Practice patterns and approach to kidney biopsy in lupus: a collaboration of the Midwest pediatric nephrology consortium and the childhood arthritis and rheumatology research alliance

Results from our survey show that 78 % of respondents intend to follow ACR adult guidelines for recommending kidney biopsy in their pSLE patients (Table 3), equally divided by pediatric subspecialty. However, when asked about specific scenarios, only 28.8 % actually followed the guidelines. Hematuria alone warranted biopsy according to 25.0 %; proteinuria alone (defined as >150 mg/24h or a spot urine protein/creatinine ratio > 0.2 mg/mg), according to 58.2 %. Two pediatric rheumatologists and 2 pediatric nephrologists (each at separate institutions) routinely recommend or perform kidney biopsy in any patient with a new diagnosis of pSLE, regardless of the clinical findings.

Most respondents (66.5 %) have their patients’ biopsies read by a dedicated renal pathologist at their primary institution, although 21.2 % use a local general pathologist. The remaining 12.3 % have their biopsies sent out to a renal pathologist at another institution. Pathologists utilize the WHO grading system (33.2 %), the ISN/RPS grading system (22.4 %), or both (33.0 %). A total of 9.5 % of respondents were unsure of which grading system their pathologists were using (11.6 % of nephrologists vs. 8.2 % of rheumatologists). For diffuse proliferative GN, the activity and chronicity indices were provided routinely to 80.9 % of respondents, whereas 2.3 % received only the activity index, 1.1 % received only the chronicity index, and 15.2 % worked with pathologists who did not provide either.

More than 90 % of pediatric rheumatologists and nephrologists refer their patients with pSLE to the other provider after a diagnosis of nephritis is made. There are very few pediatric nephrologists who do not have access to a pediatric rheumatologist at their primary institution (3.9 %, n = 3), and 2 of those 3 have an internal medicine rheumatologist at their institution who care for children. Similarly, only 3.6 % of pediatric rheumatologists (n = 4) do not have access to a pediatric nephrologist at their primary institution. The decision to perform a kidney biopsy in a patient with pSLE is made by 62.5 % of respondents only after discussions are held between nephrology and rheumatology. Only 29.6 % of respondents stated that the decision was made by the pediatric nephrologists alone; and only 7.8 %, by the pediatric rheumatologist alone.

Pediatric rheumatologists report viewing the tissue slides together with their pediatric nephrologist and pathologist 49.1 % of the time and without a pathologist 2.7 % of the time. Pediatric rheumatologists report reviewing tissue section slides 14.6 % of the time with the pathologist separately from nephrology, and 33.4 % of pediatric rheumatologists review only the biopsy report.

There were no significant differences in the proportion of pediatric rheumatologists or nephrologists that would recommend/perform repeat renal biopsies under specific instances (Table 4). For proliferative LN (class III or IV), there were 10 % (n = 18) who obtain true “protocol” biopsies in order to guide the dosing of induction therapy, and 5 % who re-biopsy patients in remission in order to guide withdrawal of immunosuppression. However, 41.7 % would obtain repeat kidney biopsy after induction therapy in patients with no response and 25 % would in patients with only partial response. For patients in renal remission from a proliferative nephritis who develop a renal flare, 69 % would obtain another biopsy to guide therapy. Only 2.7 % (n = 5) obtain repeat biopsy for non-renal lupus flares.

There were 29 % of respondents who stated that they do not routinely repeat kidney biopsy after an initial pathologic diagnosis of proliferative LN. Of these 17 were nephrologists and 35 were rheumatologists. Of the nephrologists who do not re-biopsy, most stated that they made biopsy decisions alone without consultation of their rheumatology colleagues. Of the rheumatologists who do not re-biopsy, only 2 stated that they made biopsy decisions alone, and most did so in consultation with their local pediatric nephrologists. The pediatric rheumatologists who do not recommend routine repeat biopsy tended to have been practicing longer, whereas the pediatric nephrologists who do not perform routine repeat kidney biopsy tended to have been practicing for < 10 years. All 16 pediatric nephrologists practicing their specialty >20 years stated that they will routinely re-biopsy patients with proliferative LN for specific indications.

