The quality of life (QOL) of long-term survivors of bladder cancer in a population-based registry was assessed.
The quality of life (QOL) of long-term survivors of bladder cancer in a population-based registry was assessed.
The Functional Assessment of Cancer Therapy (FACT-BL) instrument was used to evaluate QOL in a population-based sample of bladder cancer patients. QOL scores were compared between those undergoing radical cystectomy (RC) or those with an intact bladder (BI) and between continent and conduit urinary diversion groups. The influence of current age and time since diagnosis of cancer on QOL were also examined. Multivariate regression analyses were performed to examine the influence of age, time since diagnosis, current condition, treatment, stage of cancer, and comorbid conditions on QOL.
A total of 259 patients participated in the study who had undergone RC (n = 82) or other therapy (BI) (n = 177). There were no differences in general QOL scores between RC and BI groups and between the 2 urinary diversion groups, but patients undergoing RC had worse sexual function scores. QOL scores for BI patients tended to decrease with increasing age (P = .01). Presence of comorbid conditions lowered QOL (P<.05).
General QOL does not vary among long-term bladder cancer survivors regardless of treatment, but sexual functioning can be adversely affected in those undergoing cystectomy. Long-term QOL declines even in those with intact bladders, particularly in those with comorbidities. Cancer 2006. © 2006 American Cancer Society.
Bladder cancer is the fourth most common cancer among US males and the eighth most common cancer among US females.1 In the US, approximately 63,000 new cases of bladder cancer will be diagnosed in the year 2005 and 13,180 people will die of the disease.2 Although most bladder cancer cases are superficial at the time of diagnosis, repeated transurethral resections and intravesical chemo/immunotherapy may be required to treat recurrences. Each of these interventions, particularly intravesical therapy, is accompanied by side effects.3 Health-related quality of life (HRQOL) can deteriorate because of multiple interventions and the accompanying side effects. The other major treatment for bladder cancer, radical cystectomy with urinary diversion, can also have a significant impact on HRQOL. But it is possible that patients adjust over time to their diversion and their QOL could stabilize or even improve after the initial decline. A number of studies have reported on QOL of patients with bladder cancer but have focused on the relation between types of urinary tract diversion after total cystectomy.4, 5 Not much is known about the long-term impact of various treatment modalities for superficial or invasive disease on the QOL of bladder cancer patients.4 The impact of treatment for superficial disease or invasive disease on longitudinal QOL is also unclear. Our primary study objective was to examine QOL in long-term survivors of bladder cancer. We intended to compare differences in QOL between patients who had undergone radical cystectomy and those who still had their bladder intact and also to assess if type of urinary diversion influences QOL scores among those patients who had undergone cystectomy. We expected patients with their bladders intact to have better long-term QOL than patients who had undergone radical cystectomy, that QOL improves as time from diagnosis of cancer increases, and patients who have undergone continent diversion after radical cystectomy have better long-term QOL than patients who underwent conduit diversion.
For purposes of this study we identified a cohort of bladder cancer patients through the State Health Registry of Iowa who had been diagnosed between 1990 and 1999. The State Health Registry of Iowa is 1 of 14 population-based cancer registries reporting to the Surveillance, Epidemiology, and End Results (SEER) Program of the National Cancer Institute. Each of the patients was directly contacted by mail to solicit participation in the study and to provide consent for release of medical record information. Two repeated mailings, each 2 weeks apart, were used to try and contact nonresponders. A total of 840 patient solicitations were mailed. In all, 259 patients consented to participate in the study, whereas 302 declined to participate. The reasons for not volunteering to participate in the study were: deceased (n = 35), dementia (n = 10), not confirmed case of cancer (n = 15), not a good candidate (n = 6), too ill (n = 13), too old (n = 6), and not stated (n = 217). There were 249 patients who never responded to our multiple mailings. The remaining patients (n = 30) were deemed not locatable.
