| |
Colorectal Cancer Screening in HIV-Infected Patients 50 Years of Age and Older: Missed Opportunities for Prevention
|
| |
| |
The American Journal of Gastroenterology
Volume 101 Page 907 - April 2006
Massimiliano Berretta, M.D., Ph.D.
Umberto Tirelli, M.D.
Division of Medical Oncology A
National Cancer Institute
Centro di Riferimento
Oncologico, Aviano (PN), Italy
TO THE EDITOR: The article by Dr. Reinhold et al. (1) provides very appropriate and timely considerations with regard to the management of HIV-infected patients 50 yr of age and older in the highly active antiretroviral therapy (HAART) era.
Within the HIV-related activities at our Institution, since 1986, we are following the HIV-related tumor aspects of the HIV epidemic. While the HIV-related tumors are definitely declining since the advent of HAART in the clinics (2), the number of non-HIV-related tumors as a result of chronicity of HIV infection and aging of the HIV-infected population is increasing. We have observed 21 patients with advanced colorectal cancer at our Institution and at the Institutions working within the Italian Cooperative Group on AIDS and Tumors (GICAT-Gruppo Italiano Cooperativo AIDS e Tumori) activities. The analysis of this series demonstrated that the 21 HIV-positive colorectal cancer patients developed cancer at an earlier age (median age 48 yr) and the disease was much more advanced than in the general population (3). Actually, at the time of diagnosis, most of the patients presented with more poorly differentiated tumors, advanced disease with frequent liver and lung involvement, and overall poor prognosis (3).
It is evident from our data that the early diagnosis of colorectal cancer is needed within the HIV-positive population. Unfortunately, there is no program developed in such population so far, but for gastrointestinal tumors it is now mandatory. In fact, it is now recommended in HIV patients older than 45 yr at our Institution.
Massimiliano Berretta, M.D., Ph.D.
Umberto Tirelli, M.D.
Colorectal Cancer Screening in HIV-Infected Patients 50 Years of Age and Older: Missed Opportunities for Prevention
Jean-Pierre Reinhold, M.D.1, Marianne Moon, M.D.1, Craig T. Tenner, M.D.1, Michael A. Poles, M.D., Ph.D.1, and Edmund J. Bini, M.D., M.P.H.1
1Department of Medicine and Division of Gastroenterology, VA New York Harbor Healthcare System and NYU School of Medicine, New York, New York
OBJECTIVES: Although human immunodeficiency virus (HIV)-infected patients are now living longer, there are no published data on colorectal cancer (CRC) screening in this population. We hypothesized that HIV-infected patients were less likely to be screened for CRC compared to patients without HIV.
METHODS: Consecutive HIV-infected patients ≥50 yr old seen in our outpatient clinic from 1/1/01 to 6/30/02 were identified. For each HIV-infected patient, we selected one age- and gender-matched control subject without HIV infection who was seen during the same time period. The electronic medical records were reviewed to determine the proportion of patients that had a fecal occult blood test (FOBT), flexible sigmoidoscopy, air-contrast barium enema (ACBE), or colonoscopy.
RESULTS: During the 18-month study period, 538 HIV-infected outpatients were seen and 302 (56.1%) were ≥50 yr old. Despite significantly more visits with their primary care provider, HIV-infected patients were less likely to have ever had at least one CRC screening test (55.6%vs 77.8%, p< 0.001). The proportion of HIV-infected patients who ever had a FOBT (43.0%vs 66.6%, p< 0.001), flexible sigmoidoscopy (5.3%vs 17.5%, p< 0.001), ACBE (2.6%vs 7.9%, p= 0.004), or colonoscopy (17.2%vs 27.5%, p= 0.002) was significantly lower than in control subjects. In addition, HIV-infected patients were significantly less likely to be up-to-date with at least one CRC screening test according to current guidelines (49.3%vs 65.6%, p< 0.001).
CONCLUSIONS: A substantial number of HIV-infected patients are ≥50 yr of age and CRC screening is underutilized in this population. Public health strategies to improve CRC screening in HIV-infected patients are needed.
