Faculty Bibliography

Insulin-like growth factor I (IGF-I) is a systemic hormone with potent mitogenic and anti-apoptotic properties, which could influence the proliferative behavior of normal breast cells. Limited epidemiological observations suggest that the hormone may play a role in the etiology of breast cancer, especially at pre-menopausal ages. In a prospective case-control study nested within a cohort of New York City women, IGF-I, IGF-binding protein 3 (IGFBP-3) and C peptide were measured in frozen serum samples from 172 pre-menopausal and 115 post-menopausal subjects who were subsequently diagnosed with breast cancer. Subjects were eligible if diagnosed 6 months or more after recruitment into the study (7 to 120 months). Cohort members who matched the cases on age, menopausal status, date of blood sampling and day of menstrual cycle at blood collection served as controls. Post-menopausal breast cancer was not associated with serum IGF-I, IGFBP-3 or C-peptide levels. However, the risk of breast cancer increased with increasing serum concentrations of IGF-I in pre-menopausal women. The odds ratio (OR) for the highest quartile of IGF-I (>256 ng/ml) compared to the lowest (<168 ng/ml) was 1.60 [95% confidence interval (CI) 0.91-2. 81]. The OR decreased to 1.49 (95% CI 0.80-2.79) after adjustment for IGFBP-3. In analyses restricted to subjects who were pre-menopausal at the time of blood sampling and whose cancer was diagnosed before age 50, the top vs. bottom quartile OR increased appreciably to 2.30 (95% CI 1.07-4.94). Adjustment for IGFBP-3 reduced the OR to 1.90 (95% CI 0.82-4.42). There was no association between pre-menopausal breast cancer and IGFBP-3, IGF-I:IGFBP-3 ratio or non-fasting levels of C peptide. Elevated circulating levels of IGF-I may be an indicator of increased risk of breast cancer occurring before age 50

The lack of tumor models that can reliably predict for response to anticancer agents remains a major deficiency in the field of experimental cancer therapy. Although heterotransplants of certain human solid tumors can be successfully grown in nude mice, they have never been appropriately explored for prediction of in vivo chemosensitivity to anticancer agents. We determined the tumor response rate and studied the influence of several biological and molecular tumor parameters on the in vivo sensitivity to paclitaxel in a series of heterotransplanted human non-small cell lung cancer (NSCLC) tumors. One hundred consecutive resected NSCLC tumors were heterotransplanted s.c. in nude mice. The in vivo sensitivity to i.v. paclitaxel (60 mg/kg every 3 weeks) was studied in 34 successfully grown heterotransplants. Treatment started when the tumors reached a size of 5 mm in diameter, and strict standard clinical criteria (>50% shrinkage in tumor weight or cross-sectional surface) were used to define tumor response. Baseline multidrug resistance protein (MRP), Her-2/neu, and epidermal growth factor receptor (EGFR) expression, and pre- and posttherapy bax and bcl-2 expression were determined by Western blot analysis. p53 status was determined by sequencing. The overall take rate was 46% (95% confidence interval, 36-56%) and was significantly higher (P < 0.05) for squamous carcinoma tumors (75%) than for adenocarcinoma tumors (30%) and bronchoalveolar tumors (23%). The heterotransplants were morphologically very similar to the original tumors. The response rate to paclitaxel was 21% (95% confidence interval, 9-38%). Baseline tumor parameters associated with response were no Her-2/neu expression (none of the responding tumors expressed Her-2/neu versus 48% of the nonresponding tumors, P = 0.05) and baseline bcl-2 expression (all responding tumors expressed bcl-2 versus only 43% of the nonresponding tumors, P = 0.02). There was a trend toward a higher response rate in bax-positive tumors, and MRP- and EGFR-negative tumors, but it was not statistically significant. The response was independent of baseline p53 status and baseline mitotic index. Responding tumors had a higher bax/bcl-2 ratio 24 h after therapy, but the difference was only marginally significant (2.8 for responding tumors versus 1.1 for nonresponding tumors, P = 0.07). The extent of mitotic arrest at 24 h after therapy was not associated with response. Human NSCLC heterotransplants are morphologically identical to the original tumors and have a response rate to paclitaxel that is equivalent to that reported in Phase II studies in patients with advanced NSCLC treated with single-agent paclitaxel. NSCLC heterotransplants deserve to be explored to evaluate new agents for lung cancer and to predict clinical response on an individual basis in selected groups of patients

