Faculty Bibliography

Background/Purpose: Treatment of lupus nephritis relies on renal histopathological features. However, renal biopsies do not capture patient-specific active biological pathways. Urine proteomic biomarkers could revolutionize the diagnosis and management of lupus nephritis by predicting active intrarenal biological pathways and can be noninvasively monitored over time.
Method(s): One thousand proteins were quantified (RayBiotech) in a total of 112 longitudinal urine samples from 30 SLE patients with active lupus nephritis and 7 healthy controls (HC). The proteins and molecular pathways detected in the urine proteome at the time of biopsy were then analyzed with respect to lupus nephritis class, response to treatment after 1 year, histopathological features (activity and chronicity indeces), and trajectory over time (baseline and week 12, 26, and 52). The intrarenal expression of candidate biomarkers was evaluated using single cell transcriptomics of renal biopsies from patients with active lupus nephritis.
Result(s): There were 237 proteins (FDR < 10%) enriched in the urine of patients with lupus nephritis reflecting several molecular pathways involving chemotaxis, extracellular matrix remodeling, and activation of neutrophils and platelets. Hierarchical clustering using urine proteomics segregated SLE patients into 2 groups, with 80% of complete responders clustering together. This finding could not be similarly reproduced using standard features including baseline proteinuria, creatinine, histologic activity or chronicity scores, or class, indicating unique informative features of urine proteomics (Fig. 1). Patients with proliferative lupus nephritis (class III or IV) had stronger activation of chemotaxis pathways. IL-16 was the urinary protein most significantly increased in proliferative disease compared to membranous (FC 6, p=0.002) (Fig. 2A). Assessment of urine proteins that correlated with histologic activity kidney highlighted IL-16 as the single most strongly correlated protein with histologic activity (r=0.69, p=9.5.10-5; Fig. 2B). IL-16 concentration was independent of the amount of proteinuria and progressively diminished over time in patients who were responding to immunosuppression (Fig. 2C). Single cell RNA sequencing revealed significant intrarenal expression of IL16 by all infiltrating immune cells and highlighted IL16 as the second most expressed cytokine in lupus nephritis kidneys out of a compendium of 236 cytokines (Fig. 3A-B).
Conclusion(s): Urine proteomics can noninvasively identify active and biologically relevant pathways in lupus nephritis. Integrated urine proteomics and renal single cell transcriptomics revealed that IL-16, a CD4 ligand with chemotactic and proinflammatory functions, was one of the most expressed cytokine in lupus nephritis. As a urine proteomic biomarker, IL-16 may predict renal histological activity and could be monitored over time to assess response to immunosuppression. Urinary IL-16 is independent of proteinuria thus potentially providing actionable clinical information that is not captured by currently used biomarkers. Further studies are ongoing to validate these findings

Background/Purpose: Patients with systemic lupus erythematosus (SLE) are at increased risk of premature atherosclerosis and thrombosis. Hydroxychloroquine (HCQ) is widely used in the treatment of SLE and has been considered of benefit for overall vascular health albeit studies to address this benefit at the cellular level have been limited. Accordingly, this study was initiated to investigate the relationship between HCQ use and dose with platelet activity, the platelet transcriptome, and vascular functional readouts.
Method(s): Patients fulfilling ACR or SLICC criteria for SLE were consecutively recruited for platelet evaluation with the only exclusion being on nonsteroidal anti-inflammatory medications, aspirin or anticoagulants. At enrollment, blood was collected for hematology analysis using the Sysmex XN-1000 analyzer, platelet aggregation via the Helena AggRAMTM system, and platelet RNA isolation and storage. Microvascular function was assessed via sublingual sidestream darkfield imaging. Brachial artery reactivity testing was used to evaluate large vessel function. Stored platelet RNA was isolated and analyzed by RNA sequencing (Illumina HiSeq4000 Sequencing).
