Quiz Page October 2011:
An Under-recognized Cause of CKD
Article Outline
Clinical Presentation
A 47-year-old woman was evaluated for decreasing kidney function. In 2008, serum creatinine level was 0.89 mg/dL (79 μmol/L; estimated glomerular filtration rate [eGFR] >60 mL/min/1.73 m2 [>1 mL/s/1.73 m2], calculated using the 4-variable MDRD [Modification of Diet in Renal Disease] Study equation) and had been stable for 5 years. In 2009, it increased to 1.26 mg/dL (111 μmol/L; eGFR, 49 mL/min/1.73 m2 [0.82 mL/s/1.73 m2]), 1.62 mg/dL (143 μmol/L; eGFR, 37 mL/min/1.73 m2 [0.62 mL/s/1.73 m2]), and then 2.15 mg/dL (190 μmol/L; eGFR, 26 mL/min/1.73 m2 [0.43 mL/s/1.73 m2]) by March 2010. Medical history included anorexia nervosa, deliberate self-harm, left hip fracture, and chronic pain syndrome. Prescribed medications were gabapentin, naproxen, ranitidine, cyclizine, and acetaminophen.
On evaluation, the patient was concerned about a 2-week history of lethargy with decreased oral intake. She appeared cachectic and malnourished on examination. Weight was 33 kg, body mass index was 14.1 kg/m2, blood pressure was 137/96 mm Hg, and she was clinically volume depleted. Initial blood tests (Table 1) showed decreased kidney function with creatinine level of 2.9 mg/dL (256 μmol/L; eGFR, 18 mL/min/1.73 m2 [0.3 mL/s/1.73 m2]). Urinalysis showed protein (+), and spot urine protein-creatinine ratio was 5.86 mg/mg. A kidney ultrasound scan showed 9-cm kidneys bilaterally. There was marked prominence of the kidney pyramids, but collecting systems were not dilated. A kidney biopsy was performed (Fig 1).
Table 1. Laboratory Data at Presentation
| Parameter | Value | Ref Range |
|---|---|---|
| Hemoglobin (g/dL) | 11.5 | 12.0-16.0 |
| WBC count (× 103/μL) | 4.4 | 4.0-11.0 |
| Platelets (× 103/μL) | 265 | 150-350 |
| SUN (mg/dL) | 8.3 | 7.0-18.0 |
| Sodium (mEq/L) | 140 | 133-146 |
| Potassium (mEq/L) | 3.3 | 3.5-5.4 |
| Creatinine (mg/dL) | 2.90 | 0.60-1.20 |
| eGFR (mL/min/1.73 m2) | 18 | >60 |
| Bicarbonate (mEq/L) | 10 | 18-23 |
| Albumin (g/dL) | 3.2 | 3.5-5.0 |
| Calcium (mg/dL) | 9.00 | 8.20-10.60 |
| Phosphate (mg/dL) | 3.25 | 3.00-4.50 |
| Magnesium (mEq/L) | 2.19 | 1.90-2.70 |
| Bilirubin (mg/dL) | 0.23 | Up to 1.00 |
| ALT (U/L) | 9 | 1-21 |
| ALP (U/L) | 355 | 30-200 |
| Glucose (mg/dL) | 23.4 | 70.0-110.0 |
| Thyrotropin (mIU/L) | 7.70 | 0.5-4.94 |
| Free T4 (ng/dL) | 9 | 0.7-1.5 |
| 24-h proteinuria (g/d) | 2.6 | <0.03 |
| 24-h CCr (mL/min)a | 14 | |
| Urine PCR (mg/mg) | 5.86 |
aBased on serum creatinine level of 2.24 mg/dL. |

Figure 1.
Kidney biopsy specimen. (A) Periodic acid–Schiff; original magnification, ×40. (B) Hematoxylin and eosin; original magnification, ×200. (C) Hematoxylin and eosin; original magnification, ×400.
■ What are the kidney biopsy findings and what is the pathologic diagnosis?
■ What processes may cause these findings?
■ What is the diagnosis?
■ How might this patient be treated and what is the long-term kidney prognosis?
Discussion
What are the kidney biopsy findings and what is the pathologic diagnosis?
Figure 1A is a low-power view showing cortex with focal global glomerulosclerosis and patchy interstitial scarring with tubular atrophy. Figure 1B and C are representative of the tubulointerstitium and show injured tubules with widespread tubular macrovacuolation and epithelial attenuation (Fig 1B) and brown cytoplasmic granular pigment in occasional tubular profiles (Fig 1C). The pathologic diagnosis is chronic tubulointerstitial nephropathy.
The most striking changes were in the tubulointerstitium of the cortex and outer medulla, in which epithelial attenuation and cytoplasmic vacuolation occurred. Most cortical tubules showed a mild decrease in diameter with mild thickening of the tubular basement membrane, but ∼10% showed more severe atrophy with marked thickening of the tubular basement membrane and proteinaceous luminal casts. Scattered tubules showed brownish intracytoplasmic granules of lipofuscin or similar material (negative for iron stains). The interstitium was expanded, with peritubular fibrosis affecting ∼30% of the cortex. There was no significant interstitial or tubular inflammation, and there was no nephrocalcinosis. Of 21 glomeruli sampled, 5 were globally sclerosed; nonsclerosed glomeruli looked normal. An interlobular-caliber artery profile showed mild intimal fibrosis. Arterioles were normal. Immunofluorescence (5 glomeruli) was negative for immunoprecipitants.
