Kidney Biopsy in Hospitalized Patients with Acute Kidney Disease

Percutaneous kidney biopsy is an essential clinical tool for the assessment of a multitude of kidney disorders. In addition to providing key diagnostic information that is commonly unattainable from blood, urine, and radiographic testing, the findings of kidney biopsies inform clinical decisions regarding therapeutic interventions and commonly provide key prognostic data to clinicians and patients. Kidney biopsies have also played an important role in research that has expanded our understanding of kidney pathophysiology. However, as is the case for all invasive procedures, kidney biopsy is not without risk. Bleeding is the primary complication that can, if significant, result in the need for angiographic or surgical intervention, and very rarely nephrectomy, and even death (1). Past research has elucidated patient- and procedure-related factors that increase the risk for postbiopsy complications (2). Recognition and mitigation of these risk factors along with advancements in percutaneous biopsy technique have improved both diagnostic yield and the safety of this procedure. Few contemporary studies of kidney biopsy have sought to characterize complication rates solely among patients with a specific disease-related indication for the procedure or on the basis of inpatient/outpatient status.

In this issue of the Clinical Journal of the American Society of Nephrology, Moledina et al. report the results of a retrospective study that investigated complications of kidney biopsies among patients with acute kidney disease (AKD), which by definition, included those with AKI (3). Using a database of patients who underwent native kidney biopsy at two Yale University–affiliated hospitals between January 2015 and December 2017, the investigators identified those patients who underwent this procedure for the evaluation of AKD. They extracted data on bleeding that required red blood cell transfusion and angiographic interventions, hematomas identified on radiographic imaging within 7 days after the biopsy, and death within 30 days. The need for red blood cell transfusion was considered the primary study outcome. The investigators utilized patient chart review to confirm that postbiopsy complications were related to the procedure. The primary analyses focused on characterizing biopsy-related complications and elucidating patient- and procedure-related factors independently associated with complications among inpatients. The investigators also compared complications between hospitalized and nonhospitalized patients. Finally, using the National Inpatient Sample (NIS), the investigators identified a cohort of hospitalized patients with a diagnosis code of AKI who underwent kidney biopsy. Rates of biopsy-related complications among Yale University inpatients with AKD were compared with those among 53,325 patients from the NIS.

Overall, 256 patients from the Yale University database underwent kidney biopsy for the evaluation of AKD, 159 of whom were inpatients and 97 of whom were outpatients at the time of the biopsy. Among the 159 inpatients, 12 (8%) received a blood transfusion, three (2%) underwent angiographic intervention, 11 (7%) developed a hematoma, and four (3%) died, although none of the deaths were attributed to the kidney biopsy. Independent variables associated with postbiopsy transfusion included lower platelet count, higher BUN level, and female sex. Larger gauge needle (i.e., 16 G versus 18 G) and nephrology versus radiology proceduralist were associated with a trend toward more complications, although neither association was statistically significant after covariate adjustment. Compared with outpatients, hospitalized patients were more likely to receive blood transfusion (8% versus 0%; P=0.003), but were not more likely to undergo angiographic intervention (2% versus 0%; P=0.17), develop a hematoma (7% versus 5%; P=0.52), or die (3% versus 0%; P=0.12). Finally, compared with patients from the NIS, hospitalized patients from the Yale University cohort had similar rates of interventions to stop bleeding (2% versus 2%; P=0.88) and deaths in the hospital (3% versus 2%; P=0.74).

