None of our patients were on sulindac, nabumetone or selective COX2 inhibitors. Few patients were on gold or d-penicillamine. Nephrotoxicity of methotrexate ...
Original Article
A Longitudinal Study of Serum Creatinine Levels in Patients of Rheumatoid Arthritis on Long Term NSAID Therapy E Pathan*, Shilpa Gaitonde*, S Rajadhyaksha**, A Sule***, Gayatri Mittal+, VR Joshi++
Abstract Aim : Primary : To study the effect of long term NSAID therapy on serum creatinine in patients of rheumatoid arthritis. Secondary : To study the effect of discontinuation, reduction in the dose or continuation of NSAID and of rechallenge. Material and Methods : Case records of RA patients with a minimum two years of follow up were analysed. Age, sex, duration of RA, type, dose and duration of NSAID and DMARD therapy, co-morbid conditions and serial serum creatinine levels were charted. Results : Ninety nine case records were studied. Incidence of abnormal creatinine level (renal insufficiency) defined as rise in creatinine equal to or above the upper limit of normal was 27.7%. This rise was asymptomatic in all patients. No NSAID was particularly associated with an increased risk in renal insufficiency. The rise of serum creatinine was reversible in most patients irrespective of discontinuation or continuation of NSAID but settled at a higher level. Rechallenge resulted in rise of serum creatinine in 50% patients. Hypertension, DM, IHD and diuretics carried a higher but not statistically significant risk of renal insufficiency. Conclusion : NSAID-induced asymptomatic rise of creatinine in patients of RA on long term NSAIDs is common. It is mostly reversible. Regular monitoring of serum creatinine is essential.
INTRODUCTION
N
on-steroidal anti-inflammatory drugs (NSAIDs) are one of the most commonly prescribed drugs. Their use carries a high risk of gastrointestinal and renal side effects. Gastrointestinal toxicity is well studied, but there is paucity of data on renal effects of long term NSAID therapy. NSAID-induced nephrotoxicity includes acute tubular necrosis, acute tubulointerstitial nephritis, glomerulonephritis, renal papillary necrosis, salt and water retention, hypertension, hyperkalaemia and hyporeninaemic hypoaldosteronism.1 Subclinical nephrotoxicity has also been reported.2 Chronic renal failure on long term NSAID treatment is debatable, some studies supporting3-5 while others refuting it.6-7 These studies have either analyzed NSAID effect in those who had an already existing chronic renal failure3-5 or were one point studies.2-6 Questions like reversibility, time to *Research Fellow; **Clinical Observer; ***Clinical Fellow; +Clinical Assistant; ++Consultant in Rheumatology; Division of Rheumatology, PD Hinduja National Hospital and Medical Research Centre, Veer Savarkar Marg, Mahim, Mumbai 400 016. Received : 31.10.2002; Revised : 3.4.2003; Accepted : 26.8.2003
JAPI • VOL. 51 • NOVEMBER 2003
recovery, effect of rechallenge, and its advisibility remain unanswered. These answers need long-term studies. A few recent studies have reported on longitudinal serum creatinine levels in patients of early rheumatoid arthritis (RA), but the duration of follow up has been short.8-9 We have analysed the serum creatinine levels in patients of RA on long term NSAID therapy. The effects of discontinuation, reduction or continuation and of rechallenge have also been analysed.
