Epidemiological differences between adenocarcinoma of the oesophagus and adenocarcinoma of the gastric cardia in the USA.

BACKGROUND AND AIMS: It has been suggested that gastric cardia adenocarcinoma (GCA) is a distinct entity from oesophageal adenocarcinoma (OA). We examined several epidemiological features of GCA and OA in the USA to elucidate differences/similarities between these malignancies. METHODS: Using the database of Surveillance, Epidemiology, and End Results (SEER) program, we examined incidence rates for temporal changes, and ethnic and age distributions, and performed birth cohort analyses for cases with morphologically and histologically confirmed OA or GCA. RESULTS: The age adjusted incidence rates of OA rose progressively, reaching 1.8 per 100 000 (95% confidence interval 1.7-1.9) during 1987-1991 and 2.5 per 100 000 (2.3-2.6) during 1992-1996. In 1992-1996, Whites were affected five times more than Blacks, and men eight times more than women. A significant increase in incidence occurred among younger persons aged 45-65 years. Irrespective of age, OA was characterised by higher incidence rates among more recent birth cohorts: a 40% increase in incidence for each five year increase in the date of birth--a "birth cohort effect". On the other hand, the incidence rates of GCA reached their highest level of 3.3 per 100 000 (3.2-3.4) in 1987-1991 and subsequently declined during 1992-1996 to 3.1 per 100 000 (3.0-3.3). Whites were affected twice more than blacks and men five times more than women. Most patients with GCA were older than 60 years with no increase among younger persons and no birth cohort effect (p=0.99). CONCLUSION: Several significant epidemiological differences exist between OA and GCA. These differences suggest that these two malignancies are separate entities with different risk factors.

Population-based study of microinvasive adenocarcinoma of the uterine cervix.

OBJECTIVE: To analyze lymph node status and survival rates of women with microinvasive cervical adenocarcinoma (International Federation of Gynecology and Obstetrics stages IA(1) and IA(2)). METHODS: The Surveillance, Epidemiology, and End Results (SEER) Public-Use Database was used to identify cases of microinvasive cervical adenocarcinoma diagnosed between 1988 and 1997. Variables analyzed included stage, extent of surgery, lymph node status, radiation therapy, and age. Statistics included analysis of trends, analysis of variance, log-rank test, one-sided binomial confidence interval estimation, and power analysis. RESULTS: Among 301 reported cases, 131 had stage IA(1) and 170 IA(2) disease. Simple hysterectomies were done in 54 women with IA(1) and 64 with IA(2) disease and radical hysterectomies were done in 50 and 83 women, respectively. Only one of 140 women who had lymphadenectomy had a single positive lymph node. There were four tumor-related deaths (one with IA(1), and three with IA(2) disease). There were no deaths among 96 women (47, IA(1); 49, IA(2)) treated by simple hysterectomy alone. The mean follow-up was 46.5 months (range 1--119). The censored survival rate was 98.7% overall (99.2% IA(1), 98.2% IA(2)). Power analysis estimated that 720 patients would be required in each group to detect a 2% difference in survival. Using one-sided 95% confidence interval estimations, the risk-adverse events rate for IA(1) was no more than 3.57%, and 4.50% for IA(2) disease. CONCLUSION: Prognosis is excellent for microinvasive adenocarcinoma of the uterine cervix. In 96 cases (31.9%), simple hysterectomy alone proved adequate.

Trends in Gleason score for prostate cancer diagnosed between 1983 and 1993.

