Bladder Cancer Mortality after Spinal Cord Injury over 4 Decades Laura S. Nahm, Yuying Chen,* Michael J. DeVivo and L. Keith Lloyd From the Department of Physical Medicine and Rehabilitation, Stanford University, Stanford, California (LSN), Department of Physical Medicine and Rehabilitation (YC, MJD), and Department of Urology (LKL), University of Alabama at Birmingham, Birmingham, Alabama
Purpose: We estimate bladder cancer mortality in people with spinal cord injury compared to the general population. Materials and Methods: Data and statistics were retrieved from the National Spinal Cord Injury Statistical Center and the National Center for Health Statistics. The mortality experience of the 45,486 patients with traumatic spinal cord injury treated at a Spinal Cord Injury Model System or Shriners Hospital was compared to the general population using a standardized mortality ratio. The standardized mortality ratio data were further stratified by age, gender, race, time since injury and injury severity. Results: Our study included 566,532 person-years of followup between 1960 and 2009, identified 10,575 deaths and categorized 99 deaths from bladder cancer. The expected number of deaths from bladder cancer would have been 14.8 if patients with spinal cord injury had the same bladder cancer mortality as the general population. Thus, the standardized mortality ratio is 6.7 (95% CI 5.4e8.1). Increased mortality risk from bladder cancer was observed for various ages, races and genders, as well as for those injured for 10 or more years and with motor complete injuries. Bladder cancer mortality was not significantly increased for ventilator users, those with motor incomplete injuries or those injured less than 10 years. Conclusions: Individuals with a spinal cord injury can potentially live healthier and longer by reducing the incidence and mortality of bladder cancer. Study findings highlight the need to identify at risk groups and contributing factors for bladder cancer death, leading to the development of prevention, screening and management strategies. Key Words: spinal cord injuries, urinary bladder neoplasms, mortality
SPINAL cord injury is associated with significant secondary medical and psychosocial conditions throughout life. Renal failure used to be the leading cause of death among patients with SCI,1,2 but advances in neurogenic bladder management in the last 5 decades have made genitourinary system diseases a more common cause of morbidity rather than mortality.2,3
Now cancer ranks among the top 3 causes of death.2 Bladder cancer specifically is the third leading cause of cancer death in this population, while it is ranked tenth in the general population.4 The current estimate of bladder cancer incidence, calculated as the number of new bladder cancer cases identified during the study period
0022-5347/15/1936-1923/0 THE JOURNAL OF UROLOGY® © 2015 by AMERICAN UROLOGICAL ASSOCIATION EDUCATION AND RESEARCH, INC.
http://dx.doi.org/10.1016/j.juro.2015.01.070 Vol. 193, 1923-1928, June 2015 Printed in U.S.A.
Abbreviations and Acronyms AIS ¼ American Spinal Injury Association Impairment Scale NDI ¼ National Death Index NSCID ¼ National Spinal Cord Injury Database NSSCID ¼ National Shriners Spinal Cord Injury Database SCI ¼ spinal cord injury SCIMS ¼ Spinal Cord Injury Model Systems SMR ¼ standardized mortality ratio SSDI ¼ Social Security Death Index Accepted for publication January 14, 2015. Study received institutional review board approval. Nothing to disclose. Supported by the National Institute on Disability and Rehabilitation Research (Grant H133A110002), U.S. Department of Education, Paralyzed Veterans of America Research Foundation and South Carolina Spinal Cord Injury Research Fund. * Correspondence: 515 Spain Rehabilitation Center, 1717 6th Ave. South, Birmingham, Alabama 35249-7730 (telephone: 205-934-3320; FAX: 205-934-2709; e-mail: [email protected]).
