- Email: [email protected]
Role of chronic catheterization in the development of bladder cancer in patients with spinal cord injury
ADULT UROLOGY
ROLE OF CHRONIC CATHETERIZATION IN THE DEVELOPMENT OF BLADDER CANCER IN PATIENTS WITH SPINAL CORD INJURY DOUGLAS A. WEST, JAMES M. CUMMINGS, WALTER E. LONGO, KATHERINE S. VIRGO, FRANK E. JOHNSON, AND RAUL O. PARRA
ABSTRACT Objectives. Patients with spinal cord injury (SCI) and chronic indwelling catheters are known to be at increased risk of bladder malignancy. “Decatheterization” by clean intermittent catheterization, external condom catheterization, or spontaneous voiding is thought to reduce the risk by decreasing the chronic mucosal irritation and rate of infection. We examined two Department of Veterans Affairs (DVA) data bases to test this theory. Methods. A population-based retrospective analysis of invasive treatments for carcinoma of the bladder in all DVA hospitals was conducted using computerized inpatient files from fiscal years 1988 to 1992. Results. One hundred thirty patients with bladder malignancy were identified from a pool of 33,565 patients with SCI (0.39%). All 130 patients underwent either radical cystectomy (n 5 63, 48%) or transurethral resection of bladder tumor (n 5 67, 52%). The 30-day perioperative mortality and overall 5-year survival rates were 2 (1.5%) and 49 (38%) of 130, respectively. Of the 130 patients analyzed, 42 (32%) had adequate data available regarding tumor pathologic findings and method of bladder management for analysis. The average age at diagnosis was 57.3 years. The histologic finding was transitional cell carcinoma in 23 (55%), squamous cell carcinoma in 14 (33%), and adenocarcinoma in 4 (10%) of 42. Bladder management was an indwelling urethral catheter in 18 (43%), suprapubic catheter in 8 (19%), clean intermittent catheterization in 8 (19%), and condom catheter in 6 (14%) of 42 patients. Squamous cell carcinoma was more common in patients with indwelling urethral catheters and suprapubic tubes (11 of 26, 42%) than in those using clean intermittent catheterization, condom catheterization, or spontaneous voiding (3 of 16, 19%). Conclusions. Bladder cancer was diagnosed in approximately 0.39% of this large SCI population during a 5-year period. Most cancers (55%) were transitional cell carcinomas. Squamous cell carcinoma was more common in patients with SCI and indwelling catheters than those without chronic catheterization. These data continue to suggest that avoidance of indwelling catheters, when feasible, is the preferred method of bladder management in patients with SCI. UROLOGY 53: 292–297, 1999. © 1999, Elsevier Science Inc. All rights reserved.
S
pinal cord injury (SCI) is a difficult and common problem. It is estimated that 15,000 SCIs occur in the United States each year.1 Integral to the successful treatment of patients with SCI is preservation of renal function. Current strategies From the Division of Urology, Department of Surgery, St. Louis University School of Medicine, and the John Cochran Veterans Affairs Medical Center, St. Louis, Missouri Reprint requests: Raul O. Parra, M.D., Division of Urology, St. Louis University Medical Center, 3635 Vista at Grand, P.O. Box 15250, St. Louis, MO 63110-0250 Submitted: July 9, 1998, accepted (with revisions): August 31, 1998
292
© 1999, ELSEVIER SCIENCE INC. ALL RIGHTS RESERVED
emphasize maintaining a low-pressure urinary reservoir that is easily drained, free of infection, and reasonably continent. Various forms of bladder catheterization are used. The success of this management plan has led to improved survival among this population. The altered natural history of SCI has permitted long-term complications of bladder catheterization to emerge as a major clinical challenge. These include the development of bladder cancer. The natural history of bladder cancer in patients with SCI is poorly understood. One important factor leading to the malignant transformation of 0090-4295/99/$20.00 PII S0090-4295(98)00517-2
bladder urothelium is thought to be the chronic irritation of indwelling catheters and recurrent urinary tract infections.2– 4 To minimize the risk of bladder cancer, “decatheterization” is currently considered the method of choice for long-term management.3 We performed a retrospective analysis of patients with SCI and bladder malignancy in a large population of U.S. veterans to better characterize the development of these malignancies and their relationship to bladder catheterization. MATERIAL AND METHODS The U.S. Department of Veterans Affairs (DVA) operates a system of 159 hospitals in 49 states, ranging in hospital size from 120 to 1742 beds. Approximately 600,000 to 700,000 inpatients and 2 million outpatients are treated in DVA facilities yearly. Because information about these patients is available in a standardized format through a single, accessible computer system, we surveyed this data set for patients with SCI and invasive treatment for bladder cancer. The Veterans Affairs Data Processing Center in Austin, Texas consists of an IBM mainframe computer (model 9021982) linked with several direct access IBM-emulating EMC storage computers (model 5500). The Beneficiary Identification and Records Location System, which is the primary veteran/beneficiary information source in the DVA, and the Patient Treatment Files, which contains a statistical record for each episode of inpatient care provided under Veterans Affairs auspices, were the two primary sources of data for this analysis. Access is limited to those with the requisite programming knowledge and security clearance to use this enormous data set. Admission data from the DVA hospital system for fiscal years 1988 to 1992 were abstracted for patients with a diagnosis of SCI and bladder cancer requiring invasive treatment. The International Classification of Diseases, 9th Edition, Clinical Modification (ICD-9-CM) diagnostic codes of 344 –344.99 (patients with SCI), 188.0 –188.99 (patients with bladder cancer), and 57.49 –57.799 (patients undergoing bladder resection) were used to compile a complete list of patients with both SCI and bladder cancer that specifically encoded all patients treated with either radical cystectomy (57.71) or transurethral bladder tumor resection (57.49). Institutional tumor registrars and medical record departments in all DVA hospitals that reported patients with SCI and treatment for bladder cancer from 1988 to 1992 were then contacted to request records of tumor staging information, discharge summaries, operative reports, and pathology reports. Data regarding method of bladder management, duration of catheterization, and duration of SCI were obtained by chart review. Duration of catheterization indicates the immediate preoperative form of bladder management and the approximate number of years before surgery it had been used. Patients were eliminated from the study if insufficient data regarding method of catheterization or prolonged use of alternating forms of bladder management were indicated in the medical records. In most instances, the duration of catheterization represented greater than 80% of the elapsed time from SCI. From the data available, no patient in the study group had previously received either pelvic radiation or chemotherapy for other malignancies. Statistical analysis of continuous variables (age, duration of SCI, duration of catheterization) used either the Student t test or, when appropriate, one-way analysis of variance. Frequency data (race, level of SCI, surgical intervention, tumor histologic findings) were analyzed using either Fisher’s exact UROLOGY 53 (2), 1999