Result(s): Among 132 SLE subjects, 108 were on HCQ. Mean age was 39.9 +/- 13.0 and 97% were female. Lupus disease activity at the time of blood draw assessed by the SELENA-SLEDAI activity index was 3.44 (range 0-20). Demographics and SLE disease activity did not differ between those on versus off HCQ (Table 1). Platelet count and size were not different between groups (Figure 1A). Platelet aggregation in response to submaximal ADP at multiple concentrations was lower in participants on HCQ (Figure 1B). Consistently, there was an inverse relationship between HCQ dosing and platelet aggregation in response to ADP (2uM: R=-0.213, P=0.037; 1uM: R=-0.310, P=0.0025; 0.4uM: R=-0.376, P=0.00018; Figure 1C). Since no subjects were on aspirin (or any other antiplatelet therapy at enrollment), aggregation in response to arachidonic acid (AA) was robust and similar between groups. However, after incubating platelets with aspirin (3mM) in vitro, platelet aggregation in response to AA was lower in the HCQ group compared to non-HCQ group (P=0.035, Figure 1B). To investigate the potential mechanisms of HCQ induced lower platelet aggregation, we evaluated platelet RNA sequencing in 49 subjects (8 no HCQ, 41 on HCQ). Positive regulation of pathways related to platelet activation (and in particular, P-selectin expression) was inversely related to HCQ, especially with higher doses (Figure 1E). In terms of vascular function, subjects on HCQ had improved microvascular function as noted by an increased proportion of sublingual capillaries filled with RBCs (P=0.011) and smaller perfused boundary region (PBR, P=0.010). HCQ dosing correlated with PBR (R=-0.599, P=0.002, Figure 1H) and RBC Filling (R=-0.592, P=0.002, Figure 1I). BART also trended positively with HCQ dose (R=0.385, P=0.094; Figure 1J).
Conclusion(s): These findings suggest that HCQ may provide benefit for vascular health in SLE as supported by ex vivo experiments demonstrating decreased platelet aggregation and downregulation of platelet functional pathways as well as improved vascular readouts

Background/Purpose: Current management of lupus nephritis (LN) is guided by histopathological features on kidney biopsy and measurement of proteinuria. Urine proteomics is a non-invasive source of novel biomarkers which may better reflect the complex dynamic immunobiology of LN in real time. Two composite measures include CKD273, which can predict the risk of progression of chronic kidney disease in the general population, and LN120, which was designed to diagnose LN. Both are multidimensional urine proteomic classifiers consisting of 273 or 120 peptides, respectively, with major components including collagen fragments, abundant blood-derived proteins, and proteins involved in inflammation. We investigated the ability of these classifiers to predict traditional biopsy features and disease response in LN.
Method(s): A total of 31 adults with biopsy-proven LN were included in this study. All participants met the SLICC and 2019 EULAR/ACR Classification Criteria for SLE based on a spot urine protein-to-creatinine ratio of >0.5 and class III, IV, and/or V LN on renal biopsy. Urine samples were collected at week 0 (at the time of renal biopsy) and week 12 and then subjected to peptidome analysis using a capillary electrophoresis-mass spectrometry (CE-MS) platform. This peptidome data was used to calculate CKD273 and LN120 classifiers at each time point. LN response status was determined at week 52 based on proteinuria, creatinine, and prednisone dosage (no more than 10 mg daily). Spearman's rank correlation and t-tests were used to compare proteomic classifiers with renal biopsy characteristics and response.
Result(s): At week 0, both CKD273 and LN120, but not proteinuria, exhibited a moderate to strong correlation with histological activity index on renal biopsy (Figure 1; rho = 0.65 with p = 0.00024 for CKD273; rho = 0.47 with p = 0.013 for LN120). CKD273 also correlated with chronicity index (rho = 0.54, p = 0.0037). Neither classifier significantly correlated with lupus nephritis ISN class. With respect to response, CKD273 and LN120 were not significantly different between groups at week 0. However, a reduction in LN120 was observed in 100% of complete responders, 60% of partial responders, and 50% of non-responders at week 12 (Figure 2). The magnitude of this change in LN120 in complete responders versus non-responders did not reach statistical significance (p = 0.13), though this is potentially because of the small number of responders with CE-MS data available at both time points (n = 4). CKD273 did not significantly change with time in any response group (Figure 3).
Conclusion(s): This work provides proof of concept that urine proteomic classifiers can noninvasively predict histological activity and chronicity in LN. Complete responders, but not partial responders or non-responders, exhibited an impressive numerical decrease in LN120 by week 12, suggesting that proteomic scores may track with and predict a durable treatment response. Larger studies are needed to validate these findings

Background/Purpose: Based on inhibition of viral replication and limited reports on clinical efficacy, hydroxychloroquine (HCQ) was initially considered as a prophylaxis and treatment of COVID-19. Despite this optimism, more extensive reports have significantly dampened the promise of efficacy, however cardiac toxicity has surfaced raising attention to this complication. Although HCQ is generally considered safe during pregnancy based on studies in patients with systemic lupus erythematous and other rheumatic conditions, this initiative leveraged a unique opportunity to evaluate neonatal electrocardiograms (ECGs) in the context of HCQ levels to address any potential cardiotoxicity.