PURPOSE: During the 1980s and 1990s the number, incidence rate and proportion of moderately differentiated prostate cancer cases ascertained by population based cancer registries increased substantially. The increase is thought to have resulted from the widespread use of prostate specific antigen (PSA) for screening because it occurred coincidentally with the introduction of PSA for early detection of prostate cancer. We investigate this increase in a population based study. MATERIALS AND METHODS: To report the trends in tumor grade we conducted a blinded, standardized pathological study and reviewed medial records of a stratified random sample of cases diagnosed before and after the introduction of PSA (1983 to 1984 and 1992 to 1993). Archival tumor biopsy specimens or transurethral resection of the prostate specimens were reviewed for the diagnosis of cancer and assignment of Gleason score. Medical records were reviewed to determine the method of prostate cancer detection for each case. RESULTS: We found a small but statistically insignificant shift in the distribution of Gleason scores assigned after review of biopsy or transurethral resection specimens. The proportion of Gleason score 2, 3 and 4 tumors decreased, and the proportion of 7, 8, 9 and 10 tumors as a group did not change. The shifts in Gleason score resulted in a slight statistically nonsignificant increase in mean Gleason score. There was a significant shift in the method of detection from predominately incidental detection in the earlier period to predominately screen detection in the later period. Because the proportion of screen detected tumors increased and they had a significantly higher mean Gleason score than incidentally detected tumors within each interval, the overall mean Gleason score increased. CONCLUSIONS: After a standardized pathological review a small shift in the distribution of Gleason scores occurred resulting in a small increase in mean Gleason score between 1983 and 1984, and 1992 and 1993. There was little change in the proportion of Gleason score 7, 8, 9 and 10 tumors between the 2 periods.

Patterns of mammography use among Hispanic, American Indian, and non-Hispanic White women in New Mexico, 1994-1997.

For screening efforts to maximally reduce mortality in the general population, a large proportion of women need to utilize mammography routinely. To investigate utilization of mammography in a community setting, the authors used population-based data collected by the New Mexico Mammography Project for residents of the Albuquerque, New Mexico, metropolitan area for the period 1994-1997. The authors computed screening rates and the proportion of women who routinely use mammography. The utilization of mammography was low. Only 50% of the women aged 50-74 years were screened each year. Less than one third of women aged 40-49 years or 75 years and older were screened annually. The percentage of women who routinely used mammography on an annual or biennial basis was low in all age groups, especially among Hispanics and American Indians. Women aged 50-74 years had the highest percentage of routine annual mammography use, ranging from 30% in non-Hispanic Whites to 20% in Hispanics. Current utilization of mammography in community-based screening efforts is unlikely to achieve a potential 30% reduction in breast cancer mortality. Interventions are needed to increase the routine use of mammography.

The rising incidence of adenocarcinoma relative to squamous cell carcinoma of the uterine cervix in the United States--a 24-year population-based study.

OBJECTIVE: The aim of this study was to compare the age-adjusted incidence and survival for invasive adenocarcinoma and squamous cell carcinoma of the uterine cervix using population-based data. METHODS: The SEER database was used to identify all cases of cervical cancer registered between 1973 and 1996. Stage was defined as localized, regional, or distant. Age-adjusted incidence rates were analyzed statistically using the Jonchkeere-Terpstra exact test for trends. Relative and observed survival rates, respectively, were compared using z tests and log-rank tests. RESULTS: The age-adjusted incidence rates per 100,000 for all invasive cervical cancers decreased by 36.9% over 24 years [12.35 (1973-1977) vs 7.79 (1993-1996)]. Similarly, the age-adjusted incidence rates for squamous cell carcinoma declined by 41.9% [9.45 (1973-1977) vs 5.49 (1993-1996)]. In contrast, the age-adjusted incidence rates for adenocarcinoma increased by 29.1% [1.34 (1973-1977) vs 1.73 (1993-1996)]. The proportion of adenocarcinoma increased 107.4% relative to all cervical cancer, 95.2% relative to squamous cell carcinoma, and 49.3% relative to the population of women at risk [10. 8% vs 22.4% (P < 0.001), 12.4% vs 24.0% (P < 0.001), and 1.40 vs 2. 09 per 100,000 women (P < 0.001), respectively]. Observed survival rates for adenocarcinoma vs squamous cell carcinoma were poorer for regional (P = 0.04), but not localized or distant disease. CONCLUSIONS: Over the past 24 years, the incidence of all cervical cancer and squamous cell carcinoma has continued to decline. However, the proportion of adenocarcinoma relative to squamous cell carcinoma and to all cervical cancers has doubled, and the rate of adenocarcinoma per population at risk has also increased. These results suggest that current screening practices in the United States are insufficient to detect a significant proportion of adenocarcinoma precursor lesions.

