Low rate of microsatellite instability in young patients with adenomas: reassessing the Bethesda guidelines

BACKGROUND AND AIM

Screening adenomas for microsatellite instability (MSI) in patients younger than 40 yr of age has been recommended by the Bethesda Guidelines as a means of identifying patients at risk for hereditary nonpolyposis colorectal cancer (HNPCC). We sought to determine the rate of MSI in adenomas removed from individuals under 40 yr of age over a 5-yr period in a university general gastroenterology practice.

METHODS

We identified patients between 18 and 39 yr of age with endoscopically removed adenomatous colorectal polyps. Patients with polyposis syndromes, inflammatory bowel disease, or colorectal carcinoma were excluded. A three-generation family history was obtained via telephone interview. Endoscopic and histology reports were reviewed. Adenomas were tested for MSI using the BAT26 and BAT40 microsatellite markers, and expression of the MSH2 and MLH1 proteins was assessed by immunostaining.

RESULTS

A total of 34 patients had 46 adenomas removed endoscopically. Out of 34 patients, 14 (41%) had a family history of colorectal cancer and 3 were from Amsterdam criteria positive families. A total of 28 of 46 adenomas (61%) were distal to the splenic flexure. Polyps ranged in size from 2 to 20 mm and averaged 6.6 mm. Five polyps (11%) were tubulovillous adenomas, and the remainder were tubular adenomas. None of the polyps were serrated adenomas and none demonstrated high-grade dysplasia. Among the 40 adenomas available for testing, none demonstrated MSI using either BAT26 (0/40) or BAT40 (0/21), nor did any of the polyps tested demonstrate loss of either MSH2 or MLH1 expression (0/16).

CONCLUSION

Screening adenomas from patients younger than 40 yr of age for MSI was ineffective in identifying potentially new cases of HNPCC. New strategies that improve on the current clinical and molecular screening methods should be developed so that at-risk individuals can be identified and referred for germline testing before developing their first cancer.

BRAF mutation is frequently present in sporadic colorectal cancer with methylated hMLH1, but not in hereditary nonpolyposis colorectal cancer

PURPOSE

The BRAF gene encodes a serine/threonine kinase and plays an important role in the mitogen-activated protein kinase signaling pathway. BRAF mutations in sporadic colorectal cancer with microsatellite instability (MSI) are more frequently detected than those in microsatellite stable cancer. In this study, we sought to compare the frequencies of BRAF mutations in sporadic colorectal cancer with MSI with those in hereditary nonpolyposis colorectal cancer (HNPCC).

EXPERIMENTAL DESIGN

We analyzed BRAF mutations in 26 colorectal cancer cell lines, 80 sporadic colorectal cancers, and 20 tumors from HNPCC patients by DNA sequencing and sequence-specific PCR. The methylation status of the hMLH1 gene was measured by either sequencing or restriction enzyme digestion after NaHSO(3) treatment.

RESULTS

We observed a strong correlation of BRAF mutation with hMLH1 promoter methylation. BRAF mutations were present in 13 of 15 (87%) of the colorectal cell lines and cancers with methylated hMLH1, whereas only 4 of 91 (4%) of the cell lines and cancers with unmethylated hMLH1 carried the mutations (P < 0.00001). Sixteen of 17 mutations were at residue 599 (V599E). A BRAF mutation was also identified at residue 463 (G463V) in one cell line. In addition, BRAF mutations were not found in any cancers or cell lines with K-ras mutations. In 20 MSI+ cancers from HNPCC patients, however, BRAF mutations were not detectable, including a subset of 9 tumors with negative hMLH1 immunostaining and methylated hMLH1.

CONCLUSIONS

BRAF mutations are frequently present in sporadic colorectal cancer with methylated hMLH1, but not in HNPCC-related cancers. This discrepancy of BRAF mutations between sporadic MSI+ cancer and HNPCC might be used in a strategy for the detection of HNPCC families.

