Prognostic factors in gastric cancer using log-normal censored regression model.

Background & objectives: Gastric cancer is one of the most common cancers in the world. It is rarely detected early, and the prognosis remains poor. Cox proportional hazard model is used to examine the relationship between survival and covariates. Parametric survival models such as log normal regression model can also be used for this analysis. We used log normal regression model in this study to evaluate prognostic factors in gastric cancer and compared with Cox model. Methods: We retrospectively studied the 746 patients diagnosed with gastric cancer admitted in a referral hospital in Tehran, Iran, from February 2003 through January 2007. Age at diagnosis, sex, extent of wall penetration, histology type, tumour grade, tumour size, pathologic stage, lymph node metastasis and presence of metastasis were entered into a log normal model. Hazard rate (HR) was employed to interpret the risk of death and the results were compared with Cox regression. The AIC (Akaike Information Criterion) was employed to compare the efficiency of models. Results: Univariate analysis indicated that with increasing age the risk of death increased significantly in both log normal and Cox models. Patients with grater tumour size were also in higher risk of death followed by those with poorly differentiated and moderately differentiated in tumour grade and advanced pathologic stage. The presence of metastasis was significant prognostic factor only in log normal analysis. In final multivariate model, age was still a significant prognostic factor in Cox regression but it was not significant in log normal model. Presence of metastasis followed by histology type were other prognostic features found significant in log normal results. Based on AIC, log normal model performed better than Cox. Interpretation & conclusion: Our results suggest that early detection of patients in younger age and in primary stages and grade of tumour could be important to decrease the risk of death in patients with gastric cancer. Comparison between Cox and log normal models indicated that log normal regression model can be a useful statistical model to find prognostic factors instead of Cox.

Effect of demographic and clinicopathologic factors on prognosis of early gastric cancer in Iran.

BACKGROUND: The aim of this study was to estimate some prognostic factors that affect on overall survival of patients with early gastric cancer. METHODS: A retrospective study had been done on patients diagnosed with early gastric cancer who registered in cancer registry center, Tehran, Iran, between December 21, 2001 and December 21, 2006 and all patients were followed by telephone contacts. The Kaplan-Meier method was performed to describe survival curves and log-rank test to compare the survival rate in subgroups. Cox regression was used to determine the prognosis factors. RESULTS: The mean age was 57.9 years and 72.6% of patients were male. Tumor size (>35 mm) and lymph node metastasis were established as significant factors for survival of patients with EGC in both univariate and multivariate analysis. CONCLUSION: The findings of this study indicate that lymph node metastasis and tumor size are the most independent prognostic factors in these patients.

Relation between demographic factors and type of gastrointestinal cancer using probit and logit regression.

BACKGROUND: Approximately 50,000 new cases of cancer occur each year in the Iranian population of 70.4 million. The organ system involved with more than 38% of all cancers is the gastrointestinal (GI) tract. The objective of this study was to investigate the relation between demographic factors and type of gastrointestinal cancer using probit and logit models. METHODS: This study was designed as a cross-sectional survey including all consecutive GI cancer patients admitted over a one year period in a randomly selected hospital group located in Tehran in 2006. RESULTS: The largest number of cases were colorectal cancers (40.0%), followed by gastric cancers (34.5%) and esophagus cancers (17.1%). There was a significant gender effect in the colorectal, gastric and esophagus cancer also there was a significant association between age and gastrointestinal cancers in both logit and probit regression. The factor of duration was not significant in gastric cancer. CONCLUSION: Men are more likely have colorectal cancer than women. Older people are more likely to have gastric cancer than younger people. For esophagus cancer all factors were significant. Results from probit and logit models were similar, indicating that probit analysis can be employed as a logit model to analyze relationships between demographic factors and cancer type.

Statistical comparison of survival models for analysis of cancer data.

BACKGROUND: The Cox Proportional Hazard model is the most popular technique to analysis the effects of covariates on survival time but under certain circumstances parametric models may offer advantages over Cox's model. In this study we use Cox regression and alternative parametric models such as: Weibull, Exponential and Lognormal models to evaluate prognostic factors affecting survival of patients with stomach cancer. Comparisons were made to find the best model. METHODS: To determine independent prognostic factors reducing survival time for stomach cancer, we compared parametric and semi-parametric methods applied to patients who registered in one cancer registry center located in southern Iran using the Akaike Information Criterion. RESULTS: Of a total of 442 patients, 266 (60.2%) died. The results of data analysis using Cox and parametric models were approximately similar. Patients with ages 60-75 and >75 years at diagnosis had an increased risk for death followed by those with poor differentiated grade and presence of distant metastasis (P<0.05). CONCLUSION: Although the Hazard Ratios in the Cox model and parametric ones are approximately similar, according to Akaike Information Criterion, the Weibull and Exponential models are the most favorable for survival analysis.

Quality of life in breast cancer patients--a quantile regression analysis.

