Pharmacokinetics of Prednisone in Adults

OBJECTIVES Main objective Describe the pharmacokinetic of prednisone when administered to adult patients in the setting of an internal medicine service.

Secondary objectives To identify individual vulnerability factors (renal and hepatic functions, albuminemia, age, comedications) and glucocorticoid tolerance factors in adults.

To study the association between individual exposition and glucocorticoids' frequent and clinically significative side effects.

METHOD Conduct of research Population follow up Clinical and biological data will be collected through an observational prospective mono-centric study. Patients' inclusion will be performed in the "Internal medicine service of Cochin hospital, Assistance Publique - Hôpitaux de Paris (AP-HP)" during hospitalization and will be supervised by the pharmacology service of Cochin hospital.

The evaluated drug is prednisone, used for routine care in auto-immune and other systemic inflammatory diseases. Prednisone will be administered according to the service usual practices, dosing will be determined according to the patient's disease. The study will not modify the prescription of the treatment. Blood tests will be performed at different intervals during each patient's normal biological monitoring of the treatment.

Several tubes will be collected for the needs of the study at the time of each venipuncture (blood sample or peripheral venous catheter laying). The delay between drug intake and blood sampling will be attributed to routine care hazard, samples will be then realized on different occasions. No venous or arterial punction will be specifically planned for protocol purposes, GLUCOMAD's samples will be adapted to routine care blood samplings.

1st visit : Patients recruiting phase Will occur during hospitalization. Its goal is to check for eligibility criteria.

During this visit, the investigator will explain the patient the purpose of the study. The patient will receive an information notice and a consent form.

Patient's data collection :

• Age
• Sex
• Pregnancy, menopause
• Personal history of kidney, liver or thyroid disease. Personal or familial history of adrenal pathology, impaired glucose tolerance, diabetes mellitus, gestational diabetes, high blood pressure, preeclampsia, hypertension in pregnancy, open angle glaucoma, cataract, osteoporosis/osteopenia, osteoporotic fracture, gastric ulcer, psychiatric disorder, dyslipidemia.
• Osteoporosis risk factors
• Blood pressure, body weight, height, and clinical exam results looking in particular for hepatocellular insufficiency, or uremic syndrome.
• Face photography at the beginning and the end of the study participation.

Data collection about prednisone :

• Treatment indication
• Anterior corticotherapy, and optionally, anterior treatment duration, total dose taken up (as estimated by recovered orders from the patients pharmacy), and side effects of this previous treatment.
• Name of the chosen molecule (international nonproprietary name and commercial name), daily dosing (mg/Kg), treatment's start date.
• Adherence to the treatment, evaluated by the medication possession ratio (MPR42).
• Clinical symptoms of corticotherapy side effects Actual co-occuring treatments,
• Especially for medications interacting with P450 cytochrome, or P glycoprotein. Including the non proprietary name, treatment's start date, dosage, and optionally the results of plasma rates testing.
• Side effects treatments : anti hypertensive drugs, hypolipidemic drugs, potassium supplementation, oral antidiabetic drugs/insulin, psychotropic drugs.

Blood tests Follow up

• During hospitalizations, routine care blood samples will be performed, on these occasions, blood will be collected in a dry tube for the study purposes.
• At the end of the attack phase, albumin, transcortin, HbA1c, and 8am cortisol will be assayed.
• Usual clinical and biological follow up results, including body weight and side effects monitoring, will be collected.
• Medication possession ratio assessment
• Co-occuring treatments Collection Sampling procedure, place of collection Blood samples necessary to the study will be stored in an anonymous way. Anonymity will be guaranteed all the study long.

Storage location and conditions Samples for non pharmacological tests will be tested and destroyed. All the pharmacological samples will be stored during the study at -80°C in the pharmacology service of Cochin hospital where pharmacological analyses will be performed.

Samples processing Molecules plasma assays will be performed in the drug dosing platform of clinical pharmacology service of Cochin hospital, AP-HP. Free and bound forms of prednisolone will be tested.

Samples management Pharmacological samples will be stored at the end of the study for complementary studies.

Statistical considerations Statistical analyses We propose to describe prednisone's pharmacokinetic in adults thanks to a population pharmacokinetic approach. This technique is specially adapted for populations affected by rare inflammatory diseases that are supported in internal medicine services. Indeed, it only requires a few measures in each patient and allows flexibility in data collection ; thus it enables us to assay concentrations on routine care samples. This method will allow us to describe an average subject pharmacokinetic and then to estimate interindividual variability in the population under prednisone treatment. We will the attempt to explain this variability according to covariates as age, bodyweight, size, liver or kidney insufficiency, or co-occurring treatments.

Individual pharmacokinetic parameters will be estimated from this population model through a Bayesian estimation method. Potential associations between individual expositions (i.e. residual concentration, area under the curve, maximal concentration), and prednisone side effects will be studied.

Pharmaco-statistical analysis will be made thanks to non linear mixed effects modelisation softwares (Monolix 4.2 or Nonmem 7) to calculate prednisone's pharmacokinetic parameters as well as inter-individual variability parameters on one hand and residual variability on the other hand (the latter including intra-individual variability and measurement error).

The final model will be validated by a simulation technique. Briefly, 500 simulations of the final model will be generated from the original database. Then the 5th, 50th, and 95th percentiles of the simulated predictions will be compared to observations (the simulation median has to be close to the observation's median, and 90% of the observation must be between the 5th and 95th percentiles of simulations).