Spectrum of Kidney Diseases in Patients With Hepatitis C Virus Infection

A 10-Year Study

Shunhua Guo, MD; Meghan E. Kapp, MD; Diego M. Beltran, MD; Cesar Y. Cardona, MD; Dawn J. Caster, MD; Ronald R. Reichel, MD; Agnes B. Fogo, MD


Am J Clin Pathol. 2021;156(3):399-408. 

In This Article


The kidney disease that occurs in a patient with a history of hepatitis C infection can often be challenging to diagnose, either clinically or pathologically. HCV infection history may raise high suspicion for HCV infection–associated GN. However, as the current study shows, HCV-associated GN is present in only 42.1% of biopsied HCV+ patients, in comparison to non–immune complex–mediated kidney disease in 46.5% and other immune complex–mediated glomerular disease in 11.4% of these patients. Clinically, even though heavy proteinuria is often seen in HCV-associated GN, other immune complex–mediated glomerular disease such as idiopathic membranous GN and non–immune complex–mediated kidney disease such as diabetic nephropathy and focal segmental glomerulosclerosis can show a similar level of proteinuria. Therefore, kidney biopsy is still the best way to reach the correct diagnosis.

However, the pathologic diagnosis of HCV-associated GN is not an easy task either. When pathology shows characteristic findings Image 1, such as MPGN pattern and cryo plugs, dominant IgM staining by IF and short fibrillary, or short microtubular substructure of the deposits by EM, the pathologist may be able to directly diagnose "cryoglobulinemic MPGN, most likely associated with HCV infection." However, as the current study demonstrates, most of these characteristic findings are not uniformly present, with cryo plugs seen in only 14.8%, MPGN pattern in 28.7%, and characteristic substructure of deposits by EM in 48.7%; IgM dominance by IF was the most often observed (92.8%). Therefore, the diagnosis must be reached with integrated evaluation of all the clinical, laboratory, and pathologic findings with consideration of other possible diagnosis/etiologies in the differential diagnosis. As a result, the diagnosis is not straightforward in many cases but rather often a pattern of injury with discussion of the likely underlying etiology.

Image 1.

Pathologic features of hepatitis C virus (HCV)–associated glomerulonephritis. A, The classic kidney injury typically seen in HCV-associated glomerulonephritis is characterized by mesangial and endocapillary hypercellularity and double contours of the glomerular basement membrane (Jones's silver, ×200). B, Cryoglobulin deposition can be demonstrated as periodic acid–Schiff stain-positive intraluminal cryo plugs, also called hyaline pseudothrombi in glomerular capillary lumens (×200). C-E, Immunofluorescence shows deposits with often immunoglobulin M dominance (C, ×200) and skewed light chain deposition (one light chain stronger than the other): κ light chain (D, ×200) and λ light chain (E, ×200). F, Electron microscopy shows subendothelial and mesangial deposits and intraluminal cryo plug with short fibrillary substructure (×8,900).

Cryoglobulinemia was tested or reported in only a minor portion of the HCV patients in our series. The positive rate was relatively high (56.1%) in the tested/reported patients. Of note, because a negative cryoglobulinemia test may be more likely not reported in the available clinical history, these data might have overestimated the real positive rate. Cryoglobulinemia does not necessarily translate into the presence of cryo plugs seen by LM, IF, or EM, as shown in our study. Notably, one case with positive cryoglobulinemia had only diabetic nephropathy by the renal biopsy study. Our data are limited by not knowing the cryoglobulinemia results in many patients, but we think that although it is a difficult test that requires great efforts in strictly following protocols for specimen collection, processing, and analysis, the cryoglobulinemia test is still a useful parameter for the patient's diagnosis, treatment, and follow-up and hence should be done in HCV+ patients with kidney manifestations.

