Clinical Pharmacokinetics

, Volume 46, Issue 5, pp 417–432

Exposure-Effect Population Model of Inolimomab, a Monoclonal Antibody Administered in First-Line Treatment for Acute Graft-Versus-Host Disease


  • Céline Dartois
    • EA3738 CTO, Faculté de Médecine Lyon SudUniversitéLyon 1
    • Université de Lyon
  • Gilles Freyer
    • EA3738 CTO, Faculté de Médecine Lyon SudUniversitéLyon 1
    • Université de Lyon
    • Service d’Oncologie MédicaleCentre Hospitalier Lyon Sud, Hospices Civils de Lyon
  • Mauricette Michallet
    • Service d’HématologieHôpital Edouard Herriot
  • Emilie Hénin
    • EA3738 CTO, Faculté de Médecine Lyon SudUniversitéLyon 1
    • Université de Lyon
  • Benoît You
    • EA3738 CTO, Faculté de Médecine Lyon SudUniversitéLyon 1
    • Université de Lyon
    • Service d’Oncologie MédicaleCentre Hospitalier Lyon Sud, Hospices Civils de Lyon
  • Isabelle Darlavoix
    • OPi SA
  • Claudine Vermot-Desroches
    • OPi SA
  • Brigitte Tranchand
    • EA3738 CTO, Faculté de Médecine Lyon SudUniversitéLyon 1
    • Université de Lyon
    • Centre Anticancéreux Léon Bérard
    • EA3738 CTO, Faculté de Médecine Lyon SudUniversitéLyon 1
    • Université de Lyon
Original Research Article

DOI: 10.2165/00003088-200746050-00004

Cite this article as:
Dartois, C., Freyer, G., Michallet, M. et al. Clin Pharmacokinet (2007) 46: 417. doi:10.2165/00003088-200746050-00004


Background and objective

Inolimomab, a monoclonal antibody against interleukin (IL)-2Rα (CD25) has shown promising results in the treatment of corticosteroid-resistant acute graft-versus-host disease (GvHD). The objective of the present study was to characterise the pharmacokinetic and pharmacodynamic properties of inolimomab as first-line treatment in this condition.


The data came from 21 patients with acute GvHD (8 with an International Bone Marrow Transplant Registry [IBMTR] score of B, 11 with a score of C and 2 with a score of D) following haematopoietic stem cell transplantation after a median delay of 26 days (range 10–127 days). Inolimomab was administered at 0.1, 0.2, 0.3 or 0.4 mg/kg daily in association with methylprednisolone (2 mg/kg) for 8 or 16 days depending on the status at day 9. Then, for responder patients, administrations were continued three times weekly until day 28. Inolimomab concentrations and pharmacodynamic data (acute GvHD scores) were recorded during the study. The pharmacodynamic data were assessed in four grades according to the IBMTR and Glucksberg classification in parallel with Karnofsky scores. A population analysis was developed using a nonlinear mixedeffects model to define the pharmacokinetic model, to test covariates and, when apparent, to model the exposure-effect relationship by a proportional odds model. The modelling was finally qualified by a predictive check.


The best pharmacokinetic model was two-compartmental. For each score, the most demonstrative exposure-effect graphics linked the cumulative area under the concentration-time curve to cumulated probabilities of observing a specific score. This relationship was identified as a maximum effect model for the skin (with two patient subpopulations: sensitive/less sensitive) and a linear model for the intestinal tract and liver. No covariate was identified as influencing any of these parameters.


Inolimomab exposure-effect relationships as first-line treatment for acute GvHD have been identified and modelled. The discovered dose-effect relationship allows confirmation of the treatment response, thereby establishing the first step towards optimising the inolimomab dosage in future trials.