Biologic characteristics of interval and screen-detected breast cancers.

BACKGROUND: Interval breast cancer is defined as a cancer that is detected within 12 months after a negative mammogram. The failure of mammography to detect breast cancer depends on testing procedures, radiologist interpretation, patient characteristics, and tumor properties. Although errors by radiologists explain some interval cancers, another explanation is that the tumor is rapidly growing and was too small to be detected on the last mammogram. To determine whether markers of tumor growth rate are associated with risk of an interval cancer, we conducted a population-based study with the use of data collected statewide by the New Mexico Mammography Project. METHODS: Among women who received a mammographic examination from 1991 throughout 1993, we ascertained records of all patients with breast cancer diagnosed within 12 months of a negative screening mammographic examination (interval cancers) and corresponding tumor samples, when available. We selected an age- and ethnicity-matched comparison group of control patients with screen-detected breast cancers diagnosed during the same period. In tumor samples, p53, bcl-2, and Ki-67 were examined immunologically and the apoptotic index was assessed histologically. We used logistic regression to determine whether interval cancers were associated with selected demographic, radiologic, and biologic characteristics. RESULTS: It is more likely that mammography did not detect tumors with a high proportion of proliferating cells (>20%) than tumors with a low proportion of proliferating cells (<5%) (odds ratio [OR] = 4.09; 95% confidence interval [CI] = 1.14-14.65). The OR for mammographic failure was 2.96 (95% CI = 1.07-8.20) among cancers that expressed p53 compared with cancers that did not. Interval cancers also had fewer apoptotic cells. Approximately 75% of interval cancers appear to have tumors with 5% proliferating cells or more. Younger women had a higher proportion of rapidly proliferating and aggressive cancers. CONCLUSION: Rapidly growing and aggressive tumors account for a substantial proportion of mammographic failure to detect breast cancer, especially among younger women, who have a high proportion of aggressive cancers.

Uranium mining and lung cancer among Navajo men in New Mexico and Arizona, 1969 to 1993.

Navajo men who were underground miners have excess risk of lung cancer. To further characterize the long-term consequences of uranium mining in this high-risk population, we examined lung cancer incidence among Navajo men residing in New Mexico and Arizona from 1969 to 1993 and conducted a population-based case-control study to estimate the risk of lung cancer for Navajo uranium miners. Uranium mining contributed substantially to lung cancer among Navajo men over the 25-year period following the end of mining for the Navajo Nation. Sixty-three (67%) of the 94-incident lung cancers among Navajo men occurred in former uranium miners. The relative risk for a history of mining was 28.6 (95% confidence interval, 13.2-61.7). Smoking did not account for the strong relationship between lung cancer and uranium mining. The Navajo experience with uranium mining is a unique example of exposure in a single occupation accounting for the majority of lung cancers in an entire population.

Trends in incidence and treatment for ductal carcinoma in situ in Hispanic, American Indian, and non-Hispanic white women in New Mexico, 1973-1994.