Patients' QOL was measured using the self-administered Functional Assessment of Cancer Therapy–Bladder Cancer (FACT-BL) instrument. The FACT-BL instrument includes a FACT-G portion that assesses general QOL as well as a disease-specific component that includes 12 questions specific to bladder cancer (additional concerns, AC). The FACT instruments were originally developed and validated by Cella et al.6 The 27 FACT-G core items assess the physical well being (PWB), social and family well being (SFWB), emotional well being (EWB), and functional well being (FWB). The PWB domain comprises questions such as having lack of energy, having trouble meeting needs of family, and being bothered by side effects. The SFWB assesses the closeness of the respondent to friends and family members. The EWB examines how well the respondents are coping with their illnesses and has questions specifically asking about feeling nervous or sad. The FWB assesses how well the respondents are able to work and enjoy life. The AC domain has questions that are specific to bladder cancer patients such as caring for and embarrassment due to an ostomy appliance. The AC domain also has questions pertaining to sexual interests of the respondents. All the questions in each of the domains are scored on the Likert scale, with scores ranging from 0 through 4 (“not at all” through “very much”). Responses to some of the questions have to be reverse-coded because higher QOL scores on these indicate worse HRQOL. Proration of missing values was performed as suggested in the FACT-G scoring manual. The range of possible scores for PWB, SFWB, EWB, FWB, and AC are 0–28, 0–28, 0–20, 0–28, and 0–48, respectively. FACT-G scores were obtained by adding the scores of the PWB, SFWB, EWB, and FWB domains. FACT-BL total QOL score was a sum of the PWB, SFWB, EWB, FWB, and AC domains. Higher scores indicate better QOL in each of the domains.
Information regarding patients' demographics and disease-related factors such as sex, race, date of birth, age at diagnosis, and stage of cancer at diagnosis were obtained from the tumor registry, whereas data regarding current condition of patient, comorbid conditions (current and previously present), and treatment were obtained from a patient survey, telephone interviews, and direct contact with treating physicians when necessary. Patients were defined as having undergone radical cystectomy (RC) or surviving with an intact bladder (BI) based on self-reported data from the survey. Demographic and clinical characteristics of participants and nonparticipants as well as those in RC and BI were compared using chi-square tests for categorical variables and independent sample t-tests for continuous variables.
Differences in QOL scores between RC and BI subjects were examined using the Mann–Whitney U-test. Individual domain and total scores were compared. The scores were tested for normality using the Kolmogorov–Smirnov test. Nonparametric tests were used due to the skewed data distribution.7 We also compared scores between patients who had undergone RC immediately after diagnosis of bladder cancer (IC group) and those who had radical cystectomy after other therapy (CFOT group). Comparisons between RC and BI groups and between IC and CFOT groups were also made within 4 other strata (i.e., current age ≤75 years, current age >75 years, time since diagnosis ≤100 months, and time since diagnosis >100 months). Patients were stratified into the above-mentioned 4 groups based on the median current age and median time since diagnosis for the overall study population.
Multivariate linear regression analyses were conducted to examine the influence of independent variables such as current age, time since diagnosis, sex, treatment type (radical cystectomy or therapies where the bladder is left intact including intravesical therapy), stage of cancer at diagnosis, current condition of patient, and presence of comorbid conditions on QOL scores. Current age and time since diagnosis were treated as categorical variables. Current age was categorized into ≤70, 71–75, 76–80, and >80 years groups. Time since diagnosis was divided into ≤75, 76–100, 101–125, and >125 months groups. Charlson scores were computed using the self-reported comorbidity data obtained from the survey to determine the extent of severity of comorbidity.8 Separate regression models were built for each QOL domain, FACT-G, and FACT-BL scores. To control for stage, multiple regression analysis was also conducted only on cases in which the tumor was localized to the site of origin (n = 114). Because the QOL scores were skewed, the multivariate regression analyses were performed on log-transformed scores. We developed full models that included Charlson scores and reduced models that did not adjust for Charlson scores. We followed this strategy as we could obtain information about comorbid conditions for only 210 patients.
One-way analysis of variance (ANOVA) was used to examine the influence of current age and time since diagnosis on QOL scores. For current age and time since diagnosis, the groups described in the previous paragraph were used. Separate tests were performed for RC and BI groups.
Differences in individual domain, FACT-G, and FACT-BL QOL scores between continent and conduit diversion groups were also examined using the Mann–Whitney U-test.
Informed consent was obtained from the participants and the study was approved by the University of Iowa Institutional Review Board. Statistical significance was set at a 2-tailed P value of <.05. For the ANOVA, Bonferroni corrections were applied and P<.01 was considered statistically significant for comparisons between 3 groups and P<.008 for comparison between 4 groups. All analyses were considered exploratory and hence no power calculations were possible. All statistical analyses were performed using SAS v. 9.1 (SAS Institute, Cary, NC) and SPSS v. 13.0 (SPSS, Chicago, IL) for Windows software.