Method(s): Neonatal ECGs and HCQ blood levels were available in a recently completed study evaluating the efficacy of HCQ 400mg daily to prevent the recurrence of congenital heart block associated with anti-SSA/Ro antibodies. The ECGs of affected newborns who met the primary outcome of advanced block were not included in this safety study so that the results only reflect those infants with no clinical cardiac disease. Using the Bazett formula to correct for heart rate, corrected QT (QTc) intervals were calculated and compared to age-matched normal values. For reference, the median (2nd percentile - 98th percentile) values for QTc were 413 (378-448) msec in males, and 420 (379-462) msec in females. QTc intervals were recorded in the absence of knowledge of the HCQ levels. Values exceeding 448 msec for males and 462 msec for females were considered abnormal. Levels of HCQ were assessed during each trimester of pregnancy and in the cord blood, providing unambiguous assurance of drug exposure.
Result(s): There were 45 ECGs available for interpretation within the first 4 months of life in unaffected infants. Overall, there was no correlation between cord blood levels of HCQ and the QTc (R = 0.02, P = 0.86) or the average value of HCQ levels obtained during each individual pregnancy and cord blood and the QTc (R = 0.04, P = 0.80), as shown in Figure 1A and Figure 1B. Likewise there was no correlation between the average of the maternal HCQ levels obtained at each trimester and delivery plus cord levels and the QTc on the ECGs of the 31 infants evaluated on day of life 1-4 (R = 0.08, P = 0.63) or those of the 14 children older than 4 days (R = 0.01, P = 0.95). Maternal values of HCQ were sustained throughout pregnancy and delivery (Figure 2). Mean QTc values were nearly identical between those in the highest and lowest quartiles of cord blood HCQ levels (P = 0.57) and between the highest and lowest quartiles of average HCQ levels during pregnancy (P = 0.54) (Figure 3A and 3B). Among these 45 infants, only 5 had prolongation of the QTc (11%; 95% CI: 4% - 24%), 2 marked and 3 marginal. No arrhythmias occurred in any neonate that was not known to have heart block.
Conclusion(s): In aggregate, these data provide reassurances that the maternal use of HCQ is not associated with a high incidence of QTc prolongation in the neonate

Background/Purpose: Patients with systemic lupus erythematosus (SLE) represent a unique population in considering risk for coronavirus disease 2019 (COVID-19) with biologic, genetic, demographic, clinical and treatment issues all at play. By the nature of their chronic inflammatory autoimmune condition and regular use of immunosuppressive medications, these individuals would traditionally be considered at high risk of contracting the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) and having a worse prognosis. Accordingly, we aimed to characterize patients with SLE affected by COVID-19 in New York City (NYC) and analyze associations of comorbidities and medications on outcomes.
Method(s): Patients with SLE and COVID-19 (confirmed by RT-PCR testing), were identified through a longitudinal survey of an established NYU lupus cohort, query of New York University Langone Health and Bellevue Hospitals systems and referrals from rheumatologists at those institutions. All patients were age 18 or older and met SLE classification criteria or carried a rheumatologist's diagnosis of SLE. Only English-, Spanish- or Mandarin-speaking patients were included in the study. Data were prospectively collected via a web-based questionnaire and review of electronic medical records. Baseline characteristics and medications were compared between the hospitalized and ambulatory patients with COVID-19. A logistic regression analysis was performed to identify independent predictors of hospital admission.
Result(s): A total of 41 SLE patients were confirmed COVID-19 positive by RT-PCR. The patients were predominantly female and encompassed the major racial/ethnic demographics seen in NYC. The most common symptoms of COVID-19+ patients were cough (78.4%), fever (64.9%), and shortness of breath (64.9%). Of those SLE patients with COVID-19, 24 (59%) were hospitalized, 4 required ICU level of care, and 4 died, all of hypoxic respiratory failure, Table 1. Hospitalized patients tended to be older, non-white, Hispanic, and have higher BMI, antiphospholipid syndrome, a history of lupus nephritis and at least one medical comorbidity, Table 2. There was no difference between the groups in use of hydroxychloroquine, systemic steroids or immunosuppressants. Logistic regression analysis identified the following independent predictors of being hospitalized with COVID-19: race (OR = 7.78 for non-white vs. white; 95% CI: 1.13 to 53.58; p=0.037), the presence of at least one comorbidity (OR=4.66; 95% CI: 1.02 to 21.20; p=0.047), and BMI (OR = 1.08 per increase in kg/m2; 95% CI: 0.99 to 1.18; p=0.096).