MATERIAL AND METHODS This is a retrospective study. Case records of RA patients, with a minimum 2 years follow up and who attended our clinic between March and May 2000 were screened. Patients not on NSAID or with incomplete data were excluded. Age, sex, disease duration, dose and duration of NSAIDs, weight and BP were noted. Serum creatinine as recorded at approximately 1,1.5,3,6,9,12,18,24,30,36,42,48,54,60,66,72,84,96,108 and 120 months from the first visit was noted. Almost all patients were on a DMARD necessitating regular S. creatinine monitoring. Reports of urine examination, ultrasound and kidney histology were recorded as available. Total NSAID 1045
exposure (calculated in terms of duration of illness), concomitant medications (DMARD, other nephrotoxic drugs like diuretics and ACE inhibitors) and associated disorders like hypertension (HT), diabetes mellitus (DM), congestive cardiac failure (CCF), renal calculi, vasculitis and urinary tract infection were recorded. A rise in creatinine up to or above the upper limit of normal was taken as abnormal (renal insufficiency). The effect of discontinuation or otherwise of NSAID, time to recovery of serum creatinine and effect of rechallenge were noted. Mean ± SD of lowest and highest creatinine levels and percentage rise in creatinine were calculated in those with and without renal insufficiency. The mean post-recovery creatinine was compared with the base-line level. Data measured on continuous scale was presented as mean along with standard deviation and compared using Student’s t-test. Categorical data was presented as proportion or percentage and compared using chi-square test. Odds ratio (OR) was used to indicate relative risk, p < 0.05 was taken as statistically significant.
RESULTS One hundred and eight case records were screened. Nine were excluded. Creatinine remained within normal range in 72 patients (group 1), touched the upper limit of normal in 7 and crossed it in 20 (group 2 = 20+7), giving a 27.7% (27/99) incidence of renal insufficiency (RI). The maximum number of patients belonged to the 5th decade in both the groups. Age and sex distribution and duration of illness of the patients in the two groups were comparable (p > 0.05) (Table 1). The duration of illness was more than 5 and 10 years in 75% and 33% patients respectively (Table 2). Parallel urine reports, available in six patients of group 2 at the time of elevated serum creatinine, was normal in 3, showed trace albuminuria in two and a 2g 24 hour proteinuria in one Table 1 : Age distribution in groups 1 and 2 Age (years)
Group 1 (n=72)
Group 2 (n-27)
21-30 31-40 41-50 51-60 61-70 > 70
5 (6.94%) 11 (15.2%) 20 (27.7%) 19 (26.38%) 15 (13.8%) 2 (2.7%)
— 3 (11.1%) 10 (37.03%) 7 (25.92%) 7 (25.92%) —
Table 2 : Duration of illness in groups 1 and 2
< 3 years 3-5 years 5-10 years > 10 years
Group 1 (n=72) 7 (9.7%) 9 (12.5%) 33 (45.8%) 23 (31.94%)
Group 2 (n=27) 4 3 10 10
Both groups are comparable (p - value > 0.05)
1046
Table 3 : Risk factor profile in both groups Risk factor Hypertension Diabetes mellitus IHD Cyclosporin Diuretics D-penicillamine HTN+IHD HTN+diuretics HTN+ACEI HTN+ACEI+diuretics
(14.8%) (11.1%) (37.03%) (37.03%)
Group 1 (n=72)
Group 2 (n=27)
Odds ratio
14 (19.4%) 2 (2.7%) 3 (4.16%) 0 2 (2.7%) 2 (2.7%) 2 (2.7%) 2 (2.7%) 4 (5.55%) 1 (1.38%)
7 (25.92%) 2 (7.4%) 2 (7.4%) 1 (3.7%) 2 (7.4%) 2 (7.4%) 1 (3.7%) 2 (7.4%) 1 (3.7%) 1 (3.7%)
1.45 2.87 1.8 — 2.87 2.87 1.34 2.77 0.39 1.07
P - value > 0.05 in each category. IHD - Ischaemic heart disease; HTN-Hypertension; ACEI-Angiotensin converting enzyme inhibitor
Table 4 : DMARD usage in both groups DMARD used
Both groups are comparable (p value > 0.05).