Allogeneic haematopoietic stem cell transplantation is an effective and curative treatment for many haematological malignancies[1] related to the existence of a graft-versus-malignancy effect. However, it is frequently associated with graft-versus-host disease (GvHD), which is still responsible for a high rate of treatment-related mortality.[2] Acute GvHD usually involves the skin, liver and intestinal tract, but lymphoid and haematopoietic tissues can also be affected. It is induced by alloreactive T cells from the donor, which react against the recipient’s tissues and organs. Standard GvHD prophylaxis consists of administering a combination of immunosuppressive drugs, ciclosporin and methotrexate being the classical combination.[3] Other therapeutic approaches have been tested, including T-cell depletion, which can be performed in vivo or ex vivo. The incidence of GvHD decreases, but this is at the cost of a high risk of rejection and relapse associated with delayed immune reconstitution.[4] The first-line treatment of acute GvHD is based on corticosteroids, usually methylprednisolone at the dose of 2–2.5 mg/kg/day.[5] However, corticosteroid resistance is observed in ≊40% of patients and therefore requires alternative treatment.[6,7] No standard therapy really exists for corticosteroid-refractory acute GvHD. Therapy could be based on high-dose corticosteroids (10–15 mg/kg/day) either alone or in combination with antithymocyte globulins or monoclonal anti-T-cell preparations.[810] In some cases, although it cures acute GvHD, this treatment is responsible for strong immunosuppression, leading to an increased incidence of severe bacterial infections, viral infections, and an increased risk of Epstein-Barr virus-related lymphoproliferative disorders.[11,12] Because of its inhibitory effects on activated T cells, inolimomab (Leukotac®1, OPi SA, Limonest, France) could be useful for the treatment of acute GvHD. This murine monoclonal antibody specifically targets the α chain (CD25) of the interleukin-2 (IL-2) receptor. Activated T cells express the inducible IL-2Rα chain, whereas resting cells and their precursors do not. Consequently, fewer adverse experiences are expected owing to lower and more targeted immunosuppressive activity. Some clinical trials have already been performed in corticosteroid-resistant acute GvHD patients and have shown some promising results in terms of response and survival.[1317]

A clinical trial of inolimomab, given in combination with corticosteroids, was conducted as initial therapy for acute GvHD. As the compound was well tolerated,[15] this clinical trial was expected to show a better, longer and less heterogeneous overall response than would be expected in corticosteroid-resistant patients. This study presents the original population pharmacokinetic-pharmacodynamic modelling of these clinical trial data, showing an exposure-effect relationship of a monoclonal antibody for the first time in this indication. The specific aims were: (i) to model the pharmacokinetics of inolimomab given as a repeated dose; (ii) to identify inolimomab exposure-effect relationships on different efficacy markers in acute GvHD; and (iii) to propose a model in order to help future dose optimisation of this treatment.


Patients and Treatment

The data were collected from an open-label, dose-escalating, nonrandomised phase I–II study of inolimomab in combination with corticosteroids (methylprednisolone 2 mg/kg) as first-line therapy for grade II–IV acute GvHD following allogeneic haematopoietic stem cell transplantation. The main objective of this trial was to establish the pharmacokinetics of four dosages of inolimomab. Six French institutions participated in the study (Hôpital Edouard Herriot, Lyon; Centre Jean Perrin, Clermont-Ferrand; Institut Paoli Calmettes, Marseille; Hôpital Claude Hurriez, Lille; Hôtel Dieu, Nantes; Hôpital Henri Mondor, Créteil). The study protocol was approved by the Independent Ethics Committee of Lyon — Centre Léon Bérard. The inclusion criteria included: age ≥18 years, grade II–IV acute GvHD following haematopoietic stem cell transplantation using either allogeneic bone marrow or allogeneic peripheral blood progenitor cells, and provision of informed consent to participate in the study. Patients were excluded if they were receiving corticosteroids for prophylaxis against acute GvHD or had acute GvHD after donor lymphocyte infusions. Diagnosis and classification of acute GvHD was done according to the Seattle criteria (the Glucksberg classification)[18,19] and the International Bone Marrow Transplant Registry (IBMTR) classification.[20]

After their eligibility was confirmed, 21 patients were registered and assigned to one of four cohorts to receive a 30-minute intravenous inolimomab infusion (0.1, 0.2, 0.3 or 0.4 mg/kg), with five patients in each dose group except for the 0.3 mg/kg group, which had six patients. The treatment was divided into the induction and maintenance regimen phases. The induction regimen was given from day 1 to day 8 and consisted of a once-daily intravenous infusion of inolimomab at the patient’s assigned dose level. The clinical response assessed at day 9 determined subsequent treatment. Patients with a complete response were assigned to receive the maintenance regimen. Patients with a partial response, mixed response, no response or disease progression were reassigned to the induction regimen for 1 week. The maintenance regimen consisted of administration of intravenous inolimomab three times weekly at the patient’s induction dose level. During both the induction and maintenance regimen phases, all patients received a concomitant intravenous infusion of methylprednisolone. Patients received between 6 and 22 administrations of inolimomab depending on the duration of their induction and maintenance regimen phases. The entire treatment period lasted a maximum of 4 weeks.