Genotype and phenotype of patients with both familial adenomatous polyposis and thyroid carcinoma

The incidence of thyroid carcinoma in familial adenomatous polyposis (FAP) is thought to be 1%-2%, with the majority of cases being female. We have investigated the phenotype and genotype of 16 patients with FAP associated thyroid carcinoma. Among 1194 FAP patients studied in two high risk registries in North America (Familial Gastrointestinal Cancer Registry, Toronto and University California, San Francisco), 16 (1.3%) unrelated patients with FAP associated thyroid cancers were identified. Adenomatous polyposis coli (APC) gene testing was performed in 14 of the 16 cases. The average age of diagnosis for FAP and thyroid carcinoma was 29 years (range 17-52 years) and 33 years (range 17-55 years), respectively. All FAP patients except 1 had more than 100 colonic adenomas. Extracolonic manifestations, beside thyroid cancer, were presented in 81% (n = 13) of the patients, including gastric and duodenal polyps, desmoid tumor, osteoma, epidermoid cyst, sebaceous cyst and lipoma. Colorectal cancer was diagnosed in 38% (n = 6) of the patients. The pathology of the FAP associated thyroid cancer was predominantly papillary carcinoma. Germline mutations were identified in 12 of 14 patients tested. Mutations proximal to the mutation cluster region (1286-1513) were detected in 9 cases. Thyroid cancer in our FAP population was rare, predominantly in females and showed papillary carcinoma histology. Additionally, thyroid cancer in our patients occurred in the setting of classic FAP phenotype. Germline mutations were located predominantly outside the APC mutation cluster region.

Hereditary polyposis syndromes and hereditary non-polyposis colorectal cancer

Colorectal cancer due to hereditary syndromes comprises approximately 5% of the overall colorectal cancer burden. Conditions fall into two distinct categories, the polyposis syndromes and hereditary non-polyposis colorectal cancer. It is important for the clinician to have a working knowledge of both as screening and surveillance recommendations differ significantly from those applicable to the general population. The polyposis syndromes include familial adenomatous polyposis, Peutz-Jeghers syndrome, juvenile polyposis, and Cowden syndrome. For each condition, a review of both the intestinal and extra-intestinal clinical findings is presented as well as the genetic basis, genetic testing, screening, surveillance and treatment options. As genetic testing for several of these conditions has recently become both commercially available and standard practice, special attention is given to indications and strategies for genetic testing in hereditary colorectal cancer syndromes.

Comparison of selection strategies for genetic testing of patients with hereditary nonpolyposis colorectal carcinoma: effectiveness and cost-effectiveness

BACKGROUND

Molecular testing for hereditary nonpolyposis colorectal carcinoma (HNPCC) is becoming standard care and it is cost-effective compared with no genetic testing. However, the best strategy for detection of HNPCC gene carriers is unknown.

METHODS

We use a decision analytic model to evaluate the effectiveness and incremental cost-effectiveness of four commonly used testing strategies to detect HNPCC gene carriers. The model starts with a population of colorectal carcinoma (CRC) patients and measures costs, the number of gene carriers detected, and incremental costs per gene carrier detected.

RESULTS

We found that germline testing on only those CRC probands who meet the Amsterdam criteria detects the fewest gene carriers and has the lowest cost whereas tumor microsatellite instability (MSI) testing of all CRC patients and families has the highest cost and detects the most gene carriers. When cost-effectiveness is considered, the mixed strategy (MSH2 and MLH1 testing on those who meet the Amsterdam criteria and germline testing for the remainder who meet less stringent modified criteria and are MSI-High) seems superior. The mixed strategy detects 59.6 mutation carriers per 1000 CRC cases and costs much less than the test all strategy, which has an incremental cost-effectiveness of $51,151. The mixed strategy often other strategies and when compared to the Amsterdam strategy, has a cost-effectiveness of only $6441 per gene carrier detected.

CONCLUSIONS

It is not very effective to limit genetic testing to only individuals who meet the Amsterdam criteria, as many gene carriers are missed. However, testing all CRC patients for tumor MSI-H, although effective, may be prohibitively expensive. A mixed strategy is the more cost-effective approach.

Hereditary nonpolyposis colorectal cancer in young colorectal cancer patients: high-risk clinic versus population-based registry

BACKGROUND & AIMS

Early onset colorectal cancer (CRC) is an important feature of hereditary nonpolyposis colorectal cancer (HNPCC). We sought to compare rates of genetically defined HNPCC among individuals with early onset CRC drawn from a high-risk clinic and a population-based cancer registry.

METHODS

Probands with CRC diagnosed before 36 years of age were enrolled from a high-risk CRC clinic at the University of California, San Francisco (UCSF), and a population-based Kaiser Permanente (KP) Health Plan cancer registry. Probands provided cancer family histories and tumors for microsatellite instability (MSI) testing and MSH2/MLH1 protein immunostaining. Germline MSH2 and MLH1 mutational analysis was performed.

