1
1CORRESPONDENCE
EFFECTSOF SULINDACON RENAL PROSTAGLANDINS AND GLOMERULAR FILTRATIONRATE To the Editor: The article by Eriksson et al [l] extends the evidence that sulindac has different effects from other nonsteroidal anti-inflammatory drugs on renal prostaglandins and glomerular filtration rate (GFR) in damaged kidneys. I wonder, however, why they measured only urinary 6keto-PGFi, (the breakdown product of prostacyclin). Both thromboxane (TX) As and prostaglandin (PG) Es have important effects on glomerular blood flow, GFR, and tubular electrolyte handling [2]. The renal production of these two eicosanoids can be assessed by urinary PGEz and TXBz (the stable breakdown product of TXAs), at least in females, in whom prostatic prostanoid production is not a confounding factor. Eriksson et al observed that proteinuria decreased with sulindac despite the absence of any change in GFR or urinary 6-ketoPGFi, excretion. It would be of interest to know the urinary TXBz excretion, as some animal models have demonstrated that thromboxane inhibition can reduce proteinuria in the absence of hemodynamic changes [3]. The other problem with the report is the calculation of fractional excretions (their Table III). A fractional excretion of sodium of around 10% is huge and, with the mean GFR of 30 ml/minute, would lead to a sodium excretion of 25 mmol/hour (three times that reported). Averaged out over the day, this would imply a sodium excretion of 600 mmol/day on a 144-mmol sodium diet. The fractional excretions of potassium seem to have similar errors in calculation.
1. Eriksson L-C. Sturfelt G. Thysell H, Wollheim FA. Effects of sulindac and naproxen on prostaglandin excretion in patients with impaired renal function and rheumatoid arthritis. Am J Med 1990; 89: 313-21. 2. Hart D. Lifschitz MD. Renal physiology of the prostaglandins and the effects of nonsteroidal antiinflammatory agents on the kidney. Am J Nephrol 1987; 7: 408-18. 3. Remuui G. lmberti L, Rossini M. eta/. Increased glomerular thromboxane synthesis as a possible cause of proteinuria in experimental nephrosis. J Clin Invest 1985; 75: 94-101. Submitted
October
3, 1990, and accepted
May 1, 1991
The Reply: Dr. Magner asks for additional eicosanoid measurements in our patients with impaired renal function and emphasizes that both PGEz and TXAz have important effects on renal blood flow and GFR and that these two eicosanoids can be assessed by urinary PGEz and TXB2. A number of methodologic as well as biologic variables can affect urinary prostaglandin measurements in human studies. Dr. Magner mentions contamination of urine by prostatic prostanoid production, which is usually circumvented by inclusion of only women in the studies. Equally important are the sampling procedure, storage conditions, and purification and assay methods. Measurements of urinary eicosanoids and their metabolites by radioimmunoassay (RIA) are highly subject to problems of sensitivity and specificity. Performance of RIA on unextracted or insufficiently purified urine and the use of antibodies with crossreactivity to structurally related compounds have been common sources of error. Unfortunately, immunologic characterization of commercial anti-eicosanoid sera is often limited to primary prostaglandins and TXB2, with limited information on the cross-reactivity of other major urinary metabolites such as the 2,3-dinor derivatives of PGIz and TXB2. PETERMAGNER,M.D., F.R.c.P.c., Thus, it is important to carefully F.R.A.C.P. University of Calgary validate the assay procedure and Calgary, Alberta, Canada use rigorous sample purification
206
August 1991 The American Journal of Medicine
Volume 91
in combination with highly specific antibodies to be able to draw firm conclusions of urinary eicosanoid data. In our study, we used an extensively validated RIA procedure for the stable hydration product of prostacyclin involving careful purification of the urine and the use of a specific and well-characterized antibody with a high-affinity constant towards 6-ketoPGFl,. The method gives urinary levels well in agreement with those obtained by gas chromatography-mass spectrometry [1,2]. The major reason why we did not measure PGE2 and TXBl was that we did not have access to fully assessed and reliable methods for these eicosanoids. Furthermore, there is convincing evidence for the renal origin of urinary 6-keto-PGFi, from studies employing different approaches [2-4], although this may be true also for PGE2 and TXB2. It also appears from several previous studies that RIA of 6keto-PGFI, in urine gives more reliable and reproducible results than assay of PGEz during sulindac treatment [5]. Since the main aim of the study was to investigate the influence of sulindac and naproxen on GFR and renal blood flow, i.e., cortical functions, it is logical that the excretion of a prostaglandin synthesized primarily in the cortex was monitored rather than one that is synthesized and acts predominantly in the medulla, such as PGE2. However, we agree with Dr. Magner that, in addition to urinary 6-keto-PGFi,, it would have been of great interest to know the excretion rates of both PGEz and TXB2 with respect to glomerular and tubular functions in this clinical setting. We observed that proteinuria decreased with sulindac treatment in the absence of any change in GFR and urinary 6keto-PGFl,. One explanation may be that sulindac induces a
