Vancomycin-Associated Cast Nephropathy: Reality or Fantasy?

Vancomycin is a glycopeptide antibiotic commonly used for severe gram-positive bacterial infections, notably methicillin-resistant Staphylococcus aureus. AKI occurs in 5%–43% of patients who are vancomycin treated (1–3), and is commonly (59%) attributable to vancomycin nephrotoxicity (VANT) (4). The true incidence of VANT is unknown, given the absence of randomized controlled trials and the high prevalence of confounding causes of AKI. However, a meta-analysis comparing vancomycin with a nonglycopeptide antibiotic suggests this risk is low (relative risk 2.45, 95% confidence interval, 1.69 to 3.55) with standard vancomycin doses (<4 g/day) (4). The risk of nephrotoxicity with higher doses used for severe infections is undoubtedly higher. Fortunately, AKI usually resolves after drug discontinuation (4). Because kidney biopsy is rarely performed, the pathologic features of VANT are poorly characterized, but both acute tubular injury and acute interstitial nephritis have been described (5). More recently, a novel “cast nephropathy” containing vancomycin intermingled with uromodulin has been described in VANT (6–8). However, the relationship of these casts to AKI (i.e., cause or effect) remains unclear.

Vancomycin Nephrotoxicity: Clinical Features

VANT is defined clinically as AKI that improves after discontinuation of the drug. Risk factors fall into three major categories: pharmacokinetics, patient factors, and concomitant nephrotoxin exposures (9). The frequency of AKI correlates with trough levels of vancomycin (82% if >35 mg/L, versus 5% if <10 mg/L), consistent with dose-dependent toxicity (9). In one meta-analysis, the incidence of AKI was 30% with trough levels >15 mg/L versus 9% with levels <15 mg/L (odds ratio, 2.67; 95% confidence interval, 1.95 to 3.65) (10). This highlights the risk for nephrotoxicity with higher targeted trough levels (15–20 mg/L) employed for resistant strains of S. aureus (11). Longer duration of therapy (>7 days) is also a risk factor, whereas the relationship to mode of administration (intermittent bolus versus continuous infusion) is less consistent (9). Patient risk factors include obesity, CKD, and critical illness (9). Lastly, coadministration of aminoglycosides and piperacillin-tazobactam compound the risk of AKI, possibly by altered pharmacokinetics (9). The timing of AKI with VANT varies from 4 to 17 days after initiation of vancomycin (3), reflecting infrequent therapeutic drug monitoring and the poor performance of creatinine rise as a biomarker of injury (i.e., 2–3-day lag from injury to rise). Importantly, vancomycin is excreted predominantly (>90%) by glomerular filtration and its clearance is inversely proportional to creatinine clearance. Thus, a major challenge in diagnosing VANT is avoiding reverse causality bias by attributing AKI to high vancomycin levels when elevated levels may be consequent to AKI from other causes. Nonetheless, increased trough levels (or area under concentration time curve-to-minimum inhibitory concentration ratio) are important clues to the presence of VANT.

Management of VANT consists of reducing ongoing nephrotoxic exposures and monitoring for the development of severe AKI requiring RRT. Identifying patients with risk factors for the development of VANT should prompt early and frequent therapeutic drug monitoring, to prevent unnecessarily high trough concentrations. If VANT is suspected clinically, vancomycin should be discontinued until blood levels return to a therapeutic range.

Pathogenesis and Pathology

The role of acute tubular injury in VANT is supported by morphologic findings in kidney biopsies and experimental models (5). Vancomycin can enter proximal tubular epithelium via the organic acid transport system at the basolateral membrane (12), and possibly via megalin receptor-mediated transport at the apical surface (13), where it induces oxidative stress (14,15), mitochondrial damage (14), and activation of inflammatory and complement pathways (16). Distal tubular injury also occurs, as evidenced by increased urinary excretion of the distal tubule marker dimethylamine (17). Tubular toxicity is likely triggered by conditions that promote increased intratubular vancomycin concentration, including low urinary flow rates and tubular obstruction. Of note, acute interstitial nephritis due to vancomycin (5) reflects an idiosyncratic allergic response. This is usually accompanied by signs of acute tubular injury and skin findings. Rare severe allergic reactions include toxic epidermal necrolysis and drug rash with eosinophilia and systemic symptoms (5,9).

In 2017, Luque et al. (6) described the first case of AKI with vancomycin-containing casts demonstrated by immunohistochemistry, immunoelectron microscopy, and infrared spectroscopy. These casts appeared granular by light microcopy and contained uromodulin (Tamm–Horsfall protein), which was also detected in the Bowman space. Transmission electron microscopy disclosed noncrystalline spherical vancomycin aggregates (100–900 nm). Similar pathologic findings were demonstrated in eight other patients and in mice exposed to vancomycin. Uromodulin is synthesized by the thick ascending loop of Henle, and uromodulin casts are a common finding in AKI of diverse causes (18), but particularly prominent in outflow obstruction, where retrograde extension to proximal tubules and the Bowman space may be seen, sometimes accompanied by tubular rupture, extrusion, and a localized inflammatory cell reaction. Tubular dilation, casts, sloughed epithelium, and Bowman space expansion were noted in previous experimental models of VANT (14,15), providing additional indirect evidence for tubular obstruction in VANT.

