Original Investigation Gout After Living Kidney Donation: A Matched Cohort Study Ngan N. Lam, MD,1,2 Eric McArthur, MSc,3 S. Joseph Kim, MD, PhD,3,4 G.V. Ramesh Prasad, MD, MSc,4 Krista L. Lentine, MD, PhD,5,6 Peter P. Reese, MD,7,8 Bertram L. Kasiske, MD,9 Charmaine E. Lok, MD, MSc,4 Liane S. Feldman, MD,10,11 and Amit X. Garg, MD, PhD,1,2,3 on behalf of the Donor Nephrectomy Outcomes Research (DONOR) Network* Background: In the general population, high serum uric acid concentration is a risk factor for gout. It is unknown whether donating a kidney increases a living donor’s risk of gout as serum uric acid concentration increases in donors after nephrectomy. Study Design: Retrospective matched cohort study using large health care databases. Setting & Participants: We studied living kidney donors who donated in 1992 to 2010 in Ontario, Canada. Matched nondonors were selected from the healthiest segment of the general population. 1,988 donors and 19,880 matched nondonors were followed up for a median of 8.4 (maximum, 20.8) years. Predictor: Living kidney donor nephrectomy. Outcomes: The primary outcome was time to a diagnosis of gout. The secondary outcome in a subpopulation was receipt of medications typically used to treat gout (allopurinol or colchicine). Measurements: We assessed the primary outcome with health care diagnostic codes. Results: Donors compared with nondonors were more likely to be given a diagnosis of gout (3.4% vs 2.0%; 3.5 vs 2.1 events/1,000 person-years; HR, 1.6; 95% CI, 1.2-2.1; P , 0.001). Similarly, donors compared with nondonors were more likely to receive a prescription for allopurinol or colchicine (3.8% vs 1.3%; OR, 3.2; 95% CI, 1.5-6.7; P 5 0.002). Results were consistent in multiple additional analyses. Limitations: The primary outcome was assessed using diagnostic codes in health care databases. Laboratory values for serum uric acid and creatinine in follow-up were not available in our data sources. Conclusions: The findings suggest that donating a kidney modestly increases an individual’s absolute longterm incidence of gout. This unique observation should be corroborated in future studies. Am J Kidney Dis. 65(6):925-932. ª 2015 The Authors. Published by Elsevier Inc. on behalf of the National Kidney Foundation, Inc. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/ licenses/by-nc-nd/4.0/). INDEX WORDS: Cohort study; donor outcomes; gout; allopurinol; colchicine; health administrative data; diagnostic codes; health outcomes; hyperuricemia; living kidney donor; kidney donation; nephrectomy; renal transplantation; uric acid.
H
igh serum uric acid concentration is a potent risk factor for gout, for which the 10-year incidence rates are estimated to be 49%, 5%, and 1% for uric acid levels $ 9, 7.0 to 8.9, and , 7.0 mg/dL, respectively.1,2 A decline in glomerular filtration rate (GFR) results in less uric acid excretion and a higher serum uric acid concentration.3,4 These changes are evident with the 25% to 40% reduction in GFR that occurs after living kidney donation.5-9 As early as 6 months after nephrectomy, donors versus nondonor
controls demonstrate an 8.2% higher serum uric acid level (mean values of 5.3 6 1.1 [standard deviation] vs 4.9 6 1.2 mg/dL; P , 0.001) and a 20% higher serum uric acid level a mean of 7 years after nephrectomy (Table S1, available as online supplementary material); mean uric acid level is 1.0 mg/dL higher in donors compared with nondonor controls.8,10 However, whether donating a kidney appreciably increases a person’s risk of gout is unknown. We undertook this study to investigate
From the 1Department of Medicine, Division of Nephrology, and 2Department of Epidemiology and Biostatistics, Western University, London; 3Institute for Clinical Evaluative Sciences (ICES), Ontario; 4Department of Medicine, Division of Nephrology, University of Toronto, Toronto, Ontario, Canada; 5 Department of Medicine, Division of Nephrology, and 6Center for Outcomes Research, Saint Louis University, St Louis, MO; 7 Renal-Electrolyte and Hypertension Division, Perelman School of Medicine, and 8Department of Biostatistics and Epidemiology, University of Pennsylvania, Philadelphia, PA; 9Department of Medicine, Division of Renal Diseases and Hypertension, University of Minnesota, Minneapolis, MN; 10Research Institute of the McGill University Health Centre; and 11Department of Surgery, McGill University, Montreal, Quebec, Canada.
* A list of the active investigators of the DONOR Network appears in the Acknowledgements. Received September 10, 2014. Accepted in revised form January 15, 2015. Originally published online March 26, 2015. Address correspondence to Amit X. Garg, MD, PhD, London Kidney Clinical Research Unit, Room ELL-101, Westminster Tower, London Health Sciences Centre, 800 Commissioners Road East, London, Ontario, Canada N6A 4G5. E-mail: [email protected] 2015 The Authors. Published by Elsevier Inc. on behalf of the National Kidney Foundation, Inc. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/ licenses/by-nc-nd/4.0/). 0272-6386 http://dx.doi.org/10.1053/j.ajkd.2015.01.017
Am J Kidney Dis. 2015;65(6):925-932
925
Lam et al
whether kidney donors have a higher risk of gout than nondonors with similar indicators of baseline health.
METHODS Design and Setting We conducted a retrospective matched cohort study using manual chart review and linked health care databases in Ontario, Canada, where citizens have universal access to hospital care and physician services. The reporting of this study followed guidelines for observational studies.11 The research ethics board approved the prespecified protocol and waived the need for informed consent.
Data Sources We ascertained patient characteristics, covariate information, and outcome data from records in 7 linked databases. Trillium Gift of Life Network is Ontario’s organ and tissue donation registry and captures information for all living kidney donors in the province at the time of donation. We supplemented the data at this registry by manually reviewing perioperative charts of all living kidney donors who underwent donor nephrectomy at 1 of 5 major transplantation centers in Ontario in 1992 through 2010 to ensure data accuracy and completeness. Demographics and vital status information were retrieved from Ontario’s Registered Persons Database. We used the Ontario Drug Benefit database to identify prescription drug use. This database contains highly accurate records of outpatient prescriptions dispensed to all patients 65 years or older, with a basic error rate , 1%.12 Diagnostic and procedural information during hospital admissions was gathered from the Canadian Institute for Health Information Discharge Abstract Database (CIHI-DAD) and Same Day Surgery (CIHI-SDS), whereas information regarding emergency department visits was gathered from the National Ambulatory Care Reporting System. The Ontario Health Insurance Plan database contains health claims for both inpatient and outpatient physician services. These databases have been used extensively for epidemiologic and health services research, including the study of living kidney donor outcomes.13-19
Population Donors We included all permanent residents who donated a kidney from July 1, 1992, through April 30, 2010, at any of the 5 major transplantation centers in Ontario. The date of nephrectomy served as the start date for follow-up and was designated the index date. Prior to matching, 35 living kidney donors with a predonation diagnosis of gout or a prescription for a medication typically used for the treatment of gout (eg, allopurinol or colchicine) were excluded. This was done to assess de novo gout in follow-up. There were too few donors with predonation gout to permit meaningful analyses of their outcomes after donation. We also excluded donors with recognized risk factors for gout, such as those with a predonation diagnosis of alcoholism (n 5 10) or those with evidence of a prescription for medications that may increase uric acid levels (n 5 86), such as thiazides.
Healthy Nondonors Selecting the appropriate nondonors with whom donors can be compared is central to any study reporting risks associated with donor nephrectomy.20 Donors undergo a rigorous medical screening and selection process and thus are inherently healthier than the general population. To address this issue, we used techniques of restriction and matching to select the healthiest segment of the general population. We randomly assigned an index date to the entire Ontario adult general population according to the distribution of index dates in the donors. We then identified comorbid conditions and measures of health care access from the beginning of 926
available database records (July 1, 1991) to the index date. This provided an average of 12 years of medical records for baseline assessment, with 98.4% of individuals having at least 2 years of data for review. Among the general population, we excluded adults with a history of gout (n 5 210,814) or any medical conditions prior to the index date that could preclude donation, including diabetes and hypertension (Table S2). Those who had a nephrectomy, kidney transplantation, kidney biopsy, dialysis, or a previous nephrology consultation also were excluded. Furthermore, we excluded any individual who had evidence of frequent physician visits (.4 visits in the previous 2 years) or who failed to see a physician at least once in the previous 2 years. The latter criterion was applied to ensure that healthy nondonors in our study were accessing physicians for routine health care needs, including preventive health measures. From a total of 9,484,623 Ontarians during the period of interest, this selection process resulted in the exclusion of 45.2% of adults (n 5 4,291,484) as eligible nondonors. From the remaining adults, we matched 10 healthy nondonors to each donor based on age (within 2 years), sex, index date (within 6 months), rural (population , 10,000) or urban residence, and income (categorized into fifths of average neighborhood income).