TABLE I. Comparison between those patients with bladder cancer and spinal cord injury included in the study group versus those not in the study group Study Group (n 5 42)
Nonstudy Group P (n 5 88) Value*
Age (yr, mean 6 SD) 57.3 6 11.5 62.7 6 11.0 0.003 Race Black 4 (10) 8 (9) White 38 (90) 80 (91) NS Level of SCI Quadriplegia 17 (40) 19 (22) Paraplegia 25 (60) 69 (78) 0.035 Procedure Radical cystectomy 25 (60) 38 (43) TURBT 17 (40) 50 (57) NS Survival† Alive 18 (43) 25 (28) Deceased 24 (57) 63 (72) NS KEY: Age 5 age at diagnosis of cancer; SD 5 standard deviation; NS 5 not significant; SCI 5 spinal cord injury; TURBT 5 transurethral resection of bladder tumor. Numbers in parentheses are percentages. * Comparison between the study and nonstudy groups; only age and level of SCI are significantly different. † Survival data tabulated as of September 30, 1997.
test or chi-square test. A Kaplan-Meier curve was constructed to depict overall survival of the 130 patients studied as of September 30, 1997.
RESULTS A total of 83,241 inpatient admissions for 33,565 patients with SCI were recorded in the DVA data base for fiscal years 1988 to 1992. A computerized search of the Patient Treatment Files identified 130 patients (129 men, 1 woman) with a diagnosis of SCI who later underwent treatment for bladder cancer (0.39%). Of the 159 DVA facilities, 57 (36%) reported having at least 1 patient with the simultaneous diagnoses of SCI and bladder cancer. Specific information regarding inpatient treatment of these patients was available from 25 (44%) of 57 hospitals. These institutions graciously provided records for 66 (50%) of the 130 patients identified by computer search. Of these, 42 (64%) had adequate data available regarding the method of bladder management and tumor histologic findings to be considered evaluable. These constitute the study population. Table I summarizes data regarding age, race, level of paralysis, surgical procedure, and survival for all 130 patients as reported in the computerized Patient Treatment Files and Beneficiary Identification and Records Location System. Overall, most patients were paraplegic and underwent transurethral resection of their bladder tumors. Patient characteristics of the study group were similar to those in the nonstudy group. Only age at diagnosis 293
adenocarcinomas in 5 (12%). The IND group had 12 (46%) of 26 patients with TCC, 11 patients (42%) with SCC, and 3 patients (12%) with adenocarcinoma of the bladder. By contrast, 11 (68%) of 16 patients in the decatheterized group had TCC, 3 (19%) had SCC, 1 (6%) had adenocarcinoma, and 1 (6%) had mixed tumor on pathologic examination. COMMENT
FIGURE 1. Kaplan-Meier curve depicting overall survival of 130 patients with SCI and bladder cancer dating from time of treatment.
of bladder cancer and level of SCI were significantly different. Two patients (1.5%) died within 30 days of their operative procedure. Figure 1 displays a Kaplan-Meier overall survival curve for all 130 patients in the study, with 1- and 5-year survival rates of 61% (79 of 130) and 38% (50 of 130), respectively. Mean follow-up duration was 5.5 years. Overall, the predominant method of bladder management in the study group (n 5 42) was either permanent suprapubic cystostomy tube (n 5 8, 19%) or chronic indwelling urethral catheter (n 5 18, 43%). These patients were classified as a single group (indwelling catheter group [IND]) representing 26 (62%) of the 42 patients evaluated. Bladder management in the remaining decatheterized (16 of 42) patients was by clean intermittent catheterization in 8 (19%), condom catheterization in 6 (14%), and spontaneous voiding in 2 (5%). Table II displays data regarding age, duration of SCI, duration of catheterization, and type of invasive procedure used to treat bladder cancer in the two groups. Average patient age at diagnosis of cancer for the study group was 57.3 years; mean time from SCI and mean duration of catheterization were 23.9 and 20.0 years, respectively. There was no significant difference in mean age at diagnosis or duration of catheterization among the two groups (IND versus decatheterized). However, patients in the decatheterized group were diagnosed with bladder cancer approximately 18.1 years after SCI versus 26.5 years for the IND group (P 5 0.054). Evaluation of tumor type, stratified against method of bladder management, is presented in Table III. Most tumors (23 of 42, 55%) were transitional cell carcinomas (TCC), and squamous cell carcinomas (SCC) were found in 14 (33%) and 294
This study represents a retrospective analysis of patients with known SCI and a documented method of bladder management who subsequently developed bladder carcinoma requiring invasive treatment. Previously published studies2,4 –7 (Table IV) of such patients have been hampered by the small number of patients identified with both SCI and bladder cancer. The DVA institutions across the country recorded data on 33,565 patients with SCI during the fiscal years 1988 through 1992, of whom 130 had bladder cancer. In 42, adequate information regarding duration of SCI, duration and method of bladder catheterization immediately before surgery, and therapy for bladder cancer was available for analysis, making this the largest series thus far reported. Previous reports have indicated an increased incidence of bladder cancer, ranging from 2.5% to 10%, among the population with SCI. The low incidence in our analysis may reflect an improvement in urologic treatment for these patients, with earlier and more aggressive treatment of the complications of chronic catheterization or the non-referral-based population evaluated. This also presumably underestimates the incidence because only a 5-year period was assessed. Bickel et al.5 reported an incidence of 0.32% in a study of 2900 patients with SCI and attribute this decreased incidence to improved lower tract management in these patients. Although improved detection and treatment of urinary tract infections, bladder calculi, and other complications of chronic catheterization may contribute to a declining incidence of bladder malignancy in this population, other factors may play a role as well. These include errors in diagnostic coding and the possible treatment of Veterans Affairs’ patients in non-VA hospitals. The mean age at diagnosis was 57.3 years (range 36 to 84), which is approximately 10 years sooner than that expected in the general population.8 Although TCC was the predominant histologic finding (55%) identified in this study, it has been previously reported that SCC is most common.2,4,9 Several possible explanations exist for this finding. First, like all retrospective analyses, the study is potentially biased. We report on the histopathologic findings from 42 patients (32%) of the 130 UROLOGY 53 (2), 1999
TABLE II.