Subsequently, Tantranont et al. (7) identified vancomycin- and uromodulin-containing casts in most (25 of 28) patients with VANT (i.e., whose kidney function improved after vancomycin was discontinued), versus one of nine patients without clinical nephrotoxicity. Casts localized to the distal tubule and had a variable appearance, ranging from periodic acid Schiff red, granular aggregates to variegated spherules intermingled with uromodulin and necrotic epithelial cells, accompanied by acute tubular injury and/or interstitial nephritis (7). In mild cases, casts consisted predominantly of uromodulin and vancomycin was only detected by immunohistochemistry. By electron microscopy, some casts formed globular aggregates with a concentric multilamellated appearance, suggestive of crystallization. Other investigators have shown the presence of calcium phosphate apatite in vancomycin-containing casts (19). Tantranont et al. speculated that vancomycin- and uromodulin-containing casts developed after direct tubular toxicity or single nephron obstruction from sloughed epithelial cells, setting off a vicious cycle of tubular obstruction and increased local vancomycin concentration that promoted crystallization and further tubular injury (7). Of note, casts with a similar morphologic and ultrastructural appearance derived from the plasma membrane of degenerating epithelial cells have also been described in cast nephropathy with delayed graft function associated with use of tacrolimus plus sirolimus (20). This raises the possibility that vancomycin-containing casts may be derived from necrotic/apoptotic epithelial cells that are shed after severe acute tubular injury.

Vancomycin-Associated Cast Nephropathy: Reality or Fantasy?

There is little doubt that vancomycin-containing casts are real and not “fantastical,” and are a reliable indicator of VANT (6,7). But whether these casts cause AKI (e.g., via tubular obstruction) or simply reflect reduced “wash-out” in the setting of AKI remains unclear (Figure 1). This question is probably impossible to answer because serial kidney biopsies are not ethical in human subjects or feasible in animal models. Moreover, because vancomycin encounters the proximal tubule before reaching the distal tubule, it is very difficult to prove a primary pathogenetic role for distal tubular obstruction, independent of proximal tubular injury. Similar uncertainty surrounds many other kidney diseases and drug toxicities characterized by cast formation and/or intratubular drug crystallization (21), in which the pathophysiology of AKI may involve both direct cytotoxic injury to tubular epithelium and outflow obstruction (22).

Figure 1.

Vancomycin- and uromodulin-containing casts in the distal tubule are a common kidney biopsy finding in patients with AKI attributed to vancomycin nephrotoxicity. Whether these casts are the cause or the result of AKI remains unknown.

The formation of casts may involve interactions between cast constituents and uromodulin (as in myeloma cast nephropathy), in addition to altered conditions in the distal tubule microenvironment that enhance uromodulin gel formation (23) and promote supersaturation and precipitation, such as increased concentration and urinary pH (24). In vancomycin-associated cast nephropathy, it is notable that kidney function recovered completely in most reported patients, even those requiring RRT (7), indicating these casts are reversible, or at least excretable. This contrasts with other forms of nephropathy associated with drug precipitates, which typically have residual kidney dysfunction, presumably due to their insolubility (24,25). Further studies are needed to understand the nature of vancomycin-uromodulin interactions, and to determine if vancomycin undergoes crystallization in the kidney tubule in VANT. These findings might inform strategies to prevent or ameliorate this toxicity. The detection of vancomycin-containing casts in the urine, if reproducible, might eliminate the need for an invasive diagnostic biopsy in VANT and shed more light on the dynamics of cast formation and its relationship to AKI.

Should vancomycin immunostaining now be added to the renal pathologist’s diagnostic armamentarium? Given the rarity of this toxicity and the limited added value of identifying vancomycin-containing casts for managing patients with suspected VANT, this extra effort seems hard to justify. However, the findings of numerous uromodulin-containing casts should certainly raise the possibility of VANT in the appropriate clinical setting. Uromodulin casts have distinct tinctorial properties by light microscopy and a filamentous ultrastructural appearance that are easily recognizable. The ultrastructural finding of nanospherical casts of variable electron density, admixed with uromodulin filaments, may offer additional evidence of possible vancomycin-associated cast nephropathy. However, the latter will require increased attention to distal tubules, which are rarely targeted by routine kidney biopsy examination.

The Kidney Medicine Personalized Medicine Project has drawn renewed attention to the clinical effect of AKI and the need to identify differences in pathomechanisms that affect patient outcomes. The discovery of vancomycin-containing casts in patients with VANT suggests a potential novel form of drug-related AKI. However, further study is needed to determine if this distinctive pathologic finding is the cause or consequence of VANT.

Disclosures

All authors have nothing to disclose.

Funding

None.

Author Contributions

J. Stevens and M.B. Stokes conceptualized the study, wrote the original draft, and reviewed and edited the manuscript.