Outcomes All patients were followed up until death, emigration from the province, or the end of the study period (March 31, 2013). The primary outcome was time to the first health care encounter (physician visit, emergency department visit, or hospitalization) in follow-up at which a diagnostic code of gout was recorded in a health care database by medical coders or those submitting claims for physician reimbursement (Table S2). The secondary outcome was receipt of a prescription for allopurinol or colchicine, 2 medications typically used to treat gout. This secondary outcome was examined in the subpopulation that reached 65 years or older in follow-up.
Statistical Analysis We compared standardized differences in baseline characteristics between donors and healthy nondonors. Differences . 10% suggest meaningful imbalance.21 We used Cox proportional hazards regression models, stratified on matched sets, to calculate the hazard ratio (HR) with 95% confidence interval (CI) for the time to first gout diagnosis. The proportional hazards assumption was not violated (nonsignificant donor 3 log [follow-up time] interaction term; P 5 0.5). We expressed the risk of developing an outcome in both relative and absolute terms. Absolute risk also was expressed as the number needed to harm (ie, the reciprocal of the absolute risk increase). This measure indicates how many individuals need to donate a kidney for 1 patient to experience an event, who otherwise would not have been harmed if all individuals did not donate a kidney (a lower number indicating greater harm). The number needed to harm was calculated for ease of interpretation and not to imply causality. In the subpopulation in which all individuals in a matched set reached 65 years or older in follow-up, we used conditional logistic regression accounting for matched sets to calculate the odds ratio (OR) and 95% CI for receipt of a prescription for allopurinol or colchicine. We repeated the primary analysis in 4 prespecified subgroups defined by the presence or absence of a family history of kidney disease, median age at index date (#43 vs .43 years), sex, and index date (1992-2003 [median follow-up, 13.6 (interquartile range [IQR], 11.4-16.4) years] vs 2004-2010 [median follow-up, 5.8 (IQR, 4.3-7.3) years]). Information for family history of kidney disease was available for only donors (living related vs living unrelated donor), and in each subgroup analysis, sets of nondonors were categorized according to the characteristic of their matched donor (so as not to break matched sets in subgroup analysis). We examined whether subgroup-specific rate ratios differed among subgroups using a series of pairwise standard z tests. We examined the characteristics Am J Kidney Dis. 2015;65(6):925-932
Gout in Living Kidney Donors associated with a diagnosis of gout separately in donors and healthy nondonors using Cox proportional hazards regression models. These characteristics were age (per 5 years older), sex, rural or urban residence, income quintile, and index date (per 1 year later). In donors, we also examined whether a family history of kidney disease or the predonation estimated GFR (eGFR) was associated with a diagnosis of gout. We conducted all analyses with SAS, version 9.3 (SAS Institute Inc). In all outcome analyses, we interpreted 2-tailed P , 0.05 as statistically significant.
RESULTS Baseline Characteristics Baseline characteristics for 1,988 living kidney donors and 19,880 matched healthy nondonors are shown in Table 1. Median age was 43 (IQR, 35-51) years and 61% were women. Unsurprisingly, donors had more physician visits during the year prior to the index date compared with nondonors because these visits are a required part of the donor evaluation process. Most living kidney donors were siblings of the recipients (34.4%), followed by spouses (19.2%), parents (13.7%), and children (13.0%). Before donation, median serum creatinine level was 0.8 (IQR, 0.7-1.0) mg/dL and eGFR was Table 1. Characteristics of Kidney Donors and Matched Healthy Nondonors at Time of Cohort Entry Donors (n 5 1,988)
Age (y) Female sex Rural town Income quintile 1st: lowest 2nd 3rd: middle 4th 5th: highest Physician visits in prior yeara Year of cohort entry 1992-1997 1998-2003 2004-2010
43 [35-51] 1,217 (61.2) 276 (13.9) 282 342 427 474 463
(14.2) (17.2) (21.5) (23.8) (23.3)
5 [2-8] 354 (17.8) 648 (32.6) 986 (49.6)
Healthy Nondonors (n 5 19,880)
43 [35-50] 12,170 (61.2) 2,760 (13.9) 2,802 3,484 4,401 4,656 4,537
(14.1) (17.5) (22.1) (23.4) (22.8)
1 [0-2] 3,550 (17.9) 6,456 (32.5) 9,874 (49.7)
Note: Values for categorical variables are given as number (percentage); values for continuous variables, as median [interquartile range]. Time of cohort entry also is referred to as the index date. For living kidney donors, this was the date of nephrectomy and was randomly assigned to nondonors to establish the time that follow-up began. a Indicates a standardized difference between donors and healthy nondonors . 10%. Standardized differences are less sensitive to sample size than traditional hypothesis tests. They provide a measure of difference between groups divided by pooled standard deviation; a value . 10% is interpreted as a meaningful difference between groups. As expected, donors had more physician visits in the year prior to the index date compared with healthy nondonors; such visits are a necessary part of the donor evaluation process. Am J Kidney Dis. 2015;65(6):925-932
99 (IQR, 87-109) mL/min/1.73 m2 (calculated using the CKD-EPI [Chronic Kidney Disease Epidemiology Collaboration] creatinine equation22). Median follow-up was 8.4 years (in donors, 8.8 years; in healthy nondonors, 8.4 years; maximum, 20.8 years), with 844 donors and 8,145 healthy nondonors having more than 10 years of follow-up. Of 21,868 total individuals, 429 (2.0%) were censored at the time of death; 1,507 (6.9%), at the time of provincial emigration; 19,475 (89.1%), at the time they reached last follow-up (March 31, 2013); and the rest, at the time of their first diagnosis of gout. Total person-years of follow-up were 205,432 (donors, 19,063; nondonors, 186,369). Outcomes There were 457 episodes of gout (in donors, 67; in healthy nondonors, 390; Table 2; Fig 1). The risk of gout was higher in donors than matched healthy nondonors (3.4% vs 2.0%; 3.5 vs 2.1 events/1,000 person-years; HR, 1.6; 95% CI, 1.2-2.1; P , 0.001). At an 8-year median follow-up, the absolute risk difference was 1.4% (95% CI, 0.6%-2.2%) and number needed to harm was 71 (95% CI, 45-170). Median time to first gout diagnosis was 5.7 (IQR, 2.89.9; range, 0.04-20.3) years. We assessed prescription medications in the subpopulation in which all individuals in a matched set reached 65 years or older in follow-up (262 donors and 2,620 healthy nondonors; Table S3). The risk of receiving allopurinol or colchicine was higher in living kidney donors than in matched healthy nondonors (10/262 [3.8%] vs 33/2,620 [1.3%]; OR, 3.2; 95% CI, 1.5-6.7; P 5 0.002). Figure 2 shows subgroup analyses defined by a family history of kidney disease, age at index date, sex, and date of index date. The overall association of a higher risk of gout in donors versus matched healthy nondonors did not differ significantly across any of these subgroups (P for interaction range 5 0.05-0.3). When donors and matched healthy nondonors were examined separately, in both groups, older age and male sex were associated with higher risk of gout (Table 3). In donors, family history of kidney disease was not associated with higher risk of gout, and the association with different strata of lower predonation eGFRs was inconsistent (Table 4). Additional Analyses The primary association proved robust in additional analyses. First, we redefined the primary outcome by the presence of 2 separate days with diagnostic codes for gout (vs 1 day as done in the primary analysis), for which the date of the second code defined an event. There were 222 episodes of gout (in donors, 38; in healthy nondonors, 184). The risk of gout remained 927
Lam et al Table 2. Gout Among Kidney Donors and Matched Healthy Nondonors Receipt of Gout Medicationsa
Gout Diagnosis
No. of events Events per 1,000 person-y Model-based rate ratiob
Donors (n 5 1,988)
Healthy Nondonors (n 5 19,880)
Donors (n 5 262)
Healthy Nondonors (n 5 2,620)
67 (3.4) 3.5 1.6 (1.2-2.1)c
390 (2.0) 2.1 1.00 (reference)
10 (3.8)
33 (1.3)
3.2 (1.5-6.7)d
1.00 (reference)
—
—
Note: Values are given as number, number (percentage), or value (95% CI). Abbreviation: CI, confidence interval. a This subpopulation represents matched sets in which all individuals in the set reached 65 years or older in follow-up (when all Ontario citizens participate in the universal prescription drug plan). b Cox proportional hazard regression models, stratified on matched sets, were used to calculate the hazard ratio with 95% CI for the primary outcome of gout diagnosis, and conditional logistic regression for matched sets was used to calculate the odds ratio and 95% CI for the secondary outcome of receipt of gout medications. c P , 0.001. d P 5 0.002.
