Dietary risk factors for the incidence and recurrence of colorectal adenomatous polyps. A case-control study

OBJECTIVE

To investigate the association of dietary factors with the incidence and recurrence of colorectal adenomatous polyps.

DESIGN

Two case-control studies.

SETTING

Three university-based colonoscopy practices in New York City.

PATIENTS

For the incidence study, 286 patients with pathologically confirmed incident adenomatous polyps (162 men and 124 women) were compared with 480 controls (210 men and 270 women) with no current or previous neoplasia. For the recurrence study, 186 patients with recurrent polyps (130 men and 56 women) were compared with 330 controls (187 men and 143 women). These patients had a history of polyps but no current neoplasia.

MEASUREMENTS

Structured interviews using the Block food frequency questionnaire were conducted on all participants and were compared over quartiles of crude nutrient intake of total and saturated fat; fiber; protein; carbohydrates; carotene; vitamins A, C, and E; and various food groups. Data were adjusted for age, Quetelet index, and caloric intake by multiple logistic regression for men and women separately.

RESULTS

For incident polyps, elevated adjusted odds ratios (high to low quartile) for women were found for saturated fat (odds ratio, 2.3; 95% CI, 0.9 to 5.8) and the ratio of red meat to chicken and fish intake (odds ratio, 1.9; CI, 1.0 to 3.6). Protective associations were observed for fish and chicken (odds ratio, 0.6; CI, 0.3 to 1.2) and vitamin A intake (odds ratio, 0.4; CI, 0.2 to 0.9). Among women, recurrent polyps showed an association with total dietary fat (odds ratio, 4.4; CI, 1.0 to 19.5), saturated fat (odds ratio, 3.8; CI, 1.0 to 14.9; P = 0.15 for the trend), and total fiber (odds ratio, 0.2; CI, 0.1 to 0.5; P = 0.01) and a borderline association with carbohydrates (odds ratio, 0.3; CI, 0.1 to 1.3; P = 0.10). No consistent relations were observed for men, although increased caloric intake increased the risk for incidence and recurrence in both men and women.

CONCLUSIONS

These findings are consistent with previously described relations between diet and the incidence of colon cancer and suggest that, at least in women, dietary modification may be warranted in patients diagnosed with polyps.

The pattern of diagnosis of a second primary tumor in the breast

One hundred and sixty-seven patients with metachronous bilateral breast cancer were diagnosed at the Northern Israel Oncology Center during the years 1950-1989. The group at high risk to develop a second breast tumor included Jewish women born in Europe whose first tumor was diagnosed when the patient was under the age of 55. The mean time interval between tumors was 88 months. Seventy percent of the patients were diagnosed within nine years of the diagnosis of the first tumor. The characteristics of the 27% of patients with single breast cancers who did not comply with follow-up recommendations were compared to those who did comply. Patients who were under follow-up had smaller tumor and less lymph node involvement. Nevertheless, their survival rate did not differ from those who did not keep their follow-up appointments. This was ascribed to the fact that follow-up procedures for many years used mainly clinical examination and this was not enough to decrease mortality. The diagnosis of non-palpable breast cancer by routine yearly mammography has proved to reduce mortality in patients over the age of 50 with single breast cancers. Therefore, yearly mammography of the contralateral breast in patients with single breast tumors must be done in order to increase the cure rate of contralateral breast cancer. Less patient delay in diagnosis was found before the diagnosis of the second tumor than in patients with a single tumor.

Survival of first and second primary breast cancer

BACKGROUND

A second primary tumor (SPT) in the breast is the most common one seen in clinical practice. There are conflicting reports regarding the incidence and survival of patients with SPT in the breast.

METHODS

To elucidate this, data on 139,932 patients with primary breast cancer, reported to the Surveillance, Epidemiology, and End Results Program between 1973 and 1986, were analyzed. Of these, 3431 patients had a contralateral metachronous breast cancer (interval, > 6 months).

RESULTS

Survival from the date of diagnosis of the metachronous breast tumors was compared with that of patients with single breast tumors, controlling for age at diagnosis, stage, race, and treatment. In the multivariate analysis, age and stage at diagnosis of first and second tumors were the only variables that significantly influenced survival. The survival of patients with localized SPT was similar to that of patients with a localized single breast tumor. Patients with regionally advanced SPT lived for a shorter time than did corresponding patients with only one tumor.

CONCLUSIONS

Patients with single breast cancer were at increased risk of having a SPT in the breast. As a result, efforts for early detection by physical examination and mammography of survivors of single breast cancer are advocated because the survival of patients with a localized SPT in the breast is as good as that of those with a single localized tumor.