Duration of illness (years)
patient. One group 1 patient on d-penicillamine had nephrotic range proteinuria. In group 2 ultrasound was normal in two patients, showed prostatic enlargement in two others and a renal calculus in one patient. Kidney biopsy in the patient with proteinuria (group 2) revealed tubulointerstitial nephritis with membranous glomerulopathy. The frequency of HT, DM and ischaemic heart disease (IHD) was similar in the two groups (Table 3) (p > 0.05), the odds ratio (OR) for development of RI being 1.45, 2.87, 1.8 and 2.87 for patients of HT, DM, IHD and diuretic therapy respectively. Combined OR for HT with IHD was 1.34 and for HT with diuretic therapy 2.7. ACE inhibitors (ACEI) seemed to lower the risk (OR 0.39). Combined ACEI-diuretic therapy did not decrease the risk of renal insufficiency (OR 1.07). None of these trends however attained statistical significance. DMARD use was comparable in the two groups (Table 4). No patient on gold therapy had a rise in creatinine. Artaminerelated proteinuria with normal s. creatinine (Gr I) was seen in one patient. Cyclosporine and diuertic-related rise of serum creatinine was seen in one patient each. It normalized on drug discontinuation. In one patient, methotrexate withdrawal resulted in recovery, but, subsequently there was a rise in creatinine which responded to reduction of NSAID dose. NSAIDs used were indomethacin, ibuprofen, diclofenac, piroxican, naproxen and nimesulide. None was particularly associated with a rise in serum creatinine. Cumulative NSAID intake could not be calculated as patients had used different
Methotrexate Methotrexate + SSZ Sulfasalazine (SSZ) d-penicillamine Cyclosporine Colchicine Gold salts Azathioprine Levamisole
Group 1 (n=72)
Group 2 (n=27)
56 (77.7%) 4 (5.5%) 5 (6.94%) 6 (8.3%) 0 1 (1.38%) 2 (2.7%) 3 (4.16%) 1 (1.38%)
21 (77.7%) 4 (14.8%) 0 1 (3.7%) 2 (7.4%) 2 (7.4%) 0 0 0
JAPI • VOL. 51 • NOVEMBER 2003
NSAIDs at different periods of time. Hence, the total burden of NSAID was equated to the disease duration. There was no difference in the two groups. Fig. 1 shows the outcome of the actions taken in group 2 patients. NSAID had to be permanently discontinued in 9/27 patients. No particular action seemed to be of specific value. In all the 27 patients creatinine rise was asymptomatic and detected because of regular testing. Creatinine normalised between 2 to 36 weeks of NSAID discontinuation. The exact time to recovery could not be ascertained. Duration of illness (NSAID exposure) or the presence of additional risk factors did not affect the time to recovery. The lowest creatinine of the two groups, the degree of rise recorded in each group, and the inter-group difference in creatinine rise were compared (Table 5). The post-recovery creatinine was compared to initial creatinine levels in group 2. The differences were statistically significant (p < 0.05).
DISCUSSION We chose to study the effect of NSAIDs on kidney in RA patients, as it is the commonest inflammatory rheumatic disorder needing long term NSAIDs. Further, renal disease primarily due to RA is extremely rare.10-13 Serum creatinine is the most commonly used marker of RI.
