Cancer history and other personal factors affect quality of life in patients with hepatitis C

Patients who tested positive for hepatitis C virus (HCV) by PCR (polymerase chain reaction) were approached at outpatient clinics at Memorial Sloan-Kettering Cancer Center (MSKCC), North General Hospital (NGH), and Mt. Sinai Hospital. We began the study at these institutions in June 2000, October 2001, and May 2003, respectively, following approval by the Institutional Review Boards at each site. Those eligible for the study were aged 18 years or older, English-speaking, and approved for the study by their physician. The overall response rate to the study was 66% of those approached. There were 188 patients who completed both the main questionnaire and the SF-36. Twenty-three were excluded from this analysis because they had sustained response to treatment, defined as having negative PCR for at least two measurements at least 6 months apart. Also excluded were 23 patients who completed the SF-36 while on treatment for HCV and two who completed it while being treated for cancer; they were excluded because treatment is likely to lead to short-term reduction in quality of life. This analysis is based on 140 patients.

After obtaining informed consent, the medical interviewer administered a questionnaire to determine demographic characteristics, medical history, the probable route of infection, and other lifestyle factors. Information on health-related quality of life was collected by use of the SF-36. This is a validated and frequently-used instrument that includes 36 questions on quality of life, grouped into eight domains: physical functioning, general health perception, pain, social functioning, role limitations-emotional, role limitations-physical, vitality, and mental health. These eight domains can be summarized in two overall measures, the Physical Component Scale (PCS) and the Mental Component Scale (MCS). Using standardized methods [7, 8], we recalibrated raw scores as required, imputed missing values for individual questionnaire items where possible, and transformed scores for each domain to a scale of 0 to 100.

We calculated summary measures for the PCS and MCS and transformed them using norm-based scoring, which results in a mean of 50 and SD of 10 in the general U.S. population [8]. Because of missing data on some of the domains, the number of patients with PCS and MCS scores was 136. Higher scores on these measures indicate better health. Information on clinical factors, such as treatment, stage, and ALT level was abstracted from medical records. Data on ALT were missing for two patients.

Data were analyzed using SAS. Normative data on the PCS and MCS summary scales have been published for the general U.S. population, based on a survey conducted in 1990 [8]. We compared mean scores of our patients to these norms, in total and for men and women separately. We used t-tests for independent samples to compare means on the eight domains and on the PCS and MCS summary scales in subgroups of our patients defined by demographic factors (e.g., gender) and health factors (e.g., presence of chronic conditions). For those variables for which statistically significant differences were found in the PCS and MCS between subgroups of patients, we used multivariate general linear models to determine which of these variables were independently related to the PCS and MCS scores. Because of our interest in the effects of cancer history, we also included this factor in these models.

About half of the patients in this study were men (53%) and the mean age was 53.4 years (SD 11.4, range 23 to 87). Men were younger than women (50.8 vs 56.3, p < 0.01). Eighty-six percent were from MSKCC, with most of the rest from NGH. About half were Caucasian and about one-third were African-American. While more than half had at least some college education, 20% had not graduated from high school. Forty-seven percent were currently employed. Two-thirds were physically active when interviewed, participating in physical activities such as walking, sports, aerobics, running or other activity more than once a week. Forty-eight percent had a previous diagnosis of cancer. These are mainly long-term cancer survivors, with a mean of 9.4 years since diagnosis (median 6.6 years, range 4 months to 38 years).

Diagnosis of cancer and years since diagnosis were the same in men and women. Among these patients, most had breast cancer (n = 17), followed by lymphoma (n = 14), colon cancer (n = 7), testicular cancer (n = 6), prostate cancer (n = 5) and leukemia (n = 5). Among all patients, other conditions were common: 72% had at least one chronic medical condition (heart disease, diabetes, hypertension, lung disease, thyroid disease, arthritis, rheumatoid arthritis, asthma, stroke or transitional ischemic attack, Crohn's disease, colitis, ulcers, or psoriasis); the highest prevalences were for hypertension (37%) and arthritis (22%). Forty-three percent had two or more of these conditions. Women were more likely than men to have heart disease and thyroid disease and to have two or more conditions.