Study group by method of bladder management
Age (yr) Years from injury Years catheterized† Surgical intervention (n) Radical cystectomy TURBT
Total (n 5 42)
IND (n 5 26)
Decatheterized (n 5 16)
57.3 6 11.5 23.9 6 11.1 20.0 6 11.5
56.7 6 10.7 26.5 6 10.1 23.0 6 11.4
58.3 6 12.9 18.1 6 11.6* 14.5 6 10.1
28 (66) 14 (33)
20 (77) 6 (23)
8 (50) 8 (50)
KEY: IND 5 patients with an indwelling Foley catheter; Decatheterized 5 patients without an indwelling catheter; TURBT 5 transurethral resection of bladder tumor. Data presented as mean 6 SD, unless otherwise noted. Numbers in parentheses are percentages. *P 5 0.054 vs. IND. † Number of years before surgery that patients were using a given form of bladder catheterization.
TABLE III.
Histopathologic findings of bladder cancer versus method of bladder catheterization Total (n 5 42)
Transitional cell carcinoma Squamous cell carcinoma Adenocarcinoma Mixed
23 14 4 1
IND (n 5 26)
(55) (33) (9) (2)
12 (46) 11 (42) 3 (12) —
Decatheterized (n 5 16) 11 3 1 1
(68) (19) (6) (6)
Abbreviations as in Table II. No significant differences between groups were identified (P 5 0.3). Numbers in parentheses are percentages.
TABLE IV.
Reported incidence of bladder cancer among patients with spinal cord injury
Study Current study Bickel et al.,5 1991 Broecker et al.,7 1981 Bejany et al.,4 1987 Locke et al.,6 1985 Kaufman et al.,2 1977
Pts. with SCI Surveyed (n) 33,565 2,900 1,052 300 25 62
Pts. with CA (n) 131 8 10 11 2 6
(0.39) (0.28) (0.95) (3.6) (8) (10)
SCC (n) 14/42 2/8 2/10 9/11 2/2 6/6
(33) (25) (20) (82) (100) (100)
KEY: Pts. 5 patients; SCI 5 spinal cord injury; CA 5 cancer; SCC 5 squamous cell carcinoma. Numbers in parentheses are percentages.
identified by computer search. Unfortunately, pathologic findings are not recorded in the Patient Treatment Files and can only be obtained from chart review. Therefore, we cannot rule out the possibility that most of the remaining patients had SCC of the bladder. Second, the study population itself (U.S. veterans) may have an inherent bias toward the development of TCC, because the prevalence of smoking is higher in veterans than in the population at large, and cigarette smokers are known to have up to a fourfold higher incidence of bladder cancer,10 –12 predominantly of the TCC type. Although exact smoking histories were not routinely recorded in the charts reviewed, the incidence of smoking in veterans with SCI is about UROLOGY 53 (2), 1999
80% compared with approximately 50% in the general veteran population.13 Finally, the age at diagnosis in the study group was significantly younger than that for the remaining patients. The average age at diagnosis in the 88 patients not included in the study was 62.7 6 11.0. The possibility exists that patients included in the analysis were diagnosed earlier in the natural progression of their disease, before the “dedifferentiation” or “metaplasia” of their TCC to either SCC or adenocarcinoma. We report a nearly equal incidence of TCC and SCC in patients with indwelling catheters, whether transurethral catheters or chronic suprapubic tubes. The mean duration of SCI in this subset of patients was 23.9 years, which is consistent with 295
previous reports, indicating that most tumors are diagnosed after at least 10 years of catheterization for SCI.2 Interestingly, of the 12 patients in the IND group with a final diagnosis of TCC, 8 (66%) had elements of squamous metaplasia without SCC in the final pathology specimen. Of the patients in the IND group with a diagnosis of SCC, 6 (55%) of 11 also had squamous metaplasia. The significance of urothelial metaplasia in this setting is unknown. However, Sakamoto et al.14 analyzed 31 cases of bladder carcinoma, including 3 cases of “pure” SCC and 28 cases of TCC associated with squamous metaplasia. They proposed two forms of histogenesis for SCC of the bladder: (a) malignant transformation on the basis of squamous metaplasia alone, and (b) extensive differentiation toward a squamous form in a pre-existing TCC. According to this concept, perhaps SCC in patients with SCI often arise by metaplastic change in TCC. Certainly, delays in diagnosis of TCC in patients with SCI with continued exposure to inflammatory processes would be consistent with this hypothesis. Additional pathologic analysis is required to establish the pattern of tumorigenesis in this unique cohort of patients. In the decatheterized group, two interesting findings were noted. First, only 3 of the patients had a diagnosis of SCC (1 using clean intermittent catheterization and 2 using condom catheterization), but TCC was found in 11 of 16 patients. We cannot conclude from the available data that patients with decatheterization are less likely to develop a bladder malignancy than patients with indwelling catheters. However, it appears that fewer of these patients will develop SCC of the bladder. One theory behind the development of bladder malignancies is malignant transformation of the urothelium