Medicines such as methotrexate suppress the immune-system and are used to treat conditions like rheumatoid arthritis (RA) that result from an overactive immune-system. Patients treated with these medicines may develop side-effects such as blood, liver or kidney damage, and, fortnightly to monthly blood tests are performed to detect them early, before any permanent damage can occur. However, these side-effects become less common after the first few months of treatment. Nevertheless, monitoring with periodic blood tests is continued indefinitely for all patients treated with these medicines. The benefit from such long-term monitoring is not known, and, whether all patients should undergo regular blood-tests indefinitely needs to be determined.
We want to find out the best strategy for long-term monitoring of people with inflammatory conditions (e.g. RA, psoriasis, inflammatory bowel disease etc.) treated with immune-suppressing medicines, specifically, methotrexate, sulfasalazine, aminosalicylates (e.g. mesalazine), azathioprine, and mycophenolate.
For this, we will use anonymized information from a healthcare database (Clinical Practice Research Datalink), which includes information on the treatment and monitoring of adults registered with a large number of general practices in the UK. We will use information about risk factors such as age, sex, body mass index, other illnesses, lifestyle factors (such as alcohol intake, smoking) and prescription of other medicines to develop a score to calculate the risk of blood, liver, or kidney damage over a 5-year time-period for people treated with these medicines. A separate risk score will be calculated for each medicine of interest named in the previous paragraph.
Background: The optimal monitoring strategy for adults with common inflammatory conditions treated with long-term immune-suppressing drugs is not known.
Question: What is the optimum strategy of blood testing after the first six months of primary-care immune-suppressing drug prescription?
Objectives: [1] Develop and validate a prediction model for estimating an individual’s risk of target-organ damage conditional on their values of multiple prognostic factors available at 6 months of primary-care immune-suppressing drug prescription, [2] Identify incidence of drug discontinuation for any reason, and due to abnormal monitoring blood test results in each year of prescription.
Methods: Data from Clinical Practice Research Datalink (CPRD) Gold and Aurum will be used. We will utilise CPRD-HES linked data to validate the sensitivity of outcome definition in primary-care data alone.
People with common inflammatory conditions treated with immune-suppressing drugs, will be followed up electronically from day 180 of the first primary-care prescription of a drug of interest until the earliest of target-organ adverse event, drug discontinuation, death, moving to a different practice, 5-year follow-up time, or 31/12/2019 (study end date).
A penalized flexible parametric survival model will be used to develop a risk prediction score for drug discontinuation due to target organ damage using CPRD Aurum given its larger sample size, accounting for competing risks as necessary, and with internal validation using bootstrapping to quantify and adjust for optimism. The predictive performance of the score will be validated in CPRD Gold. GP practices contributing data to both CPRD Aurum and GOLD will be excluded from the validation cohort. Separate risk scores will be developed and validated for each drug of interest. Risk thresholds will be agreed using decision curve analysis and with patient and clinician involvement. Incidence (per 1,000 person-years) of treatment discontinuation in each 12-month of primary-care prescription will be estimated.
Primary outcome: Drug discontinuation due to target organ specific adverse event. This will be defined as drug discontinuation with either (a) abnormal blood test results due to target-organ damage, or (b) target organ damage based on clinical code allocation.
Drug discontinuation due to abnormal monitoring blood test results will be defined as present when there is:
(a) a gap between two prescriptions, or between the last prescription date and earliest of death date, moving to a different practice, 5-year follow-up time, or 31/12/2019 (study end date) of ≥ 90 days, and
(b) the last prescription is preceded in up to 15 days by abnormal blood test results (to account for any delay in acting on abnormal results) or followed by an abnormal test result in the next 45 days.
Abnormal blood test results will be defined according to the British Society for Rheumatology DMARD monitoring guidelines1, and CKD progression will be defined as per the Kidney Disease Improving Global Outcomes guideline 20122:
• White blood cells <3.5x109/L,
• Neutrophils <1.6x109/L,
• Platelets <140x109/L,
• Alanine Transaminase and/or Aspartate Transaminase >100 units/L,
• CKD progression defined as creatinine difference > 26 micro mol/L above the first or second preceding blood-test result or chronic kidney disease stage progression.
In addition to the above blood-test abnormalities, in people with psoriasis, methotrexate discontinuation will be defined to be due to blood-test abnormalities if (i) serum amino-terminal pro-peptide of type III procollagen (PIIINP) is > 8 mg/L on two previous occasions; (ii) if three previous PIIINP measurements were >4.2 mg/ L in the previous 12-month period; or (iii) if previous PIIINP measurement was > 10 mg/L as per guidelines for the British Association of Dermatologists [3].
Secondary outcome: Drug discontinuation due to any reason.
Defined as present if there is a gap between two prescriptions, or between last prescription date and earliest of death date, moving to a different practice, 5-year follow-up time, or 31/12/2019 (study end date) of ≥ 90 days.
Abhishek Abhishek - Chief Investigator - University of Nottingham
Georgina Nakafero - Corresponding Applicant - University of Nottingham
Christian Mallen - Collaborator - Keele University
Danielle van der Windt - Collaborator - Keele University
Matthew Grainge - Collaborator - University of Nottingham
Richard Riley - Collaborator - Keele University
Timothy Card - Collaborator - University of Nottingham
HES Admitted Patient Care