higher in donors versus healthy nondonors (1.9% vs 0.9%; 2.0 vs 1.0 events/1,000 person-years; HR, 2.1; 95% CI, 1.5-3.0; P , 0.001). Second, to test the specificity of our findings, we examined whether the risk of rheumatoid arthritis differed in donors and healthy nondonors. Because there is no plausible reason why living donation would increase the risk of rheumatoid arthritis, we reasoned that a null association with this outcome would enhance causal inference in our gout analyses. The risk of rheumatoid arthritis was no different in donors than in healthy nondonors (2.7% vs 2.1%; 2.7 vs 2.2 events/1,000 person-years; HR, 1.2; 95% CI, 0.9-1.7; P 5 0.2). Third, we were concerned that donors would see their physicians in follow-up for routine care more
Figure 1. Cumulative percentage of living kidney donors and matched healthy nondonors with a diagnosis of gout. Abbreviation: CI, confidence interval. 928
often than healthy nondonors, which might bias the ascertainment of gout. Thus, we restricted the analysis to matched sets of patients who were gout free for the first 3 years after their index date and followed up this subpopulation for a diagnosis of gout, adjusting for the number of physician visits in the prior 2 years. Donors compared with healthy nondonors had more physician visits during this period (median visits, 7.5 [IQR, 3-14] vs 4 [IQR, 1-8]). However, there was no substantial attenuation of the higher risk of gout in donors compared with healthy nondonors after adjustment for the number of physician visits (2.4% vs 1.5%; 2.3 vs 1.5 events/1,000 person-years; HR, 1.5; 95% CI, 1.0-2.1; P 5 0.04). In another analysis, we assessed the receipt of prescription nonsteroidal anti-inflammatory drugs (NSAIDs) to consider the possibility that donors were instructed to avoid NSAIDs in favor of other gout medications. The likelihood of receiving an NSAID was no different in donors and healthy nondonors (32.8% vs 29.5%; OR, 1.2; 95% CI, 0.9-1.6; P 5 0.2). We also performed other analyses to put the results into context. First, rather than matching donors to the healthiest segment of the general population, we matched donors to the general population (for which the only exclusion was a prior episode of gout; Tables S4 and S5). The risk of gout was not statistically different in donors compared with general population nondonors (67/1,988 [3.4%] vs 540/ 19,880 [2.7%]; 3.5 vs 2.9 events/1,000 person-years; HR, 1.2; 95% CI, 0.9-1.6; P 5 0.1; Fig S1). Similarly, the risk of receiving a prescription for allopurinol or colchicine was not statistically different in donors compared with general population nondonors. Second, we compared the cumulative probability of gout in the 20 years after the index date, stratifying results by donation status and age at index date (20, 40, or 60 years). The 20-year cumulative probability of gout was higher in donors versus healthy Am J Kidney Dis. 2015;65(6):925-932
Gout in Living Kidney Donors
Figure 2. Influence of family history of kidney failure, age at index date, sex, and date of index date (duration of follow-up) on primary outcome of a diagnosis of gout in living kidney donors versus healthy nondonors. Information for family history of kidney disease was available for only donors (living related vs living unrelated donor) and in each subgroup analysis, sets of nondonors were categorized according to the characteristic of their matched donor (so as not to break matched sets in subgroup analysis). Although 61% of donors had a first-degree relative with kidney failure, we assume that few nondonors had a similar family history, although this information was not available for nondonors. Individuals with an index date of 1992 to 2003 had median follow-up of 13.6 (interquartile range [IQR], 11.4 to 16.4) years, and individuals with an index date of 2004 to 2010 had median follow-up of 5.8 (IQR, 4.3 to 7.3) years. Abbreviation: CI, confidence interval.
nondonors and highest in the oldest age strata (60 years at index date; Table S6). We also considered the lifetime cumulative probability of gout, assuming a maximum lifespan of 80 years and adjusting for the competing risk of death (using the technique of modifying the survival analysis for which survival age was used as the time scale as opposed to time in study; Table S6).23,24 The lifetime probability of gout was higher in donors versus healthy nondonors and highest in the youngest age strata (20 years at index date). For example, 1 in 5 (20.4%) donors who Table 3. Risk Factors for Diagnosis of Gout in Living Kidney Donors and Healthy Nondonors When Each Group Was Analyzed Separately Donors
Age, per 5-y older Female sex Rural vs urban residence Income, per each quintile higher Index date, per each year more recent
1.13 0.45 0.78 0.94
(1.01-1.26) (0.28-0.74) (0.37-1.63) (0.79-1.12)
0.98 (0.92-1.04)
Healthy Nondonors
1.14 0.28 0.98 1.01
(1.09-1.20) (0.23-0.35) (0.75-1.30) (0.94-1.08)
0.99 (0.97-1.02)
Note: Separate Cox proportional hazards regression models were created for living kidney donors and healthy nondonors. Values are given as adjusted hazard ratio (95% confidence interval). Am J Kidney Dis. 2015;65(6):925-932
donated at the age of 20 years were estimated to receive a diagnosis of gout in their lifetime compared to 1 in 10 healthy nondonors (10.9%) of the same age. The estimated lifetime cumulative probability of gout was similar in donors and nondonors from the general population.
DISCUSSION The increase in serum uric acid concentration after donor nephrectomy is described in 2 recent studies.7,8 This prompted the current study, which to our knowledge is the first to describe the long-term risk of gout in living kidney donors. We found that donors have a higher relative risk of gout compared with nondonors selected to have a similar level of baseline health as donors. In absolute terms, the estimated increase in the 20-year incidence of gout was small. It is reassuring that a donor’s likelihood of developing gout was not statistically greater than for the general population. Among our study’s strengths, it was made possible because of Ontario’s universal health care benefits, with the collection of all health care encounters for all citizens; concerns about selection biases are minimized as a result. In addition, we manually reviewed more than 2,000 perioperative medical charts, thereby ensuring the accuracy of donor information presented. 929
Lam et al Table 4. Risk Factors for Diagnosis of Gout in Living Kidney Donors, Considering Also Family History of Kidney Disease and Predonation eGFR Donors
Age, per 5-y older Female sex Rural vs urban residence Income, per each quintile higher Index date, per each year more recent Family history of kidney disease Predonation eGFRa ,80 mL/min/1.73 m2 80-89 mL/min/1.73 m2
1.09 0.52 0.71 0.93 0.99 1.48
(0.97-1.23) (0.32-0.86) (0.34-1.50) (0.78-1.10) (0.93-1.05) (0.84-2.62)
1.63 (0.85-3.11) 1.91 (1.03-3.52)
Note: Cox proportional hazards regression model. Values are given as adjusted hazard ratio (95% confidence interval). Abbreviation: eGFR, estimated glomerular filtration rate. a Reference category is $ 90 mL/min/1.73 m2.