BACKGROUND: Quality of life study has an important role in health care especially in chronic diseases, in clinical judgment and in medical resources supplying. Statistical tools like linear regression are widely used to assess the predictors of quality of life. But when the response is not normal the results are misleading. The aim of this study is to determine the predictors of quality of life in breast cancer patients, using quantile regression model and compare to linear regression. METHODS: A cross-sectional study conducted on 119 breast cancer patients that admitted and treated in chemotherapy ward of Namazi hospital in Shiraz. We used QLQ-C30 questionnaire to assessment quality of life in these patients. A quantile regression was employed to assess the assocciated factors and the results were compared to linear regression. All analysis carried out using SAS. RESULTS: The mean score for the global health status for breast cancer patients was 64.92+/-11.42. Linear regression showed that only grade of tumor, occupational status, menopausal status, financial difficulties and dyspnea were statistically significant. In spite of linear regression, financial difficulties were not significant in quantile regression analysis and dyspnea was only significant for first quartile. Also emotion functioning and duration of disease statistically predicted the QOL score in the third quartile. CONCLUSION: The results have demonstrated that using quantile regression leads to better interpretation and richer inference about predictors of the breast cancer patient quality of life.

Association between risk of breast cancer and fertility factors--a latent variable approach.

BACKGROUND: Breast cancer is the most common malignant tumor in females. Many studies have been carried out in order to assess the reproductive risk factors. Particular attention has focused on information regarding fertility, including breastfeeding, age at first birth and number of live births. These factors are highly correlated with each other. The objective of this study was to employ latent variables to reduce the confounding effect of this correlation with a logistic regression analysis. METHODS: The investigation drew upon results from a dataset belonged to a hospital based case-control study covering 303 breast cancer patients and 303 hospital controls. Data were collected through interview and reproductive variables included age at first full-term pregnancy and live birth, number of pregnancies and live births, and total length of breast feeding. Latent variables were generated using factor analysis and principal components analysis. RESULTS: The study revealed that for both latent variable approaches the odds ratios of two latent variables significantly indicated a protective impact of number of pregnancy and live birth and breastfeeding and a prognostic relation with age at first pregnancy or live birth. CONCLUSION: The findings suggest that breastfeeding and decreasing age at first live birth have protective influences on breast cancer risk. Also using statistical model with latent variables in the presence of collinear data leads to reliable results.

Survival rates and prognosis of gastric cancer using an actuarial life-table method.

BACKGROUND AND AIMS: The aim of this study was to calculate survival rates and analyze patterns of survival in gastric cancer. METHODS: A total number of 746 patients with gastric cancer registered in the Cancer Registry Center of Research Center of Gastroenterology and Liver Disease of Shahid Beheshti University of Medical Sciences, Iran, from Dec 21, 2001 to Dec 21, 2006 were investigated. 1- to 5-year survival rates were estimated using life-table method and compared by Wilcoxon (Gehan) test. P<0.05 was considered as statistically significant. All calculations were carried out with SPSS (version 13.0) statistical software. RESULTS: There were 530 male patients with a mean age of 60.5+/-12.6 years and 216 females with a mean age of 57.5+/-13.5 years. Of the total, 454 died and 285 were censored during the investigation. The median survival time was 24.2 months and survival rates at one, two, third, fourth and five years after diagnosis were 73.6, 50.2, 40.6, 33.2 and 29.7%, respectively. Stages of tumor, histology grade, histologic type of cancer, tumor size, age at diagnosis and surgery approach were independent prognostic factors . However, variables such as sex (P=.533), body mass index (P=.214), ethnicity (P=.092), and level of education (P=.762) did not shown significant effects on survival. CONCLUSION: Early detection of patients at lower age and with primary stages and grades of tumor is important to increase patient's life expectancy.

Comparing Cox regression and parametric models for survival of patients with gastric carcinoma.

BACKGROUND: Researchers in medical sciences often tend to prefer Cox semi-parametric instead of parametric models for survival analysis because of fewer assumptions but under certain circumstances, parametric models give more precise estimates. The objective of this study was to compare two survival regression methods - Cox regression and parametric models - in patients with gastric adenocarcinomas who registered at Taleghani hospital, Tehran. METHODS: We retrospectively studied 746 cases from February 2003 through January 2007. Gender, age at diagnosis, family history of cancer, tumor size and pathologic distant of metastasis were selected as potential prognostic factors and entered into the parametric and semi parametric models. Weibull, exponential and lognormal regression were performed as parametric models with the Akaike Information Criterion (AIC) and standardized of parameter estimates to compare the efficiency of models. RESULTS: The survival results from both Cox and Parametric models showed that patients who were older than 45 years at diagnosis had an increased risk for death, followed by greater tumor size and presence of pathologic distant metastasis. CONCLUSION: In multivariate analysis Cox and Exponential are similar. Although it seems that there may not be a single model that is substantially better than others, in univariate analysis the data strongly supported the log normal regression among parametric models and it can be lead to more precise results as an alternative to Cox.