The comparative profile of participants and nonparticipants is summarized in Table 1. The only significant differences were mean age at diagnosis (nonparticipants were older) and mean current age (nonparticipants were older).
|Profile||Participants N (%)||Nonparticipants N (%)||P|
|No. of subjects||259||302|
|Sex||Male||200 (77.2)||220 (72.8)||NS|
|Female||58 (22.4)||81 (26.9)|
|Missing||1 (0.4)||1 (0.3)|
|Race||White||249 (96.1)||297 (98.3)||NS|
|Black||0 (0)||0 (0)|
|Unknown||10 (3.9)||5 (1.7)|
|Age at diagnosis (yrs)||Mean||64.4||71||.001|
|Current age (yrs)||Mean||73.3||79.7||.001|
|Time since diagnosis (mos)||Mean||99.8||104.9||NS|
|Stage of cancer at diagnosis||In situ||124 (47.9)||172 (56.9)||NS|
|Localized to site of origin||114 (44)||109 (36.1)|
|Regional, direct extension to adjacent organs or tissue only||11 (4.2)||15 (5)|
|Missing||10 (3.9)||6 (2)|
|Current condition||Cured||162 (62.5)||NAV||NA|
|Disease in remission||66 (25.5)||NAV|
|Disease recurrence||8 (3.1)||NAV|
Of the 259 patients who agreed to participate in the study, 82 patients had undergone RC for bladder cancer and 177 patients had BI. Of the 82 RC patients, 62 were classified as IC and 20 as CFOT. The demographic and clinical characteristics of the RC and BI groups are summarized in Table 2. There were no statistically significant differences between the 2 groups except for stage of cancer at diagnosis and current condition of disease. The majority of the BI cases (117 out of 170 cases) were in situ tumors.
|Profile||RC (%)||BI (%)||P|
|No. of subjects||82||177|
|Sex||Male||61 (74.4)||139 (78.5)||NS|
|Female||21 (25.6)||37 (20.9)|
|Missing||0 (0)||1 (0.6)|
|Race||White||79 (96.3)||170 (96)||NS|
|Unknown||3 (3.7)||7 (4)|
|Age at diagnosis (yrs)||Mean||64.4||64.3||NS|
|Time since diagnosis (mos)||Mean||100.6||99.3||NS|
|Stage of cancer at diagnosis||In situ||7 (8.5)||117 (66.1)||.001|
|Localized to site of origin||61 (74.4)||53 (29.9)|
|Regional, direct extension to adjacent organs or tissue only||11 (13.4)||0|
|Missing||3 (3.7)||7 (4)|
|Current condition||Cured||62 (75.6)||100 (56.5)||.03|
|Disease in remission||15 (18.3)||51 (28.8)|
|Disease recurrence||1 (1.2)||7 (4)|
|Missing||4 (4.9)||19 (10.7)|
Table 3 summarizes the QOL scores of the RC and BI groups. There were no statistically significant differences in QOL scores between the 2 groups when all cases were considered.
|Domain (score range)||Group||N||Mean (SD)||P|
|Physical well being (0–28)||RC||82||26 (3)||.84|
|Social and family well being (0–28)||RC||82||22 (5)||.75|
|Emotional well being (0–20)||RC||82||18 (2)||.35|
|Functional well being (0–28)||RC||82||23 (4)||.58|
|FACT-G total score (0–104)||RC||82||89 (11)||.93|
|Additional concerns (0–48)||RC||82||35 (6)||.17|
|FACT-BL total score (0–152)||RC||82||124 (15)||.55|
Comparison of QOL scores between IC and CFOT groups revealed no statistically significant difference in QOL scores between the groups for the whole sample.
