A Prospective Study of Risk Factors for Nephrolithiasis After Roux-en-Y Gastric Bypass Surgery

A Prospective Study of Risk Factors for Nephrolithiasis After Roux-en-Y Gastric Bypass Surgery Alyssa M. Park,* Douglas W. Storm, Brant R. Fulmer, Christopher D. Still, G. Craig Wood and James E. Hartle, II From Geisinger Medical Center, Danville, Pennsylvania Abbreviations and Acronyms BMI body mass index JI jejunoileal RYGB Roux-en-Y gastric bypass SS supersaturation Submitted for publication March 16, 2009. Supported by Mission Pharmacal, San Antonio, Texas. Study received institutional review board approval. * Correspondence: Geisinger Medical Center, 100 North Academy Ave., Danville, Pennsylvania 17822 (e-mail: ampark@geisinger.edu). See Editorial on page 2105. For another article on a related topic see page 2490. Purpose: Roux-en-Y gastric bypass surgery has become an increasingly common form of weight management. Early retrospective reviews have suggested that new onset nephrolithiasis develops in some patients after undergoing Roux-en-Y gastric bypass. We present a prospective longitudinal study to assess risk factors for nephrolithiasis after Roux-en-Y gastric bypass. Materials and Methods: A total of 45 morbidly obese patients scheduled to undergo Roux-en-Y gastric bypass surgery were enrolled in this prospective study between November 2006 and November 2007. Exclusion criteria included history of nephrolithiasis or inflammatory bowel disease. Serum uric acid, parathyroid hormone, calcium, albumin, and creatinine and 24-hour urine collections were obtained within 6 months before Roux-en-Y gastric bypass, and at 6 to 12 months postoperatively. A Wilcoxon signed-rank test was used to compare preoperative and postoperative serum laboratory values and 24-hour urine values. McNemar s test was used to determine if the percent of abnormal values underwent a statistically significant change after Roux-en-Y gastric bypass. For both statistical methods a p value was calculated for the change in each variable with p 0.05 considered statistically significant. Results: Statistically significant changes included increased urinary oxalate and calcium oxalate supersaturation, and decreased urinary citrate and total urinary volume postoperatively. A statistically significant percentage of patients exhibited decreased urinary calcium, while a statistically significant percentage of patients experienced increased urinary oxalate and calcium oxalate supersaturation. Conclusions: Our prospective study demonstrated multiple factors that increase the relative risk of nephrolithiasis after Roux-en-Y gastric bypass. These changes may make stone formation after Roux-en-Y gastric bypass increasingly likely and pose an ongoing challenge in the realm of urology. Key Words: gastric bypass; nephrolithiasis; hyperoxaluria; obesity, morbid MORBID obesity has become an increasingly common health problem in the United States. Citizens of Western nations are much more likely to be overweight than they were 20 years ago. From 2000 to 2005 the prevalence of BMI greater than 30 kg/m 2 has increased by 24%. In addition the prevalence of BMI greater than 40 kg/m 2 has increased by 50% and the prevalence of BMI greater than 50 kg/m 2 has increased by 75%. 1 Due to this rapid increase in severe morbid obesity there has also been an increase in associated health problems such as hypertension, diabetes and heart disease. With a 2334 www.jurology.com 0022-5347/09/1825-2334/0 Vol. 182, 2334-2339, November 2009 THE JOURNAL OF UROLOGY Printed in U.S.A. Copyright 2009 by AMERICAN UROLOGICAL ASSOCIATION DOI:10.1016/j.juro.2009.07.044

RISK FACTORS FOR NEPHROLITHIASIS AFTER ROUX-EN-Y GASTRIC BYPASS SURGERY 2335 large population of patients attempting to lose weight, bariatric surgery has become an increasingly common weight loss treatment in the United States. The birth of bariatric surgery began in the 1950s with the advent of jejunoileal bypass. Although successful weight reduction was achieved, the procedure was fraught with complications. Liver disease, electrolyte abnormalities, renal calculi, renal failure and cholelithiasis were among the severe postoperative complications, and the procedure was subsequently abandoned. 