Persistent Hematuria in ANCA Vasculitis

Management of ANCA vasculitis is an area of much ongoing research, but it remains a challenge for clinicians caring for these patients. With modern therapy, remission can be achieved in 70%–90% of patients (1–3), but about 40%–50% will relapse at some point in the course of their disease. To date, we do not have a good way of determining who will relapse, and when—and it is this central question that complicates management. Although treatment with immunosuppressive therapy can protect patients from a disease that is organ and life threatening, we know that continued immunosuppression is associated with short- and long-term adverse effects and costs.

Lung involvement, upper airway disease, and proteinase 3 ANCA serotype have been identified as disease characteristics and markers that increase the risk of relapse (4). However, these risk factors do not reliably differentiate patients destined to relapse from those who are unlikely to do so, and they perform poorly as predictors of imminent relapse. Changes in ANCA titer have been studied as a risk factor for relapse. Although relapses are frequently preceded by a rise in ANCA titer, the interval between rise in titer and relapse is highly variable, and it is often many months in duration. In the recent analysis by Kemna et al. (5), only 50% of patients had a relapse within 18 months of a rise in titer. In many individuals, a rise in titer may not be followed by relapse at all (6). Thus, although these data can be used in conjunction with other clinical information, they have limited utility in clinical management.

In this issue of the Clinical Journal of the American Society of Nephrology, Rhee et al. (7) ask fundamentally important questions not only for ANCA GN but indeed, for many glomerulonephritides: when is persistent hematuria (at the end of induction therapy or thereafter) a sign of ongoing glomerular inflammation and active disease and when is it simply a sequela of glomerular damage not associated with active inflammation? Of course, one could ask the same question at any point in a patient’s course. When hematuria is detected at disease presentation, physicians take it more or less for granted that it is indeed a sign of active GN, and ideally, they confirm this interpretation with a kidney biopsy. The fact that only one third of patients in the Rituximab in ANCA-Associated Vasculitis (RAVE) Trial underwent a kidney biopsy suggests that the treating physicians accepted this assumption and did not see the need for histologic confirmation (3).

It is frequently suggested that persistent hematuria in a patient otherwise in clinical remission is due to scarring, but it is difficult to be certain without biopsy confirmation. The authors must be commended for their effort in tackling this challenging question. As acknowledged in the manuscript, their analysis is made difficult by an attempt to study the predictive value for relapse of a sign that is itself used as a measure of disease activity affecting the kidneys. The detection of hematuria is fully subject to interpretation (and a priori bias) in determining whether it represents active glomerular inflammation versus residual scarring. This problem is most salient in the analysis of time to relapse, especially in trying to evaluate whether an increase in hematuria is a predictor of subsequent relapse. The authors report a median time to relapse of 22 months from the time of increase in hematuria to the time that a patient is deemed to be in relapse of GN, defined as new or worsening red blood cell casts and/or rise in serum creatinine attributed to vasculitis. However, the presence of red blood cell casts is an insensitive measure of ongoing glomerular injury and itself subject to variability in specimen collection, and it is urinalysis operator dependent. Likewise, a rise in serum creatinine may be a delayed finding when the glomerular inflammation is not severe. These criteria likely lead to an underestimation of the frequency and an overestimation of time to relapse. Nevertheless and despite the choice of these more specific (and less sensitive) measures of active GN, the investigators did indeed find a significant positive association between the persistence of hematuria and subsequent nephritis relapse. One may posit that a careful microscopic examination of the urinary sediment (by a nephrologist) with enumeration of dysmorphic erythrocyturia would have increased the sensitivity and specificity of persistent hematuria in detecting glomerular inflammation and further improved its predictive value of subsequent relapse.

Another challenge to this investigation is the lack of a gold standard in assessing the presence of active GN. Today, the closest thing to a gold standard for diagnosing active GN is a kidney biopsy. Of course, one must remain cognizant of the limitations of this invasive technique, especially in a disease with asynchronous focal lesions. The results and interpretation of kidney biopsies are not binomial (active versus inactive), and they are subject to limitations by possible sampling error and tissue sample size. Nevertheless, histologic examination provides information about kidney disease activity that cannot be gleaned from other clinical and laboratory evaluation. In a small series of nine patients with ANCA vasculitis in clinical remission and either persistent or new hematuria, repeat kidney biopsies revealed a variety of histopathologic findings, including other types of active GN (e.g., IgA nephropathy), FSGS, and more global glomerulosclerosis, as well as no glomerular pathology (8). Although limited by size, this study highlights that a range of glomerular pathophysiology can be seen in patients with ANCA vasculitis and persistent or new/increased hematuria. Repeat kidney biopsy is infrequently performed in clinical practice or clinical research, and the study by Rhee et al. (7) underscores the need for clinical trials to consider systematic repeat biopsy at the end of induction therapy. In the setting of clinical trials, for repeat biopsies to be most beneficial, they should be performed on all patients with kidney disease, regardless of the persistence of clinical signs of glomerular injury. Only then can one formally assess the sensitivity and specificity of urinalysis in detecting active ongoing glomerular injury.

Despite the association between persistent hematuria and increased risk of relapse reported in the study by Rhee et al. (7), the key question as to whether persistent hematuria represents ongoing glomerular injury that requires continuation or escalation of immunotherapy remains unanswered. In that respect, the investigators could have leveraged other available possible markers of inflammation and disease activity to help discern between persistent inflammation and a noninflammatory damage, such as C-reactive protein, or especially, the persistence of a positive ANCA. Persistence of a positive ANCA (especially proteinase 3 ANCA) at the end of induction therapy has previously been reported, and it was recently confirmed to be strongly associated with increased risk of subsequent relapse (9). Does the concurrence of persistent hematuria and positive ANCA and/or other biomarkers better differentiate hematuria from active GN versus glomerular or interstitial scarring than isolated hematuria alone? Would it improve the predictive value of either marker alone with respect to subsequent relapse? Are there other biomarkers that can be incorporated in this measure of disease activity?

Ultimately, we are still missing a measure of true remission at the immunologic and tissue levels. In diseases, such as ANCA vasculitis or SLE, where relapses are common, the yet unanswered question is whether these relapses are true recurrences of disease after a period of unequivocal remission or instead, mere increases in inflammation in the setting of relative clinical quiescence but ongoing autoimmune activity. To date, we still do not have robust or validated measures of “true complete remission.” The Birmingham Vasculitis Activity Score, useful as it is in enumerating and semiquantifying clinically apparent signs that are attributed to active vasculitis (sometimes with uncertainty), does not measure the activity of the autoimmune process or subclinical inflammation at the tissue level. Just as health is more than the absence of overt disease, we suspect that a complete remission in ANCA vasculitis is more than the absence of overt clinical signs or symptoms. Clearly, we need better measures of complete remission. Given the heterogeneity in disease course, studying immunologic, molecular, genetic, or epigenetic characteristics of patients who attain sustained durable remission off of all immunosuppressants may provide a means of identifying a “molecular signature” of complete remission.

In summary, this study by Rhee et al. (7) reports an association of persistent hematuria with increased likelihood of relapse in ANCA vasculitis but could not directly address the fundamental question of whether or when it represents a sign of ongoing active GN. This study should serve as a call for developing more sensitive and specific measures of treatment response to therapy and remission that correlate with kidney histologic findings. Until better markers of remission emerge from clinical studies and interventional trials, nephrologists should consider repeating a kidney biopsy whenever uncertainty about full remission remains at the end of the induction phase of treatment.

Disclosures

None.