What processes may cause these findings?
The tubular macrovacuolation merits consideration of hypokalemic nephropathy, although it is noted that cortical tubular atrophy with lipofuscin can occur with analgesic nephropathy. Chronic hypokalemia produces characteristic macrovacuolar lesions in epithelial cells in the proximal tubule and occasionally distal tubule.1 This finding takes weeks to develop and is reversible with potassium repletion. However, prolonged hypokalemia can lead to more severe changes, including interstitial nephritis and fibrosis, tubular atrophy, and occasionally cyst formation that is most prominent in the renal medulla.2, 3 Acute kidney injury also has been described with hypokalemia.4 Correction of hypokalemia can lead to a decrease in number and size of cysts, although the tubulointerstitial lesions and GFR changes may be irreversible.2, 3
The pathogenesis of hypokalemic nephropathy is not well understood. Animal models suggest that hypokalemia stimulates an increase in renal ammoniagenesis that accumulates in the interstitium.5 Ammonia activates complement, leading to tubular damage and interstitial scarring. Another possible explanation for kidney injury is alterations in growth factor and cytokine levels in response to hypokalemia. These include VEGF (vascular endothelial growth factor), IGF-1 (insulinlike growth factor 1), IGFBP-1 (insulinlike growth factor binding protein 1), angiotensin II, MCP-1 (monocyte chemoattractant protein 1), and/or TGFβ (transforming growth factor β).6, 7, 8
What is the diagnosis?
The clinical presentation and pathologic findings would be in keeping with a diagnosis of anorexia nervosa with associated hypokalemic nephropathy. Additional pathologic effects of nonsteroidal anti-inflammatory drugs also would need to be considered. Although it is recognized that anorexia nervosa may be associated with chronic kidney disease, the mechanisms, histologic patterns, and natural history are not well known, perhaps in part because these patients often are reluctant to seek medical care and rarely come to the attention of a nephrologist.
Patients with anorexia nervosa deliberately restrict their food and water intake, often show purgative behavior, and may abuse laxatives and diuretics. These activities may lead to fluid and electrolyte imbalances, the most common of which is hypokalemia. Hypokalemia often occurs in the presence of metabolic alkalosis resulting from diuretic use or self-induced vomiting, and it also may coexist with metabolic acidosis when laxatives are misused. Importantly, hypokalemic nephropathy may exist when serum potassium level appears only moderately depressed (3.3 mEq/L [3.3 mmol/L] in this patient) due to the cellular shifts that occur in severe metabolic acidosis.
How might this patient be treated and what is the long-term kidney prognosis?
Naproxen and gabapentin therapy were stopped, and intravenous fluids and potassium supplementation were prescribed; other deficient minerals and electrolytes also were replaced. Long-term counseling for the anorexia nervosa should be provided. This patient had histopathologic features consistent with both reversible and irreversible hypokalemic nephropathy. Kidney function improved with potassium replacement: 6 weeks later, serum creatinine level was 0.8 mg/dL (71 μmol/L), corresponding to eGFR of 47 mL/min/1.73 m2 (0.78 mL/s/1.73 m2). Proteinuria was decreased, with spot urine protein-creatinine ratio of 0.67 mg/mg.
Restoration of circulating volume and cessation of naproxen use likely contributed to the significant decrease in serum creatinine level from 2.9 to 0.8 mg/dL (256 to 71 μmol/L; eGFR improved from 14 to 47 mL/min/1.73 m2 [0.23 to 0.78 mL/s/1.73 m2]). However, it is noteworthy that although serum creatinine level returned to the reference range, the patient still had significant renal impairment, shown by histologic features of tubular atrophy and interstitial fibrosis. It is important for those treating patients with anorexia nervosa to recognize that significant renal impairment may be masked by a “normal” serum creatinine level, and kidney function may be assessed better using a 24-hour urine collection for creatinine clearance.
Final Diagnosis
Anorexia nervosa with associated hypokalemic nephropathy.
References
- . Effects of electrolyte disorders on renal structure and function. N Engl J Med. 1967;276(7):383–389
- . Association of hypokalemia, aldosteronism, and renal cysts. N Engl J Med. 1990;322(6):345–351
- . Morphologic aspects of low-potassium and low-sodium nephropathy. Clin Nephrol. 1983;19(6):271–279
- . Hypokalaemia-induced acute renal failure. Nephrol Dial Transplant. 1999;14(9):2216–2218
- . Hypokalemic nephropathy in the rat (Role of ammonia in chronic tubular injury). J Clin Invest. 1987;79(5):1447–1458
- Expression of insulin-like growth factor-I and transforming growth factor-beta in hypokalemic nephropathy in the rat. Kidney Int. 2001;59(1):96–105
- . Angiotensin II type 1 receptor blockade ameliorates tubulointerstitial injury induced by chronic potassium deficiency. Kidney Int. 2002;61(3):951–958
- Hypokalemic nephropathy is associated with impaired angiogenesis. J Am Soc Nephrol. 2008;9(1):125–134
Support: None.
Financial Disclosure: The authors declare that they have no relevant financial interests.
PII: S0272-6386(11)00940-1
doi:10.1053/j.ajkd.2011.04.026
© 2011 National Kidney Foundation, Inc. Published by Elsevier Inc All rights reserved.
