Crowson, Cynthia S.

ObjectivePublished predictive models of disease outcomes in idiopathic inflammatory myopathies (IIMs) are sparse and of limited accuracy due to disease heterogeneity. Computational methods may address this heterogeneity by partitioning patients based on clinical and biological phenotype.MethodsTo identify new patient groups, we applied similarity network fusion (SNF) to clinical and biological data from 168 patients with myositis (64 adult polymyositis [PM], 65 adult dermatomyositis [DM], and 39 juvenile DM [JDM]) in the Rituximab in Myositis trial. We generated a sparse proof‐of‐concept bedside classifier using multinomial regression and identified characteristics that distinguished these groups. We conducted χ2 tests to link new patient groups with the myositis subtypes.ResultsSNF identified five patient groups in the discovery cohort that subdivided the myositis subtypes. The sparse multinomial regressor to predict patient group assignments (areas under the receiver operating characteristic curve = [0.78, 0.97]; areas under the precision‐recall curve = [0.55, 0.96]) found that autoantibody enrichment defined four of these groups: anti–Mi‐2, anti–signal recognition peptide (SRP), anti–nuclear matrix protein 2 (NXP2), and anti‐synthetase (Syn). Depletion of immunoglobulin M (IgM) defined the fifth group. Each group was associated with one subtype, with adult DM being associated with anti–Mi‐2 and anti‐Syn autoantibodies, JDM being associated with anti‐NXP2 autoantibodies, and adult PM being associated with IgM depletion and anti‐SRP autoantibodies. These associations enabled us to further resolve the current myositis subtypes.ConclusionUsing unsupervised machine learning, we identified clinically and biologically homogeneous groups of patients with IIMs, forming the basis of an integrated disease classification based on both clinical and biological phenotype, thus validating other approaches and what has been previously described.

Abstract Objective The study sought to determine the dependence of the Electronic Medical Records and Genomics (eMERGE) rheumatoid arthritis (RA) algorithm on both RA and electronic health record (EHR) duration. Materials and Methods Using a population-based cohort from the Mayo Clinic Biobank, we identified 497 patients with at least 1 RA diagnosis code. RA case status was manually determined using validated criteria for RA. RA duration was defined as time from first RA code to the index date of biobank enrollment. To simulate EHR duration, various years of EHR lookback were applied, starting at the index date and going backward. Model performance was determined by sensitivity, specificity, positive predictive value, negative predictive value, and area under the curve (AUC). Results The eMERGE algorithm performed well in this cohort, with overall sensitivity 53%, specificity 99%, positive predictive value 97%, negative predictive value 74%, and AUC 76%. Among patients with RA duration <2 years, sensitivity and AUC were only 9% and 54%, respectively, but increased to 71% and 85% among patients with RA duration >10 years. Longer EHR lookback also improved model performance up to a threshold of 10 years, in which sensitivity reached 52% and AUC 75%. However, optimal EHR lookback varied by RA duration; an EHR lookback of 3 years was best able to identify recently diagnosed RA cases. Conclusions eMERGE algorithm performance improves with longer RA duration as well as EHR duration up to 10 years, though shorter EHR lookback can improve identification of recently diagnosed RA cases.