2 While other weight loss surgeries have been developed, Roux-en-Y gastric bypass surgery has become the most widely used. 3 The procedure has helped morbidly obese individuals achieve long-term weight loss with a lower complication rate. 4 However, despite this improvement metabolic abnormalities have still been noted that have prompted further investigation by many researchers. Since RYGB is a relatively new procedure that has gained popularity in the last decade, attention is now being shifted to examining long-term physiological changes including renal and urological complications. Although complications after RYGB relative to bone mineral metabolism and nutrient deficiencies have been studied, 5 7 the potential for hyperoxaluria and nephrolithiasis has only recently been recognized. 8 11 In addition Asplin and Coe determined that while hyperoxaluria is less severe in patients who underwent RYGB compared to JI bypass, a relative increase is still seen compared to normal subjects. 12 Our institution is located in a rural region with a high prevalence of morbid obesity. As a result we perform a large number of bariatric procedures using open and laparoscopic approaches. Since we are now accumulating a population of patients who have undergone gastric bypass, we are focusing our attention on the potential long-term complications in this population. We present a longitudinal prospective study in which we assessed risk factors for nephrolithiasis after RYGB surgery. MATERIALS AND METHODS All patients older than 18 years who were seen in the High-Risk Obesity Clinic at our institution and who were being considered for open or laparoscopic RYGB surgery were considered for enrollment. However, patients with a history of stone disease or a history of inflammatory bowel disease were excluded from study. It was estimated that 100 patients would need to be enrolled to obtain full data from approximately 50 patients for statistical analysis. A total of 112 patients were enrolled in the study. Between November 2006 and November 2007, 52 patients underwent RYGB surgery at our institution. Of these patients 7 elected not to complete the postoperative 24-hour urinalysis. There were 60 patients who were enrolled who have not yet undergone gastric bypass surgery or were subsequently determined not to be a candidate for the surgery. Ultimately 45 morbidly obese patients who were enrolled did complete the study. Each of the 45 patients underwent a preoperative 24- hour urine collection (Mission Pharmacal, San Antonio, Texas) as well as measurement of several serum values. Serum calcium, creatinine, albumin, uric acid and parathyroid hormone were collected from each patient before undergoing RYGB surgery. A 24-hour urine collection and the serum studies were then repeated at 6 to 12 months postoperatively (median 9.6 months, range 6.1 to 14.5). The 24-hour urine collection measured urinary calcium, oxalate, uric acid, citrate, ph, sodium, phosphorus and total urinary volume. Supersaturation factors for calcium oxalate, brushite, sodium urate, struvite and uric acid were also assessed. All 24-hour urine specimens were kept refrigerated during the collection period. Laboratory studies thought to be associated with risk of stone formation were compared preoperatively vs postoperatively using a Wilcoxon signed-rank test. Median preoperative and postoperative values were calculated for each variable, in addition to the change of each value when the preoperative median was compared to the postoperative median. Interquartile range Q1, Q3 represents 25th and 75th percentile of distribution. McNemar s test was used to determine if the percent of abnormal laboratory values underwent a statistically significant change after RYGB. For both statistical methods a p value was calculated for the change in each variable, with p 0.05 considered statistically significant. All 45 patients underwent a RYGB procedure, 28 open and 17 laparoscopic. In all patients a small proximal gastric pouch was created. The proximal jejunum was divided distal to the ligament of Treitz. The distal end of the jejunum was anastomosed to the proximal gastric pouch. A Roux limb was then created with a median length of 150 cm (range 130 to 155). The integrity of the pouch and the anastomosis were verified before closure. Postoperatively all patients were prescribed 600 mg calcium citrate twice daily and 800 IU vitamin D daily. A daily chewable multivitamin containing 400 IU vitamin D and 100 mg calcium was also taken