For the period of interest, this corresponds to all living donation activity for Canada’s largest province. We used techniques of restriction and matching so that appropriate nondonors, with whom donor outcomes could be reliably compared, were selected. We matched donors and nondonors on risk factors associated with gout, including older age and male sex.4,25 We followed up many people during a decade-long span and for some, even for 20 years, with little (,7%) loss to follow-up. Our study also has some limitations. We relied on administrative data collected for nonresearch purposes, which limits the type of variables that are available for inclusion and exclusion criteria and the outcome measurements. Residual confounding may affect our observed association between living kidney donation and gout. Body mass index, overthe-counter medication use, and dietary risk factors for gout, such as red meat consumption, were unavailable in our data sources.26 Accurate racial data were not available, but 71% of Ontario citizens are white, as were w70% of donors. In the general population, African Americans have a higher incidence of gout compared with whites,27 and a recent study suggests that this also may be true in the donor population.28 Complete laboratory values for serum uric acid and creatinine in follow-up were not available in our data sources. We relied on physician diagnoses, diagnostic codes, and outpatient prescriptions to define our outcomes, which may be subject to misclassification. For the secondary analysis assessing the use of medications typically used to treat gout, we did not have complete information for donors younger than 65 years and thus our analysis is limited to matched sets of donors and nondonors who reached 65 years or older in follow-up. It remains possible that an event of gout more likely was ascertained in donors than 930
nondonors, although an additional analysis did not support this finding. A prospective study comparing living kidney donors with appropriate nondonor controls in which all participants receive objective assessments of gout is a preferred future methodology to corroborate our study findings.29 When deciding how to apply the study findings to the more than 27,000 living kidney donations that occur worldwide each year and to the care of approximately half a million prior kidney donors, this new information for the risk of gout should not deter support for living donor kidney transplantation.30,31 Living kidney donation is an important and necessary treatment option for those with kidney failure. Although the finding of gout risk is biologically plausible in living kidney donors, the study is not without limitations, raising the possibility that the true risk differs from our estimates of risk. We recommend that this unique observation be corroborated in future studies before it is shared routinely with potential donors and their recipients as a step in the informed consent process.32,33 Reassuringly, the absolute increase in incidence of gout is small and is unlikely to deter potential living kidney donors, recipients, or transplantation programs from proceeding with donation.31 If high-quality evidence that lifestyle modifications prevent gout becomes available, living donors also can be encouraged to follow such recommendations.34 For donors who lack prescription drug coverage, reimbursing the costs of medications used to manage gout could be considered.35
ACKNOWLEDGEMENTS The authors acknowledge the active investigators of the Donor Nephrectomy Outcomes Research (DONOR) Network, who, in addition to authors Liane Feldman, Amit Garg, Joseph Kim, Ngan Lam, Charmaine Lok, and G.V. Ramesh Prasad, are: Jennifer Arnold, Neil Boudville, Ann Bugeja, Christine Dipchand, Mona Doshi, John Gill, Martin Karpinski, Scott Klarenbach, Greg Knoll, Mauricio Monroy-Cuadros, Christopher Y. Nguan, Jessica Sontrop, Leroy Storsley, Darin Treleaven, and Ann Young. Aspects of this work were presented in abstract form at the World Transplant Congress, San Francisco, CA, July 26-31, 2014. Support: This project was conducted at the Institute for Clinical Evaluative Sciences (ICES) Western Site. ICES is funded by an annual grant from the Ontario Ministry of Health and Long-Term Care. ICES Western is funded by an operating grant from the Academic Medical Organization of Southwestern Ontario. This project was conducted with members of the provincial ICES Kidney, Dialysis and Transplantation Research Program (www. ices.on.ca), which receives programmatic grant funding from the Canadian Institutes of Health Research. Dr Lam was supported by the Clinical Investigator Program at Western University and by a Kidney Research Scientist Core Education and National Training Program (KRESCENT) postdoctoral fellowship award. Dr Garg was supported by a Dr Adam Linton Chair in Kidney Analytics and Application. The opinions, results, and conclusions reported in this article are those of the authors and are independent of the Am J Kidney Dis. 2015;65(6):925-932
Gout in Living Kidney Donors funding sources. The funding sources did not influence any aspect of this study. Financial Disclosure: Dr Garg received an investigatorinitiated grant from Astellas and Roche to support a Canadian Institutes of Health Research study in living kidney donors, and his institution received unrestricted research funding from Pfizer. The other authors declare that they have no other relevant financial interests. Contributions: NNL, SJK, GVRP, KLL, PPR, BLK, and AXG contributed to the research idea and study design. SJK, GVRP, CEL, LSF, and AXG assisted with the data acquisition. EM acquired the data and performed the statistical analyses. NNL drafted and revised the manuscript for publication. AXG provided supervision and mentorship. Each author contributed important intellectual content during manuscript drafting and revision and accepts accountability for the overall work by ensuring that questions pertaining to the accuracy or integrity of any portion of the work are appropriately investigated and resolved. NNL, EM, and AXG take responsibility that this study has been reported honestly, accurately, and transparently; that no important aspects of the study have been omitted, and that any discrepancies from the study as planned have been explained.
SUPPLEMENTARY MATERIAL Table S1: Characteristics of donors and nondonors from 20042008 study. Table S2: Databases and coding definitions for exclusion criteria, baseline characteristics, and outcome measurements. Table S3: Characteristics at cohort entry of kidney donors and matched nondonors eligible for drug coverage. Table S4: Characteristics of kidney donors and matched nondonors from general population at cohort entry. Table S5: Gout events in kidney donors and matched nondonors from general population. Table S6: 20-year cumulative probability of gout and lifetime risk of gout. Figure S1: Cumulative percentage of living kidney donors and matched nondonors from the general population with a diagnosis of gout. Note: The supplementary material accompanying this article (http://dx.doi.org/10.1053/j.ajkd.2015.01.017) is available at www.ajkd.org
REFERENCES 1. Campion EW, Glynn RJ, DeLabry LO. Asymptomatic hyperuricemia. Risks and consequences in the Normative Aging Study. Am J Med. 1987;82(3):421-426. 2. Zhang W, Doherty M, Pascual E, et al. EULAR evidence based recommendations for gout. Part I: diagnosis. Report of a task force of the Standing Committee for International Clinical Studies Including Therapeutics (ESCISIT). Ann Rheum Dis. 2006;65(10):1301-1311. 3. Lipkowitz MS. Regulation of uric acid excretion by the kidney. Curr Rheumatol Rep. 2012;14(2):179-188. 4. Krishnan E. Reduced glomerular function and prevalence of gout: NHANES 2009-10. PLoS One. 2012;7(11):e50046. 5. Garg AX, Muirhead N, Knoll G, et al. Proteinuria and reduced kidney function in living kidney donors: a systematic review, meta-analysis, and meta-regression. Kidney Int. 2006;70(10): 1801-1810. 6. Hida M, Iida T, Shimbo T, et al. Renal function after nephrectomy in renal donors. Tokai J Exp Clin Med. 1982;7(4): 511-516.
Am J Kidney Dis. 2015;65(6):925-932
7. Rossi M, Campbell KL, Johnson DW, et al. Uremic toxin development in living kidney donors: a longitudinal study. Transplantation. 2014;97(5):548-554. 8. Kasiske BL, Anderson-Haag T, Ibrahim HN, et al. A prospective controlled study of kidney donors: baseline and 6month follow-up. Am J Kidney Dis. 2013;62(3):577-586. 9. Young A, Nevis IFP, Geddes C, et al. Do biochemical measures change in living kidney donors? A systematic review. Nephron Clin Pract. 2007;107(3):c82-c89. 10. Young A, Hodsman AB, Boudville N, et al. Bone and mineral metabolism and fibroblast growth factor 23 levels after kidney donation. Am J Kidney Dis. 2012;59(6):761-769. 11. Von Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC, Vandenbroucke JP. The Strengthening of the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. J Clin Epidemiol. 2008;61(4):344-349. 12. Levy AR, O’Brien BJ, Sellors C, Grootendorst P, Willison D. Coding accuracy of administrative drug claims in the Ontario Drug Benefit database. Can J Clin Pharmacol. 2003;10(2):67-71. 13. Lam N, Huang A, Feldman LS, et al. Acute dialysis risk in living kidney donors. Nephrol Dial Transplant. 2012;27(8):3291-3295. 14. Garg AX, Prasad GVR, Thiessen-Philbrook HR, et al. Cardiovascular disease and hypertension risk in living kidney donors: an analysis of health administrative data in Ontario, Canada. Transplantation. 2008;86(3):399-406. 15. Garg AX, Pouget J, Young A, et al. Fracture risk in living kidney donors: a matched cohort study. Am J Kidney Dis. 2012;59(6):770-776. 16. Garg AX, Meirambayeva A, Huang A, et al. Cardiovascular disease in kidney donors: matched cohort study. BMJ. 2012;344:e1203. 17. Thomas SM, Lam NN, Welk BK, et al. Risk of kidney stones with surgical intervention in living kidney donors. Am J Transplant. 2013;13(11):2935-2944. 18. Thomas SM, Lam NN, Huang A, et al. Risk of serious gastrointestinal bleeding in living kidney donors. Clin Transplant. 2014;28(5):530-539. 19. Garg AX, Nevis IF, McArthur E, et al. Gestational hypertension and preeclampsia in living kidney donors. N Engl J Med. 2015;372(2):124-133. 20. Lin J, Kramer H, Chandraker AK. Mortality among living kidney donors and comparison populations. N Engl J Med. 2010;363(8):797-798. 21. Austin PC. Using the standardized difference to compare the prevalence of a binary variable between two groups in observational research. Commun Stat Simul Comput. 2009;38:1228-1234. 22. Levey AS, Stevens LA, Schmid CH, et al. A new equation to estimate glomerular filtration rate. Ann Intern Med. 2009;150(9):604-612. 23. Beiser A, D’Agostino RB, Seshadri S, Sullivan LM, Wolf PA. Computing estimates of incidence, including lifetime risk: Alzheimer’s disease in the Framingham Study. The Practical Incidence Estimators (PIE) macro. Stat Med. 2000;19(11-12): 1495-1522. 24. Berry JD, Dyer A, Cai X, et al. Lifetime risks of cardiovascular disease. N Engl J Med. 2012;366(4):321-329. 25. Chen J-H, Yeh W-T, Chuang S-Y, Wu Y-Y, Pan W-H. Gender-specific risk factors for incident gout: a prospective cohort study. Clin Rheumatol. 2012;31(2):239-245. 26. Torralba KD, De Jesus E, Rachabattula S. The interplay between diet, urate transporters and the risk for gout and
931
Lam et al hyperuricemia: current and future directions. Int J Rheum Dis. 2012;15(6):499-506. 27. Maynard JW, McAdams-DeMarco MA, Law A, et al. Racial differences in gout incidence in a population-based cohort: Atherosclerosis Risk in Communities Study. Am J Epidemiol. 2014;179(5):576-583. 28. Lam NN, Garg AX, Segev DL, et al. Gout after living kidney donation: Correlations with demographic traits and renal complications. Am J Nephrol., in press. 29. Singh JA, Taylor WJ, Dalbeth N, et al. OMERACT endorsement of measures of outcome for studies of acute gout. J Rheumatol. 2014;41(3):569-573. 30. Horvat LD, Shariff SZ, Garg AX. Global trends in the rates of living kidney donation. Kidney Int. 2009;75(10):10881098. 31. Young A, Karpinski M, Treleaven D, et al. Differences in tolerance for health risk to the living donor among potential
932
donors, recipients, and transplant professionals. Kidney Int. 2008;73(10):1159-1166. 32. Housawi AA, Young A, Boudville N, et al. Transplant professionals vary in the long-term medical risks they communicate to potential living kidney donors: an international survey. Nephrol Dial Transplant. 2007;22(10):3040-3045. 33. Moody EM, Clemens KK, Storsley L, Waterman A, Parikh CR, Garg AX. Improving on-line information for potential living kidney donors. Kidney Int. 2007;71(10):10621070. 34. Moi JHY, Sriranganathan MK, Edwards CJ, Buchbinder R. Lifestyle interventions for acute gout. Cochrane Database Syst Rev. 2013;11:CD010519. 35. Sickand M, Cuerden MS, Klarenbach SW, et al. Reimbursing live organ donors for incurred non-medical expenses: a global perspective on policies and programs. Am J Transplant. 2009;9(12):2825-2836.