Conclusion(s): Patients with SLE and COVID-19 have a high rate of hospitalization but similar mortality rate to the general population in NYC. Risk factors such as non-white race, higher BMI, and the presence of one or more comorbidities were identified as independent predictors of hospitalization in SLE patients who develop COVID-19. The use of hydroxychloroquine and immunosuppressants did not appear to influence the outcomes of patients with SLE in the setting of COVID-19. Further studies are needed to understand additional risk factors for poor COVID-19 outcomes in patients with SLE

Background - Based on inhibition of viral replication and limited reports on clinical efficacy, hydroxychloroquine (HCQ) is being considered as prophylaxis and treatment of COVID-19. Although HCQ is generally considered safe during pregnancy based on studies in patients with systemic lupus erythematous and other rheumatic conditions, there may still be reluctance to institute this antimalarial during pregnancy for the sole purpose of antiviral therapy. Methods - To provide data regarding any potential fetal/neonatal cardiotoxicity, we leveraged a unique opportunity in which neonatal electrocardiograms (ECGs) and HCQ blood levels were available in a recently completed study evaluating the efficacy of HCQ 400mg daily to prevent the recurrence of congenital heart block associated with anti-SSA/Ro antibodies. Results - Forty-five ECGs were available for QTc measurement, and levels of HCQ were assessed during each trimester of pregnancy and in the cord blood, providing unambiguous assurance of drug exposure. Overall, there was no correlation between cord blood levels of HCQ and the neonatal QTc (R = 0.02, P = 0.86) or the mean of HCQ values obtained throughout each individual pregnancy and the QTc (R = 0.04, P = 0.80). In total 5 (11%; 95% CI: 4% - 24%) neonates had prolongation of the QTc > 2SD above historical healthy controls (2 markedly and 3 marginally) but ECGs were otherwise normal. Conclusions - In aggregate, these data provide reassurances that the maternal use of HCQ is associated with a low incidence of infant QTc prolongation. However, if included in clinical COVID-19 studies, early postnatal ECGs should be considered.

Background/Purpose: Disparities have been reported during the coronavirus disease (COVID-19) outbreak. Systemic lupus erythematosus (SLE) patients represent a unique group that is affected by clinical, treatment, demographic, and socioeconomic (SES) risk factors for severe COVID-19 disease. The Neighborhood Deprivation Index has been associated with non-communicable disease as well as communicable disease outcomes. We conducted this study to identify neighborhood SES factors influencing SLE COVID-19 outcomes.
Method(s): Patients with SLE and COVID-19 (confirmed by RT-PCR testing), were identified through a longitudinal survey of an established NYU lupus cohort, query of NYU Langone Health and Bellevue Hospitals systems and referrals from rheumatologists at those institutions. All patients were age 18 or older and met SLE classification criteria or carried a clinical diagnosis of SLE. Baseline characteristics along with zip code neighborhood data including COVID-19 case rates and neighborhood characteristics were obtained using the Hopkins COVID database and the American Community Surveys (ACS 2014-2018) respectively. A principal component analysis was performed to identify contributory neighborhood characteristics. Then a logistic regression analysis identified predictors of testing positive for COVID-19 and COVID-19 hospitalization.
Result(s): A total of 59 SLE patients (41+ and 18-) were tested for COVID-19 by RT-PCR. The patients were predominantly female, aged 46+/-16, and racially/ethnically diverse. Roughly 140 neighborhood data points were recorded and categorized as follows: population density, race and ethnicity, household type, household size, education level, employment type and status, income and poverty, transportation method, and insurance status. COVID-19 positive patients tended to live in neighborhoods with more single parent households, households with >4 residents, higher unemployment rate, higher high school dropout rate, more public transit use, and more employment in retail, construction, and personal care services. These variables were directly proportional to principal component 1 (PC1) and accounted for 88% of the variance in neighborhood characteristics. A logistic regression model identified that PC1 (OR= 1.3; 95% CI: 1.0-1.8) and taking immune suppressants (IS) (taking vs not taking OR= 2.1; 95% CI: 1.5 to 23.3) independently correlated with having a positive COVID-19 test when controlling for hydroxychloroquine (HCQ), glucocorticoids (GC), and previous lupus nephritis (LN). Only PC1 independently correlated with COVID-19 hospitalization (OR= 1.4; 95% CI: 1.1-1.9) upon controlling for taking IS, HCQ, GCs, and LN. PC1 associated with African American (AA) or Hispanic patient race/ethnicity (OR= 1.6, 95% CI: 1.2-2.2).