twice daily. Patients were advised to drink a minimum of 60 to 80 ounces of fluid per day. They were also prescribed 60 to 80 gm protein intake daily. No specific recommendations regarding sodium or caloric intake were provided. Calcium and vitamin D supplementation was initiated at 2 months postoperatively. The study received approval from the Geisinger Medical Center institutional review board. RESULTS A total of 45 patients who underwent RYGB completed preoperative and postoperative testing with serum chemistry studies and 24-hour urine collections. These patients consisted of 37 women and 8 men, and all had a preoperative BMI of 35 kg/m 2 or greater. Median preoperative BMI was 44.2 kg/m 2 (range 35 to 66.7) and median age was 48 years (range 28 to 70). Several 24-hour urine values asso-

2336 RISK FACTORS FOR NEPHROLITHIASIS AFTER ROUX-EN-Y GASTRIC BYPASS SURGERY ciated with a risk of stone formation showed a postoperative change that reached statistical significance (table 1). In particular urinary oxalate increased from a median of 32 to 40 postoperatively (p 0.01). A statistically significant decrease in median urinary citrate was also observed (p 0.0006). Urinary total volume also decreased from 1.8 to 1.44 l per day (p 0.002). SS for calcium oxalate increased from a median of 1.27 to a median of 2.23 (p 0.001). No other SS values underwent statistically significant increases. Additional findings included a statistically significant decrease in urinary uric acid levels (p 0.0001) and a statistically significant decrease in urinary calcium (p 0.0001). Postoperative decreases were also seen in urinary sodium and phosphorus. Decreases, considered clinically insignificant, were observed in serum calcium, creatinine, uric acid and albumin postoperatively (table 2). A statistically significant change in the percent of patients with abnormal 24-hour urine parameters from the preoperative period to the postoperative period was also noted (table 3 and figure). In particular 20% of the patients had hypercalciuria (greater than 250 mg per 24 hours) preoperatively with a range of 261 to 745 mg. Postoperatively this decreased to 4% of the population with a range of 372 to 378 mg (p 0.0047). Of the patients 11% demonstrated hyperoxaluria (greater than 45 mg per 24 hours) preoperatively with a range of 47 to 80 mg. Only 1 patient (2%) had a urinary oxalate level greater than 60 mg preoperatively. This patient did not demonstrate hyperoxaluria in the postoperative 24-hour urine collection. Postoperatively hyperoxaluria increased to 42% of the population (19 patients), with a range of 46 to 126 mg (p 0.0017). There were 6 patients (13%) with urinary oxalate levels greater than 60 mg and 2 (4%) with urinary oxalate levels greater than 100 mg postoperatively. Of the postoperative hyperoxaluria that was noted 90% was de novo. Hyperuricosuria (greater than 700 mg per 24 hours) was present in 33% (15 patients) preoperatively (range 715 to 1,229 mg). This decreased to only 2% (1 patient) of the population postoperatively (p 0.0001). Hypocitraturia (less than 320 mg per 24 hours) was present in 13% (6 patients) preoperatively (range 58 to 297 mg). This increased to 31% (14 patients) postoperatively (range 36 to 302 mg). Of the hypocitraturia noted postoperatively 71% was de novo. Relative SS for calcium oxalate (greater than 2.0) was found in 20% (9 patients) preoperatively (range 2.07 to 4.48). This increased to 60% (27 patients) postoperatively (range 2.01 to 8.75, p 0.001). Of interest 70% of these patients were found to have a de novo increase postoperatively. Relative SS for uric acid (greater than 2.0) was found in 35% (16 patients) preoperatively (range 2.12 to 5.66). This increased to 55% (25 patients) postoperatively (range 2.02 to 6.95). Again a de novo increase was found in 60% of these patients postoperatively. Low urine volume of less than2lin24hours was found in 53% (24 patients) and of less than 1 l in 24 hours was found in 36% (16 patients) postoperatively. No demonstrable association was found linking preoperative weight or net loss of weight at the postoperative collection to any of the previously noted laboratory changes (data not shown). Finally it should be noted that no symptomatic episodes of urolithiasis were found in this patient population during followup. DISCUSSION We performed a prospective longitudinal