David F. Jarrard, 7037 Wisconsin Institutes of Medical Research, 1111 Highland Avenue, Madison, WI 53792, USA. e-mail: email@example.com
What's known on the subject? and What does the study add?
Very few studies have examined end-of-life urological studies in men with prostate cancer. These studies reported fewer procedures in men who received primary therapy for prostate cancer. However, these studies were typically single institution or had a short follow-up period.
The present study is the first population-based study examining end-of-life urological procedures and uses a geographic region encompassing 385 000 patients. Furthermore, this study incorporates both hospital- and office-based procedures. This approach has not been previously undertaken.
• To determine using a population-based approach whether men with end-stage prostate cancer who had definitive primary therapy might require fewer urological interventions. Repeated urological procedures can impact health-related quality of life in patients dying from prostate cancer.
PATIENTS AND METHODS
• Using the Marshfield Epidemiological Study Area (MESA) database and tumour registry, we compared end-of-life interventions in men who died from prostate cancer between 1991 and 2009.
• Patient charts were queried for urological procedures using International Classification of Disease Modification, 9th edition (ICD9) codes for 3 years before death.
• Clinicopathological information was examined including whether the patient had a history of primary therapy (radiation or radical prostatectomy).
• Among 280 patients dying from prostate cancer, 52 (19%) required 153 urological procedures during the last 3 years of life. The frequency of procedures increased closer to death. The most common procedures involved nephrostomy tube (56%), Foley catheter (24%) and transurethral resection of the prostate (10%).
• Clinicopathological features did not predict the need for an end-of-life urological procedure.
• There was no difference in the frequency of upper or lower tract procedures in surgery or radiation patients compared with patients without primary therapy (P= 0.556 and P= 0.508).
• Using a Kaplan–Meier analysis, there were no differences between groups in the proportion of patients not requiring a procedure (n= 280; P= 0.179).
• This is the first population-based study to examine the frequency of urological procedures in patients with end-stage prostate cancer.
• A minority of patients (19%) required urological procedures during the final 3 years of life.
• A history of surgery or radiation did not influence the overall risk for urological intervention.
Prostate cancer is the second leading cause of cancer death in men in the USA. The American Cancer Society projects that 28 170 men will die from this disease within the next year . The short-term complications of prostate cancer treatments, including radical prostatectomy (RP), radiation, and watchful waiting have been well-characterised [2–5]. However, less emphasis has been placed on complications of end-stage prostate cancer, subsequent surgical interventions, and other health-related quality of life (HRQL) issues that arise in the last few years of life. Furthermore, it is unclear whether these primary therapies can subsequently influence the HRQL in men dying from prostate cancer. Population-based studies examining the healthcare costs related to prostate cancer management conclude that the highest costs occur around the time of cancer diagnosis and at cancer death . Urological procedures at the end of life can contribute to high costs and may be an important HRQL measure for men dying from prostate cancer.
Patients who are treated with non-curative intent can undergo significant morbidity due to local and distant tumour progression in the final years of life [7,8]. Metastatic disease can cause anaemia, fractures, spinal cord compression, bone pain and other clinical sequelae. Local progression of end-stage prostate cancer results in LUTS, obstructive uropathy, haematuria and pain that result in hospital admissions, invasive procedures, and surgical interventions . Interventions required to treat local progression include TURP, urethral or suprapubic catheterisation, nephrostomy tube placement, ureteric stent placement, and clot evacuation . While RP and radiation are the standards for curative treatment, they may function primarily to decrease the tumour volume . To date, studies have not yet examined whether debulking actually reduces the need for urological procedures at the end of life.
The frequency of urological procedures that patients with cancer undergo in the final years of life is one contributing factor to the HRQL. Notably, the limited studies to date are based on single institutional data making an accurate assessment of the generalizability of procedural and surgical interventions difficult . In addition, the identification of clinicopathological factors placing patients at risk for these procedures is unknown. In the present population-based study, we reported on the number and type of urological procedures that were required in the last 3 years of life in men who died from prostate cancer. Furthermore, we compared urological procedures and surgical interventions rates in men who had a history of primary therapy (either RP or radiation) to those who did not undergo primary therapy.
PATIENTS AND METHODS
This study was conducted using the population from the Marshfield Clinic (MC), one of the largest multispecialty clinics in the USA, serving 385 000 unique patients in Wisconsin and Michigan. This includes the Marshfield Epidemiological Study Area (MESA) database clinic tumour registry, which was initiated in 1991 as a dynamic, population-based cohort for disease surveillance and epidemiological research . The MC maintains an electronic medical record system including computerised diagnostic files dating back to 1963. Within the area surveyed for this study ≈90% of the population receives their entire health care from this system. It represents a geographically stable population with granular capture of residents, hospital discharges, outpatient visits, and deaths.
The database was queried between the years 1991–2009 for patients with a cause of death identified as prostate cancer by death certificate. Data was collected on patients that had no local therapy, RP, or radiation treatment (external beam radiation, brachytherapy, or a combined modality). The distribution of the years in which patients were treated was similar across all groups.
