Towards the Incidence of Acute Phosphate Nephropathy

In 2003, Desmeules et al.¹ reported a 71-yr-old female who developed acute renal failure after the use of an oral sodium phosphate (OSP) solution as bowel purgative before colonoscopy. Renal biopsy revealed tubular injury and abundant tubular calcium phosphate deposits, suggesting a pathophysiological link between phosphate ingestion and the pattern of renal injury. The authors proposed the term acute phosphate nephropathy (APhN). We subsequently reported 21 cases of APhN,² and additional case reports have followed.^3–6 APhN most commonly occurs in older white females, particularly those with history of hypertension and receiving therapy with renin-angiotensin inhibitors. It is estimated that approximately 14 million screening colonoscopies are performed each year in the United States,⁷ yet the incidence of APhN remains unknown.

Two large, observational, retrospective studies^8,9 in this issue of JASN aim to define that incidence and identify risk factors. The article by Brunelli et al.⁸ is a retrospective, case-control study of patients undergoing outpatient colonoscopy at three University of Pennsylvania Health System affiliates. It compares patients who subsequently developed acute kidney injury (AKI) to those who maintained normal renal function. AKI was defined by either a 25% or a 0.5-mg/dl increase in serum creatinine over the 6 mo after colonoscopy. Using these criteria, 141 of 2237 patients (6.3%) had AKI. Complete data were available for 116 of the subjects, who were then compared with 349 controls. Risk factors for the development of AKI included female gender, heart failure, and diuretic use. An association between OSP and AKI could only be demonstrated in patients who were simultaneously receiving inhibitors of the renin-angiotensin system.

The study by Brunelli et al.⁸ is limited by the small sample size and the permissive definition of AKI. It is likely that a significant but small percentage of patients who undergo colonoscopy develop mild, transient elevations in serum creatinine related to hypovolemia. In this study, the mean rise in serum creatinine among patients with AKI was only 0.41 mg/dl and no data on reversibility were available. In contrast, among the 21 cases of APhN we previously reported, the mean increase in serum creatinine was 2.9 mg/dl (mean baseline, 1.0 mg/dl; mean peak, 3.9 mg/dl).² The renal insufficiency was largely irreversible, with 4 patients progressing to end-stage renal disease and a mean final creatinine of 2.4 mg/dl in the remaining 17 patients.² When Brunelli et al. re-examined their data using a stricter definition of AKI that requires a 1 mg/dl increase in serum creatinine, the number of patients with AKI declined from 116 to only 3.⁸ Thus, their study design is biased toward detection of relatively small increments in serum creatinine that are not likely to represent APhN. Although the incidence of APhN remains unknown, it is likely to be relatively low and quite possibly <1 in 1000, making a study of <1500 patients exposed to OSP underpowered to detect APhN.

In this same issue of JASN, Hurst et al.⁹ report a larger, observational, retrospective cohort study on 9799 patients over the age of 50 yr who underwent outpatient colonoscopy. Patients were identified through the electronic medical records of Department of Defense beneficiaries in the Washington, DC area. This study applied a more appropriate, stricter definition of AKI of ≥50% increase in serum creatinine from baseline levels over the 12 mo after colonoscopy. Among the cohort, 65.6% of patients received an OSP purgative. A total of 114 patients (1.16%) developed AKI, including 1.29% of the patients who received OSP and 0.92% of the patients who received polyethylene glycol (PEG). By univariate analysis, no association between OSP and AKI was identified because the PEG group included high-risk patients who were significantly older and had a higher incidence of diabetes mellitus, hypertension, atherosclerotic cardiovascular disease, heart failure, chronic kidney disease, diuretic therapy, and use of renin-angiotensin inhibitors (all P < 0.05). When multiple logistic regression models were applied to adjust for covariates and suspected risk factors, OSP emerged as the strongest risk factor for the development of AKI after colonoscopy (odds ratio 2.35; P < 0.001).⁹ When the authors applied a more stringent definition for AKI that required doubling of serum creatinine, multiple logistic regressions revealed an even stronger association between OSP use and AKI (odds ratio 3.52; P = 0.03). Follow-up data were provided on 99 of the 114 patients with AKI, and only 16 patients (16%) returned to baseline creatinine levels. For the 99 patients with AKI, the mean precolonoscopy, postcolonoscopy, and follow-up creatinine values were 0.98, 1.78, and 1.38 mg/dl, respectively. We commend the authors on their excellent study and agree with their conclusion that PEG purgatives are preferable in older patients and in patients with comorbidities.

We live in an era of informed consent and evidence-based medicine. The two studies in this issue of JASN suggest that at least 50% of patients undergoing screening colonoscopy receive an OSP purgative. Extrapolating from the 14 million screening colonoscopies performed annually in the United States, at least 7 million individuals each year are exposed to an agent that can cause APhN. Hypothetically, if the incidence of APhN falls somewhere between 1 in 1000 and 1 in 5000, then the number of new cases of APhN occurring annually would be in the range of 1400 to 7000. Based on the paper by Hurst et al.,⁹ these numbers may be underestimates of the true incidence. Clearly, additional prospective studies are needed to more precisely determine the incidence of APhN so that physicians and patients can make informed decisions regarding the choice of bowel purgative.

Multiple recent steps have been taken to curtail the incidence of APhN. Heightened physician awareness followed the alert issued by the United States Food and Drug Administration in May 2006 about the potential for OSP products to cause acute renal failure and APhN.¹⁰ This warning was added to a consensus document on bowel preparation issued jointly by the American Society of Colon and Rectal Surgeons, the American Society of Gastrointestinal Endoscopy, and the Society of American Gastrointestinal and Endoscopic Surgeons.¹¹ Hopefully this will lead to increased awareness of the risk factors for APhN and more judicious, individualized selection of bowel purgatives.

Based on our understanding of the pathophysiology of APhN, recent changes in dosing and hydration should have a positive impact. A new generation of OSP solution and tablets are now being marketed in which the phosphate content has been reduced by 16.7% and 20%, respectively. Greater emphasis has been placed on the importance of adequate hydration before, during, and after colonoscopy, with current recommendations increased to a minimum of 72 ounces of clear liquids to accompany a regimen of 45 and 30 ml of OSP. Future recommendations may include increasing the interval between OSP administrations, avoidance of anesthesia regimens that require no oral intake for 4 to 6 h, alternative dosing in females, and dose reduction or avoidance in the elderly and in patients with significant comorbidities.

• © 2007 American Society of Nephrology