One-quarter of the patients had been diagnosed with depression at some time and 15% were currently being treated for depression; more women than men ever had depression. Using responses to questions on the amount of beer, wine, and hard liquor drunk and the frequency of drinking each type of alcohol, we determined that 36% of respondents had a history of high alcohol consumption, defined as drinking an average of >2 drinks per day of beer, wine, or hard liquor for men and >1 drink per day for women [9]. Among those with heavy use of alcohol, 22% continued to drink in the last year, while among those who had stopped drinking, the mean number of years since drinking was 8.8 ± 7.4 (range, 1–30 years). The most common route of infection was IV drug use, for 40%, with 30% infected through transfusions and 30% infected through other or unknown routes. The mean number of years since using IV drugs was 19, and none claimed to have used IV drugs in the past 1.5 years. Sixteen patients (11%) had a positive test for human immunodeficiency virus (HIV) by self-report. Four percent were treated for hepatitis C before completing the SF-36; the mean number of months between completing treatment and filling out the questionnaire was 11.7 ± 10.1 (range 1–27 months). An additional 18% were treated after filling out the SF-36. ALT levels close to the time of completing the SF-36 questionnaire (mean 19.7 days, range 0–163) were available for 137 patients: the median was 56 and 20% had ALT levels of 100 or above. Among the 81 patients who had liver biopsy, 42% had stage III-IV disease, including 12% with stage IV.

Table 1 shows the results for the summary measures, the PCS and MCS. Compared to the norm of 50 among the general U.S. population, patients with CHC scored lower on these measures, particularly the PCS. Compared to men in the general population, men with CHC scored below the norm for men on the PCS, but were equal to the norm on the MCS. In contrast, women with CHC scored below the norms for women on both these summary scales.

Table 2 shows scores for the individual domains. The first column shows transformed results, scaled to the norm of 50; transformation accounts for differences in the structure of the scales and allows for comparison between scales. The second column shows scores before transformation, to facilitate comparison to other studies. On the transformed scores, the CHC patients in our study had the highest scores for vitality and role-emotional (i.e., the extent to which emotional problems interfere with daily activities), and the lowest for general health.

In univariate analysis, shown in Table 3, female gender, lower physical activity, history of heavy alcohol use and history of depression led to lower scores on both the PCS and MCS. Other variables that were associated with significantly lower PCS scores on univariate analysis were having less than a high school education, not working currently, and having two or more chronic conditions. Those with a high ALT level (i.e., ALT >100) close to the time of completing the SF-36 had lower scores on the PCS that were marginally significant. We included these variables in multivariate analyses of factors affecting the PCS score. In addition, because of our interest in quality of life among cancer patients with CHC, we included history of cancer in the models.

Results of multivariate models for the PCS incorporating these variables are shown in Table 4. History of cancer and having two or more other conditions were strongly associated with the PCS, and high ALT level, having fewer years of education, and low physical activity also had a significant effect on the PCS score. Although gender had an important effect in univariate analysis, it was strongly related to education and depression, with women having less education and higher prevalence of depression; with all these variables in the model, only education had an independent effect. Current employment and alcohol use were not significant in multivariate analysis. The variables in the model explained 39% of the variation in the PCS. Results of multivariate analysis differed by gender: among men, only a history of cancer was associated with the PCS, while among women, only chronic conditions affected the score. The variables included in the models explained more of the variance for women than for men.

In univariate analysis (Table 3), the factors associated with the MCS were depression, level of physical activity, gender, and alcohol consumption. These factors and cancer history were included in multivariate models. Results, shown in Table 5, indicate that each of these factors, except history of cancer, was independently associated with MCS, explaining 23% of the variance. Among men, only physical activity was significant, and among women, depression and alcohol use were significant.

We also investigated several other factors for their association with the PCS and MCS, in both univariate and multivariate analyses. Variables not related to these quality of life measures included: age, race, marital status, hospital, HIV infection, route of transmission, treatment for CHC, stage of fibrosis, and measures of social support (attendance at religious services, attendance at meetings of community groups, and number of friends).