Conclusion(s): In addition to SLE disease, neighborhood characteristics and SES are important risk factors both for contracting COVID-19 and developing severe disease. Neighborhood deprivation may mediate the reported relationship between AA and Hispanic race/ethnicity and COVID-19. Given that a plurality of SLE patients are of AA and/or Hispanic backgrounds, care teams must formulate strategies to address socioeconomic stress in our patients

Background/Purpose: Poor therapeutic response rates contribute to the increased morbidity and mortality associated with lupus nephritis. Early identification of patients likely to respond is crucial as delays in treatment associate with worse outcomes. This study evaluated response using prospectively collected data obtained from the multi-ethnic/racial, multi-center Accelerating Medicines Partnership (AMP) lupus nephritis cohort. This cohort represents a real-world clinical setting using provider chosen standard of care and uniform collection of data.
Method(s): This study included SLE patients based on ACR or SLICC classification enrolled in AMP who met the following criteria: urine protein-creatine ratio (UPCR) > 1 at entry, and histologic biopsy Class III, IV, V, or mixed. Patients were followed at 3, 12, 26 and 52 wks with demographics, history, laboratory results, disease activity, and medica-tions recorded at each visit. Follow up data were available for 136 patients at 26 wks and 118 at 52 wks. Complete response was defined as a reduction in UPCR to <.5, a normal serum creatinine or no greater than 125% of baseline, and < 10 mg prednisone at time of response assessment. Patients were partial responders if UPCR decreased > 50% but remained >.5 and nonresponders if < 50% reduction in UPCR and/or did not meet the other response criteria.
Result(s): Medications were reported at 12 wks (Table 1). The complete response rate was 26% at both 26 and 52 wks. For patients undergoing a first biopsy, the rates were 37% and 40% and for those with repeat biopsies, the rates were lower at 21% and 19% respectively (p=0.042 at 26 wks; p=0.015 at 52 wks). The complete response at 26 wks was generally sustained with only 4 of 27 patients experiencing a relapse at 52 wks. At 26 wks, patients with membranous histology were less likely to be complete responders than patients with proliferative histology. This trend was observed regardless of biopsy number and persisted for response status at 52 wks. Although baseline activity score did not predict responder status, complete responders had a significantly lower chronicity index than nonresponders (mean + SD, 2.26 + 2.22 vs 3.83 + 2.57, p=0.016) at 26 wks with similar results at 52 wks. Responder status at 26 and 52 wks whether first or repeat biopsy was independent of extrarenal disease at entry (Table 2). Complete responder status was associated with positive anti-dsDNA serology at baseline for repeat biopsy patients. Complete responders had a greater change in C3, hemoglobin, lymphocyte count, albumin, and UPCR at 12 wks compared to baseline values than nonresponders (Table 3). Similar trends were observed when considering response status at 52 wks.
Conclusion(s): The low complete response rates reported in the AMP cohort are consistent with findings in blinded controlled trials of standard-of-care therapies and support the critical need for new therapeutics particularly in patients undergoing repeat biopsies and those with increased chronicity

Background/Purpose: Neonatal Lupus (NL) is a model of passively acquired autoimmunity conferred by exposure to maternal anti-Ro antibodies with major manifestations being congenital heart block (CHB) and/or cutaneous disease. This study was initiated to address the development of de novo autoimmunity in these children and identify associated clinical and genetic risk factors.
Method(s): In a retrospective cohort study of enrollees in the Research Registry for Neonatal Lupus (RRNL), 511 children exposed to anti-Ro in utero responded to a follow up questionnaire focused on symptoms of autoimmunity. Self-reported diseases were confirmed via medical record review. Bivariate analyses were performed with potential risk factors for the development of autoimmune disease (AD) and included the NL status per se, a disease severity score based on mortality risk factors, and maternal AD (inclusive of lupus, Sjogren's syndrome, psoriasis, rheumatoid arthritis, or thyroid disease). A subset of 99 CHB, 9 cutaneous, and 55 unaffected anti-Ro exposed RRNL individuals were genotyped at Class II HLA DRB1 and DQB1 four-digit alleles, which were assigned by imputation (HIBAG) or sequencing. Generalized estimating equations (logit link, exchangeable correlation) were used to test for associations between HLA alleles and the development of AD.