(Am J Gastroenterol Aug 2005;100:1805-1812)
INTRODUCTION
Since the introduction of highly active antiretroviral therapy (HAART) in 1995, there has been a marked decline in the incidence of acquired immunodeficiency syndrome (AIDS)-defining conditions and AIDS-related deaths (1-5). The widespread use of HAART has resulted in a significant increase in the life expectancy of patients infected with human immunodeficiency virus (HIV) and those diagnosed with AIDS, and this has substantial implications for the long-term care of these individuals (6-9).
Colorectal cancer (CRC) is the second leading cause of cancer-related death in the United States; each year approximately 150,000 new cases of CRC are diagnosed and over 57,000 persons die from this disease (10). Current screening guidelines from the U.S. Preventive Services Task Force (11), the American Cancer Society (12), the Gastrointestinal Consortium (13), and others (14, 15) recommend CRC screening starting at 50 yr of age for all persons at average-risk for CRC.
Although many HIV-infected patients are living well beyond 50 yr of age, there are no published data on CRC screening in this population. Therefore, the aims of this study were to evaluate the proportion of HIV-infected patients seen in the outpatient clinics at our medical center who were ≥50 yr of age and to determine whether these individuals had been screened for CRC. We hypothesized that HIV-infected patients were less likely to be screened for CRC compared to age- and gender-matched patients without HIV, representing a missed opportunity for cancer prevention.
Study Population
All HIV-infected outpatients who were seen in the Infectious Diseases Clinic at the VA New York Harbor Healthcare System in New York City from 1/1/01 to 6/30/02 were identified using our computerized patient record system. At our medical center, infectious diseases physicians also serve as the primary care providers for patients with HIV. Patients were included in this study if they were HIV positive, were at least 50 yr of age, and were seen as an outpatient at least once during the study period. Patients were excluded from this study if they were less than 50 yr old or if they did not have documentation of HIV infection (HIV antibody positive with confirmation by Western blot). For patients who had multiple clinic visits during the study period, the index visit was defined as the most recent time that the patient was seen.
For each HIV-infected patient, we identified one control subject without HIV. The HIV negative control group was selected from the primary care clinic at the same medical center and was individually matched to each HIV-positive patient by age (within 1 yr), gender, and the time of the index visit (within 1 month). Matching was performed by reviewing a list of all patients without HIV who were seen in the primary care clinic during the same month as the HIV patients and selecting the first patient of the same age and gender on the list. The list of patients from which the control subjects were selected was printed by month and arranged in alphabetical order. Selection of control subjects was blinded to any demographic data, clinical information, and knowledge of prior CRC screening. The study protocol was reviewed and approved by the Institutional Review Board at our medical center.
DISCUSSION
Screening for CRC is cost-effective compared with no screening (16), and is associated with a significant decrease in mortality (17-23). However, data from several population-based studies demonstrated that screening for CRC is underutilized in the United States (24-29). In the 2001 Behavioral Risk Factor Surveillance System survey of 87,729 adults ≥50 yr of age in the United States, only 53.1% of persons were up-to-date with CRC screening according to published guidelines (26). In contrast, a higher proportion of our subjects without HIV were up-to-date with CRC screening (65.6%), and this was similar to the 68% of patients who were up-to-date with CRC screening (FOBT in the past 12 months, sigmoidoscopy in the past 5 yr, or colonoscopy in the past 10 yr) reported nationally in the VA health-care system (30).
Although our rates of CRC screening among patients without HIV were higher than in the general population of the United States, we found that HIV-infected individuals were significantly less likely to be up-to-date or to have ever had one or more CRC screening tests. This difference was noted for all types of CRC tests and remained statistically significant after adjusting for potential confounding variables. It is likely that patient-related factors were partially responsible for the low screening rates in patients with HIV since nearly half of the patients who never had FOBT performed had documentation in the electronic medical record that they refused the test. In addition, provider barriers and other factors likely contributed to the low CRC screening rates in this population.