BACKGROUND: Incidence rates of ductal carcinoma in situ (DCIS) breast carcinoma and the use of breast-conserving surgery (BCS) for its treatment show substantial geographic and ethnic variations nationwide. To the authors' knowledge, few studies have investigated incidence rates and treatment patterns in Hispanics and American Indians. METHODS: The authors used data from the population-based New Mexico Tumor Registry to describe trends in DCIS incidence rates between 1973-1994 and investigate patient and physician characteristics related to BCS in a multiethnic population between 1985-1994. Multiple logistic regression was used to evaluate patient and physician factors related to receiving BCS. RESULTS: Incidence rates for DCIS in Hispanics were approximately 50% lower compared with non-Hispanic whites. American Indians had the lowest incidence rate. Beginning in 1985, incidence rates for Hispanics and non-Hispanic whites showed a 21% annual increase. Between 1990-1994, incidence rates in American Indians increased more than twofold. BCS increased 5.8% per year between 1985-1994, with 50% of Hispanic and non-Hispanic white patients treated with BCS in 1994. The strongest factor associated with receiving BCS was geographic location of treatment (P < 0.001). The odds of receiving BCS were 5.8 times higher in the northern third of the state compared with the southern third. No significant variation in BCS was found by ethnicity, rural/urban residency, socioeconomic status, or physician characteristics. CONCLUSIONS: Incidence rates for DCIS increased substantially in all three ethnic groups. The use of BCS was associated most strongly with the location of treatment, most likely reflecting differences in physician practices and treatment recommendations. Further research is needed to investigate the increasing incidence rates of DCIS and the determinants of BCS for the treatment of DCIS.

Effects of age, breast density, ethnicity, and estrogen replacement therapy on screening mammographic sensitivity and cancer stage at diagnosis: review of 183,134 screening mammograms in Albuquerque, New Mexico.

PURPOSE: To examine how common patient factors affect screening mammographic sensitivity and cancer stage at diagnosis. MATERIALS AND METHODS: The authors used a population-based database of 183,134 screening mammograms and a statewide tumor registry to identify 807 breast cancers detected at screening mammography. RESULTS: Sensitivity varied significantly with ethnicity, use of estrogen replacement therapy, mammographic breast density, and age. Sensitivity was 54% (13 of 24) in women younger than 40 years, 77% (121 of 157) in women aged 40-49 years, 78% (224 of 286) in women aged 50-64 years, and 81% (277 of 340) in women older than 64 years. Sensitivity was 68% (162 of 237) for dense breasts and 85% (302 of 356) for nondense breasts and 74% (180 of 244) in estrogen replacement therapy users and 81% (417 of 513) in nonusers. Sensitivity was most markedly reduced with the combination of dense breasts and estrogen replacement therapy use; there was little difference when only one factor was present. Median cancer size and the percentage of early cancers showed little change with any factors. CONCLUSION: Age is a minor determinant of mammographic sensitivity in women aged 40 years or older. Sensitivity is substantially decreased with the combination of higher breast density and estrogen replacement therapy use. There was not a notable shift in cancer outcomes in the groups with lower mammographic sensitivity. These data do not support different screening recommendations in women aged 40-49 years or in estrogen replacement therapy users.

Evaluating cluster alarms: a space-time scan statistic and brain cancer in Los Alamos, New Mexico.

OBJECTIVES: This article presents a space-time scan statistic, useful for evaluating space-time cluster alarms, and illustrates the method on a recent brain cancer cluster alarms in Los Alamos, NM. METHODS: The space-time scan statistic accounts for the preselection bias and multiple testing inherent in a cluster alarm. Confounders and time trends can be adjusted for. RESULTS: The observed excess of brain cancer in Los Alamos was not statistically significant. CONCLUSIONS: The space-time scan statistic is useful as a screening tool for evaluating which cluster alarms merit further investigation and which clusters are probably chance occurrences.

Increasing incidence of colon and rectal cancer among Hispanics and American Indians in New Mexico (United States), 1969-94.

OBJECTIVES: This study evaluates time trends in colon and rectal cancer incidence and mortality among the three major race/ethnic groups (Hispanics, American Indians, and non-Hispanic Whites) in New Mexico (United States). METHODS: We used data from the New Mexico Tumor Registry (NMTR) and computed average annual age-standardized incidence and mortality rates. Colon cancer incidence rates were further examined by anatomical subsite. Estimated annual percent change (EAPC) in incidence and mortality over time were computed using Poisson regression. RESULTS: Invasive colorectal cancer incidence rates increased from 1969-89 in all three race/ethnic groups, but decreased among non-Hispanic Whites in 1990-94, while rates continued to increase among minority populations, especially among minority men. Over the 26-year period, EAPC in colon cancer incidence among men was 3.6 percent for Hispanics, 4.7 percent for American Indians, and 0.7 percent for non-Hispanic Whites. Right-sided colon cancers were more common among American Indian women, and among all women aged 65 years and older. Mortality rates decreased steadily among non-Hispanic Whites over the study period, especially among women. CONCLUSIONS: Studies are needed to identify important modifiable risk factors and to develop strategies to increase the use of colorectal cancer screening-procedures among the minority populations.

