Kidney biopsies often show accumulation of lipids or lipidlike material. Evidence has been provided that lipids can directly initiate and contribute to the progression of glomerular and tubulointerstitial lesions. In this study we describe a renal lipidosis occurring in patients with a positive history of narcotic abuse who were enrolled in a methadone substitution program. All 3 patients presented with proteinuria (2.5–20 g/d) and impaired renal function. Renal biopsy revealed a pronounced extracellular and intracellular deposition of lipidlike material in the glomerular, interstitial, and tubular compartments. Known causes of lipid storage could be excluded clinically and morphologically. We consider this to be a distinct renal lipidosis associated with narcotic abuse, methadone intake, or intravenous abuse thereof.
REPORT OF CASES
Patient 1
A 42-year-old white woman was referred to a nephrologist for evaluation of a decline in kidney function (creatinine, 2.2 mg/dL [to convert creatinine values to micromoles per liter, multiply by 88.4]; estimated glomerular filtration rate [eGFR], 24.5 mL/min per 1.73 m2) and proteinuria (2.5 g/d) in January 2012 (Table 1). She had a history of intravenous drug abuse (heroin) until 2004 and had been in a methadone substitution program since that time. Furthermore, she had arterial hypertension, hypothyroidism, and a history of hepatitis B and C virus (HBV and HCV) infection, which were nonproductive at presentation. There was no evidence of liver disease. Kidney biopsy showed 7 glomeruli, of which 4 had an adhesion of the capillary tuft to Bowman capsule. All glomeruli showed an extensive lipid deposition in podocytes, endothelial cells, and mesangial cells (Figure). In the interstitium, numerous foam cells could be observed. Focally, proximal tubular cells also showed a vacuolization of their cytoplasm. There was mild chronic tubulointerstitial damage comprising 30% of the cortex. Immunohistochemistry demonstrated mesangial granular positivity for immunoglobulin (Ig) M, complement factor (C) 1q, and C3, but not IgA or IgG. A portion of the glomerular and interstitial foam cells could be identified as CD68-positive macrophages (Figure, A, inset). By electron microscopy numerous lipid droplets were seen in podocytes and in mesangial, endothelial, tubular, and interstitial cells (Figure, B). No osmiophilic deposits could be demonstrated by electron microscopy.
Patient 2
A 50-year-old white man was originally referred to a hematologist for evaluation of fatigue and anemia (hemoglobin, 5.8 g/dL). The patient had a history of intravenous drug abuse and had been in a methadone substitution program since February 2011. The patient received a diagnosis of chronic HCV infection (genotype IA) and had been treated with pegylated interferon and ribavirin since July 2011. No viral load could be detected by polymerase chain reaction in his blood. Bone marrow aspiration showed solely reactive changes. A diagnosis of ribavirin-associated anemia was made, and hemoglobin levels improved after cessation of the ribavirin therapy. Liver function was normal. On admission in February 2012, decreased kidney function (creatinine, 2 mg/dL; eGFR, 35 mL/min per 1.73 m2) and nonselective proteinuria (2.6 g/d) were apparent (Table 1), whereas 8 weeks earlier creatinine levels and urine status had been unremarkable. In the following days, kidney function deteriorated further (creatinine, 4.4 mg/dL; eGFR, 14.3 mL/min per 1.73 m2). A kidney biopsy revealed 11 glomeruli, of which 1 was globally and 1 was segmentally sclerosed. The remaining glomeruli showed vacuolization and lipid droplets in endothelial and mesangial cells, along with mesangial extracellular lipid deposition (Figure, C and D). Focally, vacuolization of tubular epithelial cells and interstitial CD68-positive foam cells also could be observed. There was slight chronic tubulointerstitial damage comprising 10% of the cortex. By immunohistochemistry mesangial granular staining for IgA, IgM, C1q, and C3 could be detected. No label was detected for IgG. Electron microscopy revealed droplets of lipidlike material in mesangial cells, in the mesangial matrix, and to a lesser extent in podocytes and endothelial cells. Electron microscopy also showed sparse mesangial but no subendothelial or intramembranous electron-dense deposits.
Patient 3
This patient was a 32-year-old white man who was referred to a nephrologist because of massive edema and dyspnea. The patient had a history of a variety of drug abuses, mainly cocaine, heroin, amphetamines, and cannabis; he had been in a methadone substitution program since 2006. There was no evidence of liver disease. On admission in February 2012 he presented with nephrotic-range proteinuria (20 g/d) and impaired kidney function (creatinine, 1.5 mg/dL; eGFR, 58 mL/min per 1.73 m2; Table 1). Kidney biopsy revealed segmental adhesions of the capillary tuft to Bowman capsule in 11 of 22 glomeruli. There was a diffuse dilation of the glomerular capillaries with an extensive deposition of predominantly extracellularly localized lipidlike material in capillary lumina and basement membranes (Figure, E and F). Sparse glomerular CD68-positive cells could be identified by immunohistochemistry. Focally, lipid droplets were observed in tubular epithelia. The tubulointerstitium showed moderate chronic damage comprising 35% of the cortex. Immunohistochemistry revealed granular deposition of IgA, IgG, IgM, C1q, and C3 in the mesangium and segmentally along the glomerular basement membranes. Ultrastructurally, no electron-dense deposits were seen. Electron microscopy showed dilated glomerular capillaries filled with lipidlike material. In congruence with the 2 previously described biopsies, this lipidlike material could also be depicted in the proximal tubular epithelium.