Result(s): Of the respondents, 182 offspring had CHB, 95 had cutaneous only NL and 234 were siblings without NL. Females comprised 53% and 80% were Caucasian. The mean age was 14.2+/-9.7; 4% age 0-2 years, 48% 2-13 years, and 47% > 13 years. An AD developed in 38 offspring (20 CHB, 7 cutaneous NL, 11 non-NL siblings; Table 1). The most prevalent AD was thyroid disease. The development of an AD was significantly associated with presence of CHB vs. cutaneous only or non-NL siblings (11% vs. 5%, p=0.033). The maternal health status did not influence the development of an AD in the child (7% mothers with AD vs. 6% asymptomatic mothers, p=0.67). Mean NL severity score was higher in offspring with AD (3.8+/-4.8 vs. 2.2+/-4.0, p= 0.031). Other markers of fetal CHB disease severity were associated with subsequent AD development, including in-utero exposure to fluorinated steroids (15% vs. 6%, p=0.088) and beta agonists such as terbutaline (23% vs. 9%, p=0.043). In the study of 163 RRNL cases with HLA data (20 with AD, 143 without), HLA DRB1*03:01 (OR 3.4, CI 1.46-7.90, p=0.0045), DQA1*05:01 (OR 3.39, CI 1.16-9.92, p=0.0262), and DQB1*02:01 (OR 4.28, CI 1.73-10.62, p=0.0017) were associated with increased risk of AD (of note, these loci are in high linkage disequilibrium). In contrast, these alleles were not significantly associated with development of CHB (99 CHB vs. 64 without).
Conclusion(s): The development of an autoimmune disease was more common in anti-Ro exposed children with CHB, greater NL severity, and MHC Class II haplotypes. These factors may relate to an inherent susceptibility to inflammation and fibrosis, occuring in utero and later in life

Background/Purpose: The Accelerating Medicines Partnership (AMP) will use multi-omics modalities including single cell RNA sequencing to understand lupus nephritis with the ultimate goal to devise novel and personalized treatment strategies. African-Americans have more lupus nephritis and worse outcomes in terms of end stage renal disease. We report here the clinical findings to date on African-American patients in the AMP cohort.
Method(s): We included 118 patients with urine protein-to-creatinine ratio (UPCR) >= 1 and biopsy proven class III, IV, V or mixed lupus nephritis at time of enrollment. All patients met revised ACR or SLICC classification criteria. Clinical data were obtained at baseline, 12, 26, and 52 weeks after the renal biopsy. Response status at week 52 was defined as follows. Complete: UPCR <= 0.5, normal serum creatinine (sCr) or < 25% increase from baseline if abnormal, and prednisone < 10mg daily; partial: UPCR > 0.5 but <= 50% of the baseline value and same sCr and prednisone rules as complete response; no response: UPCR > 50% of baseline value or new abnormal elevation of sCr or >= 25% from baseline or prednisone >= 10mg daily.
Result(s): Table 1 shows that African Americans were more likely to have class V lupus nephritis (38% vs 22.5%, p=0.06), were less serologically active (low C3 50% vs 77.5%, p=0.002; anti-dsDNA 63% vs 79%, p=0.006), and were more likely to have elevated serum creatinine (55% vs 30%, p=0.03). Caucasians were older (47 vs 34 years, p=< 0.001) and more likely to be at their first biopsy (64% vs 31%, p=0.04). Table 2 shows the differences based on the first biopsy versus a repeat biopsy. African-Americans were significantly less likely to have a treatment response at the first biopsy. Regardless of first or later biopsy, they were less likely to have low C3. Table 3 shows multi-variate models. African-American patients at their first episode of lupus nephritis were less likely to respond to treatment (37.5% vs 75%, p=0.018) independently of histological features including class, activity and chronicity.
Conclusion(s): The AMP cohort demonstrates the current unmet clinical need to improve treatment of lupus nephritis in the United States. African-American lupus nephritis is different in terms of ISN class, serologies, first biopsy, and worse in terms of response status even after adjusting for activity and chronicity. Personalized treatments should be developed to improve outcomes in high risk populations such as African-Americans.Table 1. Patients characteristics by race/ethnicity. Data are presented as n (%) or mean (SD). Two patients identified as "Other" and are not shown in this Table. P values > 0.1 are indicated as ns