Table 5 : Comparison of creatinine values in both groups Value studied Lowest creat (a) Highest creat (b) % rise (c) Stabilized creat (d)
Group 1 (n=72) 0.73 ± 0.12 1.11 ± 0.20 55.8 ± 39.7% —
Group 2 (n=27) 0.82 1.64 105.33 0.90
± ± ± ±
0.13 0.56 81.39% 0.22
P < 0.05 < 0.05 < 0.05 < 0.05*
*Stabilized vs initial creatinine value in Group 2. Values Mean ± SD
Table 6 : Distribution of NSAID usage in both groups NSAID used
Group 1 (n=72)
Group 2 (n=27)
Flurbiprofen Diclofenac Ibuprofen Piroxican Indomethacin Naproxen Nimesulide
5 (6.94%) 35 (48.6%) 10 (13.8%) 16 (22.2%) 18 (25%) 7 (9.72%) 20 (27.7%)
0 19 (70.3%) 3 (11.1%) 6 (22.2%) 9 (33.3%) 2 (7.4%) 11 (40.7%)
Serum creatinine can be low in RA patients due to muscle wasting.14 Therefore, the upper limit of normal was taken as abnormal. Creatinine clearance the other measure of renal
Fig. 1 : Patients in whom creatinine touched upper limit or crossed upper limit of normal (n=27) JAPI • VOL. 51 • NOVEMBER 2003
1047
function is time consuming, often unreliable, and not routinely practiced.15 Creatinine elevation was asymptomatic in all patients. Asymptomatic reversible impairment of renal function has been reported.16 This emphasizes the importance of regular serum creatinine monitoring while on NSAIDs. Koseki et al8 (235 early RA cases, mean follow up 42 months) and van Jaarsveld et al9 (419 cases, mean follow up 90 weeks) have reported 1.5% and 3.6% incidence of renal toxicity respectively. The high figure of 27.7% in the present study is possibly due to a longer follow up period (at least 5 years in 75% of the cases) and regular s. ceratinine monitoring. Without this, reversible asymptomatic creatinine elevation would have been missed. The exact underlying pathology could not be ascertained in our study as renal biopsy was not carried out as a routine and not indicated as creatinine levels returned to normal in majority of the patients. Calvo Alen et al2 suggest that NSAIDs produce a pattern of renal damage that is consistent with subclinical interstitial nephropathy. Boers et al found that although subclinical renal dysfunction occurs frequently in patient with RA, no single pathological process emerged as a cause of renal dysfunction. 17 Others have found evidence of tubulointerstitial nephritis associated with proteinuria.18-19 Multiple risk factors have been implicated. A 13% incidence of renal toxicity with short term NSAID therapy has been reported in the very elderly (mean age 87 years) by Gurwitz et al.20 None of our patients was over 80 years of age. We noted an increased propensity to nephrotoxicity in male patients. Sandler et al5 have reported a two-fold risk of chronic renal disease in men older than 65 yeas on daily NSAIDs. Studies co-relating total NSAID intake to chronic renal disease have reported both no effect in those who consumed one or more analgesics for more than three years in some21 as against a two-fold risk in previous daily users in others.5 An increased risk of end-stage renal disease with life-time consumption of more than 5000 pills has been reported.22 Though the relationship of total NSAID consumption to renal toxicity could not be ascertained in our study, there was no significant difference between the two groups based on the duration of presumed NSAID consumption i.e. duration of RA. Both long and short half-life NSAIDs have been implicated in the pathogenesis of nephrotoxicity.3,23-24 In our study, there was no evidence of increased risk with any particular NSAID. None of our patients were on sulindac, nabumetone or selective COX2 inhibitors. Few patients were on gold or d-penicillamine. Nephrotoxicity of methotrexate has been described in literature.25 In one patient, reducing the dose of methotrexate resulted in reversal of creatinine to normal. However, subsequently the patient developed a rise in creatinine that required reduction in the NSAID dose. Cyclosporine-related creatinine elevation26 was seen in one patient. Thus in most patients the creatinine rise could be attributed to NSAIDs alone. We found no significant difference in the incidence of 1048
HT, DM and IHD between the two groups (Table 3). However the number of patients with these disorders is small. Eight (group 2) patients could not discontinue or reduce the dose of NSAID though all were so advised. This permitted us to study the effect of continuation of NSAID on serum creatinine. Surprisingly serum creatinine returned to normal range in seven of them (Fig. 1). Reduction or discontinuation of NSAID resulted in normalisation of serum creatinine in all but two cases. This suggests reversibility of renal toxicity. Similar observations have been reported by van Jaarsveld.9 Rechallenge with NSAID in 16 patients resulted in creatinine elevation in eight. Whelton et al have reported reversible worsening of renal function in patients with asymptomatic RI (s. creatinine between 130-270 mmol/L).23 Asymptomatic RI may be an important risk factor and may have been operative in our patients since the mean lowest creatinine and the degree of creatinine rise in group two patients was significantly higher than that of group 1 (Table 5). Barter et al27 have reported that the decline in renal function with each dose of NSAID is transient and no overall adverse effect on renal function occurred with chronic dosing. Recovery time ranged from 2 to 36 weeks. Recovery within 24-72 hours of cessation of NSAID therapy has been reported.28 Neither duration of exposure, nor associated HT, DM or IHD, affected the time to recovery. The mean value at which creatinine stabilized after recovery of renal impairment, was significantly higher than the lowest value in group 2, suggesting that the effect of NSAID on renal function may be partially reversible27 but an additive phenomenon. Similar findings have been reported by Caspi et al29 in elderly patients on mini-dose aspirin. A higher base-line serum creatinine seems to predispose to acute episodes of renal insufficiency.