Trends in the survival of American Indian, Hispanic, and Non-Hispanic white cancer patients in New Mexico and Arizona, 1969-1994.

BACKGROUND: The burden of cancer mortality falls disproportionately on cancer patients belonging to ethnic minority groups. In the U.S., African American, Hispanic, and American Indian cancer patients are diagnosed at a more advanced stage and receive less appropriate treatment, resulting in poorer outcomes and higher mortality, than white cancer patients. The authors hypothesized that cancer control strategies based on earlier detection and more effective treatment may be most effective in increasing survival in groups with more advanced disease at diagnosis, less appropriate treatment, and lower survival. METHODS: Data collected by the New Mexico Tumor Registry, a member organization of the Surveillance, Epidemiology, and End Results (SEER) Program of the National Cancer Institute, afforded the authors an opportunity to investigate this hypothesis by studying trends in cancer survival for American Indians, Hispanics, and non-Hispanic whites. The authors examined temporal trends and ethnic disparities in survival for in situ and invasive incident cancer cases at 25 sites diagnosed from 1969 through 1994 in New Mexico residents and in American Indians residing in Arizona. RESULTS: The distribution of stage became more favorable and the percentage of patients receiving appropriate treatment increased for all three ethnic groups during the study period. Survival improved for patients with cancer at most sites in each ethnic group; however, because the increase in survival was greater for non-Hispanic whites than for American Indians or Hispanics, the number of sites associated with disparities in survival among non-Hispanic whites, American Indians, and Hispanics increased. Differences in the distribution of age, gender, stage at diagnosis, histologic grade, and treatment did not completely explain the improved survival or the ethnic disparities regarding survival. CONCLUSIONS: Increased cancer control efforts were associated with earlier diagnosis, more patients receiving appropriate therapy, and improved survival for non-Hispanic whites, American Indians, and Hispanics. However, the improvement was greatest for non-Hispanic whites, and disparities in survival results for the different ethnic groups widened over the period of study. Cancer control strategies need to address the specific social, cultural, and biologic prognostic factors that affect different ethnic groups if disparities in outcomes are to be reduced.

Prostate cancer in American Indians, New Mexico, 1969 to 1994.

PURPOSE: Prostate cancer is the most frequently diagnosed cancer as well as the leading cause of cancer death among American Indian men. MATERIALS AND METHODS: To describe further the occurrence of prostate cancer among American Indian men, we examined population based incidence, treatment, survival and mortality data for American Indians in New Mexico during the 25-year period 1969 to 1994. RESULTS: Although American Indian men have a lower risk of prostate cancer than nonHispanic white men, the incidence and mortality rates are rising for American Indians, and mortality rates are now equal to those for nonHispanic white men. During the 25-year period age adjusted incidence rates for American Indians increased from 42.2/100,000 (95% confidence interval 27.1 to 57.3) to 64.6/100,000 (95% confidence interval 46.2 to 83.0). The burden of prostate cancer among American Indian men compared with nonHispanic white men was reflected in disproportionately high mortality rates in relation to incidence rates. The mortality rates were high because American Indian cases were more advanced at diagnosis, 23.3% of prostate cancers were diagnosed after distant spread had occurred compared with 11.6% for nonHispanic white men and the 5-year relative survival rate was poorer (57.1% compared with 77.6% for nonHispanic white men). CONCLUSIONS: Effective and culturally sensitive cancer control efforts for prostate cancer in American Indian communities are urgently needed.