Of the 82 patients who underwent RC, 56 patients had a conduit diversion and 26 patients had continent diversion (continent cutaneous diversion or neobladder). The Mann–Whitney test suggested that there were no significant differences in QOL scores between the 2 groups (Table 4).
|Domain (score range)||Group||N||Mean (SD)||P|
|Physical well being (0–28)||Conduit diversion||56||25 (4)||.76|
|Continent diversion||26||26 (2)|
|Social and family well being (0–28)||Conduit diversion||56||23 (6)||.24|
|Continent diversion||26||22 (4)|
|Emotional well being (0–20)||Conduit diversion||56||18 (2)||.73|
|Continent diversion||26||18 (3)|
|Functional well being (0–28)||Conduit diversion||56||23 (5)||.94|
|FACT-G total score (0–104)||Conduit diversion||56||89 (12)||.60|
|Continent diversion||26||89 (9)|
|Additional concerns (0–48)||Conduit diversion||56||35 (6)||.81|
|Continent diversion||26||35 (6)|
|FACT-BL total score (0–152)||Conduit diversion||56||124 (15)||.75|
|Continent diversion||26||124 (13)|
In the BI group, the AC and FACT-BL scores were lower (worse QOL) for the >80 years age group when compared with the ≤70 age group (P<.005). A similar pattern was also seen for the RC group, but this was not statistically significant. In both RC and BI groups there were no statistically significant differences in QOL scores between the time since diagnosis groups.
After adjusting for multiple comparisons, there were no statistically significant differences in QOL scores among IC, CFOT, and BI groups with current age (≤75 and >75 years) and the time since diagnosis (≤100 and >100 months) subsets.
Table 5 summarizes the results of the multivariate regression analyses. Considering all 259 cases, males had higher EWB scores (better QOL) when compared with females (P = .004). PWB, FWB, AC, FACT-G, and FACT-BL scores decreased (worse QOL) as Charlson scores increased (P<.05). Patients who underwent radical cystectomy had lower AC scores when compared with bladder intact patients (P<.05). When examining only Stage II cases (localized tumors, n = 114), males had low SFWB scores (worse QOL) when compared with females (P = .001) and subjects whose current age was >80 years had higher SFWB scores (better QOL) when compared with subjects ≤70 years (P = .04). Patients who underwent radical cystectomy had lower SFWB, AC, and FACT-BL scores when compared with bladder intact patients (P<.05). EWB, FWB, FACT-G, and FACT-BL scores decreased (worse QOL) as comorbidity scores increased (P<.05).
|Time since diagnosis (mos)|
|101–125||1.17%||−1.98%||−0.04% to 1.24%||3.34%||−0.25%||1.42%||0.9%|
|Current age (yrs)|
|Stage of cancer at diagnosis|
|Stage II (localized) cancer cases only|
|Time since diagnosis (mos)|
|Current age (yrs)|
Fifty-nine of the 82 patients who underwent RC responded to the question “Are you embarrassed by an ostomy appliance?”. Of these, only 17 (29%) expressed at least a little embarrassment with the ostomy appliance. Twenty-seven of 59 patients (46%) felt some degree of difficulty in caring for the ostomy appliance. Patients with their bladders preserved had more interest in sex when compared with patients who had their bladders removed. Twenty-one percent of respondents who had their bladders preserved were not interested in sex as compared with 39% who underwent radical cystectomy (P = .002). In the RC group, about 89% of the respondents could not have and keep an erection, whereas in the BI group only 32% of respondents could not have and keep an erection (P<.001). Among patients who had undergone RC, 8.5% of the responders reported being unhappy with their body image, compared with the 5% reported by patients in BI group (P = .28).
The study participants generally had QOL scores closer to the highest possible score for PWB (26 out of 28) and EWB (18 of 20). Similar results were observed in previous studies on QOL in bladder cancer patients.9, 10 There were no significant differences in QOL scores that could be identified by using the FACT-BL instrument between the RC and BI groups. These results suggest that the general QOL, at least as determined by the FACT-BL instrument, is well maintained in patients with bladder cancer regardless of type of therapy. However, patients who had undergone radical cystectomy had a higher likelihood of complaining of sexual dysfunction and were characterized by having a lack of interest in sex. Several other studies have demonstrated similar findings.9–14 Based on the study by Litwin,15 erectile dysfunction is prevalent in 43% of older men. Even accounting for such baseline erectile dysfunction due to age, a greater proportion of men who had undergone radical cystectomy had erectile dysfunction most likely due to surgical damage to cavernous nerves. After controlling for other factors, comorbidity independently affects QOL scores. The effect on QOL scores was small yet significant and may be more obvious with larger sample sizes. Disease- and treatment-related factors appear to have a larger effect on specific domain scores, indicating that they may be more important than comorbidity in influencing QOL scores.