Repeat biopsy after initial diagnosis of membranous LN (class V) differed very little from the practice following initial diagnosis of proliferative LN (Table 3). Only 32.0 % would obtain repeat kidney biopsy after induction therapy in patients with no response and 14.7 % would in patients with only partial response. There were no significant differences to the responses received from nephrologists compared to rheumatologists.

The majority (71.3 %) of respondents reported that at their institutions, renal biopsies on pSLE patients with nephritis were primarily performed by pediatric nephrologists. There were 19.9 % of respondents from institutions where interventional radiology primarily performed the biopsies, and respondents from the remaining 8.8 % of institutions had both pediatric nephrologists and radiologists performing proportionate numbers of biopsies.

Pediatric nephrologists were also questioned about their preferences regarding the biopsy procedure itself. When asked about pre-procedural laboratory testing, complete blood count (CBC) is usually performed by 96 % of nephrologists, partial thromboplastin time (PTT) by 94 %, prothrombin time (PT) by 90 %, and INR by 76 %. Less common laboratory testing included: type and screen (36 %), platelet function testing (15 %), lupus anticoagulant (12.6 %), serum chemistries (2.8 %), urine beta-HCG (2.8 %), urinalysis (1.4 %), and bleeding time (1.4 %). Significant anemia (<7 g/dL) is thought to be a contraindication to kidney biopsy by 35.4 % of nephrologists, whereas 5.4 % delay biopsy for Hgb < 8 g/dL and 1.4 % delay biopsy for Hgb < 9 g/dL until anemia can be corrected. For significant thrombocytopenia (<50x10^3/mcL) 58 % of nephrologists will defer biopsy. A few specifically commented that they would transfuse PRBC, platelets, or FFP in order to perform kidney biopsy without delay. Other contraindications to renal biopsy included active infection (58 %), uncontrolled hypertension (59 %), and solitary kidney (17 %).

Non-steroidal anti-inflammatories (NSAIDs), including aspirin, are routinely held by 85 % of pediatric nephrologists prior to biopsy. Most recommended holding these drugs for 3–7 days prior to kidney biopsy (54.4 %), whereas 1.5 % held for 1 day, 11.8 % held for 2 days, 11.8 % held 8–14 days, and 5.9 % held >14 days. NSAID use shortly before kidney biopsy is considered an absolute contra-indication by 27 % of nephrologists. However, 14.7 % of pediatric nephrologists would proceed with biopsy despite NSAID use even for routine non-emergent kidney biopsies. Some pediatric nephrologists consider recent NSAID use a higher risk procedure warranting referral to interventional radiology. Others start immunosuppressive therapy and delay biopsy because of the NSAID use.

Of 68 pediatric nephrologists, 54.4 % routinely observe patients overnight after kidney biopsy, whereas 38.2 % routinely send patients without complications home on the same day. Some nephrologists have particular concerns about pSLE patients after biopsy, as suggested by the fact that 7.4 % routinely observed their pSLE patients, but not patients with non-SLE kidney disease, overnight before discharge (Fig. 1).

Pediatric rheumatologists and nephrologists appear to value guidelines such as those published by the ACR, based on the high proportion who stated that they use them. However, the number of respondents who deviated based on pediatric definitions of proteinuria suggests that pediatric guidelines might be useful. The numbers of respondents who deviated were comparable between specialties. The differences in opinion on when to biopsy relate to number of years in practice in a more complex manner. There were approximately equal proportions of both pediatric nephrologists and pediatric rheumatologists in each of the 4 groups (0–5 years, 6–10 years, 11–20 years, and > 20 years) (p-value = 0.114). The only significant correlation between years in practice and practice patterns was between the proportion of individuals who had been practicing 11–20 years and adherence to ACR guidelines (93 % compared to 70-76 % in the other groups, p-value = 0.029).

The results of the survey show that it is uncommon for pediatric nephrologists to manage LN without input from pediatric rheumatologists. Of pediatric nephrologists responding, 90 % refer even their patients with renal-limited pSLE to rheumatology before or after a diagnosis of LN is made. If extra-renal manifestations are also present, 97 % refer. Moreover, rheumatologists are consulted by over 2/3rds of pediatric nephrologists in making the decision to perform kidney biopsy, and over half view the biopsy slides along with their pediatric rheumatology colleagues.