BACKGROUND: Leading a Western lifestyle, being overweight, and being sedentary are associated with an increased risk of colorectal cancer. Recent theories propose that the effects of these risk factors may be mediated by increases in circulating insulin levels and in the bioactivity of insulin-like growth factor (IGF)-I. To test this hypothesis, we conducted a case-control study nested within a cohort of 14 275 women in New York. METHODS: We used blood samples that had been obtained from these women from March 1985 through June 1991 and stored in a biorepository. C-peptide (a marker for insulin secretion), IGF-I, and IGF-binding proteins (IGFBPs)-1, -2, and -3 were assayed in the serum of 102 women who subsequently developed colorectal cancer and 200 matched control subjects. Logistic regression was used to relate cancer risk to these peptide levels, by adjustment for other risk factors. All statistical tests used are two-sided. RESULTS: Colorectal cancer risk increased with increasing levels of C-peptide (P:(trend) =.001), up to an odds ratio (OR) of 2. 92 (95% confidence interval [CI] = 1.26-6.75) for the highest versus the lowest quintiles, after adjustment for smoking. For colon cancer alone (75 case subjects and 146 control subjects), ORs increased up to 3.96 (95% CI = 1.49-10.50; P:(trend) <.001) for the highest versus the lowest quintiles. A statistically significant decrease in colorectal cancer risk was observed for increasing levels of IGFBP-1 (P:(trend) =.02; OR in the upper quintile = 0.48 [95% CI = 0.23-1. 00]), as well as for the highest quintile of IGFBP-2 levels (P:(trend) =.06; OR = 0.38 [95% CI = 0.15-0.94]). Colorectal cancer risk showed a modest but statistically nonsignificant positive association with levels of IGF-I and was statistically significantly increased for the highest quintile of IGFBP-3 (OR = 2.46 [95% CI = 1. 09-5.57]). CONCLUSIONS: Chronically high levels of circulating insulin and IGFs associated with a Western lifestyle may increase colorectal cancer risk, possibly by decreasing IGFBP-1 and increasing the bioactivity of IGF-I

A genetic variant of luteinizing hormone (LH) characterized by two point mutations in codons 8 (TGG-->CGG) and 15 (ATC-->ACC) of the LH beta-subunit gene has been described recently. As compared with wild-type LH, the variant LH appears to have higher in vitro bioactivity but a shortened circulatory half-life, and it has been reported to affect circulating levels of sex hormones. Our purpose was to determine whether the variant form of LH is associated with an altered risk of breast cancer. This hypothesis was addressed in a case-control study nested within a prospective cohort that included 270 cases of breast cancer and twice as many matching control subjects. The study was limited to subjects diagnosed at age 50 years or older. The LH status was determined by the combination of two immunofluorometric assays of serum using monoclonal antibodies. Frequency of the variant LH was similar in breast cancer cases and controls (11.5% versus 10.7%). In conditional regression models, the presence of the variant LH was not associated with a considerable increase of breast cancer risk (odds ratio, 1.07; 95% confidence interval, 0.68-1.69). Adjustment for potential confounders did not notably change the risk estimate (odds ratio, 1.11; 95% confidence interval, 0.69-1.78). These observations do not appear to support the hypothesis that this particular variant of LH is associated with altered risk of breast cancer diagnosed at age 50 years and older

BACKGROUND: The use of psychotropic medications may increase the risk of hormone-related cancers in females through increased gonadotropin secretion, but the data from epidemiologic studies are limited to evaluate the hypothesis. METHODS: The association between the use of psychotropic medications and cancer incidence was studied in a prospective cohort study that involves 15,270 women who participated in mammographic screening. The relative risks (RR) and 95 per cent confidence intervals (CIs) for cancer associated with the use of psychotropic medications were estimated by the Cox's proportional hazard model. RESULTS: During an average of 7.3 years of follow-up, 1,130 incident cases of cancer were identified, including 566 breast, 67 endometrial and 47 ovarian cancers. The use of any type of psychotropic medication at baseline was associated with increased risks of breast [relative risk (RR) = 1.39, 95 per cent CI 1.11-1.74], endometrial (RR=1.71; 95 per cent CI 0.93-3.14) and ovarian (RR= 1.48, 95 per cent CI 0.69-3.16) cancers, whereas no increase in risk was observed for other cancers (RR = 1.06). When the subjects were divided by menopausal status at baseline, premenopausal women tended to have higher risk of all hormone-related cancers (RR = 1.73, 95 per cent CI 1.27-2.35) than postmenopausal women (RR=1.23, 95 per cent CI 0.94-1.62). The magnitude of the RR associated with the use of these medications did not change by length of follow-up. Analysis by type of medication did not find that the association was limited to specific types. CONCLUSION: The observed association needs to be confirmed in further studies based on more detailed medication history