secondary to chronic irritation and infection associated with indwelling catheters. Previous histologic studies show an eosinophilic inflammatory response with urothelial dysplasia and squamous metaplasia at the site of the catheter.15 This would suggest that decatheterization may reduce the number of infections, as well as the degree of mucosal irritation, and result in less squamous metaplasia, ultimately leading to fewer SCC. Second, it appears that patients using a decatheterized approach develop bladder cancer sooner (approximately 8 years in the current study) than those with indwelling catheters (albeit 18 years after their SCI). Is it possible that patients who have the capacity and/or the opportunity to store urine develop bladder cancer sooner than patients with continuous bladder drainage? Environmental carcinogens in urine probably play a role in the development of some bladder cancers. Perhaps noncarcinogenic substances are converted to more active carcinogenic factors when stored in the bladder, or 296
possibly the increased surface area and prolonged exposure of the urothelium to these activated carcinogens hastens the development of bladder cancer. Although the number of patients reported in this study is large by comparison with other studies, the absolute numbers are still rather small. Clearly, further studies, ideally with prospective data, are necessary to evaluate these interesting findings. More recently, focus has centered on appropriate screening for patients with SCI and chronic catheterization in an attempt to facilitate early detection of these malignancies. Clearly, cystoscopy of patients with recurrent urinary tract infections and/or gross hematuria, with biopsy of any suspicious lesions, is warranted. Perhaps yearly cystoscopy after 10 years of catheterization, as advocated recently in published reports, would be a reasonable approach to early diagnosis.16,17 CONCLUSIONS We report a 5-year risk of approximately 0.39% for the development of bladder cancer in a large population of U.S. veterans with SCI. Evaluation of a subset of this population revealed a surprisingly large percentage of TCC of the bladder. Patients with SCI and bladder cancer who used for the most part a decatheterized approach had relatively fewer SCC than those with indwelling urethral or suprapubic catheters. These data lend further support to the avoidance of indwelling catheters in patients with SCI. REFERENCES 1. Bedbrook GM, and Sedgley GI: The management of spinal injuries past and present. Int Rehabil Med 2: 45– 61, 1980. 2. Kaufman JM, Fam B, Jacobs SC, et al: Bladder cancer and squamous metaplasia in spinal cord injury patients. J Urol 118: 967–971, 1977. 3. Jacobs SC, and Kaufman JM: Complications of permanent catheter drainage in spinal cord injury patients. J Urol 119: 740 –741, 1978. 4. Bejany DE, Lockhart JL, and Rhamy RK: Malignant vesical tumors following spinal cord injury. J Urol 138: 1390 – 1392, 1987. 5. Bickel A, Culkin DJ, and Wheeler JS: Bladder cancer in spinal cord injury patients. J Urol 140: 1240 –1242, 1991. 6. Locke JR, Hill DE, and Walzer Y: Incidence of squamous cell carcinoma in patients with long-term catheter drainage. J Urol 133: 1034 –1035, 1985. 7. Broecker FH, Klein FA, and Hackler RH: Cancer of the bladder in spinal cord injury patients. J Urol 125: 196 –197, 1981. 8. 1987 Annual Cancer Statistics Review: Including Cancer Trends: 1950 –1985, NIH Publication No. 88-2789. Bethesda, Maryland, Department of Health and Human Services, National Cancer Institute. 9. Melzak J: The incidence of bladder cancer in paraplegia. Paraplegia 4: 85–96, 1967. 10. Burch JK, Rohan TE, Howe GR, et al: Risk of bladder UROLOGY 53 (2), 1999
cancer by source and type of tobacco exposure: a case-control study. Int J Cancer 44: 622– 628, 1989. 11. Clavel J, Cordier S, Boccon-Gibod L, et al: Tobacco and bladder cancer in males: increased risk for inhalers and smokers of black tobacco. Int J Cancer 44: 605– 610, 1989. 12. Morrison AS, Buring JE, Verhoek WG, et al: An international study of smoking and bladder cancer. J Urol 131: 650 – 654, 1984. 13. Spungen AM, Lesser M, Almenoff PL, et al: Prevalence of cigarette smoking in a group of male veterans with chronic spinal cord injury. Military Med 160: 308 –311, 1995. 14. Sakamoto N, Tsuneyoshi M, and Enjoji M: Urinary
UROLOGY 53 (2), 1999
bladder carcinoma with a neoplastic squamous component: a mapping study of 31 cases. Histopathology 21: 135–141, 1992. 15. Goble NM, Clarke JJ, and Hammonds JC: Histological changes in the urinary bladder secondary to urethral catheterization. Br J Urol 63: 354 –357, 1989. 16. Navon JD, Hani S, Khonsari F, et al: Screening cystoscopy and survival of spinal cord injured patients with squamous cell cancer of the bladder. J Urol 157: 2109 –2111, 1997. 17. Stonehill WH, Goldman HB, and Dmochowski RR: The use of urine cytology for diagnosing bladder cancer in spinal cord injured patients. J Urol 157: 2112–2114, 1997.