Ethnic differences in uterine corpus cancer incidence and mortality in New Mexico's American Indians, hispanics and non-Hispanic whites.

BACKGROUND: Although ethnic and radical differences in uterine corpus cancer incidence and mortality have been reported worldwide, few published data have addressed the epidemiology of uterine cancer among US American Indians and Hispanics. METHODS: We reviewed uterine corpus cancer incidence and survival data from New Mexico's population-based cancer registry collected from 1969 to 1992, and examined State vital records data for uterine cancer deaths collected from 1958 to 1992, focusing on ethnic differences in occurrence and outcomes of uterine malignancies. RESULTS: Non-Hispanic white women had age-adjusted incidence rates that were substantially higher (20.8 per 100,000) than rates for Hispanics (10.3) and American Indians (6.0) over the 24-year period. Uterine cancer mortality rates were also higher for non-Hispanic whites and Hispanics than for American Indian women, although mortality rates were substantially lower than incidence rates. Five-year survival for uterine cancer was comparable among all groups for all stages combined (87.3% for non-Hispanic whites, 81.4% for Hispanics, and 84.6% for American Indians). CONCLUSIONS: Our population-based data show ethnic differences in uterine corpus cancer incidence rates for non-Hispanic white women that were double those for Hispanics, and triple those for American Indian women. Ethnic differences in survival were comparable. Aetiologic studies are warranted to investigate the dramatic ethnic differences in occurrence of uterine cancer.

Prognostic factors for thyroid carcinoma. A population-based study of 15,698 cases from the Surveillance, Epidemiology and End Results (SEER) program 1973-1991.

BACKGROUND: A number of prognostic factors for thyroid carcinoma have been identified, including sociodemographic characteristics, such as age and gender, and tumor characteristics, such as histology and stage. The relative importance of these factors as independent predictors of survival for patients with papillary, follicular, anaplastic, and medullary thyroid carcinoma has been extensively studied but remains uncertain. METHODS: The authors used data collected by the Surveillance, Epidemiology and End Results (SEER) program of the National Cancer Institute between 1973 and 1991 to investigate prognostic factors for each of the major histologic types of thyroid carcinoma in a population-based patient series and to assess the effect of these factors as predictors of survival. RESULTS: Both tumor and sociodemographic characteristics were independently associated with survival. Patients with papillary carcinoma had the highest 10-year relative survival (0.98), followed by those with follicular carcinoma (0.92) and medullary carcinoma (0.80). Anaplastic tumors had the lowest 10-year relative survival (0.13). Stage at diagnosis and differentiation status were strong independent prognostic factors for each histologic type. Advanced stage at diagnosis was a stronger prognostic factor for medullary carcinoma than for other histologic types. Increasing age was associated with lower relative survival for each histologic type. Gender, marital status, and ethnicity were significant, but weaker, predictors of survival. CONCLUSIONS: Survival varied markedly among patients with different histologic types of thyroid carcinoma. Stage at diagnosis and tumor differentiation were important prognostic factors for each histologic type. Age at diagnosis was a stronger predictor of survival for patients with follicular and medullary carcinoma than for patients with papillary carcinoma.

Breast cancer survival among New Mexico Hispanic, American Indian, and non-Hispanic white women (1973-1992).

A study of breast cancer survival was conducted among New Mexico Hispanic and non-Hispanic white women and New Mexico and Arizona American Indian women diagnosed between 1973 and 1992. The goals were to determine whether, after adjusting for first treatment and the extent of disease at diagnosis, American Indian and Hispanic women had poorer survival than non-Hispanic whites and, if survival had improved over time, whether comparable improvements had been made for the three racial/ethnic groups. Five-year relative survival rates were calculated, and a Cox proportional hazards model was constructed to compare survival between races/ethnicities, adjusting for first treatment and the extent of disease at diagnosis. Findings indicate that during 1983-1992, breast cancer was more commonly detected at a local stage for all three groups compared to 1973-1982. Five-year relative survival improved for non-Hispanic white and American Indian women with local or regional disease, but the improvement was statistically significant only for non-Hispanic white women and for American Indian women with local disease. Despite earlier stages at diagnosis, Hispanic females showed less improvement in overall or stage-specific survival than non-Hispanic whites. The Cox model indicated that American Indian women experienced poorer survival than non-Hispanic whites during both time periods. Survival of Hispanic women with breast cancer was comparable to non-Hispanic whites during 1973-1982 but was significantly worse during 1983-1992. The significance of this lower survival is amplified by increasing breast cancer incidence among New Mexico Hispanics and American Indians.