Electronic chart abstraction was performed to determine age at diagnosis, age at death, stage, grade, presence of metastatic disease at the time of diagnosis, as well as number and type of procedures. The electronic medical record was queried using International Classification of Disease Modification, 9th edition (ICD9) codes for specific urological conditions and procedures including insertion of a nephrostomy tube (55.02, 55.03), replacement of a nephrostomy tube (55.93), Foley catheter insertion (57.94, 57.95), TURP (60.29), suprapubic tube placement or replacement (57.17, 57.18, 59.94), ureteric stent placement (59.8), cystoscopy with clot evacuation (57), cystoscopy with fulguration (57.4), and urinary diversion (56.71, 56.6). Data collection for procedures was limited to the last 3 years of life. We chose a 3-year follow-up based on the data that the median time from castrate-resistance to death is ≈3 years and that complications rarely occur before becoming castrate-resistant [14,15].
The t-test, chi-square test, and univariate and multivariate multinomial logistic regression were used to compare patients with a history of radiation or surgery against patients without primary therapy. For multinomial logistic regression, radiation and surgery were outcomes of interest. Odds ratios and CIs were reported for only variables that were statistically different between groups. The t-test, chi-square test, and univariate and multivariate binomial logistic regression were used to test the association between patient characteristics and the need for a urological procedure. For binomial logistic regression, requiring a subsequent procedure was the outcome of interest. anova was used to determine whether having primary therapy influenced the mean number of each type of urological procedure. When Levene's test for homogeneity of variances was significant at the P < 0.05 level, nonparametric statistics (Kruskal–Wallis) were used to confirm the results of the anova. Kaplan–Meier analysis was used to compare the proportion of patients who remained procedure-free among groups.
There were 3770 patients diagnosed with prostate cancer in the tumour registry, with 1303 recorded deaths in this population. The query identified 280 (7.4%) patients with prostate cancer recorded as the primary cause of death. Complete data was available for 96.68% of patients with prostate cancer (only 125 of 3770 individuals were not seen by a database provider within the 3 years of the censor date). The mean follow-up was 2.7 years. For the population that died from prostate cancer, the average age at diagnosis and death was 73.7 and 78 years, respectively (Table 1). At the time of diagnosis, 47% had poorly differentiated disease, and 46% had metastatic disease.
Table 1. Treatment group clinicopathological features at diagnosis
*Chi-square test, t-test, and multinomial logistic regression. †Grade I (well differentiated: Gleason score <6); Grade II (moderately differentiated: Gleason score 7); Grade III and IV (poorly differentiated: Gleason score 8–10).
The patients who died from prostate cancer were divided into three groups: no primary therapy, surgery, and radiation. Among the 280 patients, 29 (10%) of patients had undergone RP and 70 (25%) of patients had undergone radiation as primary therapy. There were 181 patients who had no record of primary therapy in the registry. When the clinical and pathological characteristics at diagnosis were examined between these groups, we found several significant differences. Men that got surgery or radiation at diagnosis were younger (P < 0.001 and P= 0.004, respectively) than those not receiving primary therapy (Table 1). A lower incidence of metastases at diagnosis and longer survival after diagnosis was identified (both P < 0.001) reflecting earlier stage disease in patients receiving primary therapy. There were no differences in the age at death or grade between primary and no primary therapy groups.
We assessed whether clinicopathological features at the time of diagnosis could predict the need for a urological procedure in the terminal stages of the disease. Among patients who died from prostate cancer, 52 (19%) underwent a urological procedure or surgery during the last 3 years of life. There were 228 (81%) patients who did not require a procedure. Using univariate and multivariate analyses, we found no statistical differences in the age at diagnosis, age at death, survival after diagnosis, stage, grade, or presence of metastatic disease at diagnosis between patients who subsequently underwent a urological procedure and patients who did not (Table 2). The t-test and chi-squared test were also used and confirmed no differences between these groups. Thus, clinical and pathological parameters at the time of prostate cancer diagnosis did not predict the need for a subsequent urological procedure.
Table 2. Predictors of a subsequent end-of-life urological procedure
The number of procedures per patient gradually increased approaching death, with the highest frequency (0.35) in the last year of life. However, patients were also at marked risk for undergoing a procedure in the second (0.16) and third (0.08) years before death (Fig. 1). Among the 52 (19%) patients who required a procedure during the last 3 years of life, 153 procedures were performed. Of the 52 patients, 25 (48%) patients required one procedure, 19 (37%) patients required two to five, and eight (15%) of patients required five or more.
The most common urological procedure was nephrostomy tube replacement (38%), followed by Foley insertion (24%), nephrostomy tube insertion (18%), TURP (10%), suprapubic tube insertion (8%), and ureteric stent placement (3%) (Fig. 2). Nephrostomy tube replacement was included, as these patients required a hospital-based procedure with recovery. The query did not show any patient who required cystoscopy with clot evacuation, fulguration, or urinary diversion. Our analysis included office-based procedures. We found that for patients undergoing a procedure, the frequency was highest for nephrostomy tube replacement (1.11) followed by Foley insertion (0.71) (Table 3).