CONCLUSIONS Despite the inherent limitations of a retrospective study, the data allows some valid conclusions. Mild reversible rise in s. creatinine is common on long term NSAID therapy. No particular NSAID seems more nephrotoxic. Renal impairment is usually asymptomatic warranting regular monitoring. Males and elderly may be more susceptible. Most patients recover, but not always and recovery may not be complete. Time to recovery is variable. Discontinuation, reduction or continuation of NSAID does not affect recovery. Alteration in dose of other nephrotoxic drugs helps recovery. Rechallenge mandates careful monitoring.
REFERENCES 1.
Tse WY, Adu D. Non-steroidal anti-inflammatory drugs and the kidney. In: Davison AM, Cameron S, Grunfeld JP, Kerr DNB, Ritz E, Winerals CG, eds. Oxford Textbook of Nephrology, 2nd Edition, Oxford Publication, 1998:114556.
2.
Calvo-Alen J, Angeles De Cos M, Rodriguez-Valverde V, Escalladv R, Florez J, Arias M. Subclinical renal toxicity in rheumatic patients receiving long term treatment with nonsteroidal anti-inflammatory drugs. J Rheumatol 1994;214:174247. JAPI • VOL. 51 • NOVEMBER 2003
3.
Adams DH, Howie AJ, Michael J, McConkey B, Bacon P, Adu D. Non-steroidal anti-inflammatory drugs and renal failure. Lancet 1986;1:57-59.
Can Brummelen P, Pauwels EK, Cats A. Subclinical renal dysfunction in rheumatoid arthritis. Arthritis Rheum 1990;33:95-101.
4.
Pommer W, Bronder E, Klimpel A, Molzahn M, Helmert U, Greiser E. Regular intake of analgesic mixtures and risk of end stage renal failure. Lancet 1989;1:381.
18. Schrier RW, Henrich Wl. Nonsteroidial anti-inflammatory drugs caution still indicated. JAMA 1984;251:1301-2.
5.
Sandler DP, Burr R, Weinberg CR. Non-steroidal antiinflammatory drugs and the risk for chronic renal disease. Ann Int Med 1991;115:165-72.
6.
Fox DA, Jick H. Nonsteroidal anti-inflammatory drugs and renal disease. JAMA 1984;251:1299-1300.
20. Gurwitz JH, Avron J, Ross-Degnan, Lipsitz LA. Non-steroidal anti-inflammatory drug-associated azotemia in the very old. JAMA 1990;264:471-75.
7.
Richards IM, Fraser SM, Capell HA, Fox JG, Boulton-Jones JM. A survey of renal function in outpatients with rheumatoid arthritis. Clin Rheumatol 1988;7:267-71.
21. Murray T, Stolley PD, Anthony JC, Schinnar, Helper Smith E, Jeffreys JL. Epidemiologic study of regular analgesic use and end stage renal disease. Ann Intern Med 1983;143:1687-93.
8.
Koseki Y, Terai C, Moriguchi M, Uesato M, Kamatani N. A prospective study of renal disease in patients of early rheumatoid arthritis. Ann Rheum Dis 2001;60:327-31.