OBJECTIVE: We examined the reliability of the fatty acid composition of serum phospholipids in the New York University Women's Health Study, a prospective study of sex hormones, diet and breast cancer. DESIGN: Non-fasting serum samples collected at three yearly visits, in 46 healthy women, and stored at -80 degrees C for 7-12 y, were included in the study. Serum phospholipid fatty acid composition was measured by capillary gas chromatography. RESULTS: For the 20 individual fatty acids measured, the reliability coefficients were less than 0.50 for four, between 0.50 and 0.70 for nine, and greater than 0.70 for seven. Among the major fatty acids, arachidonic and alpha-linolenic acids had high reliability coefficients (0.71 and 0. 72, respectively), palmitic, oleic, linoleic, eicosapentaenoic and docosahexaenoic acids had intermediate coefficients (0.57, 0.69, 0. 62, 0.64 and 0.66, respectively), whereas stearic acid had the lowest coefficient (0.15). The reliability coefficients for total monounsaturated fats, omega-6 and omega-3 fatty acids were moderately high (0.66, 0.53 and 0.66, respectively), whereas the coefficients for total saturated fats and total polyunsaturated fats were low (0.31 and 0.43, respectively). CONCLUSIONS: These results indicate that the fatty acid composition of serum phospholipids can be a useful tool in epidemiologic studies, although for most fatty acids a single determination is associated with some error in measurement that should be taken into account at the design and analysis stages. Storage for up to 12 y at -80 degrees C preserved polyunsaturated fatty acids from oxidation very well

Iron overload, expressed as increased body iron stores, has been recognized as a potential hazard because it promotes the generation of oxygen radicals. We analyzed factors associated with serum ferritin levels (an indicator of body iron stores) among middle-aged women with a high prevalence of nutrient supplement use. Serum ferritin concentrations were determined on automated immunoassay for 487 healthy women with the mean age of 57 years who participated in the New York University Women's Health Study. The mean serum ferritin concentration in postmenopausal women was more than twice that in premenopausal women. Serum ferritin concentrations progressively increased with advancing age, but adjustment for menopausal status considerably weakened this association. Among non-dietary factors, nonwhite ethnicity, obesity and cigarette smoking were positively associated with serum ferritin concentrations. After adjustment for these factors and for menopausal status, serum ferritin levels were positively associated with meat intake and multivitamin use and inversely associated with breakfast cereal consumption. However, none of these lifestyle factors positively associated with serum ferritin levels had a significant impact on serum ferritin levels above 100 ng/ml (approximately equal to median concentration). Our results suggest that iron overload seems unlikely among middle aged women through their diet and nutritional supplements

A prospective investigation of breast cancer and organochlorine (OC) exposures was undertaken in the New York University Women's Health Study. Cases (n = 148) and individually matched controls (n = 295) were identified among women whose blood had been obtained 6 months or more prior to breast cancer diagnosis. In addition, among 84 cases and 196 controls, two or more consecutive annual blood samples were available to estimate half-lives of 1,1-dichloro-2,2-bis(p-chlorophenyl) ethene (DDE) and polychlorinated biphenyls (PCBs). Cases and controls had similar levels of DDE (geometric mean, 6.95 versus 7.27 ng/ml; lipid-adjusted geometric mean, 977 versus 1100 ng/g) and PCBs (5.04 versus 4.97 ng/ml; lipid-adjusted geometric mean, 683 versus 663 ng/g). These differences remained nonsignificant when estrogen receptor status of tumors was considered. DDE and PCB half-lives did not differ in case versus control patients. In control patients, DDE and PCB half-lives were strongly correlated (r(s) = 0.71), and the half-life of DDE (but not that of PCB) was inversely correlated with body mass index (BMI), yet the blood serum levels of PCB (but not those of DDE) were correlated with BMI. We conclude that there is no evidence for an association of breast cancer risk with DDE or PCB levels in blood (based on samples collected during the period 1987-1992) nor with their elimination half-lives. However, changes in DDE and PCBs over time are influenced by metabolism, BMI, and current OC exposures, and each may affect interpretation of OC levels in risk assessment models

OBJECTIVE: Bone fractures are an important cause of morbidity and mortality among the elderly in the US. The present study assesses the possible role of a number of risk factors for postmenopausal bone fractures. METHODS: We analysed the relationships of anthropometric, demographic and lifestyle factors with the risk of bone fracture among 6250 postmenopausal women in a prospective cohort study, the New York University Women's Health Study. RESULTS: After an average of 7.6 years of follow-up, 1025 new incident bone fractures were reported, including 34 hip and 159 wrist fractures (incidence rates; 71.6 and 334.7 per 105 woman-years, respectively). The risk of fracture increased with increasing age, body height and total fat intake, while it was significantly lower among obese and African American women. The relative risk among African Americans was 0.45 (95% CI: 0.32-0.63) compared with non-African Americans. Women taller than 170 cm had a 64% increase in risk of fractures, as compared with those under 155 cm. These associations were generally more pronounced when fractures were limited to those at the hip and wrist. CONCLUSIONS: The present study provides an indication for a potential role of dietary fat in the development of postmenopausal fractures and further evidence to support protective effects of obesity, short stature and African American ethnicity