297
ROLE OF CHRONIC CATHETERIZATION IN THE DEVELOPMENT OF BLADDER CANCER IN PATIENTS WITH SPINAL CORD INJURY DOUGLAS A. WEST, JAMES M. CUMMINGS, WALTER E. LONGO, KATHERINE S. VIRGO, FRANK E. JOHNSON, AND RAUL O. PARRA
ABSTRACT Objectives. Patients with spinal cord injury (SCI) and chronic indwelling catheters are known to be at increased risk of bladder malignancy. “Decatheterization” by clean intermittent catheterization, external condom catheterization, or spontaneous voiding is thought to reduce the risk by decreasing the chronic mucosal irritation and rate of infection. We examined two Department of Veterans Affairs (DVA) data bases to test this theory. Methods. A population-based retrospective analysis of invasive treatments for carcinoma of the bladder in all DVA hospitals was conducted using computerized inpatient files from fiscal years 1988 to 1992. Results. One hundred thirty patients with bladder malignancy were identified from a pool of 33,565 patients with SCI (0.39%). All 130 patients underwent either radical cystectomy (n 5 63, 48%) or transurethral resection of bladder tumor (n 5 67, 52%). The 30-day perioperative mortality and overall 5-year survival rates were 2 (1.5%) and 49 (38%) of 130, respectively. Of the 130 patients analyzed, 42 (32%) had adequate data available regarding tumor pathologic findings and method of bladder management for analysis. The average age at diagnosis was 57.3 years. The histologic finding was transitional cell carcinoma in 23 (55%), squamous cell carcinoma in 14 (33%), and adenocarcinoma in 4 (10%) of 42. Bladder management was an indwelling urethral catheter in 18 (43%), suprapubic catheter in 8 (19%), clean intermittent catheterization in 8 (19%), and condom catheter in 6 (14%) of 42 patients. Squamous cell carcinoma was more common in patients with indwelling urethral catheters and suprapubic tubes (11 of 26, 42%) than in those using clean intermittent catheterization, condom catheterization, or spontaneous voiding (3 of 16, 19%). Conclusions. Bladder cancer was diagnosed in approximately 0.39% of this large SCI population during a 5-year period. Most cancers (55%) were transitional cell carcinomas. Squamous cell carcinoma was more common in patients with SCI and indwelling catheters than those without chronic catheterization. These data continue to suggest that avoidance of indwelling catheters, when feasible, is the preferred method of bladder management in patients with SCI. UROLOGY 53: 292–297, 1999. © 1999, Elsevier Science Inc. All rights reserved.
S
pinal cord injury (SCI) is a difficult and common problem. It is estimated that 15,000 SCIs occur in the United States each year.1 Integral to the successful treatment of patients with SCI is preservation of renal function. Current strategies From the Division of Urology, Department of Surgery, St. Louis University School of Medicine, and the John Cochran Veterans Affairs Medical Center, St. Louis, Missouri Reprint requests: Raul O. Parra, M.D., Division of Urology, St. Louis University Medical Center, 3635 Vista at Grand, P.O. Box 15250, St. Louis, MO 63110-0250 Submitted: July 9, 1998, accepted (with revisions): August 31, 1998
292
© 1999, ELSEVIER SCIENCE INC. ALL RIGHTS RESERVED
emphasize maintaining a low-pressure urinary reservoir that is easily drained, free of infection, and reasonably continent. Various forms of bladder catheterization are used. The success of this management plan has led to improved survival among this population. The altered natural history of SCI has permitted long-term complications of bladder catheterization to emerge as a major clinical challenge. These include the development of bladder cancer. The natural history of bladder cancer in patients with SCI is poorly understood. One important factor leading to the malignant transformation of 0090-4295/99/$20.00 PII S0090-4295(98)00517-2
bladder urothelium is thought to be the chronic irritation of indwelling catheters and recurrent urinary tract infections.2– 4 To minimize the risk of bladder cancer, “decatheterization” is currently considered the method of choice for long-term management.3 We performed a retrospective analysis of patients with SCI and bladder malignancy in a large population of U.S. veterans to better characterize the development of these malignancies and their relationship to bladder catheterization. MATERIAL AND METHODS The U.S. Department of Veterans Affairs (DVA) operates a system of 159 hospitals in 49 states, ranging in hospital size from 120 to 1742 beds. Approximately 600,000 to 700,000 inpatients and 2 million outpatients are treated in DVA facilities yearly. Because information about these patients is available in a standardized format through a single, accessible computer system, we surveyed this data set for patients with SCI and invasive treatment for bladder cancer. The Veterans Affairs Data Processing Center in Austin, Texas consists of an IBM mainframe computer (model 9021982) linked with several direct access IBM-emulating EMC storage computers (model 5500). The Beneficiary Identification and Records Location System, which is the primary veteran/beneficiary information source in the DVA, and the Patient Treatment Files, which contains a statistical record for each episode of inpatient care provided under Veterans Affairs auspices, were the two primary sources of data for this analysis. Access is limited to those with the requisite programming knowledge and security clearance to use this enormous data set. Admission data from the DVA hospital system for fiscal years 1988 to 1992 were abstracted for patients with a diagnosis of SCI and bladder cancer requiring invasive treatment. The International Classification of Diseases, 9th Edition, Clinical Modification (ICD-9-CM) diagnostic codes of 344 –344.99 (patients with SCI), 188.0 –188.99 (patients with bladder cancer), and 57.49 –57.799 (patients undergoing bladder resection) were used to compile a complete list of patients with both SCI and bladder cancer that specifically encoded all patients treated with either radical cystectomy (57.71) or transurethral bladder tumor resection (57.49). Institutional tumor registrars and medical record departments in all DVA hospitals that reported patients with SCI and treatment for bladder cancer from 1988 to 1992 were then contacted to request records of tumor staging information, discharge summaries, operative reports, and pathology reports. Data regarding method of bladder management, duration of catheterization, and duration of SCI were obtained by chart review. Duration of catheterization indicates the immediate preoperative form of bladder management and the approximate number of years before surgery it had been used. Patients were eliminated from the study if insufficient data regarding method of catheterization or prolonged use of alternating forms of bladder management were indicated in the medical records. In most instances, the duration of catheterization represented greater than 80% of the elapsed time from SCI. From the data available, no patient in the study group had previously received either pelvic radiation or chemotherapy for other malignancies. Statistical analysis of continuous variables (age, duration of SCI, duration of catheterization) used either the Student t test or, when appropriate, one-way analysis of variance. Frequency data (race, level of SCI, surgical intervention, tumor histologic findings) were analyzed using either Fisher’s exact UROLOGY 53 (2), 1999
