23.10.2012 Views

View PDF Version - RePub - Erasmus Universiteit Rotterdam

View PDF Version - RePub - Erasmus Universiteit Rotterdam

View PDF Version - RePub - Erasmus Universiteit Rotterdam

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

Chapter 4.3<br />

214<br />

The resulting model function V i (t) describes the viral load of individual i at time point t<br />

and depends on the virion clearance rate c i and the infected cell loss rate δ i . The total<br />

number of cells (i.e. infected target cells, I i , and uninfected target cells, T i ) is assumed to<br />

remain constant in each individual during treatment motivated by a fast liver regeneration.<br />

As usual, the infection rate βi and the viral production rate pi were substituted by<br />

the other parameters assuming that they remained unchanged from the steady state<br />

situation.<br />

Modelling and data fi tting<br />

The PROC NLMIXED procedure of SAS 9.1 (SAS Institute Inc., Cary, NC) was used to fi t<br />

the fi rst month pharmacokinetic data of all 96 patients with a non-linear mixed modelling<br />

approach. The NLME procedure of R (R Foundation for Statistical Computing)<br />

yield highly comparable results (data not shown). The prediction of the PEG-IFN α-2b<br />

concentration (equation 2) and the model for effectiveness (equation 3) was thereafter<br />

incorporated in the model of the viral load as solution from equations 4 and 5 of the<br />

patients treated with PEG-IFN α-2b monotherapy. The viral load was hereafter fi tted<br />

with non-linear mixed modelling with the NLME procedure of R including the ordinary<br />

differential equation solver LSODA from the ODESOLVE package in a nested way to<br />

estimate the infected cell loss rate δ, the baseline levels of viral load as well as the IC50 levels and the time delay t0 . Because interindividual variation could already be modelled<br />

by baseline viral load and IC50 levels, the other parameters were set constant between<br />

patients (fi xed effects). Furthermore, relatively few data points can lead to biased<br />

estimates of the viral clearance rate c. 16 Therefore, we fi xed c to 1.3 per day. Different<br />

Hill coeffi cients (n=1, n=2, n=3 and n=4) were checked and we used a coeffi cient of 1<br />

because this gave the best results.<br />

SPSS (version 14.0.1, SPSS Inc., Chicago, IL) was used for further data analyses. All tests<br />

for signifi cance and resulting P values were two-sided, with a level of signifi cance of 0.05.<br />

RESULTS<br />

Patient characteristics<br />

Demographic and baseline characteristics of the 96 included patients in this study are<br />

shown in table 1. Forty-eight patients received PEG-IFN α-2b monotherapy; the other<br />

48 patients received combination therapy consisting of PEG-IFN α-2b and lamivudine.<br />

There were no signifi cant differences between the two groups with respect to ALT, viral<br />

load, age, sex, weight and race. PEG-IFN α-2b concentration was measured in all 96

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!