22. Perneger TV, Whelton PK, Klag MJ. Risk of kidney failure with the use of acetaminophen, aspirin and non-steroidal anti-inflammatory drugs. N Eng J Med 1994;331:1675-79.
9.
van Jaarsveld CHM, Jahangier ZN, Jacobs JWG, Blauuw AAM, et al. Toxicity of anti-rheumatic drgs in a randomized clinical trial of early rheumatoid arthritis. Rheumatology 2000;39:137482.
23. Whelton A, Stout RL, Spilman PS, Klassen DK. Renal effects of ibuprofen, piroxican and sulindac in patients with asymptomatic renal failure. A prospective randomised, crossover comparison. Ann Intern Med 1990;112:568-76.
10. Solomon MI, Gallo G, Poon TP, Goldblat MV, Tchertkoff V. The kidney in rheumatoid arthritis. A study based on renal biopsies. Nephron 1974;12:297-310.
24. Cook ME, Wallin JD, Thakur VD, Kadowitz PJ, Mc Namara DB, Garcia MM, Lipani JA, Poland M. Comparative effects of nabumetone, sulindac and ibuprofen on renal function. J Rheumatol 1997;24:1137-44.
11. Ramirez G, Lamber R, Bloomer A. Renal pathology in patients with rheumatoid arthritis. Nephron 1981;29:124-26. 12. Boers M, Croonen A, Dijkmans B, Breedveld FC, Eulderink F, Cats A, Weening JJ. Renal findings in rheumatoid arthritis: clinical aspects of 132 necropsies. Ann Rheum Dis 1987;46:658-63. 13. Helin H, Korpela M, Mustonen J, Pasternack A. Renal biopsy findings and clinicopathological correlations in rheuamtoid arthritis. Arthritis Rheum 1995;2:242-47. 14. Nived O, Strutfelt G, Westling H, White T. Is serum creatinine concentration a reliable index of renal function in rheumatic disease? Br Med J 1983;286:684-85. 15. Boers M, Dijkmans BAC, Breedveld FC, Mattie H. Erros in the prediction of creatinine clearance in patients with rheumatoid arthritis. Br J Rheum 1988;27:233-35. 16. Unsworth J, Struman S, Lunec J, Blake DR. Renal impairment associated with non-steroidal anti-inflammatory drugs. Ann Rheum Dis 1987;46:233-36. 17. Boers M, Dijkmans, Breedveld FC, Camps JAJ, Chang PC.
19. Brezin J, Katz SM, Schwartz AB, Chinitz JL. Reversible renal failure and nephrotic syndrome associated with non-steroidial anti-inflammatory drugs. N Engl J Med 1979;301:1271-73.
25. Seideman P, Muller-Surr R, Ekman E. Renal effects of low dose methotrexate in rheumatoid arthritis. J Rheumatol 1993;20:1126-28. 26. Altman RD, Perez GO, Sfakianakis GN. Interaction of cyclosporine A and non-steroidal anti-inflammatory drugs on renal function in patients with rheumatoid arthritis. Am J Med 1992;93:396-402. 27. Brater DC, Anderson SA, Brown Cartwright D, Toto RD. Effects of non-steroidal antiinflammatory drugs on renal function in patients with renal insufficiency and in cirrhotics. Am J Kidney Dis 1986;8:351-5. 28. Blackshear H, Napier JS, Davidman S, Stillman MT. Renal complications of non-steroidal anti-inflammatory drugs: identification and monitoring those at risk. Semin Arthritis Rheum 1985;14:163-75. 29. Caspi D, Lubart E, Graff E, Habot B, et al. The effect of minidose aspirin on renal function and uric acid handling in elderly patients. Arthritis Rheum 2000;43:103-8.
Browse
JAPI Website www.japi.org JAPI • VOL. 51 • NOVEMBER 2003
1049