TABLE I. Comparison between those patients with bladder cancer and spinal cord injury included in the study group versus those not in the study group Study Group (n 5 42)
Nonstudy Group P (n 5 88) Value*
Age (yr, mean 6 SD) 57.3 6 11.5 62.7 6 11.0 0.003 Race Black 4 (10) 8 (9) White 38 (90) 80 (91) NS Level of SCI Quadriplegia 17 (40) 19 (22) Paraplegia 25 (60) 69 (78) 0.035 Procedure Radical cystectomy 25 (60) 38 (43) TURBT 17 (40) 50 (57) NS Survival† Alive 18 (43) 25 (28) Deceased 24 (57) 63 (72) NS KEY: Age 5 age at diagnosis of cancer; SD 5 standard deviation; NS 5 not significant; SCI 5 spinal cord injury; TURBT 5 transurethral resection of bladder tumor. Numbers in parentheses are percentages. * Comparison between the study and nonstudy groups; only age and level of SCI are significantly different. † Survival data tabulated as of September 30, 1997.
test or chi-square test. A Kaplan-Meier curve was constructed to depict overall survival of the 130 patients studied as of September 30, 1997.
RESULTS A total of 83,241 inpatient admissions for 33,565 patients with SCI were recorded in the DVA data base for fiscal years 1988 to 1992. A computerized search of the Patient Treatment Files identified 130 patients (129 men, 1 woman) with a diagnosis of SCI who later underwent treatment for bladder cancer (0.39%). Of the 159 DVA facilities, 57 (36%) reported having at least 1 patient with the simultaneous diagnoses of SCI and bladder cancer. Specific information regarding inpatient treatment of these patients was available from 25 (44%) of 57 hospitals. These institutions graciously provided records for 66 (50%) of the 130 patients identified by computer search. Of these, 42 (64%) had adequate data available regarding the method of bladder management and tumor histologic findings to be considered evaluable. These constitute the study population. Table I summarizes data regarding age, race, level of paralysis, surgical procedure, and survival for all 130 patients as reported in the computerized Patient Treatment Files and Beneficiary Identification and Records Location System. Overall, most patients were paraplegic and underwent transurethral resection of their bladder tumors. Patient characteristics of the study group were similar to those in the nonstudy group. Only age at diagnosis 293
adenocarcinomas in 5 (12%). The IND group had 12 (46%) of 26 patients with TCC, 11 patients (42%) with SCC, and 3 patients (12%) with adenocarcinoma of the bladder. By contrast, 11 (68%) of 16 patients in the decatheterized group had TCC, 3 (19%) had SCC, 1 (6%) had adenocarcinoma, and 1 (6%) had mixed tumor on pathologic examination. COMMENT
FIGURE 1. Kaplan-Meier curve depicting overall survival of 130 patients with SCI and bladder cancer dating from time of treatment.
of bladder cancer and level of SCI were significantly different. Two patients (1.5%) died within 30 days of their operative procedure. Figure 1 displays a Kaplan-Meier overall survival curve for all 130 patients in the study, with 1- and 5-year survival rates of 61% (79 of 130) and 38% (50 of 130), respectively. Mean follow-up duration was 5.5 years. Overall, the predominant method of bladder management in the study group (n 5 42) was either permanent suprapubic cystostomy tube (n 5 8, 19%) or chronic indwelling urethral catheter (n 5 18, 43%). These patients were classified as a single group (indwelling catheter group [IND]) representing 26 (62%) of the 42 patients evaluated. Bladder management in the remaining decatheterized (16 of 42) patients was by clean intermittent catheterization in 8 (19%), condom catheterization in 6 (14%), and spontaneous voiding in 2 (5%). Table II displays data regarding age, duration of SCI, duration of catheterization, and type of invasive procedure used to treat bladder cancer in the two groups. Average patient age at diagnosis of cancer for the study group was 57.3 years; mean time from SCI and mean duration of catheterization were 23.9 and 20.0 years, respectively. There was no significant difference in mean age at diagnosis or duration of catheterization among the two groups (IND versus decatheterized). However, patients in the decatheterized group were diagnosed with bladder cancer approximately 18.1 years after SCI versus 26.5 years for the IND group (P 5 0.054). Evaluation of tumor type, stratified against method of bladder management, is presented in Table III. Most tumors (23 of 42, 55%) were transitional cell carcinomas (TCC), and squamous cell carcinomas (SCC) were found in 14 (33%) and 294
This study represents a retrospective analysis of patients with known SCI and a documented method of bladder management who subsequently developed bladder carcinoma requiring invasive treatment. Previously published studies2,4 –7 (Table IV) of such patients have been hampered by the small number of patients identified with both SCI and bladder cancer. The DVA institutions across the country recorded data on 33,565 patients with SCI during the fiscal years 1988 through 1992, of whom 130 had bladder cancer. In 42, adequate information regarding duration of SCI, duration and method of bladder catheterization immediately before surgery, and therapy for bladder cancer was available for analysis, making this the largest series thus far reported. Previous reports have indicated an increased incidence of bladder cancer, ranging from 2.5% to 10%, among the population with SCI. The low incidence in our analysis may reflect an improvement in urologic treatment for these patients, with earlier and more aggressive treatment of the complications of chronic catheterization or the non-referral-based population evaluated. This also presumably underestimates the incidence because only a 5-year period was assessed. Bickel et al.5 reported an incidence of 0.32% in a study of 2900 patients with SCI and attribute this decreased incidence to improved lower tract management in these patients. Although improved detection and treatment of urinary tract infections, bladder calculi, and other complications of chronic catheterization may contribute to a declining incidence of bladder malignancy in this population, other factors may play a role as well. These include errors in diagnostic coding and the possible treatment of Veterans Affairs’ patients in non-VA hospitals. The mean age at diagnosis was 57.3 years (range 36 to 84), which is approximately 10 years sooner than that expected in the general population.8 Although TCC was the predominant histologic finding (55%) identified in this study, it has been previously reported that SCC is most common.2,4,9 Several possible explanations exist for this finding. First, like all retrospective analyses, the study is potentially biased. We report on the histopathologic findings from 42 patients (32%) of the 130 UROLOGY 53 (2), 1999
TABLE II.