There are also very few pediatric rheumatologists (<5 %) who manage nephritis in pSLE without routinely seeking the input of pediatric nephrologyOf these 6 rheumatologists, 3 make the decision to biopsy without other consultation, but the other 3 seek guidance from their nephrologist when deciding to order a kidney biopsy, and 5 have their nephrologist perform the biopsy once the decision is made. In addition, 5 of the 6 review renal biopsy results with their local pediatric nephrologist.

There is clearly no consensus regarding the indications for repeat kidney biopsy in the management of proliferative or membranous LN. Just under one third of respondents do not routinely recommend repeat biopsy in any instance. Alternatively, 1.7 % routinely recommend repeat biopsies for any lupus flare, regardless of kidney involvement. The most common practice is to repeat kidney biopsy in patients who have an exacerbation of their lupus which results in worsening of the urinary sediment (2/3rds of respondents). The second most common is to repeat biopsy at some time point during induction therapy in non-responders (32 % for membranous LN and 42 % for proliferative LN). Only a few programs use “protocol” biopsies to guide duration of induction or maintenance therapies in patients regardless of response. Interestingly, there were no significant differences in the proportion of rheumatologists or nephrologists that engaged in the practice of recommending/performing routine “protocol” repeat biopsy. This suggests that in general pediatric rheumatologists and nephrologists are similar in their approach to repeat biopsy, and that differences that exist are more common between individual physicians than between sub-specialties.

Despite the ongoing role of pediatric nephrologists in medical management of nephritis in pSLE, there might be a shift in who is actually performing the biopsies. Many nephrologists appear to consider pSLE to be a risk factor for complications after kidney biopsy, and 12 (17 %) are primarily referring their pSLE patients to interventional radiology to perform the biopsy. It was not asked whether these nephrologists perform kidney biopsies routinely on their other patients. Of rheumatologists who completed the survey, referral rates to nephrology and interventional radiology are comparable. Therefore only 70 % of pediatric nephrologists are choosing to perform biopsies of patients with pSLE and only 70 % of pediatric rheumatologists are having their local pediatric nephrologist performing the biopsies.

Finally, there is great variability in the perceived safety of kidney biopsies in pSLE. Most of this variability is likely to reflect differences in perceived safety of kidney biopsy in general. There are 38 % of pediatric nephrologists who discharge their pSLE patients the same day as the procedure, similar to their practice with non-SLE patients. Conversely, 54 % observe both their pSLE and non-SLE patients overnight. However, 8 % of pediatric nephrologists do send most patients home on the day of biopsy, but specifically observe pSLE patients overnight, due to perceived risk of complications. Identifying sources of perceived increased risk of kidney biopsy in children with lupus nephritis was beyond the scope of this study, but hypotheses can be drawn from nephrologist responses to survey questions on biopsy procedure. Possibilities include higher rates of anemia, thrombocytopenia, hypertension, NSAID use, or anticoagulant use in pSLE patients than in non-SLE patients, as well as a perceived need for performing biopsy more urgently and less electively when relative contraindications exist,

Limitations of our study include the possibility of selection bias, recall bias, or reporting bias. Results of survey-based studies can be inherently subject to significant bias, and the strategies for survey distribution as well as the response rates should always be taken into account when interpreting survey data. Questions were written by pediatric nephrologists and rheumatologists together in order to minimize leading questions or error due to differences in terminology. By surveying specialists participating in either CARRA or MWPNC, we may have biased the results towards particular types of pediatric centers. We chose to survey providers as opposed to centers, as differences in practice patterns amongst providers in the same center are often encountered. There were no restrictions on how many providers could respond to the survey from any one center, which could allow for bias. However, the large degree of voluntary reporting of the respondents’ institution allowed us to compare practice patterns within centers with the highest degree of survey participation. The degree of collaboration between sub-specialties may not be applicable to primary care pediatric centers or centers outside of the United States or Canada. However, the strengths of the survey include the >40 % response rate, the large number of institutions represented, and the high proportion of completed surveys without missing responses. Aside from reports of transplant kidney biopsy, this study represents one of the only surveys to assess approach to kidney biopsy for a particular kidney disease, lupus nephritis, and the only survey to date that compares the approaches between pediatric rheumatologists and nephrologists.