See Editorial on page 1880. Editor’s Note: This article is the second of 5 published in this issue for which category 1 CME credits can be earned. Instructions for obtaining credits are given with the questions on pages 2160 and 2161.
www.jurology.com
j
1923
1924
BLADDER CANCER AFTER SPINAL CORD INJURY
divided by the total number of SCI cases, is highly variable (0.1% to 10.0%).5 However, the mortality risk of bladder cancer has not been consistently addressed in the SCI literature. Groah et al calculated a standardized mortality ratio based on 12 bladder cancer deaths in a study of 3,670 patients with SCI, resulting in a 70 times increased mortality risk compared to the general population.6 However, given the small cohort, this SMR estimate has a wide 95% CI (36.9e123.3), and the estimate was not specifically tailored to individual age, gender and racial groups. Thus, in this study we provide a better estimate of increased bladder cancer mortality, using a large cohort of 45,486 individuals with traumatic SCI. In addition, we assess whether increased bladder cancer mortality varies by demographic and injury characteristics. Understanding this potentially fatal complication after SCI will guide the direction of future research that targets the prevention, screening and treatment of bladder cancer.
MATERIALS AND METHODS Data Sources and Study Population Participants and SCI data for this study were retrieved from 3 data sources. Institutional review board approval was obtained from the National SCI Statistical Center as well as locally at each participating center. The National SCI Database was established in 1975.7 Data were collected retrospectively from 1973 and prospectively since 1975. Since its inception 28 federally funded SCIMS throughout the United States have contributed data to the database. To be qualified for the NSCID, patients must have 1) sustained SCI due to a traumatic event, 2) had a clinically discernible degree of neurological deficit and 3) received initial hospital care from one of the SCIMS within 1 year of injury. To increase sample size and to supplement mortality information contained in the NSCID, a series of collaborative survival studies were conducted during 1973 to 1999,8 recruiting additional participants from several SCIMS who were not registered in the NSCID. While approximately 92% of participants were injured after 1970, there are participants injured as long ago as 1936 and admitted to the SCIMS as early as 1960. A parallel NSSCID that enrolls children with SCI who received care from the 3 SCI units of the Shriners Hospital for Children in California, Illinois and Pennsylvania was established in 1987. Data have been collected prospectively since 1987 using the same protocol as the NSCID,9 except that the NSSCID includes patients admitted to the system beyond 1 year of injury. A total of 45,486 individuals were eligible for this study, including 38,205 from the NSCID, 5,199 from the collaborative survival study and 2,122 from the NSSCID (table 1).
Demographic and Injury Profile Demographic and injury characteristics were collected by trained personnel during initial hospital care. Neurological
Table 1. Characteristics of study participants Total Participants Bladder Ca Deaths At baseline No. gender (%): M 36,195 (79.6) F 9,291 (20.4) No. race (%): NonHispanic white 30,498 (67.0) NonHispanic black 8,916 (19.6) Hispanic 4,457 (9.8) Asian 727 (1.6) Native American 399 (0.9) Other 248 (0.5) Unknown 241 (0.5) Mean age at injury (SD) 32.8 (16.8) No. age at injury (%): 0e29 24,321 (53.5) 30e59 17,043 (37.5) 60 or Greater 4,122 (9.1) No. calendar yr of injury (%): 1936e1970 327 (0.7) 1971e1980 8,275 (18.2) 1981e1990 12,756 (28.0) 1991e2000 14,086 (31.0) 2001e2009 10,042 (22.1) No. injury severity (%): Ventilator dependent 1,096 (2.4) C1eC4, AIS ABC 5,403 (11.9) C5eC8, AIS ABC 9,360 (20.6) T1eS3, AIS ABC 14,968 (32.9) All levels, AIS D 10,379 (22.8) Unknown 4,280 (9.4) At last followup deceased or withdrew alive Mean age (SD) 44.5 (16.7) No. current age (%): 0e29 9,662 (21.2) 30e59 27,475 (60.4) 60 or Greater 8,349 (18.4) Mean yrs since injury (SD) 12.7 (10.4) No. yrs since injury (%): Less than 1 8,040 (17.7) 1e9 12,941 (28.5) 10e19 12,216 (26.9) 20 or Greater 12,289 (27.0) No. calendar yr (%): 1973e1980 743 (1.6) 1981e1990 4,593 (10.1) 1991e2000 6,418 (14.1) 2001e2009 33,732 (73.2) No. bladder management (Form I participants only) (%):* Normal 5,814 (25.8) Catheter-free 3,196 (14.2) Intermittent catheterization 6,336 (28.1) Indwelling/suprapubic catheterization 5,802 (25.8) Other 1,378 (6.1)
78 (78.8) 21 (21.2) 80 10 6 0 1 0 2 29.5
(80.8) (10.1) (6.1) (0.0) (1.0) (0.0) (2.0) (14.2)
62 (62.6) 31 (31.3) 6 (6.1) 9 (9.1) 66 (66.7) 21 (21.2) 2 (2.0) 1 (1.0) 0 (0.0) 7 (7.1) 32 (32.3) 49 (49.5) 8 (8.1) 3 (3.0) 52.5 (12.3) 1 71 27 23.9
(1.0) (71.7) (27.3) (8.5)
0 (0.0) 7 (7.1) 18 (18.2) 74 (74.7) 1 (1.0) 7 (7.1) 37 (37.4) 54 (54.5) 3 (7.1) 13 (31.0) 6 (14.3) 18 (42.9) 2 (4.7)
* Data on 22,526 participants and 42 deaths from bladder cancer.