The most common morphologic pattern of HCV-associated GN in our series was diffuse mesangial proliferative GN (50.4%), followed by the classic MPGN pattern (28.7%). There were more cases with the MPGN pattern in 2007 to 2011 than in 2012 to 2016. In contrast, there were more biopsy specimens with the mesangial proliferative pattern in 2012 to 2016 than in 2007 to 2011. In addition, there were more cryo plugs found in 2007 to 2011 than in 2012 to 2016 and a higher percentage of positive cryoglobulinemia in tested patients in 2007 to 2011 than in 2012 to 2016. These findings may hint to a trend of less severe immunologic abnormality in the era of DAA therapy. It was possible that the overall immunologic derangements caused by the HCV infection in patients with DAA therapy had decreased due to the more effective antiviral treatment, and hence the pathologic changes appeared to be less severe than the pre-DAA era. However, global glomerulosclerosis was similar in the two groups, and interstitial fibrosis was also not significantly different, albeit numerically higher in the second period. The numerically higher interstitial fibrosis can be associated with higher serum creatinine seen in this group, likely reflecting timing of biopsies after presentation. A major shortcoming in our study was that we did not have adequate clinical and follow-up information about presentation, treatment, viral load changes, and pre- and posttherapy renal function and proteinuria results of each patient for specific correlation analysis. These limitations precluded more analysis on the potential effect of antiviral treatment on the manifestation and progression of the HCV-related kidney disease.

The third most common type of HCV-associated GN was the membranous pattern, all of which showed secondary membranous features, with mesangial or endocapillary proliferation. Most of them had mesangial deposits, two-thirds showed subendothelial deposits, and one-third showed characteristic substructures by EM. PLA2R immunofluorescence was performed in seven cases, with three (42.9%) cases positive. The three cases with positive PLA2R staining showed mesangial and/or subendothelial deposits in addition to subepithelial deposits; one case showed short fibrillary substructure of deposits by EM, and one case even had a crescentic lesion. All these features are more consistent with a secondary etiology. Positive PLA2R test is not a specific indicator of primary membranous GN, and its positivity has been reported in some secondary membranous GN cases. Our study reiterates that membranous pattern GN associated with hepatitis C infection can have PLA2R positivity.[22]

Previously, GN with crescentic lesions has occasionally been mentioned but not well studied in HCV-associated GN.[6] The current study showed that 9 (7.8%) had proliferative GN with crescentic lesions in the 115 HCV-associated GN cases in 2007 to 2016. IF showed IgM-dominant staining, and EM demonstrated characteristic substructures in most cases. Two of three cases with available ANCA results were negative. In the one case with positive ANCA, no vasculitis was identified in arteries or arterioles, IF showed dominant IgM staining, and EM showed mesangial, subendothelial, and subepithelial deposits with short fibrillary and short microtubular substructures. The patient also had low complements C3 and C4. These features supported that this was most likely HCV-associated GN. However, the possibility cannot be completely excluded that a concurrent ANCA-mediated injury presented as an additional component to the HCV-associated immune complex–mediated process.

Most of our cases with dominant IgA deposits are consistent with primary IgA nephropathy. There are occasional case reports of HCV-associated IgA nephropathy.[23,24] However, these reports did not demonstrate convincing morphologic findings of HCV-associated GN. In addition, IgA deposition in the kidney can be observed in patients with liver disease, including hepatitis or cirrhosis due to the reduced hepatic clearance of IgA. Therefore, we think mere history of HCV infection with IgA-dominant deposits in the kidney is not diagnostic of "HCV-associated IgA nephropathy." We did observe one case in our series with IgA-dominant IF staining, but other features were indicative of a cryoglobulin-associated process. IgA-dominant cryoglobulin is rare but has been reported, such as in a case of Staphylococcus infection–related GN with cryoglobulinemic features.[25]

Fibrillary GN is a rare kidney disease, comprising 0.4% to 1.4% of kidney biopsy diagnoses.[26] Studies have identified DnaJ homolog subfamily B member 9 (DNAJB9) as a highly sensitive and specific marker for this disease.[27] The association of fibrillary GN with HCV infection has been described previously.[8,28] The fibril substructures seen in fibrillary GN are thinner and longer than the short fibrils observed in HCV-associated GN and most often show IgG dominance by IF. In our group, fibrillary GN appeared slightly enriched in our HCV+ patients, with 9 (3.3%) cases in our 273 HCV+ patients. Because DNAJB9 is a recent advance for fibrillary GN, none of the cases in our series had DNAJB9 study performed.

In summary, we found a variety of kidney biopsy changes in patients with a clinical history of HCV infection, with most having kidney disease unrelated to HCV infection. HCV-associated GN had five morphologic patterns. Our data suggested a possible trend of less proliferative lesions in the more recent era in comparison to the pre-DAA era. However, our study showed that the availability of effective HCV antiviral therapy had not yet resulted in major changes in the spectrum of kidney diseases in these patients.