Am J Kidney Dis. 2015;65(6):925-932
H
igh serum uric acid concentration is a potent risk factor for gout, for which the 10-year incidence rates are estimated to be 49%, 5%, and 1% for uric acid levels $ 9, 7.0 to 8.9, and , 7.0 mg/dL, respectively.1,2 A decline in glomerular filtration rate (GFR) results in less uric acid excretion and a higher serum uric acid concentration.3,4 These changes are evident with the 25% to 40% reduction in GFR that occurs after living kidney donation.5-9 As early as 6 months after nephrectomy, donors versus nondonor
controls demonstrate an 8.2% higher serum uric acid level (mean values of 5.3 6 1.1 [standard deviation] vs 4.9 6 1.2 mg/dL; P , 0.001) and a 20% higher serum uric acid level a mean of 7 years after nephrectomy (Table S1, available as online supplementary material); mean uric acid level is 1.0 mg/dL higher in donors compared with nondonor controls.8,10 However, whether donating a kidney appreciably increases a person’s risk of gout is unknown. We undertook this study to investigate
From the 1Department of Medicine, Division of Nephrology, and 2Department of Epidemiology and Biostatistics, Western University, London; 3Institute for Clinical Evaluative Sciences (ICES), Ontario; 4Department of Medicine, Division of Nephrology, University of Toronto, Toronto, Ontario, Canada; 5 Department of Medicine, Division of Nephrology, and 6Center for Outcomes Research, Saint Louis University, St Louis, MO; 7 Renal-Electrolyte and Hypertension Division, Perelman School of Medicine, and 8Department of Biostatistics and Epidemiology, University of Pennsylvania, Philadelphia, PA; 9Department of Medicine, Division of Renal Diseases and Hypertension, University of Minnesota, Minneapolis, MN; 10Research Institute of the McGill University Health Centre; and 11Department of Surgery, McGill University, Montreal, Quebec, Canada.
* A list of the active investigators of the DONOR Network appears in the Acknowledgements. Received September 10, 2014. Accepted in revised form January 15, 2015. Originally published online March 26, 2015. Address correspondence to Amit X. Garg, MD, PhD, London Kidney Clinical Research Unit, Room ELL-101, Westminster Tower, London Health Sciences Centre, 800 Commissioners Road East, London, Ontario, Canada N6A 4G5. E-mail: [email protected] 2015 The Authors. Published by Elsevier Inc. on behalf of the National Kidney Foundation, Inc. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/ licenses/by-nc-nd/4.0/). 0272-6386 http://dx.doi.org/10.1053/j.ajkd.2015.01.017
Am J Kidney Dis. 2015;65(6):925-932
925
Lam et al
whether kidney donors have a higher risk of gout than nondonors with similar indicators of baseline health.
METHODS Design and Setting We conducted a retrospective matched cohort study using manual chart review and linked health care databases in Ontario, Canada, where citizens have universal access to hospital care and physician services. The reporting of this study followed guidelines for observational studies.11 The research ethics board approved the prespecified protocol and waived the need for informed consent.
Data Sources We ascertained patient characteristics, covariate information, and outcome data from records in 7 linked databases. Trillium Gift of Life Network is Ontario’s organ and tissue donation registry and captures information for all living kidney donors in the province at the time of donation. We supplemented the data at this registry by manually reviewing perioperative charts of all living kidney donors who underwent donor nephrectomy at 1 of 5 major transplantation centers in Ontario in 1992 through 2010 to ensure data accuracy and completeness. Demographics and vital status information were retrieved from Ontario’s Registered Persons Database. We used the Ontario Drug Benefit database to identify prescription drug use. This database contains highly accurate records of outpatient prescriptions dispensed to all patients 65 years or older, with a basic error rate , 1%.12 Diagnostic and procedural information during hospital admissions was gathered from the Canadian Institute for Health Information Discharge Abstract Database (CIHI-DAD) and Same Day Surgery (CIHI-SDS), whereas information regarding emergency department visits was gathered from the National Ambulatory Care Reporting System. The Ontario Health Insurance Plan database contains health claims for both inpatient and outpatient physician services. These databases have been used extensively for epidemiologic and health services research, including the study of living kidney donor outcomes.13-19
Population Donors We included all permanent residents who donated a kidney from July 1, 1992, through April 30, 2010, at any of the 5 major transplantation centers in Ontario. The date of nephrectomy served as the start date for follow-up and was designated the index date. Prior to matching, 35 living kidney donors with a predonation diagnosis of gout or a prescription for a medication typically used for the treatment of gout (eg, allopurinol or colchicine) were excluded. This was done to assess de novo gout in follow-up. There were too few donors with predonation gout to permit meaningful analyses of their outcomes after donation. We also excluded donors with recognized risk factors for gout, such as those with a predonation diagnosis of alcoholism (n 5 10) or those with evidence of a prescription for medications that may increase uric acid levels (n 5 86), such as thiazides.
Healthy Nondonors Selecting the appropriate nondonors with whom donors can be compared is central to any study reporting risks associated with donor nephrectomy.20 Donors undergo a rigorous medical screening and selection process and thus are inherently healthier than the general population. To address this issue, we used techniques of restriction and matching to select the healthiest segment of the general population. We randomly assigned an index date to the entire Ontario adult general population according to the distribution of index dates in the donors. We then identified comorbid conditions and measures of health care access from the beginning of 926
available database records (July 1, 1991) to the index date. This provided an average of 12 years of medical records for baseline assessment, with 98.4% of individuals having at least 2 years of data for review. Among the general population, we excluded adults with a history of gout (n 5 210,814) or any medical conditions prior to the index date that could preclude donation, including diabetes and hypertension (Table S2). Those who had a nephrectomy, kidney transplantation, kidney biopsy, dialysis, or a previous nephrology consultation also were excluded. Furthermore, we excluded any individual who had evidence of frequent physician visits (.4 visits in the previous 2 years) or who failed to see a physician at least once in the previous 2 years. The latter criterion was applied to ensure that healthy nondonors in our study were accessing physicians for routine health care needs, including preventive health measures. From a total of 9,484,623 Ontarians during the period of interest, this selection process resulted in the exclusion of 45.2% of adults (n 5 4,291,484) as eligible nondonors. From the remaining adults, we matched 10 healthy nondonors to each donor based on age (within 2 years), sex, index date (within 6 months), rural (population , 10,000) or urban residence, and income (categorized into fifths of average neighborhood income).
Outcomes All patients were followed up until death, emigration from the province, or the end of the study period (March 31, 2013). The primary outcome was time to the first health care encounter (physician visit, emergency department visit, or hospitalization) in follow-up at which a diagnostic code of gout was recorded in a health care database by medical coders or those submitting claims for physician reimbursement (Table S2). The secondary outcome was receipt of a prescription for allopurinol or colchicine, 2 medications typically used to treat gout. This secondary outcome was examined in the subpopulation that reached 65 years or older in follow-up.