Previous studies have shown that bladder cancer patients undergoing cystectomy had negative feelings about their body image.16, 17 However, in our study very few participants reported being unhappy with their body image and the number was no different from those with their bladders intact (8.5% in RC and 5% in BI). Considering the fact that this study was done in long-term survivors of bladder cancer, an explanation could be that patients could have undergone psychological adaptation and thus any negative feelings about their body image may have diminished as a consequence. The fact that they have survived despite their cancer may also generate positive feelings that counter any negative feelings regarding the altered body image generated by the treatment necessary for their cancer.
Previous studies suggest that depression and anxiety are common among cancer patients and many cite “concern about cancer” as a major factor affecting their life.18, 19 In our study, a vast majority of the patients were neither nervous nor worried about dying. Only 4 out of 259 patients (1.5%) mentioned being either nervous or worried quite a bit about dying. Overall, the patients had high QOL scores for the EWB domain, suggesting that they were emotionally stable. Similar results were also seen in a study conducted by Matsuda et al.9 Thus, a sizable proportion of long-term survivors of bladder cancer are likely to be emotionally stable. However, these results must be interpreted with caution, as there may be a respondent bias, with only the more emotionally stable responding to the instrument.
The results of our study indicate that for patients with their bladders intact the FACT-BL scores were lower in the older age groups. One possible explanation for this is that repeated conservative treatment and surveillance procedures on these patients could have lowered their QOL. Recurrence rates for superficial bladder cancers are very high and it is very likely that patients with superficial disease will need repeated interventions and recurrent monitoring that would negatively impact their WOL. Older patients may be more significantly affected by such interventions compared with younger patients. Among patients who underwent RC, there were no significant differences in QOL scores between the different age and time since diagnosis groups. Previous studies have shown that patients who undergo surgery have an initial drop in QOL, which then stabilizes at a lower level within 9 months.20–22
To our knowledge, 5 other studies used the FACT-BL instrument to assess QOL in patients with bladder cancer.9, 10, 23–25 Matsuda et al.9 evaluated the long-term effects of treatment on QOL among 95 bladder cancer patients and found that RC negatively impacts sexual functioning. Autonomy in daily life and old age were found to affect QOL. Familial situation was found to impact emotional and social/family well being. They also found that patients younger than 75 years had significantly better QOL scores when compared with patients older than 75 years. Our study results are in agreement with those of Matsuda et al.9 We found that RC negatively influenced sexual functioning and QOL scores were lower in the higher age groups. Herman et al.24 used the FACT-BL instrument to assess QOL of bladder cancer patients who had undergone transurethral resection of a bladder tumor followed by gemcitabine with concurrent radiotherapy. The study results suggest that gemcitabine and radiotherapy do not negatively influence patient-reported QOL. Two studies used FACT-BL or FACT-G to compare QOL in bladder cancer patients who had undergone different types of urinary diversion.10, 25 They have been unable to demonstrate a difference in QOL between conduit versus continent diversion25 or continent cutaneous versus orthotopic diversion.10 Our study results also suggest that there were no differences in QOL scores between patients who received a conduit diversion and continent diversion.
Limitations of our study include lack of information about patients' baseline pretreatment QOL. This makes it difficult to ascertain the effect that baseline QOL may have on QOL after therapy. This would require a prospective longitudinal study, which we are planning at this time. Another issue is that there are few patients in the CFOT group (n = 20). There were only 11, 9, 13, and 7 patients in the CFOT group with a current age ≤75 years, >75 years, time since diagnosis ≤100 months, or >100 months, respectively. These small sample sizes make it difficult for us to determine the impact of cystectomy in patients with presumed recurrent or progressive cancer. This study was conducted on long-term survivors of bladder cancer (minimum time since diagnosis was 45 months) and it is possible that patients had time to readjust to their condition . The current age of participants was significantly less than that of nonparticipants. It is possible that the QOL scores for the whole group would decrease if more of the older nonparticipants with ostensibly poorer QOL had participated in the study. Patients who were relatively healthy may have participated in the study. As mentioned earlier, some of the common reasons for nonparticipation in the study were dementia, too ill, and too old. It is possible that the QOL scores would decrease if the patients with the above-mentioned conditions had participated in the study.
Overall, there were no major differences in QOL between RC and BI patients based on the analyses of this population-based cohort. FACT-BL QOL scores for BI patients were lower for older age groups. RC negatively impacts the sexual functioning of bladder cancer patients. There are no differences in QOL scores between continent and conduit diversion groups. The presence of comorbid conditions lowers QOL independent of treatment.