Study group by method of bladder management
Age (yr) Years from injury Years catheterized† Surgical intervention (n) Radical cystectomy TURBT
Total (n 5 42)
IND (n 5 26)
Decatheterized (n 5 16)
57.3 6 11.5 23.9 6 11.1 20.0 6 11.5
56.7 6 10.7 26.5 6 10.1 23.0 6 11.4
58.3 6 12.9 18.1 6 11.6* 14.5 6 10.1
28 (66) 14 (33)
20 (77) 6 (23)
8 (50) 8 (50)
KEY: IND 5 patients with an indwelling Foley catheter; Decatheterized 5 patients without an indwelling catheter; TURBT 5 transurethral resection of bladder tumor. Data presented as mean 6 SD, unless otherwise noted. Numbers in parentheses are percentages. *P 5 0.054 vs. IND. † Number of years before surgery that patients were using a given form of bladder catheterization.
TABLE III.
Histopathologic findings of bladder cancer versus method of bladder catheterization Total (n 5 42)
Transitional cell carcinoma Squamous cell carcinoma Adenocarcinoma Mixed
23 14 4 1
IND (n 5 26)
(55) (33) (9) (2)
12 (46) 11 (42) 3 (12) —
Decatheterized (n 5 16) 11 3 1 1
(68) (19) (6) (6)
Abbreviations as in Table II. No significant differences between groups were identified (P 5 0.3). Numbers in parentheses are percentages.
TABLE IV.
Reported incidence of bladder cancer among patients with spinal cord injury
Study Current study Bickel et al.,5 1991 Broecker et al.,7 1981 Bejany et al.,4 1987 Locke et al.,6 1985 Kaufman et al.,2 1977
Pts. with SCI Surveyed (n) 33,565 2,900 1,052 300 25 62
Pts. with CA (n) 131 8 10 11 2 6
(0.39) (0.28) (0.95) (3.6) (8) (10)
SCC (n) 14/42 2/8 2/10 9/11 2/2 6/6
(33) (25) (20) (82) (100) (100)
KEY: Pts. 5 patients; SCI 5 spinal cord injury; CA 5 cancer; SCC 5 squamous cell carcinoma. Numbers in parentheses are percentages.
identified by computer search. Unfortunately, pathologic findings are not recorded in the Patient Treatment Files and can only be obtained from chart review. Therefore, we cannot rule out the possibility that most of the remaining patients had SCC of the bladder. Second, the study population itself (U.S. veterans) may have an inherent bias toward the development of TCC, because the prevalence of smoking is higher in veterans than in the population at large, and cigarette smokers are known to have up to a fourfold higher incidence of bladder cancer,10 –12 predominantly of the TCC type. Although exact smoking histories were not routinely recorded in the charts reviewed, the incidence of smoking in veterans with SCI is about UROLOGY 53 (2), 1999
80% compared with approximately 50% in the general veteran population.13 Finally, the age at diagnosis in the study group was significantly younger than that for the remaining patients. The average age at diagnosis in the 88 patients not included in the study was 62.7 6 11.0. The possibility exists that patients included in the analysis were diagnosed earlier in the natural progression of their disease, before the “dedifferentiation” or “metaplasia” of their TCC to either SCC or adenocarcinoma. We report a nearly equal incidence of TCC and SCC in patients with indwelling catheters, whether transurethral catheters or chronic suprapubic tubes. The mean duration of SCI in this subset of patients was 23.9 years, which is consistent with 295
previous reports, indicating that most tumors are diagnosed after at least 10 years of catheterization for SCI.2 Interestingly, of the 12 patients in the IND group with a final diagnosis of TCC, 8 (66%) had elements of squamous metaplasia without SCC in the final pathology specimen. Of the patients in the IND group with a diagnosis of SCC, 6 (55%) of 11 also had squamous metaplasia. The significance of urothelial metaplasia in this setting is unknown. However, Sakamoto et al.14 analyzed 31 cases of bladder carcinoma, including 3 cases of “pure” SCC and 28 cases of TCC associated with squamous metaplasia. They proposed two forms of histogenesis for SCC of the bladder: (a) malignant transformation on the basis of squamous metaplasia alone, and (b) extensive differentiation toward a squamous form in a pre-existing TCC. According to this concept, perhaps SCC in patients with SCI often arise by metaplastic change in TCC. Certainly, delays in diagnosis of TCC in patients with SCI with continued exposure to inflammatory processes would be consistent with this hypothesis. Additional pathologic analysis is required to establish the pattern of tumorigenesis in this unique cohort of patients. In the decatheterized group, two interesting findings were noted. First, only 3 of the patients had a diagnosis of SCC (1 using clean intermittent catheterization and 2 using condom catheterization), but TCC was found in 11 of 16 patients. We cannot conclude from the available data that patients with decatheterization are less likely to develop a bladder malignancy than patients with indwelling catheters. However, it appears that fewer of these patients will develop SCC of the bladder. One theory behind the development of bladder malignancies is malignant transformation of the urothelium