It is interesting to note that several of the CRC screening tests (FOBT, flexible sigmoidoscopy, and colonoscopy) were more likely to have been performed for clinically indicated reasons in our HIV-infected patients as compared with control subjects. This finding results in an even greater disparity in CRC screening between the two groups.
Among our patients with HIV infection, we found that older age, a family history of CRC, more than 10 visits with their primary care provider in the past 24 months, and an undetectable HIV viral load were the only variables that were significantly associated with ever having at least one CRC screening test. These findings were not surprising because older patients and those with a family history of CRC are at increased risk of developing cancer of the colon and rectum and were probably more likely to be offered and to accept screening (13, 14). One potential explanation for the finding that patients with more than 10 visits with their primary care provider in the past 24 months and those with an undetectable HIV viral load were more likely to have been screened for CRC is that those individuals who were more compliant with follow-up and with taking their antiretroviral medications were also more compliant with recommendations by their provider to undergo CRC screening.
The underuse of CRC screening tests in patients with HIV has important clinical implications and may represent a missed opportunity for cancer prevention. Although there has been a decline in the incidence of AIDS-defining malignancies (Kaposi's sarcoma, non-Hodgkin's lymphoma, and cervical cancer) after the introduction of HAART, recent studies have demonstrated that non-AIDS-defining malignancies account for an increasing proportion of cancers seen in HIV-infected patients (31-39). Several studies have demonstrated that CRC is one of the non-AIDS-defining malignancies that is increasing in incidence in the HIV population (32, 38, 40). For example, a prospective cohort study of 2,882 patients with HIV found that the annual incidence of CRC was 0.65 per 1,000 patient-years in the pre-HAART era and this increased to 2.34 per 1,000 patient-years between 1997 and 2002 (38).
In addition to an increase in the incidence of non-AIDS-defining malignancies, these neoplasms tend to develop at an earlier age and are more aggressive in HIV-infected patients as compared to those without HIV (36, 40-43). In a cancer registry study from Bellevue Hospital in New York City, Demopoulos et al. (43) noted that the mean age at the time of diagnosis of CRC was 51.0 yr in HIV-infected patients as compared with 68.6 yr in those without HIV. Interestingly, Yeguez and colleagues (40) found that the median age of HIV-infected patients with adenocarcinoma of the colon was even younger (46 yr), and 4 of the 6 patients were less than 40 yr old. The increasing incidence, younger age of onset, and more aggressive course of CRC in patients with HIV, coupled with our data showing underuse of screening in these individuals, highlights the need to increase patient and provider awareness of CRC screening in this population. In addition, it is important to assess the potential costs and benefits of CRC screening programs in patients with HIV and to determine whether CRC screening should begin before 50 yr of age.
The strengths of the present study include the relatively large number of HIV-infected patients ≥50 yr of age, the inclusion of a control group of patients without HIV, and the use of the computerized patient record system to obtain detailed demographic and clinical data. In addition, this study is unique because we are unaware of any published studies that have evaluated CRC screening in patients with HIV.
However, there are some limitations of this study that should be considered when interpreting our findings. The most obvious limitation of this investigation was the retrospective study design. The VA electronic medical record is very comprehensive, but is it possible that certain variables, such as a family history of CRC, may have been underestimated in our study due to lack of documentation. Although our findings need to be confirmed in well-designed prospective studies, it would be difficult to perform this study prospectively because it would likely influence patient and provider behaviors and bias the results.
Another limitation of this study is that it was performed at a single VA medical center and the majority of our patients were men. Therefore, our findings may not be generalizable to other VA medical centers, non-VA settings, or women. However, the VA health-care system has demonstrated a leading role in health-care delivery and the quality of care is significantly better than the Medicare fee-for-service program (30). Given the high compliance with CRC screening in the VA health-care system, it is possible that screening for CRC in HIV-infected patients is even less common in other practice settings.
Finally, the reasons for the low CRC screening rates in patients with HIV could not be determined in our study. Although infectious diseases specialists at our medical center serve as the primary care providers for patients with HIV, they do not serve as the primary care providers for non-HIV-infected patients. Therefore, it is difficult to know whether the low screening rates are due to selective under screening of HIV-infected patients or due to the fact that infectious disease specialists are less likely to screen their patients for CRC as compared with other primary care physicians. It is likely that the low screening rates in patients with HIV are due to a combination of factors, including patient barriers, provider barriers, and other unidentified variables.