examinations were performed in accordance with the International Standards for Neurological Classification of SCI.10 Ventilator dependency was defined as requiring partial or total respiratory support on a daily basis. For analysis, ventilator-free participants were further grouped into 1 of 4 neurological categories according to the AIS or Frankel’s classification scale of 1) high tetraplegia with motor functionally complete injury, AIS/Frankel A, B or C (C1-C4 ABC); 2) low tetraplegia with motor functionally complete injury (C5-C8 ABC); 3) paraplegia with motor functionally complete injury (T1-S3 ABC); and 4) motor
BLADDER CANCER AFTER SPINAL CORD INJURY
1925
functionally incomplete injury, AIS/Frankel D (AIS D), regardless of injury level.
Vital Status and Cause of Death Deaths were identified from routine followup conducted by data collection personnel at each hospital or periodic searches of the SSDI and other online databases. The most recent search was conducted in early 2013. The SSDI was established to be 92.4% sensitive and 99.5% specific in identifying survival status for individuals in the NSCID.11 Cause of death information was obtained from a combination of hospital records, death certificates, autopsy reports and NDI searches. The most recent NDI search was conducted in fall 2011 for deaths that occurred up to 2009. The guidelines used to determine the underlying cause of death when multiple causes have been reported are consistent with those used by the National Center for Health Statistics to create annual cause specific mortality rates for the United States, with the exception that any reference to SCI or the event that caused the injury is ignored (http://www.cdc.gov/nchs/data/dvs/2a_2014.pdf). Cause of death was considered unknown whenever SCI or the cause of the SCI was the only listed cause of death.
Statistical Analysis For analytic purposes participant followup terminated on December 31, 2009. Only deaths occurring before that date were considered. To avoid the bias caused by delayed admissions and selective survivors, the person-years of followup for each participant were counted from the date of admission to a SCIMS or Shriner Hospital to the date of death (for deceased participants) or the date last known to be alive (for “withdrew alive” participants). The person-years of followup were further stratified by participant characteristics and used to calculate expected number of deaths by applying the age-gender-race specific mortality rate of bladder cancer reported for 1997 by the National Center for Health Statistics.12 We chose the year 1997 because it was roughly the midpoint of the total person-years of followup for the study population. The number of bladder cancer deaths observed for each demographic and injury group was then divided by the expected number of bladder cancer deaths to compute the SMR. A SMR of 1.0 implies no increase in bladder cancer mortality. The corresponding 95% CI measures the precision of the SMR estimates and suggests statistical significance if 1.0 was not included in the intervals. All statistical analyses were performed by one of us (MJD) using SASÒ version 9.3 and an online website for SMR calculation (http://www.quantitativeskills.com/sisa/ statistics/smr.htm).
RESULTS From the 45,486 study participants 566,532 personyears of followup were calculated (average followup 12 years, range less than 1 to 49) and 10,575 deaths were identified, of which 9,884 (93.5%) had identifiable causes of death (see figure). Bladder cancer was reported as the cause of death in 99 cases. Demographic and injury characteristics of these
Recruitment of study participants and identification of bladder cancer deaths.