1CORRESPONDENCE
EFFECTSOF SULINDACON RENAL PROSTAGLANDINS AND GLOMERULAR FILTRATIONRATE To the Editor: The article by Eriksson et al [l] extends the evidence that sulindac has different effects from other nonsteroidal anti-inflammatory drugs on renal prostaglandins and glomerular filtration rate (GFR) in damaged kidneys. I wonder, however, why they measured only urinary 6keto-PGFi, (the breakdown product of prostacyclin). Both thromboxane (TX) As and prostaglandin (PG) Es have important effects on glomerular blood flow, GFR, and tubular electrolyte handling [2]. The renal production of these two eicosanoids can be assessed by urinary PGEz and TXBz (the stable breakdown product of TXAs), at least in females, in whom prostatic prostanoid production is not a confounding factor. Eriksson et al observed that proteinuria decreased with sulindac despite the absence of any change in GFR or urinary 6-ketoPGFi, excretion. It would be of interest to know the urinary TXBz excretion, as some animal models have demonstrated that thromboxane inhibition can reduce proteinuria in the absence of hemodynamic changes [3]. The other problem with the report is the calculation of fractional excretions (their Table III). A fractional excretion of sodium of around 10% is huge and, with the mean GFR of 30 ml/minute, would lead to a sodium excretion of 25 mmol/hour (three times that reported). Averaged out over the day, this would imply a sodium excretion of 600 mmol/day on a 144-mmol sodium diet. The fractional excretions of potassium seem to have similar errors in calculation.
1. Eriksson L-C. Sturfelt G. Thysell H, Wollheim FA. Effects of sulindac and naproxen on prostaglandin excretion in patients with impaired renal function and rheumatoid arthritis. Am J Med 1990; 89: 313-21. 2. Hart D. Lifschitz MD. Renal physiology of the prostaglandins and the effects of nonsteroidal antiinflammatory agents on the kidney. Am J Nephrol 1987; 7: 408-18. 3. Remuui G. lmberti L, Rossini M. eta/. Increased glomerular thromboxane synthesis as a possible cause of proteinuria in experimental nephrosis. J Clin Invest 1985; 75: 94-101. Submitted
October
3, 1990, and accepted
May 1, 1991
The Reply: Dr. Magner asks for additional eicosanoid measurements in our patients with impaired renal function and emphasizes that both PGEz and TXAz have important effects on renal blood flow and GFR and that these two eicosanoids can be assessed by urinary PGEz and TXB2. A number of methodologic as well as biologic variables can affect urinary prostaglandin measurements in human studies. Dr. Magner mentions contamination of urine by prostatic prostanoid production, which is usually circumvented by inclusion of only women in the studies. Equally important are the sampling procedure, storage conditions, and purification and assay methods. Measurements of urinary eicosanoids and their metabolites by radioimmunoassay (RIA) are highly subject to problems of sensitivity and specificity. Performance of RIA on unextracted or insufficiently purified urine and the use of antibodies with crossreactivity to structurally related compounds have been common sources of error. Unfortunately, immunologic characterization of commercial anti-eicosanoid sera is often limited to primary prostaglandins and TXB2, with limited information on the cross-reactivity of other major urinary metabolites such as the 2,3-dinor derivatives of PGIz and TXB2. PETERMAGNER,M.D., F.R.c.P.c., Thus, it is important to carefully F.R.A.C.P. University of Calgary validate the assay procedure and Calgary, Alberta, Canada use rigorous sample purification
206
August 1991 The American Journal of Medicine
Volume 91
in combination with highly specific antibodies to be able to draw firm conclusions of urinary eicosanoid data. In our study, we used an extensively validated RIA procedure for the stable hydration product of prostacyclin involving careful purification of the urine and the use of a specific and well-characterized antibody with a high-affinity constant towards 6-ketoPGFl,. The method gives urinary levels well in agreement with those obtained by gas chromatography-mass spectrometry [1,2]. The major reason why we did not measure PGE2 and TXBl was that we did not have access to fully assessed and reliable methods for these eicosanoids. Furthermore, there is convincing evidence for the renal origin of urinary 6-keto-PGFi, from studies employing different approaches [2-4], although this may be true also for PGE2 and TXB2. It also appears from several previous studies that RIA of 6keto-PGFI, in urine gives more reliable and reproducible results than assay of PGEz during sulindac treatment [5]. Since the main aim of the study was to investigate the influence of sulindac and naproxen on GFR and renal blood flow, i.e., cortical functions, it is logical that the excretion of a prostaglandin synthesized primarily in the cortex was monitored rather than one that is synthesized and acts predominantly in the medulla, such as PGE2. However, we agree with Dr. Magner that, in addition to urinary 6-keto-PGFi,, it would have been of great interest to know the excretion rates of both PGEz and TXB2 with respect to glomerular and tubular functions in this clinical setting. We observed that proteinuria decreased with sulindac treatment in the absence of any change in GFR and urinary 6keto-PGFl,. One explanation may be that sulindac induces a