The New Mexico Mammography Project. Screening mammography performance in Albuquerque, New Mexico, 1991 to 1993.

BACKGROUND: This project was designed to collect and link population-based mammography and breast carcinoma data to assess the performance of community mammography screening. METHODS: Computerized data were collected from all radiology practices in Albuquerque, New Mexico. The data were linked by computer match to breast carcinomas in a statewide cancer registry. Analysis is based on 126,466 screening mammogram studies performed on 87,443 female residents of New Mexico between the ages of 35 and 84 by 5 radiology groups. Sensitivity, specificity, positive predictive value, and call back rates were calculated as indicators of the discriminative performance of mammography. Carcinoma size and stage distribution were analyzed as outcome measurements. RESULTS: The computer match linked 634 breast carcinomas to the 126,466 screening mammogram series. The community-wide sensitivity was 79.9%, and specificity was 90.5%. The predictive value of an abnormal screen was 4.3%, and that of a biopsy recommendation result was 16.9%. The call back rate was 11.4%. The median invasive breast carcinoma size was 15 mm, 20.3% of carcinomas were in situ, 18.3% were lymph node positive, and 68.1% were Stage 0 or Stage 1. CONCLUSIONS: Mass screening mammography as practiced in Albuquerque, New Mexico, is able to detect breast carcinomas at early, treatable stages. The stage distribution of carcinomas is similar to that seen in successful clinical trials. However, measures of mammography performance show lower sensitivity, more additional studies, and more biopsy recommendations in this community setting than have been reported by expert mammographers.

Improving survival for patients with prostate cancer diagnosed in the prostate-specific antigen ERA.

OBJECTIVES: Although prostate-specific antigen (PSA) screening has not been demonstrated to reduce prostate cancer mortality in the clinical trial or population setting, the use of PSA for screening increased during the early 1990's. A clinical trial is currently underway to test the efficacy of PSA screening; however, the results will not be available for at least 10 years. To address concerns about the effectiveness of PSA screening in the near term, population-based assessments of PSA screenings are needed. To reduce mortality, PSA screening must first produce improved survival. In New Mexico, increased screening was associated with a stage migration from distant to earlier stages and an increase in 5-year relative survival, suggesting that PSA screening may be effective. METHODS: To further investigate early indicators of the effectiveness of PSA screening in New Mexico, we examined temporal trends in survival for the period 1983-1992, using proportional-hazard models to estimate the risk of death by year of diagnosis, adjusted for age, stage, grade, ethnicity, and initial treatment. RESULTS: We found the risk of death following the diagnosis of local or regional-stage prostate cancer decreased in the 1987-1988 (relative-risk [RR] = 0.9 [95% confidence interval (CI) 0.8, 1.1]), 1989-1990 [RR = 0.9 [0.8, 1.0]), and 1991-1992 (RR =0.7 [0.6, 0.9]) periods compared with the 1983-1984 period. Men treated with radical prostatectomy were at increased risk between 1985 and 1990, compared with those treated in the 1983-1984 period. However, for men diagnosed and treated in the 1991-1992 period, risks were lower than in the 1983-1984 period (RR =0.8 [0.4, 1.5]). CONCLUSIONS: The earlier stage at diagnosis and the improved survival during the period of increased PSA screening are consistent with changes expected from an effective screening test and treatment modality.