A comparison of the clinical characteristics of second primary and single primary sarcoma: a population based study

The clinical characteristics of 240 patients with sarcoma as a second metachronous primary neoplasm (SPN) were compared with those of 8,815 patients with sarcoma as a single tumor. The data were obtained from patients registered during the period 1973-1986 in the Surveillance, Epidemiology and End Results (SEER) Program in the United States. Seventy-four of the 240 SPN patients had postirradiation sarcoma (PIS) while the other 153 patients developed the sarcoma as a second tumor in an area which was not exposed to prior radiotherapy (non-PIS). The stage of disease at diagnosis was more advanced in patients with PIS than in those with single sarcomas but the difference did not reach statistical significance. Overall, in comparable clinical stage localized and regional disease there was no statistically significant difference in survival between PIS and non-PIS sarcoma patients after adjusting for age. The survival of patients with localized or regionally advanced sarcoma as a second tumor was significantly worse than of those with single sarcomas with the same stage. There was no difference in survival between first or second sarcomas with metastatic disease.

Colonoscopic screening for neoplasms in asymptomatic first-degree relatives of colon cancer patients. A controlled, prospective study

Individuals with a family history of colorectal cancer are believed to be at an increased risk of developing colorectal neoplasia. To estimate this risk and the potential yield of screening colonoscopy in this population, we recruited and prospectively colonoscoped 181 asymptomatic first-degree relatives (FDR) of colorectal cancer patients and 83 asymptomatic controls (without a family history of colorectal cancer). The mean ages for the FDR and control groups were 48.2 +/- 12.5 and 54.8 +/- 11.0, respectively. Adenomatous polyps were detected in 14.4 percent of FDRs and 8.4 percent of controls. Although 92 percent of our FDRs had only one FDR afflicted with colon cancer, those subjects with two or more afflicted FDRs had an even higher risk of developing colonic adenomas (23.8 percent) than those with only one afflicted FDR (13.1 percent). A greater proportion of adenomas was found to be beyond the reach of flexible sigmoidoscopy in the FDR group than in the controls (48 percent vs. 25 percent, respectively). Logistic regression analysis revealed that age, male sex, and FDR status were independent risk factors for the presence of colonic adenomatous polyps (RR = 2.32, 2.86, and 3.49, respectively; P less than 0.001). Those at greatest risk for harboring an asymptomatic colonic adenoma are male FDRs over the age of 50 (40 percent vs. 20 percent for age-matched male controls). Based on probability curves, males with one FDR afflicted with colon cancer appear to have an increased risk of developing a colonic adenoma beginning at 40 years of age. Our results document, for the first time, an increased prevalence of colonoscopically detectable adenomas in asymptomatic first-degree relatives of colon cancer patients, as compared with asymptomatic controls, and support the use of colonoscopy as a routine screening tool in this high-risk group.

Clinical, demographic, and follow-up characteristics of patients with two primary metachronous tumors, one of them being in the colon

The clinical characteristics of 152 patients diagnosed with two primary metachronous tumors--one or both of them in the colon--were studied. Nineteen patients had both primary tumors in the colon (Group I), 59 had the first primary tumor in the colon and the second tumor elsewhere (Group II), and 74 had the second primary tumor in the colon and the first primary tumor elsewhere (Group III). The group in which the second primary tumor was in the colon included significantly more female patients than did the other two groups, with a younger median age at diagnosis of first tumor. The median time interval between the two primary tumors was 44, 57, and 62 months in Groups I, II, and III, respectively. The number of clinic visits during the year before diagnosis of the second primary was similar in all groups, but only 60 percent of the patients kept their follow-up appointment. In most instances, the diagnosis was made after the patients' symptoms, although only a small percentage of the second primary tumors (15-30 percent) were diagnosed during routine follow-up. The second primary tumor occurred in the field of radiotherapy of the first primary tumor in 27 of 35 patients who received radiotherapy. To increase the number of patients diagnosed in an earlier stage of disease, they should be urged to keep their follow-up appointment, and physicians following patients with single tumors should be aware of the increased likelihood of a second tumor. To increase the cure rate of those tumors, efforts toward early diagnosis are warranted. This includes physical examination and mammography to detect breast cancer in women, annual occult blood tests and rectal examination, and sigmoidoscopy or colonoscopy at three-year intervals to detect colon cancer early.