study to assess for the potential changes in serum and urine factors that have been associated with the risk of stone formation. Serum laboratory studies and 24- hour urine collections were performed in 45 patients before and after undergoing RYGB surgery. Several Table 1. Preoperative and postoperative 24-hour urine values Preop Median (Q1, Q3) Postop Median (Q1, Q3) Change (post-pre) (Q1, Q3) p Value Calcium (mg/day) 176 (128, 232) 135 (70, 171) 63 ( 117, 8) 0.0001 Oxalate (mg/day) 32 (25, 41) 40 (31, 50) 8 ( 7, 20) 0.01 Uric acid (mg/day) 583 (411, 779) 412 (308, 477) 271 ( 438, 53) 0.0001 Citrate (mg/day) 675 (367, 1,071) 456 (302, 824) 183 ( 409, 89) 0.0006 ph 6.03 (5.59, 6.27) 5.75 (5.40, 6.18) 0.10 ( 0.72, 0.35) 0.18 Total vol (l/day) 1.8 (1.08, 2.30) 1.44 (0.86, 1.80) 0.46 ( 1.03, 0.00) 0.002 Sodium (meq/day) 188 (150, 295) 143 (100, 175) 73 ( 132, 12) 0.0001 Phosphorus (mg/day) 896 (685, 1,339) 660 (514, 1,079) 246 ( 645, 2) 0.0018 CaOx (SS) 1.27 (0.79, 1.92) 2.23 (1.34, 3.59) 0.76 (0.03, 1.54) 0.001 Brushite (SS) 1.2 (0.51, 1.79) 0.65 (0.37, 1.22) 0.31 ( 0.98, 0.27) 0.046 Sodium urate (SS) 2.56 (1.40, 5.85) 1.84 (0.90, 2.96) 0.73 ( 2.86, 0.92) 0.029 Strurite (SS) 0.41 (0.12, 1.02) 0.44 (0.09, 0.70) 0.05 ( 0.68, 0.42) 0.86 Uric acid (SS) 1.52 (0.68, 2.77) 2.14 (1.03, 4.04) 0.32 ( 0.72, 1.86) 0.2 Creatinine urine (mg/day) 1,325 (1,128, 1,570) 1,080 (936, 1,350) 318 ( 618, 47) 0.0001

RISK FACTORS FOR NEPHROLITHIASIS AFTER ROUX-EN-Y GASTRIC BYPASS SURGERY 2337 Table 2. Preoperative and postoperative serum laboratory values Preop Median (Q1, Q3) Postop Median (Q1, Q3) Change (post-pre) (Q1, Q3) p Value Parathyroid hormone (pg/ml) 38 (28, 52) 39 (28, 56) 3 ( 5, 11) 0.15 Uric acid (mg/dl) 6.1 (5.0, 7.3) 4.8 (4.2, 5.4) 1.2 ( 2.0, 0.6) 0.0001 Calcium (mg/dl) 9.6 (9.40, 9.90) 9.4 (9.1, 9.6) 0.3 ( 0.5, 0.0) 0.0008 Albumin (gm/dl) 4.3 (4.1, 4.4) 4.1 (3.9, 4.2) 0.2 ( 0.4, 0.1) 0.0001 Creatinine (mg/dl) 0.8 (0.8, 0.9) 0.7 (0.7, 0.9) 0 ( 0.1, 0.0) 0.0007 other studies have commented on stone formation in patients after undergoing RYGB surgery, although many of these patients had a history of stone disease. 12 To get a more accurate representation of stone risk in this patient population we excluded any patients with a history of stone disease before undergoing RYGB. We observed several statistically significant changes postoperatively that could contribute to an increased risk of stone formation. We found a statistically significant increase in urine oxalate excretion and in relative SS for calcium oxalate. Certainly an increase in urinary oxalate could contribute to stone formation and increased oxalate absorption is a known complication in patients undergoing small bowel resection. We also saw a significant decrease in urinary citrate in the postoperative 24-hour urine collections. Because citrate is a potent inhibitor of calcium oxalate urolithiasis, hypocitraturia could contribute to nephrolithiasis after RYGB. We also noted a statistically significant decrease in total urinary volume in this study, which can contribute to the formation of stones. Duffey et al recently published a prospective study evaluating 24 patients at 90 days after RYGB. 11 Similar to our data they found an increase in urinary oxalate and an increase in calcium oxalate SS. However, contrary to our data they did not find any decrease in urinary citrate after RYGB. Sinha et al reported on a cross-sectional study of 20 patients before gastric bypass, 8 patients 6 months after gastric bypass and 13 patients 12 months after gastric bypass. 9 Their data would suggest that patients after gastric bypass may experience decreased urinary citrate by 6 months after RYGB. Their data also displayed increased urinary oxalate and increased SS for calcium oxalate at 12 months but not at 6 months after bypass. Our data are prospective and consistent with their suggestive findings, but also showed that hyperoxaluria and increased SS for calcium oxalate occurred by 6 months after RYGB. Considering both studies it is possible that increased oxalate excretion as well as decreased citrate excretion, both known lithogenic risk factors, may worsen with time after bariatric surgery. However, our study was not designed to assess for potential changes in lithogenic risk factors over time. Asplin and Coe also studied hyperoxaluria in patients in whom kidney stones developed