Statistical Analysis We compared standardized differences in baseline characteristics between donors and healthy nondonors. Differences . 10% suggest meaningful imbalance.21 We used Cox proportional hazards regression models, stratified on matched sets, to calculate the hazard ratio (HR) with 95% confidence interval (CI) for the time to first gout diagnosis. The proportional hazards assumption was not violated (nonsignificant donor 3 log [follow-up time] interaction term; P 5 0.5). We expressed the risk of developing an outcome in both relative and absolute terms. Absolute risk also was expressed as the number needed to harm (ie, the reciprocal of the absolute risk increase). This measure indicates how many individuals need to donate a kidney for 1 patient to experience an event, who otherwise would not have been harmed if all individuals did not donate a kidney (a lower number indicating greater harm). The number needed to harm was calculated for ease of interpretation and not to imply causality. In the subpopulation in which all individuals in a matched set reached 65 years or older in follow-up, we used conditional logistic regression accounting for matched sets to calculate the odds ratio (OR) and 95% CI for receipt of a prescription for allopurinol or colchicine. We repeated the primary analysis in 4 prespecified subgroups defined by the presence or absence of a family history of kidney disease, median age at index date (#43 vs .43 years), sex, and index date (1992-2003 [median follow-up, 13.6 (interquartile range [IQR], 11.4-16.4) years] vs 2004-2010 [median follow-up, 5.8 (IQR, 4.3-7.3) years]). Information for family history of kidney disease was available for only donors (living related vs living unrelated donor), and in each subgroup analysis, sets of nondonors were categorized according to the characteristic of their matched donor (so as not to break matched sets in subgroup analysis). We examined whether subgroup-specific rate ratios differed among subgroups using a series of pairwise standard z tests. We examined the characteristics Am J Kidney Dis. 2015;65(6):925-932
Gout in Living Kidney Donors associated with a diagnosis of gout separately in donors and healthy nondonors using Cox proportional hazards regression models. These characteristics were age (per 5 years older), sex, rural or urban residence, income quintile, and index date (per 1 year later). In donors, we also examined whether a family history of kidney disease or the predonation estimated GFR (eGFR) was associated with a diagnosis of gout. We conducted all analyses with SAS, version 9.3 (SAS Institute Inc). In all outcome analyses, we interpreted 2-tailed P , 0.05 as statistically significant.
RESULTS Baseline Characteristics Baseline characteristics for 1,988 living kidney donors and 19,880 matched healthy nondonors are shown in Table 1. Median age was 43 (IQR, 35-51) years and 61% were women. Unsurprisingly, donors had more physician visits during the year prior to the index date compared with nondonors because these visits are a required part of the donor evaluation process. Most living kidney donors were siblings of the recipients (34.4%), followed by spouses (19.2%), parents (13.7%), and children (13.0%). Before donation, median serum creatinine level was 0.8 (IQR, 0.7-1.0) mg/dL and eGFR was Table 1. Characteristics of Kidney Donors and Matched Healthy Nondonors at Time of Cohort Entry Donors (n 5 1,988)
Age (y) Female sex Rural town Income quintile 1st: lowest 2nd 3rd: middle 4th 5th: highest Physician visits in prior yeara Year of cohort entry 1992-1997 1998-2003 2004-2010
43 [35-51] 1,217 (61.2) 276 (13.9) 282 342 427 474 463
(14.2) (17.2) (21.5) (23.8) (23.3)
5 [2-8] 354 (17.8) 648 (32.6) 986 (49.6)
Healthy Nondonors (n 5 19,880)
43 [35-50] 12,170 (61.2) 2,760 (13.9) 2,802 3,484 4,401 4,656 4,537
(14.1) (17.5) (22.1) (23.4) (22.8)
1 [0-2] 3,550 (17.9) 6,456 (32.5) 9,874 (49.7)
Note: Values for categorical variables are given as number (percentage); values for continuous variables, as median [interquartile range]. Time of cohort entry also is referred to as the index date. For living kidney donors, this was the date of nephrectomy and was randomly assigned to nondonors to establish the time that follow-up began. a Indicates a standardized difference between donors and healthy nondonors . 10%. Standardized differences are less sensitive to sample size than traditional hypothesis tests. They provide a measure of difference between groups divided by pooled standard deviation; a value . 10% is interpreted as a meaningful difference between groups. As expected, donors had more physician visits in the year prior to the index date compared with healthy nondonors; such visits are a necessary part of the donor evaluation process. Am J Kidney Dis. 2015;65(6):925-932
99 (IQR, 87-109) mL/min/1.73 m2 (calculated using the CKD-EPI [Chronic Kidney Disease Epidemiology Collaboration] creatinine equation22). Median follow-up was 8.4 years (in donors, 8.8 years; in healthy nondonors, 8.4 years; maximum, 20.8 years), with 844 donors and 8,145 healthy nondonors having more than 10 years of follow-up. Of 21,868 total individuals, 429 (2.0%) were censored at the time of death; 1,507 (6.9%), at the time of provincial emigration; 19,475 (89.1%), at the time they reached last follow-up (March 31, 2013); and the rest, at the time of their first diagnosis of gout. Total person-years of follow-up were 205,432 (donors, 19,063; nondonors, 186,369). Outcomes There were 457 episodes of gout (in donors, 67; in healthy nondonors, 390; Table 2; Fig 1). The risk of gout was higher in donors than matched healthy nondonors (3.4% vs 2.0%; 3.5 vs 2.1 events/1,000 person-years; HR, 1.6; 95% CI, 1.2-2.1; P , 0.001). At an 8-year median follow-up, the absolute risk difference was 1.4% (95% CI, 0.6%-2.2%) and number needed to harm was 71 (95% CI, 45-170). Median time to first gout diagnosis was 5.7 (IQR, 2.89.9; range, 0.04-20.3) years. We assessed prescription medications in the subpopulation in which all individuals in a matched set reached 65 years or older in follow-up (262 donors and 2,620 healthy nondonors; Table S3). The risk of receiving allopurinol or colchicine was higher in living kidney donors than in matched healthy nondonors (10/262 [3.8%] vs 33/2,620 [1.3%]; OR, 3.2; 95% CI, 1.5-6.7; P 5 0.002). Figure 2 shows subgroup analyses defined by a family history of kidney disease, age at index date, sex, and date of index date. The overall association of a higher risk of gout in donors versus matched healthy nondonors did not differ significantly across any of these subgroups (P for interaction range 5 0.05-0.3). When donors and matched healthy nondonors were examined separately, in both groups, older age and male sex were associated with higher risk of gout (Table 3). In donors, family history of kidney disease was not associated with higher risk of gout, and the association with different strata of lower predonation eGFRs was inconsistent (Table 4). Additional Analyses The primary association proved robust in additional analyses. First, we redefined the primary outcome by the presence of 2 separate days with diagnostic codes for gout (vs 1 day as done in the primary analysis), for which the date of the second code defined an event. There were 222 episodes of gout (in donors, 38; in healthy nondonors, 184). The risk of gout remained 927
Lam et al Table 2. Gout Among Kidney Donors and Matched Healthy Nondonors Receipt of Gout Medicationsa
Gout Diagnosis
No. of events Events per 1,000 person-y Model-based rate ratiob
Donors (n 5 1,988)
Healthy Nondonors (n 5 19,880)
Donors (n 5 262)
Healthy Nondonors (n 5 2,620)
67 (3.4) 3.5 1.6 (1.2-2.1)c
390 (2.0) 2.1 1.00 (reference)
10 (3.8)
33 (1.3)
3.2 (1.5-6.7)d
1.00 (reference)
—
—
Note: Values are given as number, number (percentage), or value (95% CI). Abbreviation: CI, confidence interval. a This subpopulation represents matched sets in which all individuals in the set reached 65 years or older in follow-up (when all Ontario citizens participate in the universal prescription drug plan). b Cox proportional hazard regression models, stratified on matched sets, were used to calculate the hazard ratio with 95% CI for the primary outcome of gout diagnosis, and conditional logistic regression for matched sets was used to calculate the odds ratio and 95% CI for the secondary outcome of receipt of gout medications. c P , 0.001. d P 5 0.002.