Relationship between breast cancer and colorectal adenomatous polyps. A case-control study

Although there is a known reciprocal association between breast and colorectal cancer in women, few studies have investigated whether a similar association exists between breast cancer and colorectal adenomatous polyps, known to be precursor lesions for colon cancer. A case-control study was conducted on patients in three colonoscopy practices in New York to determine possible risk factors for adenomatous polyps. Among women studied, there were 128 patients with incident adenomatous polyps and 284 control subjects who underwent colonoscopy and had no colorectal neoplasia. No significant association between the incidence of an adenomatous polyp in the colon or rectum and a history of breast cancer was found (odds ratio, 0.71; 95% confidence interval, 0.34 to 1.64). If shared risk factors for breast and colorectal cancer are the reasons for the concurrence of these two malignant lesions, these results suggest that these factors act at the level of promoting adenomatous polyps of the colon and rectum into colorectal cancer.

Differences in histology between first and second primary lung cancer

Data from the Surveillance, Epidemiology, and End Results (SEER) Program were used to compare the histological distribution of second lung cancer following an initial cancer of the lung, head and neck, and breast to primary lung carcinoma occurring as a first cancer. Following initial head and neck cancer or initial squamous cell carcinoma of the lung, the proportion of second primary lung cancer which was of squamous cell histology rose dramatically, while the proportion of pulmonary adenocarcinomas rose following initial adenocarcinoma of the lung. The histological distribution of lung cancer following an initial breast cancer in women was similar to the distribution of de novo lung cancer in women. These results persisted as the time interval between diagnosis of the two primaries was increased from 12 to 48 months. We conclude that the histology of a second primary lung cancer following an initial cancer of the lung or head and neck tends to repeat the histology of the initial cancer (field effect), and this observation is not likely to be due to misdiagnosis of a recurrence of the initial cancer.

The association of breast cancer and colorectal cancer in men. An analysis of surveillance, epidemiology, and end results program data

There is a known increased risk for the co-occurrence of both breast cancer and colorectal cancer in the same women, presumably as a result of either shared reproductive hormonal or environmental risk factors. Using data from the Surveillance, Epidemiology, and End Results Program of the National Cancer Institute, the authors investigated whether there is a similar relationship between breast cancer, colorectal cancer, and prostate cancer in men. The observed-expected ratio was not significantly elevated for either colorectal cancer after breast cancer or for breast cancer after colorectal cancer in men. There was an increased incidence of prostate cancer after colorectal cancer, but this relationship was not present in the opposite direction and is believed to result from detection bias. There was an increased incidence of breast and colorectal cancer in women. These results suggest that the observed breast and colorectal cancer relationship in women may be a result of shared reproductive hormonal factors.

Cholecystectomy as a risk factor for colorectal adenomatous polyps and carcinoma

Multiple studies have suggested a possible relationship between prior cholecystectomy and the occurrence of subsequent colorectal carcinoma. This relationship has been particularly noticed among female patients and for right-sided lesions of the colon. In the current study, the authors undertook a case-control study among patients who underwent colonoscopy in three private practices in New York City between April 1986 and March 1988. Over this period, 302 adenomatous polyp cases, 106 colon cancer cases, and 507 controls were interviewed regarding their prior history of cholecystectomy. Overall, no significant association was observed between cholecystectomy and either colorectal adenomatous polyps or cancer. Cholecystectomy does not appear to be a significant risk factor for colorectal neoplasia.

Epidemiologic analysis of warfare. A historical review

Although warfare is thought to be responsible for high and increasing levels of morbidity and mortality in the modern era, little comparative epidemiologic research is available on the subject. Most research on past wars has been carried out for purposes of military planning. The present report provides an overview of the direct health impacts of various wars on military and civilian populations during the last 200 years. Risk factors for injury and death are analyzed. Changes in weaponry, military strategy, and medical services are found to modify the health effects of warfare.

A comparison of the clinical characteristics of first and second primary head and neck cancers. A population-based study

The clinical characteristics of 183 patients with second primary neoplasms in the head and neck region were compared with those of 20,598 patients with one primary tumor in the same region registered during the period 1973 to 1984 in the Surveillance, Epidemiology and End-Results Program. Second primary head and neck tumors were more likely to be diagnosed in a localized stage (47%) than if diagnosed as a single primary tumor (43%), but this difference was not significant. The tumor grade distribution was comparable in both groups. Using Cox proportional hazards modeling with age, sex, race, and stage as covariates, the median survival of patients with first and second head and neck cancer was identical (50 months). The survival of patients with localized second head and neck cancer was shorter than that of patients with single localized tumors (55 versus 102 months, P less than 0.026). Survival for regional tumors was similar (18 versus 21 months, not significant). The 84 second head and neck cancers in which the first head and neck cancer received radiation therapy (RT) had a median survival of 20 months; the 98 cases without prior RT had a median survival of 35 months. The high incidence of localized second cancers was probably due to the more intense surveillance. The worse survival in this group may be a result of prior RT or biologic characteristics of the tumor.