Objective To investigate the effect of therapies on radiographic progression in patients with axial spondyloarthritis (SpA). Methods A comprehensive database search for studies assessing radiographic progression in axial SpA (particular treatment versus no treatment of interest) was performed. Study‐specific standardized mean differences in treatment outcomes at 2 and ≥4 years were estimated and combined using random‐effects models. Results Twenty‐four studies in patients with axial SpA were identified, of which 18 involved tumor necrosis factor inhibitors ( TNF i), 8 involved nonsteroidal antiinflammatory drugs ( NSAID s), and 1 involved secukinumab. Spinal radiographic progression, as measured by the modified Stoke Ankylosing Spondylitis Spine Score ( mSASSS ), was not significantly different between TNF i‐treated and biologics‐naive patients at 2 years ( mSASSS difference −0.73 [95% confidence interval (95% CI ) −1.52, 0.12], I 2 = 28%) and ≥4 years ( mSASSS difference −2.03 [95% CI −4.63, 0.72], I 2 = 63%). Sensitivity analyses restricted to studies with a low risk of bias showed a significant difference in spinal radiographic progression between TNF i‐treated and biologics‐naive patients at ≥4 years ( mSASSS difference −2.17 [95% CI −4.19, −0.15]). No significant difference in spinal radiographic progression was observed between NSAID ‐treated and control patients ( mSASSS difference −0.30 [95% CI −2.62, 1.31], I 2 = 71%) or between secukinumab‐treated and biologics‐naive patients ( mSASSS difference −0.34 [95% CI −0.85, 0.17]). With regard to treatment differences in patients with nonradiographic axial SpA or in patients with radiographic progression measured using the sacroiliac joint score, an insufficient number of studies were available for analysis. Conclusion Although no significant protective effect of TNF i treatment on spinal radiographic progression was seen over the course of 2 years or ≥4 years in patients with axial SpA, our analysis restricted to studies with a low risk of bias showed a protective effect of TNF i after ≥4 years. Therefore, long‐term TNF i exposure might confer beneficial effects on spinal radiographic progression in axial SpA. No difference in radiographic progression at 2 years was seen in either the NSAID or secukinumab treatment groups compared to their controls. Future studies should explore the effects of biologic treatment on radiographic progression, as well as the effects of long‐term biologics exposure, in patients with early axial SpA or those with nonradiographic axial SpA.

AbstractAimEosinophilic fasciitis (EF) is a rare, fibrosing disorder of skin and subcutaneous tissue. This study was undertaken to describe its clinical and laboratory features and identify prognostic factors associated with outcome.MethodsWe conducted a retrospective review of all EF patients evaluated at our institution from 1 January1997 to 30 December 2016. Kaplan‐Meier methods were used to determine treatment response rates over time. Potential associations between baseline characteristics and complete response were examined using Cox models adjusted for age and sex. Time‐dependent covariates were used to examine treatment effects.ResultsWe identified 89 EF patients, with a female‐to‐male ratio of 1:1. Clinical features included groove sign in 26 (29%), peau d'orange/dimpling in 22 (25%), inflammatory arthritis in 9 (10%) and muscle weakness in 9 (10%). Aldolase was elevated in 11/36 (31%). Complete response rate was 60% (95% confidence interval [CI]: 35‐75) at 3 years. Diagnostic delay was inversely associated with treatment response (hazards ratio: 0.84 per 1 month increase; 95% CI: 0.73‐0.98). No baseline characteristics correlated with treatment response, but a trend toward positive association of elevated aldolase, hypergammaglobulinemia and presence of hematologic disorders was noted. Methotrexate was the most commonly used immunosuppressant in 79%, hydroxychloroquine in 45%, mycophenolate mofetil in 18% and azathioprine in 8%. No single immunosuppressant agent was associated with a superior response during treatment.ConclusionsEF is characterized by relatively high response rates. Consensus diagnostic criteria, standardized management algorithms, and large prospective multi‐center cohorts are needed to develop an evidence‐directed approach to this challenging condition.

Objective Since comorbidities such as autoimmune diseases may be associated with rheumatoid arthritis (RA) risk, we hypothesized that a family history of these other conditions might also predict RA. Therefore, we aimed to determine the association between family history of 79 comorbidities and RA. Methods This case–control study identified 821 cases of RA in the Mayo Clinic Biobank (positive predictive value 95%) and matched 3 controls to each case based on age, sex, recruitment year, and location. Patients self reported family history and characteristics (adjusted). Logistic regression was used to estimate odds ratios (ORs) and 95% confidence intervals (95% CIs) for RA risk according to the presence of family history for each comorbidity, adjusted for body mass index, race, and smoking. Results Family history of several conditions was associated with developing RA, including rheumatic autoimmune diseases (OR adj 1.89 [95% CI 1.41–2.52]), pulmonary fibrosis (OR adj 2.12 [95% CI 1.16–3.80]), inflammatory bowel disease (OR adj 1.45 [95% CI 1.05–1.98]), hyper/hypothyroidism (OR adj 1.34 [95% CI 1.10–1.63]), and obstructive sleep apnea (OR adj 1.28 [95% CI 1.05–1.55]). Parkinson’s disease and type 2 diabetes mellitus were associated with a statistically decreased risk of RA that did not reach the prespecified significance threshold of P < 0.01 (OR adj 0.70 [95% CI 0.49–0.98] and OR adj 0.81 [95% CI 0.67–0.97], respectively). Analyses among 143 cases of incident RA were similar and also suggested an association with a family history of autism (OR 10.5 [95% CI 2.51–71.3]). Conclusion Family history of several autoimmune and nonautoimmune comorbidities was associated with increased risk of RA, providing an opportunity to identify novel populations at risk for RA.