after undergoing RYGB. 12 They reviewed 132 patients in whom stones developed after RYGB and only 1 had a history of stone disease. They compared a single 24- hour urine in patients after a symptomatic episode of renal lithiasis after bariatric surgery to 24-hour urines in patients who had not undergone bariatric surgery. They concluded that patients in the bariatric surgery group tended to have a lower oxalate excretion than those who had undergone JI bypass, but that urinary oxalate excretion after bariatric surgery was greater than in routine patients with stones and in normal subjects. JI bypass is an antiquated weight loss surgery that is no longer performed due to the severe side effects. 2 One of the many complications of this sur- Table 3. Abnormal laboratory and 24-hour urine values % Preop % Postop p Value* Calcium 250 mg/day or greater 22 4 0.0047 Oxalate 45 mg/day or greater 11 42 0.0017 Uric acid 700 mg/day or greater 36 2 0.0001 Citrate 320 mg/day or less 13 29 0.052 CaOx SS 2.0 or greater 22 60 0.0001 Brushite SS 2.0 or greater 18 11 0.32 Uric acid SS 2.0 or greater 36 51 0.11 * McNemar s test. Percentage of abnormal laboratory and 24-hour urine values before and after surgery.

2338 RISK FACTORS FOR NEPHROLITHIASIS AFTER ROUX-EN-Y GASTRIC BYPASS SURGERY gery was renal failure due to oxalate nephropathy. Hyperoxaluria is also a complication of RYGB, although urinary oxalate levels do not appear as drastically increased with RYGB compared to JI bypass. Two small series to date have reported on oxalate nephropathy after RYGB. Nelson et al reported on a small cohort of patients in whom enteric hyperoxaluria developed after RYGB. 8 While calcium oxalate nephrolithiasis developed in the majority of this group, 2 patients had the more concerning diagnosis of oxalate nephropathy. Neither of these patients had a history of nephrolithiasis or renal insufficiency and oxalate nephropathy was confirmed in both via kidney biopsy. Both patients ultimately required dialysis. In another recent series Nasr et al also reported oxalate nephropathy in 11 patients treated with RYGB. 13 In this series 8 patients underwent RYGB due to morbid obesity and 3 underwent RYGB due to gastric adenocarcinoma. While a large percentage of these patients had diabetes and preexisting renal insufficiency, biopsy proven oxalate nephropathy developed in all. Of these patients 8 ultimately had progression to end stage renal disease. While this appears to be a rarely reported event in patients undergoing RYGB, the outcome for the patients in these 2 series was dire. Hyperoxaluria should be recognized as a tangible risk after RYGB with the possibility of renal failure as an end result. The proposed increased risk of nephrolithiasis after RYGB is a relatively new phenomenon, only recently published in the literature. While it seems that these patients may have a predisposition for stone formation, the underlying clinical question is how we should intervene to help prevent this process. In 2007 Kleinman cited a retrospective study from the Mayo Clinic and made several suggestions on preventive measures. 14 As with other stone formers adequate hydration to produce at least 2 l of urine per day was recommended. However, it is clearly recognized that maintaining oral hydration is often difficult for these patients in the immediate months after RYGB surgery. Limiting fat intake and calcium supplementation was also recommended. These measures may aid in preventing increased oxalate absorption by ensuring adequate calcium to complex with free oxalate. Our data showing increased urinary oxalate excretion after RYGB would support these recommendations. These are simple and intuitive recommendations that physicians may give to patients after RYGB. Clearly further studies after implementing these recommendations would be necessary to confirm the efficacy of these measures. In addition to hyperoxaluria, increased supersaturation of calcium oxalate and decreased urinary volume, we also observed a statistically significant decrease in urinary citrate levels. As citrate is a potent inhibitor of stone formation, we also believe that it may be reasonable to provide citrate supplementation to this patient population. Our prospective study has shown that there are several changes seen in the 24-hour urine collections of patients who have undergone RYGB that may