Obesity and colorectal adenomatous polyps

Obesity has been investigated as a risk factor for various malignancies, including colon cancer. A case-control study was conducted on patients in three colonoscopy practices in New York City to determine possible risk factors for colorectal adenomatous polyps, a known precursor lesion for most cases of colorectal cancer. Among 301 case subjects with incidence adenomatous polyps (174 men and 127 women) and 506 control subjects (223 men and 283 women), an increased risk was observed with increasing body mass index in women (odds ratio 2.1, 95% confidence interval 1.1-4.0; for highest versus lowest quartile, linear trend P = .02). A nonsignificant trend was observed for men. The increased risk seen in women is consistent with prior observations regarding reproductive hormonal and dietary risk factors for colorectal cancer.

Poor survival of treatment-related acute nonlymphocytic leukemia

Population-based data on more than 1 million patients registered in the Surveillance, Epidemiology, and End-Results Program of the National Cancer Institute, 1973 to 1984, were analyzed to determine the survival of patients with de novo acute nonlymphocytic leukemia (ANLL) and following a first primary tumor treated (with chemotherapy and/or radiation therapy) or untreated. Cases that occurred within 12 months of the first malignant neoplasm were excluded. Survival was estimated using Cox proportional-hazards modeling, with age, sex, and specific type of ANLL as covariates. The 6271 patients with de novo ANLL had an estimated 12-month survival of 30%, while the 107 patients with treatment-related ANLL (radiation therapy, 60; chemotherapy, 29; both, 18) had an estimated 12-month survival of 10%. This is not due to lingering effects of the first tumor since ANLL following solid tumors not treated with chemotherapy or radiation therapy (118 cases) has similar survival (estimated 12-month survival, 36%) as de novo ANLL. We conclude that ANLL that occurs after chemotherapy or radiation therapy is biologically more aggressive and/or resistant to therapy than spontaneous ANLL. This provides a rationale for current studies on treatment-induced cellular changes and on more aggressive therapy for these patients.

The clinical behavior of radiation-induced thyroid cancer in patients with prior Hodgkin's disease

Second primary neoplasms are occurring with increasing frequency. Despite growing literature on the incidence and etiology of this phenomenon, very little has been documented about the clinical aspects and biological behavior of these tumors. As our experience has hinted at a worse prognosis for second primaries, we thought it of interest to study this phenomenon for one type of tumor. We reviewed 32 cases of thyroid cancer arising as a second neoplasm in patients with Hodgkin's disease and compared various clinical characteristics to reported series of de novo (non-radiation-induced) thyroid cancer. Thyroid cancer, as a second primary, occurred more frequently in males with a more malignant histology and was diagnosed at a more advanced stage. The survival appears to be worse. The behavior of radiation-induced thyroid cancer in a host with prior malignant neoplasia appears to be more aggressive than that of both de novo (non-radiation-induced) thyroid cancer and radiation-induced thyroid cancer in a host with no prior malignancy. The selected nature of the cases precludes any firm conclusions. This type of information should be collected prospectively for all secondary malignancies as it may have an impact on the clinical management of these patients.

Serum cholesterol and primary brain tumours: a case-control study

The association between serum cholesterol level and primary brain tumours in adults was studied in a case-control study. Some 150 cases were compared with 649 hospital-based controls. An association between elevated serum cholesterol level and primary brain tumours was found. This relationship was concentrated in meningioma cases and may represent a marker for an elevated socioeconomic status.

Leiomyosarcomas of the extremities

Leiomyosarcomas are uncommon malignant neoplasms that rarely arise in the extremities. We report on the clinical characteristics of 17 cases seen at the Memorial Sloan-Kettering Cancer Center over a 32 year period. Most neoplasms occurred in the lower extremities, and most of the tumors recurred after surgery alone, with several late recurrences. The overall prognosis was grim, with a median survival of 24 months. More aggressive multidisciplinary approaches to this disease are warranted.