higher in donors versus healthy nondonors (1.9% vs 0.9%; 2.0 vs 1.0 events/1,000 person-years; HR, 2.1; 95% CI, 1.5-3.0; P , 0.001). Second, to test the specificity of our findings, we examined whether the risk of rheumatoid arthritis differed in donors and healthy nondonors. Because there is no plausible reason why living donation would increase the risk of rheumatoid arthritis, we reasoned that a null association with this outcome would enhance causal inference in our gout analyses. The risk of rheumatoid arthritis was no different in donors than in healthy nondonors (2.7% vs 2.1%; 2.7 vs 2.2 events/1,000 person-years; HR, 1.2; 95% CI, 0.9-1.7; P 5 0.2). Third, we were concerned that donors would see their physicians in follow-up for routine care more
Figure 1. Cumulative percentage of living kidney donors and matched healthy nondonors with a diagnosis of gout. Abbreviation: CI, confidence interval. 928
often than healthy nondonors, which might bias the ascertainment of gout. Thus, we restricted the analysis to matched sets of patients who were gout free for the first 3 years after their index date and followed up this subpopulation for a diagnosis of gout, adjusting for the number of physician visits in the prior 2 years. Donors compared with healthy nondonors had more physician visits during this period (median visits, 7.5 [IQR, 3-14] vs 4 [IQR, 1-8]). However, there was no substantial attenuation of the higher risk of gout in donors compared with healthy nondonors after adjustment for the number of physician visits (2.4% vs 1.5%; 2.3 vs 1.5 events/1,000 person-years; HR, 1.5; 95% CI, 1.0-2.1; P 5 0.04). In another analysis, we assessed the receipt of prescription nonsteroidal anti-inflammatory drugs (NSAIDs) to consider the possibility that donors were instructed to avoid NSAIDs in favor of other gout medications. The likelihood of receiving an NSAID was no different in donors and healthy nondonors (32.8% vs 29.5%; OR, 1.2; 95% CI, 0.9-1.6; P 5 0.2). We also performed other analyses to put the results into context. First, rather than matching donors to the healthiest segment of the general population, we matched donors to the general population (for which the only exclusion was a prior episode of gout; Tables S4 and S5). The risk of gout was not statistically different in donors compared with general population nondonors (67/1,988 [3.4%] vs 540/ 19,880 [2.7%]; 3.5 vs 2.9 events/1,000 person-years; HR, 1.2; 95% CI, 0.9-1.6; P 5 0.1; Fig S1). Similarly, the risk of receiving a prescription for allopurinol or colchicine was not statistically different in donors compared with general population nondonors. Second, we compared the cumulative probability of gout in the 20 years after the index date, stratifying results by donation status and age at index date (20, 40, or 60 years). The 20-year cumulative probability of gout was higher in donors versus healthy Am J Kidney Dis. 2015;65(6):925-932
Gout in Living Kidney Donors
Figure 2. Influence of family history of kidney failure, age at index date, sex, and date of index date (duration of follow-up) on primary outcome of a diagnosis of gout in living kidney donors versus healthy nondonors. Information for family history of kidney disease was available for only donors (living related vs living unrelated donor) and in each subgroup analysis, sets of nondonors were categorized according to the characteristic of their matched donor (so as not to break matched sets in subgroup analysis). Although 61% of donors had a first-degree relative with kidney failure, we assume that few nondonors had a similar family history, although this information was not available for nondonors. Individuals with an index date of 1992 to 2003 had median follow-up of 13.6 (interquartile range [IQR], 11.4 to 16.4) years, and individuals with an index date of 2004 to 2010 had median follow-up of 5.8 (IQR, 4.3 to 7.3) years. Abbreviation: CI, confidence interval.
nondonors and highest in the oldest age strata (60 years at index date; Table S6). We also considered the lifetime cumulative probability of gout, assuming a maximum lifespan of 80 years and adjusting for the competing risk of death (using the technique of modifying the survival analysis for which survival age was used as the time scale as opposed to time in study; Table S6).23,24 The lifetime probability of gout was higher in donors versus healthy nondonors and highest in the youngest age strata (20 years at index date). For example, 1 in 5 (20.4%) donors who Table 3. Risk Factors for Diagnosis of Gout in Living Kidney Donors and Healthy Nondonors When Each Group Was Analyzed Separately Donors
Age, per 5-y older Female sex Rural vs urban residence Income, per each quintile higher Index date, per each year more recent
1.13 0.45 0.78 0.94
(1.01-1.26) (0.28-0.74) (0.37-1.63) (0.79-1.12)
0.98 (0.92-1.04)
Healthy Nondonors
1.14 0.28 0.98 1.01
(1.09-1.20) (0.23-0.35) (0.75-1.30) (0.94-1.08)
0.99 (0.97-1.02)
Note: Separate Cox proportional hazards regression models were created for living kidney donors and healthy nondonors. Values are given as adjusted hazard ratio (95% confidence interval). Am J Kidney Dis. 2015;65(6):925-932
donated at the age of 20 years were estimated to receive a diagnosis of gout in their lifetime compared to 1 in 10 healthy nondonors (10.9%) of the same age. The estimated lifetime cumulative probability of gout was similar in donors and nondonors from the general population.
DISCUSSION The increase in serum uric acid concentration after donor nephrectomy is described in 2 recent studies.7,8 This prompted the current study, which to our knowledge is the first to describe the long-term risk of gout in living kidney donors. We found that donors have a higher relative risk of gout compared with nondonors selected to have a similar level of baseline health as donors. In absolute terms, the estimated increase in the 20-year incidence of gout was small. It is reassuring that a donor’s likelihood of developing gout was not statistically greater than for the general population. Among our study’s strengths, it was made possible because of Ontario’s universal health care benefits, with the collection of all health care encounters for all citizens; concerns about selection biases are minimized as a result. In addition, we manually reviewed more than 2,000 perioperative medical charts, thereby ensuring the accuracy of donor information presented. 929
Lam et al Table 4. Risk Factors for Diagnosis of Gout in Living Kidney Donors, Considering Also Family History of Kidney Disease and Predonation eGFR Donors
Age, per 5-y older Female sex Rural vs urban residence Income, per each quintile higher Index date, per each year more recent Family history of kidney disease Predonation eGFRa ,80 mL/min/1.73 m2 80-89 mL/min/1.73 m2
1.09 0.52 0.71 0.93 0.99 1.48
(0.97-1.23) (0.32-0.86) (0.34-1.50) (0.78-1.10) (0.93-1.05) (0.84-2.62)
1.63 (0.85-3.11) 1.91 (1.03-3.52)
Note: Cox proportional hazards regression model. Values are given as adjusted hazard ratio (95% confidence interval). Abbreviation: eGFR, estimated glomerular filtration rate. a Reference category is $ 90 mL/min/1.73 m2.
For the period of interest, this corresponds to all living donation activity for Canada’s largest province. We used techniques of restriction and matching so that appropriate nondonors, with whom donor outcomes could be reliably compared, were selected. We matched donors and nondonors on risk factors associated with gout, including older age and male sex.4,25 We followed up many people during a decade-long span and for some, even for 20 years, with little (,7%) loss to follow-up. Our study also has some limitations. We relied on administrative data collected for nonresearch purposes, which limits the type of variables that are available for inclusion and exclusion criteria and the outcome measurements. Residual confounding may affect our observed association between living kidney donation and gout. Body mass index, overthe-counter medication use, and dietary risk factors for gout, such as red meat consumption, were unavailable in our data sources.26 Accurate racial data were not available, but 71% of Ontario citizens are white, as were w70% of donors. In the general population, African Americans have a higher incidence of gout compared with whites,27 and a recent study suggests that this also may be true in the donor population.28 Complete laboratory values for serum uric acid and creatinine in follow-up were not available in our data sources. We relied on physician diagnoses, diagnostic codes, and outpatient prescriptions to define our outcomes, which may be subject to misclassification. For the secondary analysis assessing the use of medications typically used to treat gout, we did not have complete information for donors younger than 65 years and thus our analysis is limited to matched sets of donors and nondonors who reached 65 years or older in follow-up. It remains possible that an event of gout more likely was ascertained in donors than 930
nondonors, although an additional analysis did not support this finding. A prospective study comparing living kidney donors with appropriate nondonor controls in which all participants receive objective assessments of gout is a preferred future methodology to corroborate our study findings.29 When deciding how to apply the study findings to the more than 27,000 living kidney donations that occur worldwide each year and to the care of approximately half a million prior kidney donors, this new information for the risk of gout should not deter support for living donor kidney transplantation.30,31 Living kidney donation is an important and necessary treatment option for those with kidney failure. Although the finding of gout risk is biologically plausible in living kidney donors, the study is not without limitations, raising the possibility that the true risk differs from our estimates of risk. We recommend that this unique observation be corroborated in future studies before it is shared routinely with potential donors and their recipients as a step in the informed consent process.32,33 Reassuringly, the absolute increase in incidence of gout is small and is unlikely to deter potential living kidney donors, recipients, or transplantation programs from proceeding with donation.31 If high-quality evidence that lifestyle modifications prevent gout becomes available, living donors also can be encouraged to follow such recommendations.34 For donors who lack prescription drug coverage, reimbursing the costs of medications used to manage gout could be considered.35
ACKNOWLEDGEMENTS The authors acknowledge the active investigators of the Donor Nephrectomy Outcomes Research (DONOR) Network, who, in addition to authors Liane Feldman, Amit Garg, Joseph Kim, Ngan Lam, Charmaine Lok, and G.V. Ramesh Prasad, are: Jennifer Arnold, Neil Boudville, Ann Bugeja, Christine Dipchand, Mona Doshi, John Gill, Martin Karpinski, Scott Klarenbach, Greg Knoll, Mauricio Monroy-Cuadros, Christopher Y. Nguan, Jessica Sontrop, Leroy Storsley, Darin Treleaven, and Ann Young. Aspects of this work were presented in abstract form at the World Transplant Congress, San Francisco, CA, July 26-31, 2014. Support: This project was conducted at the Institute for Clinical Evaluative Sciences (ICES) Western Site. ICES is funded by an annual grant from the Ontario Ministry of Health and Long-Term Care. ICES Western is funded by an operating grant from the Academic Medical Organization of Southwestern Ontario. This project was conducted with members of the provincial ICES Kidney, Dialysis and Transplantation Research Program (www. ices.on.ca), which receives programmatic grant funding from the Canadian Institutes of Health Research. Dr Lam was supported by the Clinical Investigator Program at Western University and by a Kidney Research Scientist Core Education and National Training Program (KRESCENT) postdoctoral fellowship award. Dr Garg was supported by a Dr Adam Linton Chair in Kidney Analytics and Application. The opinions, results, and conclusions reported in this article are those of the authors and are independent of the Am J Kidney Dis. 2015;65(6):925-932
Gout in Living Kidney Donors funding sources. The funding sources did not influence any aspect of this study. Financial Disclosure: Dr Garg received an investigatorinitiated grant from Astellas and Roche to support a Canadian Institutes of Health Research study in living kidney donors, and his institution received unrestricted research funding from Pfizer. The other authors declare that they have no other relevant financial interests. Contributions: NNL, SJK, GVRP, KLL, PPR, BLK, and AXG contributed to the research idea and study design. SJK, GVRP, CEL, LSF, and AXG assisted with the data acquisition. EM acquired the data and performed the statistical analyses. NNL drafted and revised the manuscript for publication. AXG provided supervision and mentorship. Each author contributed important intellectual content during manuscript drafting and revision and accepts accountability for the overall work by ensuring that questions pertaining to the accuracy or integrity of any portion of the work are appropriately investigated and resolved. NNL, EM, and AXG take responsibility that this study has been reported honestly, accurately, and transparently; that no important aspects of the study have been omitted, and that any discrepancies from the study as planned have been explained.
