Vancomycin in the Kidney—A Novel Cast Nephropathy

• vancomycin
• acute kidney injury
• casts
• nephropathy
• drug nephrotoxicity

Ever since its discovery >60 years ago, the glycopeptide antibiotic vancomycin has been a first-line drug therapy for gram-positive bacterial infections. Initial reports of AKI with vancomycin were attributed to impurities in the original formulation (Mississippi mud) that have since been largely eliminated. However, numerous reports since that time of AKI after vancomycin exposure suggest an ongoing risk of nephrotoxicity. The reported incidence of vancomycin nephrotoxicity varies widely (5%–35%),¹ reflecting different definitions of AKI and confounding factors in the populations studied.^1,2 Nephrotoxicity is rare when vancomycin is used alone in standard doses³ but may be more common with higher doses (≥4 g/d) to prevent treatment failure in methicillin-resistant Staphylococcus aureus.⁴ Other risk factors for vancomycin nephrotoxicity include exposure for >1 week, trough levels >20 μg/ml, preexisting kidney disease, intensive care unit stay, and coadministration of other nephrotoxic agents, particularly aminoglycoside antibiotics.⁵ A recent meta-analysis of randomized, controlled trials and cohort studies identified a modest relative risk for AKI of 2.45 (95% confidence interval, 1.69 to 3.55) in patients receiving standard-dose vancomycin compared with non-nephrotoxic antibiotics.⁶ Thus, vancomycin is clearly a risk factor for AKI, but the scope of this problem remains unclear.

Vancomycin is excreted mainly (90%) by glomerular filtration, and blood levels rise in the setting of kidney failure. In experimental models of nephrotoxicity, vancomycin crosses the basolateral membrane of proximal tubular epithelium via the organic acid transport system⁷ and possibly, also via megalin receptor–mediated transport at the apical surface.⁸ Vancomycin nephrotoxicity is characterized by oxidative injury to proximal tubules,⁹ and gene expression studies have shown signals of oxidative stress, mitochondrial damage, and inflammatory and complement pathways.¹⁰ There may also be damage to the distal tubule as evidenced by increased urinary levels of dimethylamine.¹¹ Taken together with the clinical features of dose-related reversible AKI and a few biopsy reports describing acute tubular injury (ATI),¹² these findings support a role for toxic proximal tubular injury in human vancomycin nephrotoxicity as well. Interestingly, the majority of biopsy reports have described acute interstitial nephritis rather than ATI.¹² Although this probably reflects publication bias, it indicates the existence of an allergic pathomechanism, presumably idiosyncratic, in some patients with vancomycin nephrotoxicity.

In this issue of the Journal of American Society of Nephrology, Luque et al.¹³ present evidence of another potential pathomechanism of vancomycin nephrotoxicity (i.e., tubular obstruction by casts containing vancomycin precipitates). The index patient showed severe ATI with granular proteinaceous casts that contained nanometric (100–900 nm) spherical aggregates by scanning electron microscopy. The presence of vancomycin in these casts was confirmed by infrared spectroscopy and immunohistochemistry and replicated in eight additional clinical patients and mouse models. Renal function recovered in all surviving patients (including one who required hemodialysis), suggesting that these casts may resolve over time. Previous reports of experimental vancomycin nephrotoxicity have described broadly similar histologic findings (i.e., the presence of casts and dilation of tubules and Bowman’s space).⁹ A novel finding in this study was the detection of uromodulin (Tamm Horsfall protein) in Bowman’s space, suggesting urinary backflow due to obstruction by vancomycin-containing casts. Macrophage infiltrates were seen in the vicinity of casts, supporting the concept of obstruction leading to tubular rupture and leakage of cast material into the interstitium. By analogy, reflux of uromodulin into glomeruli (and interstitium) is a well recognized feature of reflux nephropathy caused by urinary outflow obstruction.¹⁴ The factors that led to vancomycin precipitation are unknown. Uromodulin, synthesized by the thick ascending loop of Henle, is the main constituent of proteinaceous (hyaline) casts in the distal tubule and collecting duct of normal kidneys and has several biologic functions, including modulation of electrolyte trafficking, binding of bacteria, inhibition of stone formation, and clearance of other proteins and cellular debris.¹⁵ Therefore, the detection of vancomycin within uromodulin casts likely reflects derangement of a physiologic excretory process. Importantly, the possibility of coexistent toxic proximal tubular injury from vancomycin could not be excluded. Without proof that cast formation preceded AKI, questions remain: did vancomycin-containing casts cause AKI, or did AKI (possibly from toxic proximal tubular injury) cause reduced washout and retrograde extension of vancomycin-containing casts? Regardless of the precise pathomechanism(s) and sequence involved, Luque et al.¹³ show the existence of a novel form of cast nephropathy in vancomycin-associated AKI.

In addition to reflux nephropathy characterized by large uromodulin casts, several other kidney diseases display distinctive tubular casts, including myeloma cast nephropathy (monoclonal light chains); crystalline nephropathies caused by drugs, phosphate, oxalate, and hereditary metabolic diseases; anticoagulant-associated nephropathy (red blood cell casts); bile cast nephropathy (bilirubin and bile acids); and rhabdomyolysis (myoglobin and hemoglobin). The pathophysiology of cast formation (and kidney dysfunction) in these conditions is likely multifactorial and may include specific interactions between cast constituents and uromodulin (as in myeloma cast nephropathy) and general physicochemical conditions in the distal tubule microenvironment that promote supersaturation and precipitation, such as increased concentration and low urinary pH. Coexistent proximal tubular injury is frequent and may contribute to cast formation via reduced urinary flow leading to increased concentration. A role for tubular obstruction has been proposed in many of these conditions but has rarely been proven. The findings of Luque et al.¹³ suggest that careful attention to the presence of “misplaced” uromodulin may offer a clue to tubular obstruction.

The clinical implications of this study are limited by the fact that renal biopsy is rarely performed in patients with suspected vancomycin nephrotoxicity. Moreover, vancomycin cast nephropathy seems to have a good prognosis; hence, pathologic diagnosis will not alter patient management. Immunohistochemical staining for vancomycin is unlikely to be widely adopted given the rarity of this drug complication. However, uromodulin casts are readily recognizable using conventional histochemical stains (i.e., red with periodic acid–Schiff, weakly eosinophilic with hematoxylin and eosin, and light blue with trichrome) and can be confirmed by immunohistochemical staining. Therefore, renal pathologists should search carefully for uromodulin in tubular casts and Bowman’s space if the clinical history suggests possible vancomycin nephrotoxicity. The ultrastructural finding of microspherical aggregates within tubular casts may provide additional evidence of vancomycin nephrotoxicity.

A diagnosis of drug-induced kidney injury is frequently challenging and involves several criteria: a temporal relationship with drug exposure, a plausible pathomechanism, recovery of kidney function after the drug is discontinued, recurrence after rechallenge, and exclusion of other potential causes of kidney injury. In clinical practice, fulfillment of all of these criteria is rarely possible, but renal biopsy examination may provide supportive evidence when drug toxicity is associated with distinctive pathologic findings. In addition to offering pathologic clues to the diagnosis of vancomycin nephrotoxicity, Luque et al.¹³ should be commended for illustrating how careful attention to renal pathologic findings may uncover a hitherto unsuspected pathomechanism of drug-related renal injury.

• Copyright © 2017 by the American Society of Nephrology