Role of sigmoidoscopy in screening for colorectal cancer: a critical review

The use of sigmoidoscopy as a screening method for colorectal cancer is controversial. Evidence regarding its efficacy is reviewed critically, with special attention given to potential biases in screening studies. The vast majority of studies are uncontrolled and without follow-up information and thus shed little light on the actual benefits of sigmoidoscopy. Two uncontrolled studies with follow-up and one randomized trial suggest a colorectal cancer mortality reduction because of the use of sigmoidoscopy, but all three studies have major shortcomings. The authors conclude that the currently available data are insufficient to establish a national recommendation for screening with sigmoidoscopy. To establish such a recommendation, a properly conducted randomized trial with colorectal cancer mortality as an outcome is needed.

Vitamin supplements among women with adenomatous polyps and cancer of the colon. Preliminary findings

Chemoprevention of various epithelial cancers with vitamins or minerals has been the subject of multiple intervention trials to assess the impact of supplementation. These include several trials in patients with adenomatous polyps of the colon, a precursor lesion for colon cancer. The authors interviewed 255 women who underwent colonoscopy at Columbia Presbyterian Medical Center between 1983 and 1985 with a telephone-administered structured questionnaire. Eleven interviews were excluded for various reasons. Overall, 57.7 percent of the 244 interviewees used vitamin pills on a regular basis (at least once a week for a year); 6.6 percent of the interviewees used vitamin A, 20.7 percent used vitamin C, and 16.2 percent used vitamin E. There were no statistically significant differences in vitamin usage among women with adenomatous polyps of the colon (105 cases), women with colon cancer (56 cases), and women without colonic neoplasia (83 cases). Despite widespread use of supplementary vitamins, this study failed to demonstrate major benefits in preventing colon polyps or cancer.

Squamous cell cancers and cigarette smoke: a matter of exposure

Most of the malignancies associated with tobacco smoke are squamous cell cancers. Others have suggested that squamous cell epithelium has a particular predisposition to tobacco carcinogenesis. However, other malignancies besides squamous cell cancers also have been associated with cigarettes. I hypothesize that the key factor which predisposes an organ or tissue to carcinogenesis by cigarette smoke is exposure to smoke constituents, either directly or indirectly. Squamous cell epithelium has no special biological susceptibility to these carcinogens.

Clinical aspects of postirradiation sarcomas

The number of cancer patients who live longer and are cured of their disease is increasing. Many of them have received radiotherapy as part of their treatment. Postirradiation sarcoma (PIS), while still uncommon, is a complication which is appearing more frequently as the number of long-term survivors increases. Studies of the clinical characteristics of PIS, such as stage of disease, grade, survival, and prognosis, are therefore of increasing importance, and may lead to different strategies for early detection and prevention. In a literature review of PIS, we identified 344 cases with sufficient data for analysis of these clinical characteristics. In these selected cases, we found that: (a) PIS was most often diagnosed at an advanced stage and high grade; (b) most of the tumors were located in areas where radical surgery could not be performed; (c) the response rate to chemotherapy was almost always poor; and (d) most patients with PIS died from locally advanced and/or metastatic disease within a few months after diagnosis. PIS is a serious, usually fatal, late complication of radiotherapy and it should be weighed in the decision to use this modality of therapy. In the follow-up of cancer patients, a low threshold of suspicion for PIS is recommended in order to detect it early for possible resection. Because of the poor prognosis, more aggressive and investigative chemotherapeutic regimens are warranted.

Cholecystectomy and adenomatous polyps of the colon in women

The relationship between prior cholecystectomy and colon cancer in women has been a subject of recent research interest. No study has yet reported on cholecystectomy and adenomatous polyps, a precursor lesion for most colon cancers. This pilot case-control study interviewed 245 women who had undergone colonoscopy between 1983 and 1985 at Columbia-Presbyterian Medical Center with a structured telephone-administered questionnaire. Fifty-six were colon cancer cases, 105 were adenomatous polyp (AP) cases, and 84 were controls (without colonic neoplasia). Adjusted for age and educational attainment, the odds ratio for prior cholecystectomy among colon cancer cases compared to controls was 1.79 (95% confidence limits, 0.59 to 5.44) and 2.26 for right-sided colon cancer cases (95% confidence limits, 0.61 to 8.42). Despite a lack of statistical power, these estimates are consistent with earlier reports. The odds ratio for adenomatous polyp cases was 1.02 (95% confidence limits, 0.40 to 2.64), suggesting no association between prior cholecystectomy and adenomatous polyps. These preliminary findings are currently being explored in a large-scale case-control study.