SUPPLEMENTARY MATERIAL Table S1: Characteristics of donors and nondonors from 20042008 study. Table S2: Databases and coding definitions for exclusion criteria, baseline characteristics, and outcome measurements. Table S3: Characteristics at cohort entry of kidney donors and matched nondonors eligible for drug coverage. Table S4: Characteristics of kidney donors and matched nondonors from general population at cohort entry. Table S5: Gout events in kidney donors and matched nondonors from general population. Table S6: 20-year cumulative probability of gout and lifetime risk of gout. Figure S1: Cumulative percentage of living kidney donors and matched nondonors from the general population with a diagnosis of gout. Note: The supplementary material accompanying this article (http://dx.doi.org/10.1053/j.ajkd.2015.01.017) is available at www.ajkd.org
REFERENCES 1. Campion EW, Glynn RJ, DeLabry LO. Asymptomatic hyperuricemia. Risks and consequences in the Normative Aging Study. Am J Med. 1987;82(3):421-426. 2. Zhang W, Doherty M, Pascual E, et al. EULAR evidence based recommendations for gout. Part I: diagnosis. Report of a task force of the Standing Committee for International Clinical Studies Including Therapeutics (ESCISIT). Ann Rheum Dis. 2006;65(10):1301-1311. 3. Lipkowitz MS. Regulation of uric acid excretion by the kidney. Curr Rheumatol Rep. 2012;14(2):179-188. 4. Krishnan E. Reduced glomerular function and prevalence of gout: NHANES 2009-10. PLoS One. 2012;7(11):e50046. 5. Garg AX, Muirhead N, Knoll G, et al. Proteinuria and reduced kidney function in living kidney donors: a systematic review, meta-analysis, and meta-regression. Kidney Int. 2006;70(10): 1801-1810. 6. Hida M, Iida T, Shimbo T, et al. Renal function after nephrectomy in renal donors. Tokai J Exp Clin Med. 1982;7(4): 511-516.
Am J Kidney Dis. 2015;65(6):925-932
7. Rossi M, Campbell KL, Johnson DW, et al. Uremic toxin development in living kidney donors: a longitudinal study. Transplantation. 2014;97(5):548-554. 8. Kasiske BL, Anderson-Haag T, Ibrahim HN, et al. A prospective controlled study of kidney donors: baseline and 6month follow-up. Am J Kidney Dis. 2013;62(3):577-586. 9. Young A, Nevis IFP, Geddes C, et al. Do biochemical measures change in living kidney donors? A systematic review. Nephron Clin Pract. 2007;107(3):c82-c89. 10. Young A, Hodsman AB, Boudville N, et al. Bone and mineral metabolism and fibroblast growth factor 23 levels after kidney donation. Am J Kidney Dis. 2012;59(6):761-769. 11. Von Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC, Vandenbroucke JP. The Strengthening of the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. J Clin Epidemiol. 2008;61(4):344-349. 12. Levy AR, O’Brien BJ, Sellors C, Grootendorst P, Willison D. Coding accuracy of administrative drug claims in the Ontario Drug Benefit database. Can J Clin Pharmacol. 2003;10(2):67-71. 13. Lam N, Huang A, Feldman LS, et al. Acute dialysis risk in living kidney donors. Nephrol Dial Transplant. 2012;27(8):3291-3295. 14. Garg AX, Prasad GVR, Thiessen-Philbrook HR, et al. Cardiovascular disease and hypertension risk in living kidney donors: an analysis of health administrative data in Ontario, Canada. Transplantation. 2008;86(3):399-406. 15. Garg AX, Pouget J, Young A, et al. Fracture risk in living kidney donors: a matched cohort study. Am J Kidney Dis. 2012;59(6):770-776. 16. Garg AX, Meirambayeva A, Huang A, et al. Cardiovascular disease in kidney donors: matched cohort study. BMJ. 2012;344:e1203. 17. Thomas SM, Lam NN, Welk BK, et al. Risk of kidney stones with surgical intervention in living kidney donors. Am J Transplant. 2013;13(11):2935-2944. 18. Thomas SM, Lam NN, Huang A, et al. Risk of serious gastrointestinal bleeding in living kidney donors. Clin Transplant. 2014;28(5):530-539. 19. Garg AX, Nevis IF, McArthur E, et al. Gestational hypertension and preeclampsia in living kidney donors. N Engl J Med. 2015;372(2):124-133. 20. Lin J, Kramer H, Chandraker AK. Mortality among living kidney donors and comparison populations. N Engl J Med. 2010;363(8):797-798. 21. Austin PC. Using the standardized difference to compare the prevalence of a binary variable between two groups in observational research. Commun Stat Simul Comput. 2009;38:1228-1234. 22. Levey AS, Stevens LA, Schmid CH, et al. A new equation to estimate glomerular filtration rate. Ann Intern Med. 2009;150(9):604-612. 23. Beiser A, D’Agostino RB, Seshadri S, Sullivan LM, Wolf PA. Computing estimates of incidence, including lifetime risk: Alzheimer’s disease in the Framingham Study. The Practical Incidence Estimators (PIE) macro. Stat Med. 2000;19(11-12): 1495-1522. 24. Berry JD, Dyer A, Cai X, et al. Lifetime risks of cardiovascular disease. N Engl J Med. 2012;366(4):321-329. 25. Chen J-H, Yeh W-T, Chuang S-Y, Wu Y-Y, Pan W-H. Gender-specific risk factors for incident gout: a prospective cohort study. Clin Rheumatol. 2012;31(2):239-245. 26. Torralba KD, De Jesus E, Rachabattula S. The interplay between diet, urate transporters and the risk for gout and
931
Lam et al hyperuricemia: current and future directions. Int J Rheum Dis. 2012;15(6):499-506. 27. Maynard JW, McAdams-DeMarco MA, Law A, et al. Racial differences in gout incidence in a population-based cohort: Atherosclerosis Risk in Communities Study. Am J Epidemiol. 2014;179(5):576-583. 28. Lam NN, Garg AX, Segev DL, et al. Gout after living kidney donation: Correlations with demographic traits and renal complications. Am J Nephrol., in press. 29. Singh JA, Taylor WJ, Dalbeth N, et al. OMERACT endorsement of measures of outcome for studies of acute gout. J Rheumatol. 2014;41(3):569-573. 30. Horvat LD, Shariff SZ, Garg AX. Global trends in the rates of living kidney donation. Kidney Int. 2009;75(10):10881098. 31. Young A, Karpinski M, Treleaven D, et al. Differences in tolerance for health risk to the living donor among potential
932
donors, recipients, and transplant professionals. Kidney Int. 2008;73(10):1159-1166. 32. Housawi AA, Young A, Boudville N, et al. Transplant professionals vary in the long-term medical risks they communicate to potential living kidney donors: an international survey. Nephrol Dial Transplant. 2007;22(10):3040-3045. 33. Moody EM, Clemens KK, Storsley L, Waterman A, Parikh CR, Garg AX. Improving on-line information for potential living kidney donors. Kidney Int. 2007;71(10):10621070. 34. Moi JHY, Sriranganathan MK, Edwards CJ, Buchbinder R. Lifestyle interventions for acute gout. Cochrane Database Syst Rev. 2013;11:CD010519. 35. Sickand M, Cuerden MS, Klarenbach SW, et al. Reimbursing live organ donors for incurred non-medical expenses: a global perspective on policies and programs. Am J Transplant. 2009;9(12):2825-2836.
Am J Kidney Dis. 2015;65(6):925-932