This study has several key strengths. First, the use of individual chart review facilitated determinations of the causal nature of the associations between the biopsy procedure and subsequent events/complications. Second, the study focused on biopsy-related complications among a more narrowly defined population of patients than most prior studies (i.e., those with AKD who were hospitalized); hence providing novel findings on complication rates and biopsy-related risks in a seemingly high-risk patient group. Finally, the assessment and comparison of biopsy-related complications between Yale University inpatients and those from a nationally representative sample of hospitalized patients increases the generalizability of the study’s findings. There are also certain limitations worth noting. First, the overall size of the Yale University inpatient cohort was relatively small. Second, the majority of inpatients had AKI stage 1 and nearly 30% had diabetes or arterionephrosclerosis as the primary histologic diagnosis, neither of which is typically considered a common cause of severe AKI. Furthermore, it was not delineated when in the course of AKI biopsies were performed. These findings raise questions about the generalizability of the findings to patients with severe acute decrements in kidney function. Finally, key differences between hospitalized and nonhospitalized patients almost certainly confounded the comparison of complication rates between these two patient groups. Although the proportion of patients who experienced a drop in hemoglobin ≥2 g/dl was comparable, the baseline hemoglobin among hospitalized patients was >2 g/dl lower than among outpatients (9.2 g/dl versus 11.4 g/dl; P<0.001). Moreover, the median number of hemoglobin measurements after kidney biopsy was considerably larger among hospitalized patients (six versus one; P<0.001).

Notwithstanding these limitations, there are several important takeaways from this study. First, the incidence of serious biopsy-related complications among hospitalized patients with AKD is quite low and generally comparable to what has previously been reported in patients overall (2). A meta-analysis by Corapi et al. that included 34 studies comprising >9000 biopsies reported an overall incidence of angiographic intervention of 0.6%, which is just marginally lower than in this analysis. None of the patients in this study died as a result of the biopsy (2). This is consistent with the findings of Lin et al. (4) in a study that compared biopsy-related complications between 147 inpatients and 183 outpatients in Taiwan and reported no deaths. In fact, one could reasonably disagree with the conclusion of Moledina et al. that “hospitalized patients experience higher risk of postbiopsy complications than previously reported.” This conclusion was made on the basis of the observed transfusion rate of 8%, which was likely driven by the lower baseline hemoglobin level and substantially higher number of postbiopsy measurements of this parameter among hospitalized patients, and was not reflective of serious, life-threatening complications.

Second, this study reinforces our understanding of risk factors associated with postbiopsy complications. Prior studies have demonstrated associations of lower platelet count and lower kidney function with risk of bleeding after kidney biopsy (5,6). Although the role of prophylactic interventions to mitigate bleeding risk was not examined in this study, careful attention to patients with lower platelet counts and/or severe azotemia is important to reduce the likelihood of significant bleeding. Perhaps most interesting is the finding that women were at higher risk for bleeding than men. This finding has been shown in prior studies; however, it remains unclear whether this relates to physiologic (e.g., lower baseline hemoglobin levels) or anatomic differences (2,7). Although not modifiable for the purpose of attenuating risk, this potential risk factor should be appreciated by providers performing these procedures.

It is interesting to note that the reported rate of serious biopsy-related complications has not necessarily changed significantly over time. In one of the earliest studies on the subject over 50 years ago, Slotkin and Madsen (8) reported on complications of 5000 kidney biopsy cases, noting that 0.3% of biopsies required surgical intervention and 0.1% were complicated by death. However, this should not be construed to indicate that no progress has been made in increasing the safety of kidney biopsy. The study by Slotkin and Madsen relied on surveys of urologists for their data. Physicians who had patients who experienced severe complications of biopsies may have been less likely to respond to a request for data. Furthermore, the success in obtaining kidney tissue was considerably lower at that time, which may help explain the low rate of kidney-related bleeding complications; however, Slotkin and Madsen's description of gallbladder, pancreas, and splenic injuries, as well as pneumothorax, would be very rare complications in current practice. Advances in the biopsy technique, including the use of real-time ultrasound guidance and automated biopsy devices, along with greater understanding and mitigation of risk factors, have undoubtedly improved the safety of this procedure.

The National Institute of Diabetes and Digestive and Kidney Diseases recently funded the Kidney Precision Medicine Project (KPMP) that involves the performance of biopsies, including among patients with AKI. A key ethical question for the KPMP relates to the degree of risk associated with performing research biopsies in patients with AKI who would not otherwise have a clinical indication for this procedure. It is not clear whether this study can fully address this question because the degree to which the Yale University inpatient cohort is representative of patients who will be enrolled in KPMP is not known. Nonetheless, Moledina et al. provide important data that supports the general safety of this procedure in hospitalized patients with AKD.

Disclosures

None.