99 participants are presented in table 1. The bladder management method at recent followup was known for 22,526 individuals enrolled in the NSCID Form I data set and was also reported in table 1. Table 2 presents the SMR of bladder cancer of the SCI participants compared with the general population of comparable age, gender and race. Overall, had patients with SCI had the same bladder cancer mortality as the general population, the expected number of deaths from bladder cancer among the 45,486 study participants would have been 14.8. Given 99 actual deaths the SMR was calculated to be 6.7. The increased bladder cancer mortality was significant (p
Purpose: We estimate bladder cancer mortality in people with spinal cord injury compared to the general population. Materials and Methods: Data and statistics were retrieved from the National Spinal Cord Injury Statistical Center and the National Center for Health Statistics. The mortality experience of the 45,486 patients with traumatic spinal cord injury treated at a Spinal Cord Injury Model System or Shriners Hospital was compared to the general population using a standardized mortality ratio. The standardized mortality ratio data were further stratified by age, gender, race, time since injury and injury severity. Results: Our study included 566,532 person-years of followup between 1960 and 2009, identified 10,575 deaths and categorized 99 deaths from bladder cancer. The expected number of deaths from bladder cancer would have been 14.8 if patients with spinal cord injury had the same bladder cancer mortality as the general population. Thus, the standardized mortality ratio is 6.7 (95% CI 5.4e8.1). Increased mortality risk from bladder cancer was observed for various ages, races and genders, as well as for those injured for 10 or more years and with motor complete injuries. Bladder cancer mortality was not significantly increased for ventilator users, those with motor incomplete injuries or those injured less than 10 years. Conclusions: Individuals with a spinal cord injury can potentially live healthier and longer by reducing the incidence and mortality of bladder cancer. Study findings highlight the need to identify at risk groups and contributing factors for bladder cancer death, leading to the development of prevention, screening and management strategies. Key Words: spinal cord injuries, urinary bladder neoplasms, mortality
SPINAL cord injury is associated with significant secondary medical and psychosocial conditions throughout life. Renal failure used to be the leading cause of death among patients with SCI,1,2 but advances in neurogenic bladder management in the last 5 decades have made genitourinary system diseases a more common cause of morbidity rather than mortality.2,3
Now cancer ranks among the top 3 causes of death.2 Bladder cancer specifically is the third leading cause of cancer death in this population, while it is ranked tenth in the general population.4 The current estimate of bladder cancer incidence, calculated as the number of new bladder cancer cases identified during the study period
0022-5347/15/1936-1923/0 THE JOURNAL OF UROLOGY® © 2015 by AMERICAN UROLOGICAL ASSOCIATION EDUCATION AND RESEARCH, INC.
http://dx.doi.org/10.1016/j.juro.2015.01.070 Vol. 193, 1923-1928, June 2015 Printed in U.S.A.
Abbreviations and Acronyms AIS ¼ American Spinal Injury Association Impairment Scale NDI ¼ National Death Index NSCID ¼ National Spinal Cord Injury Database NSSCID ¼ National Shriners Spinal Cord Injury Database SCI ¼ spinal cord injury SCIMS ¼ Spinal Cord Injury Model Systems SMR ¼ standardized mortality ratio SSDI ¼ Social Security Death Index Accepted for publication January 14, 2015. Study received institutional review board approval. Nothing to disclose. Supported by the National Institute on Disability and Rehabilitation Research (Grant H133A110002), U.S. Department of Education, Paralyzed Veterans of America Research Foundation and South Carolina Spinal Cord Injury Research Fund. * Correspondence: 515 Spain Rehabilitation Center, 1717 6th Ave. South, Birmingham, Alabama 35249-7730 (telephone: 205-934-3320; FAX: 205-934-2709; e-mail: [email protected]).
See Editorial on page 1880. Editor’s Note: This article is the second of 5 published in this issue for which category 1 CME credits can be earned. Instructions for obtaining credits are given with the questions on pages 2160 and 2161.
www.jurology.com
j
1923
1924
BLADDER CANCER AFTER SPINAL CORD INJURY
divided by the total number of SCI cases, is highly variable (0.1% to 10.0%).5 However, the mortality risk of bladder cancer has not been consistently addressed in the SCI literature. Groah et al calculated a standardized mortality ratio based on 12 bladder cancer deaths in a study of 3,670 patients with SCI, resulting in a 70 times increased mortality risk compared to the general population.6 However, given the small cohort, this SMR estimate has a wide 95% CI (36.9e123.3), and the estimate was not specifically tailored to individual age, gender and racial groups. Thus, in this study we provide a better estimate of increased bladder cancer mortality, using a large cohort of 45,486 individuals with traumatic SCI. In addition, we assess whether increased bladder cancer mortality varies by demographic and injury characteristics. Understanding this potentially fatal complication after SCI will guide the direction of future research that targets the prevention, screening and treatment of bladder cancer.
