Background The burden of comorbid illness in Rheumatoid Arthritis is high. Whilst there can be coincidental existence of comorbidity, it can be attributed to the disease process itself or therapeutic agents. Recent EULAR guidelines recommend reporting, screening and prevention of six common co-morbidities (cardiovascular, malignancy, infections, gastrointestinal, osteoporosis, depression). Limited research is available evaluating the impact of comorbidity on disease response following biologic treatment.

Objectives To analyse retrospective data from the King's College Hospital (KCH) Virtual Biologic Clinic (VBC) to assess the impact of comorbidity on disease response following treatment with biologics.

Methods Retrospective patient note review data was collected for patients referred to the VBC from May 2013 to July 2016. The following data was recorded: age, sex, disease duration, smoking status, BMI, presence or absence of Anti CCP antibody +/-Rheumatoid factor and the six specified comorbidities. Disease Activity Score in 28 joints (DAS28) at time of referral for biologic and within 6 months of commencing treatment was also recorded in order to calculate treatment response.

The impact of comorbidity and disease variables on 6 month EULAR response were analysed using logistic regression (SPSS version 23).

Results The database contained 150 patients. 18 patients were excluded due to no follow-up DAS28, leaving 132 for analysis. Mean age and disease duration were 58 years and 10.45 years respectively. Comorbidity was present in 70% of patients. 70% of patients achieved a EULAR moderate response (improvement of >1.2 of DAS28) and 36% of this group achieved EULAR good response (DAS28<3.2). The most prevalent comorbidities were infection, cardiovascular disease and depression.

Logistic regression analysis was run analysing EULAR moderate response against presence of comorbidity and dataset variables (age, gender, serostatus, baseline DAS, HAQ and polypharmacy). The resulting model was not statistically significant (p=0.975).

Logistic regression analysis looking at EULAR good response against presence of comorbidity was also not statistically significiant (p=0.149). Analysis looking at EULAR good response against three variables (HAQ, baseline DAS and serostatus) was found to be statistically significant (p<0.001)

Conclusions Comorbidities were present in the majority of patients assessed (70%). Results show no significant relationship between EULAR moderate response or EULAR good response and the presence of comorbidity. Lower HAQ, lower baseline DAS28 and seronegativity contribute towards good treatment response.The limitations of this study include gaps in clinical records, absence of patient reported questionnaires, and small size of database.

Disclosure of Interest None declared