Article Text

PDF
FRI0450 Predictors of Mortality in Systemic Sclerosis: The Singapore Scleroderma Registry
  1. A. Santosa1,
  2. G.G. Teng1,
  3. C.S. Tan2,
  4. W. Fong3,
  5. W.G. Law4,
  6. G. Chan4,
  7. E. Wong3,
  8. H.Q. Teo3,
  9. P.T. Lee1,
  10. A. Low3,5
  1. 1Division of Rheumatology, University Medicine Cluster, National University Health System
  2. 2Saw Swee Hock School of Public Health, National University of Singapore
  3. 3Department of Rheumatology and Immunology, Singapore General Hospital
  4. 4Department of Rheumatology, Allergy and Immunology, Tan Tock Seng Hospital
  5. 5Duke-National University of Singapore Graduate Medical School, Singapore, Singapore

Abstract

Background Although prognosis of systemic sclerosis (SSc) has improved significantly, the disease continues to cause excess mortality. Studies assessing mortality in SSc are lacking in Asia.

Objectives We aim to describe the causes and evaluate predictors of mortality in Singapore Asian patients with SSc in a retrospective and inception cohort.

Methods From 2008, patients diagnosed with SSc fulfilling the American College of Rheumatology (ACR)/EULAR1 or Very Early Diagnosis of Systemic Sclerosis (VEDOSS)2 criteria were recruited into the Singapore Scleroderma Registry. Demographics, disease manifestations and treatment for prevalent and incident cases seen at three tertiary rheumatology centers were collected using a standardized protocol. Mortality was verified with the national death registry up to 1 December 2013. In-hospital cause of death was determined by two independent reviewers according to categories set apriori. Person-years follow-up was calculated from date of cohort entry to either death or censor date. A univariate Cox proportional hazard (PH) regression was used to examine the association between each variable and mortality. Variables reported in the literature to associate with mortality and/or P value<0.1 were considered in the multiple Cox PH regression model. The forward stepwise selection approach and the Akaike Information Criterion were used to determine the multiple regression model, where age, gender and race were included as confounders.

Results Of 349 SSc patients (87% females), 97% fulfilled the ACR/EULAR 2013 criteria (89% and 3% fulfilled 1980 ACR and VEDOSS criteria respectively). The mean age at diagnosis was 46.2 (SD 15) years, with 3216.2 person-years follow-up from initial SSc diagnosis or 740.6 person-years from cohort entry. Limited (LcSSc), diffuse (DcSSc) cutaneous SSc and SSc-overlap syndromes occurred in 35%, 38% and 26%, respectively. There were 35 deaths (10%) over 740.6 person-years follow-up, translating to a mortality rate of 47.3 per 1000 patient-years. Fifty-seven percent of deaths were SSc-related (interstitial lung disease [ILD, n=6], pulmonary arterial hypertension [PAH, n=6], gastrointestinal involvement [GI, n=3], renal crisis [n=2], cardiac [n=1] and others [n=2]). Multivariate analysis (n=275 with complete data set) (Table 1) showed that race (Indians had the highest hazards ratio [HR] for mortality compared to Chinese), smoking, disease subtype (SSc-overlap had the highest HR), baseline renal involvement, pulmonary artery systolic pressure (PASP) ≥40mmHg on echocardiography, and treatment with peripheral vasodilators or parenteral nutrition were independently associated with mortality.

Conclusions PAH and ILD are leading causes of death in this Asian cohort. Disease subtype, race, renal involvement, PASP ≥40mmHg, smoking, and treatment with peripheral vasodilators or parenteral nutrition were independent predictors of mortality.

References

  1. van den Hoogen F, et al. Ann Rheum Dis (2013), 72(11):1747-55.

  2. Matucci-Cerinic M, et al. Ann Rheum Dis (2009), 68(9):1377-80.

Disclosure of Interest None declared

Statistics from Altmetric.com

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.