ObjectiveThe objective of this study is to compare biologic drug discontinuation rates for older‐ versus younger‐onset rheumatoid arthritis (YORA) because this is a key outcome measure that could impact prescribing practices.MethodsWe performed a retrospective medical record review of all patients who fulfilled the 1987 American College of Rheumatology (ACR) criteria for adult‐onset rheumatoid arthritis (RA) in 1999‐2013 among residents of a geographically defined area, with follow‐up until death, migration, or July 1, 2017. Discontinuation rates were estimated using cumulative incidence adjusted for the competing risk of death.ResultsA total of 240 cases of elderly‐onset rheumatoid arthritis (EORA) and 366 cases of YORA were identified (65% and 73% female, respectively; P = 0.025). Cumulative incidence of biologic initiation was lower among the EORA cohort compared with the YORA cohort (18% vs 33%, respectively, at 10 years after RA incidence; P < 0.001). Among those treated with a biologic, years from RA diagnosis to first biologic treatment was not significantly different between the two groups (P = 0.62). Drug survival of first biologic was 64% at 1 year (95% confidence interval [CI]: 45%‐77%) and 53% at 2 years (95% CI: 33%‐66%) for EORA, compared with 61% at 1 year (95% CI: 50%‐69%) and 45% at 2 years (95% CI: 34%‐53%) for YORA (P = 0.75). Concurrent glucocorticoid use at initiation of first biologic was statistically and significantly associated with a lower risk of discontinuation in EORA (hazard ratio 0.21; 95% CI: 0.08‐0.53) but not in YORA (interaction P = 0.04).ConclusionDrug survival rates of biologic medications did not differ significantly between patients with EORA and YORA.

Abstract Objectives Smoking is a major risk factor for the development of both cardiovascular disease (CVD) and RA and may cause attenuated responses to anti-rheumatic treatments. Our aim was to compare disease activity, CVD risk factors and CVD event rates across smoking status in RA patients. Methods Disease characteristics, CVD risk factors and relevant medications were recorded in RA patients without prior CVD from 10 countries (Norway, UK, Netherlands, USA, Sweden, Greece, South Africa, Spain, Canada and Mexico). Information on CVD events was collected. Adjusted analysis of variance, logistic regression and Cox models were applied to compare RA disease activity (DAS28), CVD risk factors and event rates across categories of smoking status. Results Of the 3311 RA patients (1012 former, 887 current and 1412 never smokers), 235 experienced CVD events during a median follow-up of 3.5 years (interquartile range 2.5–6.1). At enrolment, current smokers were more likely to have moderate or high disease activity compared with former and never smokers (P &lt; 0.001 for both). There was a gradient of worsening CVD risk factor profiles (lipoproteins and blood pressure) from never to former to current smokers. Furthermore, former and never smokers had significantly lower CVD event rates compared with current smokers [hazard ratio 0.70 (95% CI 0.51, 0.95), P = 0.02 and 0.48 (0.34, 0.69), P &lt; 0.001, respectively]. The CVD event rates for former and never smokers were comparable. Conclusion Smoking cessation in patients with RA was associated with lower disease activity and improved lipid profiles and was a predictor of reduced rates of CVD events.