secondary to chronic irritation and infection associated with indwelling catheters. Previous histologic studies show an eosinophilic inflammatory response with urothelial dysplasia and squamous metaplasia at the site of the catheter.15 This would suggest that decatheterization may reduce the number of infections, as well as the degree of mucosal irritation, and result in less squamous metaplasia, ultimately leading to fewer SCC. Second, it appears that patients using a decatheterized approach develop bladder cancer sooner (approximately 8 years in the current study) than those with indwelling catheters (albeit 18 years after their SCI). Is it possible that patients who have the capacity and/or the opportunity to store urine develop bladder cancer sooner than patients with continuous bladder drainage? Environmental carcinogens in urine probably play a role in the development of some bladder cancers. Perhaps noncarcinogenic substances are converted to more active carcinogenic factors when stored in the bladder, or 296
possibly the increased surface area and prolonged exposure of the urothelium to these activated carcinogens hastens the development of bladder cancer. Although the number of patients reported in this study is large by comparison with other studies, the absolute numbers are still rather small. Clearly, further studies, ideally with prospective data, are necessary to evaluate these interesting findings. More recently, focus has centered on appropriate screening for patients with SCI and chronic catheterization in an attempt to facilitate early detection of these malignancies. Clearly, cystoscopy of patients with recurrent urinary tract infections and/or gross hematuria, with biopsy of any suspicious lesions, is warranted. Perhaps yearly cystoscopy after 10 years of catheterization, as advocated recently in published reports, would be a reasonable approach to early diagnosis.16,17 CONCLUSIONS We report a 5-year risk of approximately 0.39% for the development of bladder cancer in a large population of U.S. veterans with SCI. Evaluation of a subset of this population revealed a surprisingly large percentage of TCC of the bladder. Patients with SCI and bladder cancer who used for the most part a decatheterized approach had relatively fewer SCC than those with indwelling urethral or suprapubic catheters. These data lend further support to the avoidance of indwelling catheters in patients with SCI. REFERENCES 1. Bedbrook GM, and Sedgley GI: The management of spinal injuries past and present. Int Rehabil Med 2: 45– 61, 1980. 2. Kaufman JM, Fam B, Jacobs SC, et al: Bladder cancer and squamous metaplasia in spinal cord injury patients. J Urol 118: 967–971, 1977. 3. Jacobs SC, and Kaufman JM: Complications of permanent catheter drainage in spinal cord injury patients. J Urol 119: 740 –741, 1978. 4. Bejany DE, Lockhart JL, and Rhamy RK: Malignant vesical tumors following spinal cord injury. J Urol 138: 1390 – 1392, 1987. 5. Bickel A, Culkin DJ, and Wheeler JS: Bladder cancer in spinal cord injury patients. J Urol 140: 1240 –1242, 1991. 6. Locke JR, Hill DE, and Walzer Y: Incidence of squamous cell carcinoma in patients with long-term catheter drainage. J Urol 133: 1034 –1035, 1985. 7. Broecker FH, Klein FA, and Hackler RH: Cancer of the bladder in spinal cord injury patients. J Urol 125: 196 –197, 1981. 8. 1987 Annual Cancer Statistics Review: Including Cancer Trends: 1950 –1985, NIH Publication No. 88-2789. Bethesda, Maryland, Department of Health and Human Services, National Cancer Institute. 9. Melzak J: The incidence of bladder cancer in paraplegia. Paraplegia 4: 85–96, 1967. 10. Burch JK, Rohan TE, Howe GR, et al: Risk of bladder UROLOGY 53 (2), 1999
cancer by source and type of tobacco exposure: a case-control study. Int J Cancer 44: 622– 628, 1989. 11. Clavel J, Cordier S, Boccon-Gibod L, et al: Tobacco and bladder cancer in males: increased risk for inhalers and smokers of black tobacco. Int J Cancer 44: 605– 610, 1989. 12. Morrison AS, Buring JE, Verhoek WG, et al: An international study of smoking and bladder cancer. J Urol 131: 650 – 654, 1984. 13. Spungen AM, Lesser M, Almenoff PL, et al: Prevalence of cigarette smoking in a group of male veterans with chronic spinal cord injury. Military Med 160: 308 –311, 1995. 14. Sakamoto N, Tsuneyoshi M, and Enjoji M: Urinary
UROLOGY 53 (2), 1999
bladder carcinoma with a neoplastic squamous component: a mapping study of 31 cases. Histopathology 21: 135–141, 1992. 15. Goble NM, Clarke JJ, and Hammonds JC: Histological changes in the urinary bladder secondary to urethral catheterization. Br J Urol 63: 354 –357, 1989. 16. Navon JD, Hani S, Khonsari F, et al: Screening cystoscopy and survival of spinal cord injured patients with squamous cell cancer of the bladder. J Urol 157: 2109 –2111, 1997. 17. Stonehill WH, Goldman HB, and Dmochowski RR: The use of urine cytology for diagnosing bladder cancer in spinal cord injured patients. J Urol 157: 2112–2114, 1997.
297