COMMENT
The occurrence of renal disease in morphine addicts was reported by Eduard Levinstein1 in the 19th century, followed by several other authors.2–5 The main renal manifestation was nephrotic syndrome, but the underlying pathomorphologic renal changes differed substantially, and no specific lesion could be consistently associated with addiction.6 The diversity of lesions is probably due to the numerous diseases associated with drug abuse, such as viral hepatitis, human immunodeficiency virus infection, chronic cutaneous bacterial infection, or endocarditis with concomitant bacteremia or amyloidosis, all of which lead to a different renal pathology.
We describe a renal lipidosis in patients who had a history of drug abuse and were enrolled in a methadone substitution program. The clinical presentation included proteinuria (in 1 of the 3 patients in nephrotic range) and reduced renal function (Table 1). The morphologic changes were characterized by extensive renal deposition of lipidlike material in glomerular, interstitial, and tubular cells, as well as in the extracellular space. The lipid deposits differed in their opacity in electron microscopy. Frozen samples, which would be most suitable for the demonstration of the fatty character of this material, were not available for these patients. Because paraffin-embedded tissue did not allow for staining of lipids (which are eluted from the tissue during the processing of paraffin-embedded sections), in 1 patient the lipid nature of the stored substance could be shown by Sudan black stain on araldite-embedded material that had been acquired for electron microscopy (Figure, E, inset). The light microscopy and ultrastructure nevertheless were showing features generally accepted to be lipid or lipidlike material deposits.
Renal lipid storage represents a phenomenon found in different diseases. Several animal experiments have provided strong hints that lipids can directly initiate and contribute to the progression of glomerular and tubulointerstitial lesions.7 Native lipoproteins and their oxidatively modified variants, which can be locally generated in the course of degenerative and inflammatory processes, may induce fibrogenic cytokines in mesangial cells and podocytes, and may lead to focal segmental glomerulosclerosis.8–10
Renal diseases linked to lipid storage tend to exhibit a typical pattern of lipid distribution in glomeruli, tubules, and interstitium.11,12 The distribution and the specific ultrastructure of the stored lipid can help in the differential diagnosis (Table 2). In the 3 patients reported here, the lipid deposits demonstrated some similarities with lecithin cholesterol acyltransferase deficiency and lipoprotein glomerulopathy. However, in kidneys of lecithin cholesterol acyltransferase–deficient patients, lipid is located mainly in the glomerular basement membrane, subendothelially, and in the mesangium, and it exhibits a typical osmiophilic and lamellar architecture in electron microscopy. In lipoprotein glomerulopathy, foam cells are rarely seen in the interstitium. Moreover, in 2 of the patients presented here (patients 1 and 2), serum lipid analysis revealed normal values (Table 1). In the third patient (patient 3), who had nephrotic syndrome, serum cholesterol, triglyceride, and very low-density lipoprotein were increased; however, his serum lipid analysis had been normal 6 years ago, suggesting that the lipid abnormalities resulted from the nephrotic syndrome and were not causative. Thus, in all individuals a genetic case of lipid nephropathy appears unlikely, although it cannot be excluded completely because of the unavailability of an extensive serum lipid screen.
Lipid inclusions in podocytes and intracapillary lipid-loaded macrophages are regularly seen in HCV-associated membranoproliferative glomerulonephritis. However, only 1 of the 3 patients (patient 2) had chronic HCV infection, and none of them showed light microscopic or ultrastructural features of a membranoproliferative glomerulonephritis. Moreover, in all 3 patients the amount of lipids stored in glomerular cells by far exceeded the amount typically seen in HCV-associated membranoproliferative glomerulonephritis. By immunohistochemistry a positive signal was detected for IgM, C1q, and C3 in all 3 patients. However, ultrastructurally, immune complexes could not be detected in patients 1 and 3. Therefore, we interpret the positive immunohistochemical reaction as unspecific because of the excessive lipid storage in these 2 patients. In patient 2, there were only isolated segmental, mesangial electron-dense deposits. Thus, HCV-associated membranoproliferative glomerulonephritis or another type of immune-complex glomerulonephritis appears to be unlikely as a cause of the observed findings. Also, cirrhosis-associated renal lipid deposition can be excluded because none of the patients had clinical or biochemical signs for liver cirrhosis.
In summary, the lipid nephropathy in these 3 patients enrolled in a methadone substitution program did not resemble any of the established entities. A study on the prevalence of kidney diseases in asymptomatic male heroin addicts at the start of methadone therapy reported that 3 of 145 individuals had a urinary protein excretion greater than 150 mg per 24 hours, and only 1 of them had an increased serum creatinine level.13 Lynn et al14 reported that in a methadone program in New Zealand, 5 patients who were continuing to use opiates intravenously showed predominantly mesangial and, to a lesser extent, endothelial lipid accumulation. The 3 patients presented here demonstrated prominent lipid deposition in all glomerular cells, tubular epithelium, and an infiltrate of interstitial CD68-positive foam cells (Figure).
It is intriguing that all 3 instances described here occurred in a central region of Germany (Aachen, Giessen, and Marburg) and within a relatively narrow time frame (January and February 2012). One may speculate whether the clinical and renal manifestations were related to past or ongoing abuse of narcotics with unknown contaminating substances. Contaminations in the distributed methadone preparations or solvents could not be evaluated retrospectively. Furthermore, in patients in a methadone program an intravenous misuse of these methadone preparations may be considered. These drug formulas are hyperosmolar and contain—among other things—hydroxyethyl starch, glycerol, and variable adulterants, potentially contributing to lipidosis.15
In conclusion, we consider this to be a distinct renal pathology potentially observable in patients with a history of drug addiction and current enrollment in a methadone substitution program. The histologic characteristics are prominent lipid deposits in glomerular, interstitial, and tubular cells. When this form of renal pathology is observed, an association with narcotic abuse, methadone intake, or an intravenous abuse thereof may be included in the differential diagnosis.
References
Author notes
The authors have no relevant financial interest in the products or companies described in this article.