increase the risk of stone formation. These metabolic changes were seen as soon as 6 months postoperatively. None of the patients in our study have had any documentation of nephrolithiasis to date, although it is clear that even patients without a history of stone disease have risk factors for stone formation after undergoing RYGB and renal stones may subsequently develop. As previously mentioned several recommendations could be given to patients to help ameliorate these risk factors including increased fluid intake, decreased fat intake, and calcium and citrate supplementation. Our study was a prospective trial to assess for changes of known risk factors in patients undergoing RYGB surgery. Although the numbers are somewhat small, they were sufficient for us to demonstrate statistically significant changes in a number of known risk factors within 6 months after RYGB surgery. However, our study is limited by the lack of demonstration of stone development in these patients after RYGB. Although it is known that these patients are at increased risk for stone development due to various factors including hyperoxaluria, the exact incidence of stone formation is not known and, therefore, the true implications for health care costs are uncertain. 11,12,15 Our study would indicate that relatively simple and inexpensive interventions may potentially decrease the risk of stone formation in this population. A larger, more comprehensive study is indicated to assess the true incidence of this problem and the potential benefits of the therapeutic measures suggested by our study. CONCLUSIONS In this prospective trial we examined risk factors for nephrolithiasis after RYGB surgery. Several changes were noted in the postoperative 24-hour urine collections that may increase the risk of stone formation. There are several clinical instructions that may be easily performed that could possibly help to counteract this increased risk. Further investigation should be performed to assess the actual incidence of stone formation in this population as well as the role of these clinical instructions.

RISK FACTORS FOR NEPHROLITHIASIS AFTER ROUX-EN-Y GASTRIC BYPASS SURGERY 2339 REFERENCES 1. Sturm R: Increases in morbid obesity in the USA: 2000-2005. Public Health 2007; 121: 492. 2. Griffen WO Jr, Bivins BA and Bell RM: The decline and fall of the jejunoileal bypass. Surg Gynecol Obstet 1983; 157: 301. 3. Samuel I, Mason EE, Renquist KE et al: Bariatric surgery trends: an 18-year report from the International Bariatric Surgery Registry. Am J Surg 2006; 192: 657. 4. Kendrick ML and Dakin GF: Surgical approaches to obesity. Mayo Clin Proc 2006; 81: S18. 5. De Prisco C and Levine SN: Metabolic bone disease after gastric bypass surgery for obesity. Am J Med Sci 2005; 329: 57. 6. Collazo-Clavell ML, Jimenez A, Hodgson SF et al: Osteomalacia after Roux-en-Y gastric bypass. Endocr Pract 2004; 10: 195. 7. Chaves LC, Faintuch J, Kahwage S et al: A cluster of polyneuropathy and Wernicke-Korsakoff syndrome in a bariatric unit. Obes Surg 2002; 12: 328. 8. Nelson WK, Houghton SG, Milliner DS et al: Enteric hyperoxaluria, nephrolithiasis, and oxalate nephropathy: potentially serious and unappreciated complications of Roux-en-Y gastric bypass. Surg Obes Relat Dis 2005; 1: 481. 9. Sinha MK, Collazo-Clavell ML, Rule A et al: Hyperoxaluric nephrolithiasis is a complication of Roux-en-Y gastric bypass surgery. Kidney Int 2007; 72: 100. 10. Durrani O, Morrisroe S, Jackman S et al: Analysis of stone disease in morbidly obese patients undergoing gastric bypass surgery. J Endourol 2006; 20: 749. 11. Duffey BG, Pedro RN, Makhlouf A et al: Rouxen-Y gastric bypass is associated with early increased risk factors for development of calcium oxalate nephrolithiasis. J Am Coll Surg 2008; 206: 1145. 12. Asplin JR and Coe FL: Hyperoxaluria in kidney stone formers treated with modern bariatric surgery. J Urol 2007; 177: 565. 13. Nasr SH, D Agati VD, Said SM et al: Oxalate nephropathy complicating Roux-en-Y gastric bypass: an underrecognized cause of irreversible renal failure. Clin J Am Soc Nephrol 2008; 3: 1676. 14. Kleinman JG: Bariatric surgery, hyperoxaluria, and nephrolithiasis: a plea for close postoperative management of risk factors. Kidney Int 2007; 72: 8. 15. Bhavin PN, Passman CM, Fernandez A et al: Prevalence of hyperoxaluria after bariatric surgery. J Urol 2009; 181: 161.