RESULTS

Forty-three probands were enrolled from UCSF and 23 from KP. The UCSF and KP probands had similar median age of onset of CRC (30 vs. 31 years) and the percentage with any personal or family history of another HNPCC-related cancer (70% vs. 74%). However, 28 of 40 (70%) of the UCSF tumors were MSI-H compared with 6 of 18 (33%) of KP tumors (P = 0.01), and 13 germline MSH2 or MLH1 mutations were found in the UCSF group compared with 0 in the KP group (P = 0.0001). In a multivariate analysis, institution (P = 0.002) and the total number of colorectal cancers in the family (P = 0.0001) were independent predictors of MSH2 or MLH1 mutation.

CONCLUSIONS

Family history of cancer is an important feature of HNPCC, even among individuals with early onset CRC. Caution must be undertaken when extrapolating data regarding HNPCC from high-risk clinic populations to the general population.

Prisoner of war experience: effects on wives

The aim of this study was to compare the wives of Australian soldiers who had been imprisoned during World War II (POWs) with a control group of non-POWs' wives and also to compare the POWs and non-POWs themselves in respect to several psychological and family life characteristics on which differences might be expected to arise from the long-term effects of imprisonment. A random sample of 145 of these veterans and their wives completed several self-administered mood and family life scales, an inventory of somatic symptoms, questions about the impact of the war on the veteran in the postwar decades, and several social background questions. The POWs themselves were more depressed and reported more somatic symptoms and a greater postwar impact of the war than the non-POWs. However, these differences were not accompanied by concomitant differences among their wives. There was some evidence of an influence of the POW's mood on his wife's mood in significant correlations between husbands' and wives' depression and anxiety scale scores in POW couples alone. Otherwise, there was very little indication that the POW experience had any long-term effect on the marriage relationship as measured by the variables included in this study.

A decision-supported outpatient practice system

We describe a Decision-supported Outpatient Practice (DOP) system developed and now in use at the Columbia-Presbyterian Medical Center. DOP is an automated ambulatory medical record system that integrates in-patient and ambulatory care data, and incorporates active and passive decision support mechanisms with a view towards improving the quality of primary care. Active decision support occurs in the form of event-driven reminders created within a remote clinical information system with its central data repository and decision support system (DSS). Novel features of DOP include patient specific health maintenance task lists calculated by the remote DSS. uses of a semantically structured controlled medical vocabulary to support clinical results review and provider data entry, and exploitation of an underlying ambulatory data model that provides for an explicit record of evolution of insight regarding patient management. Benefits, challenges, and plans are discussed.

Lymphatic air embolism: a new hypothesis regarding the pathogenesis of neonatal systemic air embolism

Objective. Neonatal systemic air embolism (NSAE) has been thought to result from introduction of air into the pulmonary veins through hypothesized alveolar-capillary fistula. The objective of this paper is to reassess the distribution of intravascular air visualized radiographically in this entity. Based on these data, an alternative theory for the pathogenesis of NSAE is proposed. Materials and methods. Four cases from our institutions and 21 additional reviewable published radiographs of NSAE were evaluated for the presence and location of intravascular air. Nonparametric statistical analysis was performed to determine if the predominance of intravascular air was venous or arterial in location, and to determine the presence or absence of pulmonary interstitial emphysema (PIE), pneumothorax, pneumomediastinum, and pneumopericardium. Results. Isolated systemic venous air was present significantly more often than isolated systemic arterial air (p < 0.0005). In addition, the presence of isolated right-sided cardiac air was found significantly more often than air within left heart chambers alone (p < 0.0005). PIE in patients with NSAE was found to be statistically more prevalent than pneumomediastinum or pneumopericardium, though similar in prevalence to pneumothorax. Conclusion. Our data support the hypothesis that intravascular air in NSAE is predominantly venous in location. PIE was found to be a very common associated finding in NSAE. Based on our current knowledge of pulmonary lymphatics, radiographic anatomy, and the lymphatic location of PIE, we propose that air within the pulmonary lymphatic system (PIE) gains access to the systemic venous system via lymphatic ducts, which results in the clinical entity NSAE. This readily explains the venous predominance of air in NSAE.

Medical Informatics Training at Columbia University and the Columbia-Presbyterian Medical Center

The Department of Medical Informatics at Columbia University College of Physicians and Surgeons consists of a faculty of 17 full- and part-time faculty. The Department faculty collaborate with the Department of Computer Science and several clinical departments of the medical center. We offer courses in medical informatics, formal degrees (M.A., M.Phil. and Ph.D.) and a postdoctoral training program. In addition to academic offerings, the close affiliation with the Columbia-Presbyterian Medical Center and the primary responsibilities for clinical information systems offers trainees unique opportunities to work with and develop real-world applications. Faculty research programs include work on the Integrated Advanced Information Management System (IAIMS), Unified Medical Language System (UMLS), High-Performance Computing and Communications (HPCC), Electronic Medical Records, automated decision support and technology transfer through the Center for Advanced Technology.