In conclusion, our data show that CRC screening tests are underutilized in HIV-infected patients 50 yr of age and older. Population-based studies are needed to confirm our findings and to determine patient and provider barriers to CRC screening in this population. As the life expectancy of persons with HIV continues to increase, a better understanding of the etiology, epidemiology, and natural history of colorectal polyps and cancers will be necessary in order to implement age-appropriate CRC prevention, screening, and treatment recommendations.
RESULTS
Patient Demographic and Clinical Characteristics
During the 18-month study period, a total of 538 HIV-infected outpatients were seen at our medical center. Of these 538 patients, 302 (56.1%; 95% CI, 51.8-60.4%) were 50 yr of age or older and were included in this study.
The demographic and clinical characteristics of the 302 HIV-infected patients and the 302 age- and gender-matched control subjects are shown in Table 1. HIV-positive patients were more likely to be Black and were less likely to use nonsteroidal antiinflammatory drugs or aspirin. In addition, the type of primary care providers differed between the two groups, and HIV-infected patients were seen by their primary care provider significantly more often than control subjects.
Among the 302 patients with HIV, 262 (86.8%) were receiving antiretroviral therapy at the time of the index visit and 144 (47.7%) had undetectable plasma levels of HIV RNA. The median CD4 count was 331.5 cells/mm3 (interquartile range, 195.8-512.3 cells/mm3), 77 (25.5%) had a CD4 lymphocyte count below 200 cells/mm3, and 22 (7.3%) had a CD4 lymphocyte count below 100 cells/mm3.
Proportion of Patients Who Have Ever Had CRC Screening
HIV-infected patients were significantly less likely to have ever had at least one CRC screening test performed compared with control subjects (55.6%vs 77.8%, p< 0.001), a difference of 22.2% (95% CI, 14.9-29.5%).
Since there were statistically significant differences in the baseline characteristics of the HIV-infected patients and control subjects, we performed multiple logistic regression analysis to adjust for potential confounding variables. After adjusting for age, gender, race, nonsteroidal antiinflammatory drug or aspirin use, the type of primary care provider, and the number of visits with their primary care provider in the 24 months prior to the index visit, the odds of ever having at least one CRC screening test performed was significantly lower in HIV-infected patients compared with control subjects (OR = 0.18; 95% CI, 0.11-0.30; p< 0.001).
As shown in Figure 1, HIV-infected patients were significantly less likely to have ever had each type of CRC screening test (FOBT, flexible sigmoidoscopy, ACBE, or colonoscopy) as compared with control subjects. In addition to these procedures, HIV-infected patients were also less likely to have ever had a digital rectal examination (9.6%vs 59.3%, p< 0.001). Of the 172 HIV-infected patients and 101 control subjects who never had FOBT, there was no difference in the proportion of individuals who refused the test (47.7%vs 40.6%, p= 0.26).
Among patients who had one or more CRC screening tests, we calculated the proportion of procedures that were performed for screening versus clinically indicated reasons (Table 2). The proportion of HIV-positive patients who had FOBT, flexible sigmoidoscopy, and colonoscopy for screening purposes was significantly lower than in control subjects. In contrast, the proportion of patients that had ACBE for screening purposes did not differ between the two groups.
Factors Associated with CRC Screening in HIV-Infected Patients
We determined the proportion of HIV-infected patients who ever had at least one CRC screening test stratified according to select demographic and clinical characteristics (Table 3). In the univariate analyses, older age, a family history of CRC, more than 10 visits with their primary care provider in the past 24 months, current use of antiretroviral therapy, and an undetectable HIV viral load were associated with a significantly increased odds of ever having at least one CRC screening test, while the odds were significantly lower in those with a CD4 lymphocyte count <100 cells/mm3.