MATERIALS AND METHODS Data Sources and Study Population Participants and SCI data for this study were retrieved from 3 data sources. Institutional review board approval was obtained from the National SCI Statistical Center as well as locally at each participating center. The National SCI Database was established in 1975.7 Data were collected retrospectively from 1973 and prospectively since 1975. Since its inception 28 federally funded SCIMS throughout the United States have contributed data to the database. To be qualified for the NSCID, patients must have 1) sustained SCI due to a traumatic event, 2) had a clinically discernible degree of neurological deficit and 3) received initial hospital care from one of the SCIMS within 1 year of injury. To increase sample size and to supplement mortality information contained in the NSCID, a series of collaborative survival studies were conducted during 1973 to 1999,8 recruiting additional participants from several SCIMS who were not registered in the NSCID. While approximately 92% of participants were injured after 1970, there are participants injured as long ago as 1936 and admitted to the SCIMS as early as 1960. A parallel NSSCID that enrolls children with SCI who received care from the 3 SCI units of the Shriners Hospital for Children in California, Illinois and Pennsylvania was established in 1987. Data have been collected prospectively since 1987 using the same protocol as the NSCID,9 except that the NSSCID includes patients admitted to the system beyond 1 year of injury. A total of 45,486 individuals were eligible for this study, including 38,205 from the NSCID, 5,199 from the collaborative survival study and 2,122 from the NSSCID (table 1).
Demographic and Injury Profile Demographic and injury characteristics were collected by trained personnel during initial hospital care. Neurological
Table 1. Characteristics of study participants Total Participants Bladder Ca Deaths At baseline No. gender (%): M 36,195 (79.6) F 9,291 (20.4) No. race (%): NonHispanic white 30,498 (67.0) NonHispanic black 8,916 (19.6) Hispanic 4,457 (9.8) Asian 727 (1.6) Native American 399 (0.9) Other 248 (0.5) Unknown 241 (0.5) Mean age at injury (SD) 32.8 (16.8) No. age at injury (%): 0e29 24,321 (53.5) 30e59 17,043 (37.5) 60 or Greater 4,122 (9.1) No. calendar yr of injury (%): 1936e1970 327 (0.7) 1971e1980 8,275 (18.2) 1981e1990 12,756 (28.0) 1991e2000 14,086 (31.0) 2001e2009 10,042 (22.1) No. injury severity (%): Ventilator dependent 1,096 (2.4) C1eC4, AIS ABC 5,403 (11.9) C5eC8, AIS ABC 9,360 (20.6) T1eS3, AIS ABC 14,968 (32.9) All levels, AIS D 10,379 (22.8) Unknown 4,280 (9.4) At last followup deceased or withdrew alive Mean age (SD) 44.5 (16.7) No. current age (%): 0e29 9,662 (21.2) 30e59 27,475 (60.4) 60 or Greater 8,349 (18.4) Mean yrs since injury (SD) 12.7 (10.4) No. yrs since injury (%): Less than 1 8,040 (17.7) 1e9 12,941 (28.5) 10e19 12,216 (26.9) 20 or Greater 12,289 (27.0) No. calendar yr (%): 1973e1980 743 (1.6) 1981e1990 4,593 (10.1) 1991e2000 6,418 (14.1) 2001e2009 33,732 (73.2) No. bladder management (Form I participants only) (%):* Normal 5,814 (25.8) Catheter-free 3,196 (14.2) Intermittent catheterization 6,336 (28.1) Indwelling/suprapubic catheterization 5,802 (25.8) Other 1,378 (6.1)
78 (78.8) 21 (21.2) 80 10 6 0 1 0 2 29.5
(80.8) (10.1) (6.1) (0.0) (1.0) (0.0) (2.0) (14.2)
62 (62.6) 31 (31.3) 6 (6.1) 9 (9.1) 66 (66.7) 21 (21.2) 2 (2.0) 1 (1.0) 0 (0.0) 7 (7.1) 32 (32.3) 49 (49.5) 8 (8.1) 3 (3.0) 52.5 (12.3) 1 71 27 23.9
(1.0) (71.7) (27.3) (8.5)
0 (0.0) 7 (7.1) 18 (18.2) 74 (74.7) 1 (1.0) 7 (7.1) 37 (37.4) 54 (54.5) 3 (7.1) 13 (31.0) 6 (14.3) 18 (42.9) 2 (4.7)
* Data on 22,526 participants and 42 deaths from bladder cancer.