Ventricular fibrillation complicating endomyocardial biopsy of a cardiac allograft

Transvenous endomyocardial biopsy remains the most useful diagnostic aid in assessing rejection in the transplanted heart. Although invasive, the complications associated with endomyocardial biopsy are few, and the procedure is generally regarded as safe. We report a case of apparent ventricular fibrillation complicating transvenous endomyocardial biopsy. Histologic section revealed evidence of moderate acute rejection. This case represents the first report of a life threatening ventricular dysrhythmia following routine endomyocardial biopsy in a cardiac transplant recipient.

CARTOS II semi-automated nerve tracing: three-dimensional reconstruction from serial section micrographs

The paper describes the history of tracing activities in the Levinthal lab from 1965 to the present. It also focuses on the evolution of electron micrograph (EM) tracing from early hand tracing reconstruction to the current, nearly fully automated CARTOS II tracing system. The present hardware systems and tracing algorithms are described in detail. Also presented are the partial tracing results of one EM data set in two and three dimensions.

Rhythmic-movement facilitation of learning in working-class Afro-American children

In this study, we examined whether the cognitive performance of working-class Afro-American children could be enhanced based upon a presumed knowledge of the Black cultural experience. Children (6 and 9 years old) learned to match each of 12 pairs of pictures via a rhythmic movement-coordinated procedure accompanied by a musical beat or through a rote recitation procedure. It was expected and found that subsequent retention of the picture matches was superior in the rhythmic movement condition. Moreover, for the younger children, the effect was stronger for those who came from homes providing high stimulation and who had high levels of movement expressiveness.

Lifestyle assessment: just asking makes a difference

A 10 minute assessment of 180 family practice patients showed that 11% indicated a problem with drinking alcohol, 20% with cigarette smoking, 36% with consumption of coffee or tea, and 3% with non-medical drug use, while 11% wanted to discuss their use of medications. Moreover, being asked questions resulted in a twofold or threefold increase in the patients' intentions of discussing such a problem with their doctor. Although there was good overall agreement in recognising a problem between the patient and doctor, in roughly 40% of instances where the patient indicated a problem the doctor was unaware of it. These patients tended to be young, well educated, and employed in professional occupations, and were on their first visit to the doctor. Such brief assessments of lifestyle should be routinely conducted in family practice for both case finding and prevention.

Lifestyle assessment: applying microcomputers in family practice

A randomised trial of assessment by computer was conducted with 180 patients in a family practice clinic. Histories of alcohol, tobacco, and drug use were obtained by computer (n = 60), interview (n = 60), or self completed questionnaire (n = 60). The results of previous research suggest that some patients may provide more accurate information about "sensitive" problems to a computer. No significant differences, however, in levels of consumption or problems were reported for the three methods of assessment. Patients gave differential ratings about the method of assessment, with the computer rated as more interesting but also more mechanical, cold, and impersonal. Although the interview was initially preferred by most, patients who completed the assessment by computer showed a significant increase (13% to 43%) in their preference for the computer after the assessment. The results of our study indicate that patients' acceptance of computers in family practice may be favourably influenced by direct experience with a microcomputer.

Differential assessment of alcoholism. Evaluation of the Alcohol Use Inventory

The biases of response styles, the internal-consistency reliability, concurrent validity and construct validity of the Alcohol Use Inventory were evaluated.

Correlation between medical and behavioral data in the assessment of alcoholism

Using the Munich Alcoholism Test, this study examined the prevalence and interrelationships of medical and sociobehavioral disorders in a nonhospitalized sample of 106 males who had been referred to a medical officer for "suspected" problems related to drinking. A further objective was to evaluate measurement properties of the Munich Alcoholism Test (MALT), a new diagnostic instrument for alcoholism. Self-report items focusing on the recognition of drinking problems formed a homogeneous and quite reliable scale. However, clinical signs and symptoms of disorders related to alcohol abuse occurred with relative independence of each other. In this predominantly young group of subjects, many individuals recognized that they had sociobehavioral problems, but few had clinical or laboratory manifestations of diseases associated with chronic alcohol abuse. These findings underscored the advantages of including both biomedical and sociobehavioral data for the early identification of alcohol abuse and dependence.