Abstract Objectives Multicentric reticulohistiocytosis (MRH), a rare histiocytic disease that can mimic other rheumatic conditions, may be associated with cancer and other autoimmune disorders. To better understand the disorder and its other associations, we aimed to evaluate clinical correlates and outcomes of all patients with MRH seen at Mayo Clinic, Rochester between 1980 and 2017. Methods A retrospective medical record review was conducted to identify all patients with MRH between 1 January 1980 and 30 April 2017. Results We identified 24 patients with biopsy-proven MRH (58% female, 75% Caucasian, median age at diagnosis 52 years, median follow-up of 2.3 years). All patients had cutaneous and articular involvement; 23 (96%) patients had papulonodular skin lesions (87% periungual and dorsal hand) and seven (30%) mucosal nodules; and 22 (92%) patients had arthralgias, 21 (88%) joint effusions and 13 (54%) synovitis. Most frequently used therapies included corticosteroids, cyclophosphamide, methotrexate and bisphosphonates. Biologics were used in four patients. Nine patients had symptomatic resolution at 1 year and 12 partial improvement. Radiological findings included erosive changes in three (60%) patients and arthritis mutilans in two patients (40%). Twenty-nine per cent of patients had a concomitant autoimmune disease and 25% malignancy including melanoma, endometrial, peritoneal and lung carcinoma. The 5-year survival rate was 85% (95% CI: 74, 100%). Conclusion To our knowledge, this is the largest single-centre series of patients with MRH highlighting the rarity of the condition and an unmet need for treatment options that can allow sustained disease remission. It also highlights the need for a high vigilance for malignancy and autoimmune diseases.

Abstract Background: Non-Hodgkin lymphoma (NHL) is a relatively uncommon cancer with annual incidence of ~70,000 cases but with over 50 distinct subtypes. The goal of this study was to determine the extent to which the number of NHL patients treated annually in a facility (facility volume) affects overall survival (OS). This study used the National Cancer Data Base (NCDB), a nationwide oncology database covering 70% of the US cancer population, to address this question. Methods: We used the NCDB to identify patients with NHL diagnosed from 1998 to 2006. Year 2006 was used as a cut-off in order to allow a minimum of five years of follow-up for all patients. Only patients treated at facilities with continuous annual reporting to NCDB were included. We classified treatment facilities by quartiles based on facility volume (mean patients/year): Quartile 1 (Q1: 2-13), Quartile 2 (Q2: 14-20), Quartile 3 (Q3: 21-32) and Quartile 4 (Q4: ≥33). We used Pearson correlation methods to examine collinearity, unadjusted Kaplan-Meier methods to estimate OS rates, log rank test to compare survival distributions, and multivariable Cox proportional hazards model to examine the associations between hospital volume and OS adjusting for other covariates of interest. We also included random effects for hospital to more fully adjust for clustering of outcomes within hospitals. To examine non-linear effects of hospital volume, we utilized smoothing splines. Results: There were 278,985 NHL patients cared for at 1,151 facilities. The distribution of patients according to facility volume was Q1 (10.7%), Q2 (13.5%), Q3 (23.3%) and Q4 (52.5%) and according to facility type was academic (31.2%), comprehensive community (55.9%), community (10.6%) and other (2.3%) centers. The unadjusted median OS by facility volume was: Q1: 61.8 months, Q2: 65.9 months, Q3: 71.4 months and Q4: 83.6 months. After multivariable analysis adjusting for demographic (sex, age, race, ethnicity), socioeconomic (income, insurance type), geographic (area of residence), disease-specific (NHL subtype, stage) and facility-specific (type and location) factors, we show that facility volume remains an independent predictor of all-cause mortality. Compared to patients treated at Q4 facilities, patients treated at lower quartile facilities had a worse OS (Q3HR: 1.05 [95% CI, 1.04-1.06]; Q2HR: 1.08 [1.07-1.10]; Q1HR: 1.14 [1.11-1.17]). We adjusted for hospital as a random effect, performed sensitivity analyses removing primary payor and facility type (due to collinearity with age and facility volume, respectively), and adjusted for Charlson-Deyo co-morbidity score (available only for patients diagnosed after 2003) in secondary models and found similar results. Using smoothing splines, we found a significant non-linear effect of hospital volume on OS (P &lt;0.001). This is depicted in the Figure wherein the hazard ratio of 1.0 corresponded to the average predicted hazard, which occurred at a hospital volume of 59 patients per year. Conclusions: Patients who were treated for NHL at higher volume facilities had longer OS than those who were treated at facilities with a lower volume. This is the first study in the US using a national sample to show that a volume-outcome relationship exists in the medical management of cancer. Figure 1. Figure 1. Disclosures No relevant conflicts of interest to declare.