We then evaluated procedures individually and grouped by lower tract (Foley, TURP, and suprapubic tube) or upper tract (stent and nephrostomy tube). An argument for the primary treatment of advanced disease is a decrease in the need for lower urinary tract procedures, e.g. TURP . There were 33 (12%) patients who required an upper tract procedure and 41 (15%) that required a lower tract procedure. We did not find a statistical difference in the mean number of urological procedures per patient when comparing surgery and radiation groups vs the no primary therapy group (P= 0.512). There was no difference between groups when procedures were grouped according to upper or lower urinary tract procedures (P= 0.556 and P= 0.508, respectively) (Table 3). Although radiation therapy patients might be expected to undergo TURP more frequently, TURP was generally a rare procedure in all groups at the end of life.
As some men survived <36 months after diagnosis, we performed a Kaplan–Meier analysis to account for censored data. Consistent with previous results, there were no differences among groups in the proportion of patients who remained procedure-free in the final 36 months of life (n= 280, log-rank P= 0.179) (Fig. 3).
HRQL is a major consideration for patients with terminal prostate cancer. The HRQL decreases in the final 12 months of life in men dying of this disease [16–19]. The number of urological procedures is an important measure of morbidity and costs in men dying from prostate cancer. To date, few studies have attempted to characterise the number of men who required urological procedures at the end of life (Table 4) [8,10]. The present study represents the first population-based analysis to describe specific urological procedure rates in the final years of life in men who died from prostate cancer. There were no clinicopathological factors that predicted the need for a urological procedure in patients dying from prostate cancer. In addition, we specifically show that a history of prostate radiotherapy or surgery (e.g. primary therapy) did not influence the need for more interventions in patients dying from prostate cancer.
Table 4. Urological procedures in men dying from prostate cancer*
Manuscripts that evaluated urological procedures in men dying from prostate cancer.
301 who died from prostate cancer
228 who died from prostate cancer
280 who died from prostate cancer
Swedish cancer registry
Diagnosis until death (≈5 years)
Last year of life
Last 3 years of life
Patients who underwent urological procedure, n (%):
Upper urinary tract
Lower urinary tract
Urological procedures performed, n
In the present study, we showed that the number of urological procedures increases toward the end of life. This increase in urological interventions is probably a reflection of complications and the clinical sequelae of advanced prostate cancer at the end of life. However, the number of procedures required in these patients is not insignificant in the years before death. We found that a substantial number of procedures (37%) occurred before the final year of life. Most studies to date have focused on the final year of life and have only evaluated the percentage of men affected by procedures, rather than specific urological procedural rates [6,10,17]. Even though the present study examined a longer period, we found even fewer patients (19%) who required a urological intervention than seen in previous studies (Table 1). For example, Khafagy et al.  reported that 25% required intervention. In the present study, nephrostomy tube procedures were among the most common, as well as Foley and suprapubic tube insertion similar to other studies.
Previous single institution data has suggested that RP may have benefits when compared with more conservative measures in men with advanced disease and is associated with extremely low morbidity and mortality rates . Tomlinson et al.  in 1977 compared patients with cT3 prostate cancer treated by perineal prostatectomy to TURP or simple prostatectomy. They found that 75% of the latter group eventually developed BOO and were more likely to develop ureteric obstruction (40%), infection (80%), and gross haematuria (45%) compared with the former group (4%, 26%, and 9%, respectively). Other research has found that men with untreated prostate cancer seem to be more affected by physical symptoms than expected from the ageing process alone . The present study is important in that we capture both hospital-based and out-patient procedures within a mean follow-up period of 2.7 years before death. There is a small possibility that a selection bias exists, as men with primary therapy were diagnosed younger and lived longer with the disease. We minimised this potential bias by examining procedures during the final 3 years of life across all groups and found the number of procedures increased uniformly during this time. We found that primary therapy with either RP or radiation had no influence on overall procedure rates and when procedures were stratified according to upper and lower tract procedures. In addition, there was no difference in the proportion of men requiring procedures. Compared with previous studies, we provide a longer follow-up period and account for censored data (due to death) using Kaplan–Meier analysis. Although the present survival analysis may be underpowered to detect a small difference, the study does capture a large study area encompassing 385 000 residents. The present study questions the assumption that patients with advanced prostate cancer should undergo a RP or radiation simply as a debulking therapy to prevent the morbidity of local tumour progression.
There are inherent weaknesses with any claims research. We may not have captured all men that died from prostate cancer or all urological procedures for which the database was queried. However, the MC has a remarkably stable medical population with complete medical follow-up making it unique . The MESA database is periodically validated and captures 97% of residents, 95% of discharges, 90% of outpatient visits, and 99% of deaths. We have not captured factors that may influence resource use, e.g. income, education, marital status, and geographic density. There may be practice patterns regionally that may not apply to other populations across the country. However, the population-based approach used does have distinct advantages compared with single institution studies, in that it limits bias in the treatment and follow-up of patients.
In conclusion, the prevalence of urological procedures at the end of life in men who died from prostate cancer is lower in the present study than other reports. In this population-based study with detailed follow-up, men who had a history of a RP or radiation underwent a similar number of procedures as men who had no definitive primary therapy.