We then performed multivariable logistic regression analysis to identify independent variables that were associated with CRC screening. In the multivariable analysis, older age, a family history of CRC, more than 10 visits with their primary care provider in the past 24 months, and an undetectable HIV viral load were the only variables that were significantly associated with ever having at least one CRC screening test (Table 3).
Proportion of Patients Who Were Up-to-Date with CRC Screening
In addition to determining whether patients ever had a CRC screening test, we also determined the proportion of patients that were up-to-date with CRC screening. HIV-infected patients were significantly less likely to be up-to-date with at least one CRC screening test as compared with control subjects (49.3%vs 65.6%, p< 0.001), a difference of 16.3% (95% CI, 8.4-24.0%).
As shown in Figure 2, HIV-infected patients were significantly less likely to be up-to-date with each type of CRC screening test (FOBT, flexible sigmoidoscopy, ACBE, or colonoscopy) as compared with control subjects. In addition to these procedures, HIV-infected patients were also less likely to have had a digital rectal examination performed with in the last year (6.0%vs 38.7%, p< 0.001).
Study Design
We reviewed the electronic medical records of each patient and data were abstracted on standardized data-collection sheets. Data collected on each patient included age, gender, race, medical and other comorbid conditions, use of nonsteroidal antiinflammatory drugs or aspirin, warfarin use, and family history of CRC in a first-degree relative. In addition, we collected information about the patient's primary care provider (attending physician vs house staff vs nurse practitioner) and the number of clinic visits with their primary care provider during the 24 months prior to the index visit. For HIV-infected patients, we also collected information about current use of antiretroviral medications and the most recent CD4 lymphocyte count and HIV viral load.
The electronic medical records and our endoscopy records were reviewed to determine whether patients had any prior CRC screening tests, including fecal occult blood testing (FOBT), flexible sigmoidoscopy, air-contrast barium enema (ACBE), and colonoscopy. For each test, we recorded whether the examination was performed for clinical indications or screening. At our medical center, radiographic data are available as far back as 1984 and endoscopy results are available from approximately 1975.
Study Outcomes
The primary outcome measure was the proportion of patients 50 yr of age and older who have ever had at least one CRC screening test (FOBT, flexible sigmoidoscopy, ACBE, or colonoscopy). The proportion of patients who have ever had at least one CRC screening test was compared between individuals with HIV and control subjects without HIV.
Sample size calculations for the primary outcome of this study were estimated based on preliminary data from our medical center showing that approximately 70% of patients 50 yr of age and older have ever received at least one CRC screening test in the past. In order to detect a minimum difference of 15% (55% in the HIV-positive group and 70% in the control subjects) at a 2-tailed alpha level of 0.05 with a power of 90%, 217 patients per group were needed. We chose an 18-month period for this study because approximately 300 HIV-infected patients 50 yr of age and older were seen during that time period.
Secondary outcomes of this study included factors associated with CRC screening in HIV-infected subjects, and the proportion of HIV-infected patients and control subjects that were up-to-date with CRC screening. According to published guidelines, patients were considered to be up-to-date with CRC screening if they had FOBT within 1 yr, flexible sigmoidoscopy within 5 yr, ACBE within 5 yr, or colonoscopy within 10 yr (11-15).
Statistical Analysis
Continuous variables were compared using an unpaired t-test or the Mann-Whitney U test as appropriate. Data are expressed as means ± SD for those variables that were normally distributed, and medians and interquartile range (25th-75th percentile) for those with a nonnormal distribution. Categorical variables are expressed as proportions and were compared using the x2 test or Fisher's exact test.
Univariate logistic regression analysis was utilized to identify factors associated with ever having at least one CRC screening test. Subsequently, a multivariable logistic regression model was created involving all variables included in the univariate analyses. The multivariable analysis could then assess the independent effect of each variable in the model adjusted for the effects of each of the other variables. The strength of the association between covariates and prior CRC screening are expressed as odds ratios (OR) with 95% confidence intervals (CI). Statistical analysis was performed using SPSS software version 12.0 for Windows (SPSS Inc., Chicago, IL) and a 2-tailed p-value of <0.05 was considered statistically significant.
|
|
| |
| |
|
|
|