examinations were performed in accordance with the International Standards for Neurological Classification of SCI.10 Ventilator dependency was defined as requiring partial or total respiratory support on a daily basis. For analysis, ventilator-free participants were further grouped into 1 of 4 neurological categories according to the AIS or Frankel’s classification scale of 1) high tetraplegia with motor functionally complete injury, AIS/Frankel A, B or C (C1-C4 ABC); 2) low tetraplegia with motor functionally complete injury (C5-C8 ABC); 3) paraplegia with motor functionally complete injury (T1-S3 ABC); and 4) motor
BLADDER CANCER AFTER SPINAL CORD INJURY
1925
functionally incomplete injury, AIS/Frankel D (AIS D), regardless of injury level.
Vital Status and Cause of Death Deaths were identified from routine followup conducted by data collection personnel at each hospital or periodic searches of the SSDI and other online databases. The most recent search was conducted in early 2013. The SSDI was established to be 92.4% sensitive and 99.5% specific in identifying survival status for individuals in the NSCID.11 Cause of death information was obtained from a combination of hospital records, death certificates, autopsy reports and NDI searches. The most recent NDI search was conducted in fall 2011 for deaths that occurred up to 2009. The guidelines used to determine the underlying cause of death when multiple causes have been reported are consistent with those used by the National Center for Health Statistics to create annual cause specific mortality rates for the United States, with the exception that any reference to SCI or the event that caused the injury is ignored (http://www.cdc.gov/nchs/data/dvs/2a_2014.pdf). Cause of death was considered unknown whenever SCI or the cause of the SCI was the only listed cause of death.
Statistical Analysis For analytic purposes participant followup terminated on December 31, 2009. Only deaths occurring before that date were considered. To avoid the bias caused by delayed admissions and selective survivors, the person-years of followup for each participant were counted from the date of admission to a SCIMS or Shriner Hospital to the date of death (for deceased participants) or the date last known to be alive (for “withdrew alive” participants). The person-years of followup were further stratified by participant characteristics and used to calculate expected number of deaths by applying the age-gender-race specific mortality rate of bladder cancer reported for 1997 by the National Center for Health Statistics.12 We chose the year 1997 because it was roughly the midpoint of the total person-years of followup for the study population. The number of bladder cancer deaths observed for each demographic and injury group was then divided by the expected number of bladder cancer deaths to compute the SMR. A SMR of 1.0 implies no increase in bladder cancer mortality. The corresponding 95% CI measures the precision of the SMR estimates and suggests statistical significance if 1.0 was not included in the intervals. All statistical analyses were performed by one of us (MJD) using SASÒ version 9.3 and an online website for SMR calculation (http://www.quantitativeskills.com/sisa/ statistics/smr.htm).
RESULTS From the 45,486 study participants 566,532 personyears of followup were calculated (average followup 12 years, range less than 1 to 49) and 10,575 deaths were identified, of which 9,884 (93.5%) had identifiable causes of death (see figure). Bladder cancer was reported as the cause of death in 99 cases. Demographic and injury characteristics of these
Recruitment of study participants and identification of bladder cancer deaths.
99 participants are presented in table 1. The bladder management method at recent followup was known for 22,526 individuals enrolled in the NSCID Form I data set and was also reported in table 1. Table 2 presents the SMR of bladder cancer of the SCI participants compared with the general population of comparable age, gender and race. Overall, had patients with SCI had the same bladder cancer mortality as the general population, the expected number of deaths from bladder cancer among the 45,486 study participants would have been 14.8. Given 99 actual deaths the SMR was calculated to be 6.7. The increased bladder cancer mortality was significant (p