A total of 21 patients (12 men and 9 women, age range 29–61 years, bodyweight range 49–93kg) were enrolled in the trial (table I). Based on their initial disease, they were classified into three groups: six patients with a good prognosis after successful transplantation (for chronic myeloid leukaemia in the chronic phase, acute myeloblastic leukaemia, or acute lymphoblastic leukaemia in the first complete remission), nine with an intermediate prognosis (for acute myeloblastic leukaemia, acute lymphoblastic leukaemia in the second complete remission or above, non-Hodgkin’s lymphoma, myeloma, or Hodgkin’s disease with a partial response), and six with a poor prognosis (refractory disease or in relapse).
Table I

Patient disease, pretreatment and donor characteristics

Pharmacokinetic Sampling

The median number of pharmacokinetic samples per patient was 12 (range 7–23) and the total number was 318. Blood samples for pharmacokinetic analysis (including peak and trough concentrations of inolimomab) were collected prior to infusion of the study medication and then 30 minutes, 2 hours, 8 hours and 16 hours after cessation of the infusion on day 1; prior to the infusion and 30 minutes after cessation of the infusion on days 2, 3 and 8; and prior to the infusion and 30 minutes after cessation of the infusion from day 9 to day 28 for the first three infusions. After collection, the blood samples were centrifuged and the serum samples were stored at −20°C until analysis.


Quantification was carried out in the OPi Research Department by a validated inolimomab ELISA according to Good Laboratory Practice. To trap inolimomab, serum samples were put onto a coated plate with goat polyclonal anti-mouse immunoglobulin antibodies. Next, sheep polyclonal anti-mouse IgG1 antibodies used as tracer antibodies were added to the mixture. After incubation with 3,3′,5,5′-tetramethylbenzidine substrate, the reaction was stopped by the addition of sulphuric acid and the absorption was read photometrically to quantify the samples. The range of the immunoassay was 0.15–10 µg/mL, with a sensitivity of >100 ng/mL, a sample intra-assay variation of 8% and a sample inter-assay variation of 11%.

Pharmacodynamic Assessments

Acute GvHD grades and performance status were evaluated daily until day 9 and then at each administration from day 10 to day 28, as well as at follow-up (day 60 and day 100) using the Glucksberg, International Bone Marrow Transplantation Registry (IBMTR) and Karnofsky classifications[21] (defined in the study as composite scores). The Glucksberg criteria determine acute GvHD severity (from grades 0, corresponding to no GvHD, to 4, maximum severity) by a combination of different organ scores (skin, intestinal tract and liver) and a decrease in clinical performance. These organs are graded independently from 0 to 4, which correspond respectively to the extent of a skin rash, a diarrhoea volume, and total bilirubin concentrations. The detailed definition of the IBMTR score is given in table II. Briefly, it involves the same organs but assigns the score based on maximum involvement in an individual organ system. Therefore, it tends to assign a higher overall grade for acute GvHD severity than the Glucksberg score.[20] The Karnofsky scale defines the overall performance status of a patient from 100% for a healthy person with no complaint and no sign of disease to 0% for a moribund patient. Independent organs as well as composite scores were considered for pharmacodynamics analysis.
Table II

Criteria for International Bone Marrow Transplant Registry severity index for acute graft-versus-host disease

Pharmacokinetic Analysis

Pharmacokinetic and pharmacodynamic analyses were carried out with nonlinear mixed-effect modelling using NONMEM software Version V.[22] Different pharmacokinetic models were tested, including one-, two- and three-compartment models, coupled with linear or nonlinear processes, such as saturable elimination. Interindividual pharmacokinetic parameter variability was assumed to follow log-normal distribution with non-zero correlations. Residual unexplained variability was modelled as multiplicative. The first-order conditional estimation (FOCE) INTERACTION method was used to fit all pharmacokinetic models. Models were evaluated through goodness-of-fit plots[2325] and the parameter precision was estimated by an asymptotic covariance matrix. Nested models were compared according to likelihood ratio tests (decrease of the NONMEM objective function between the reduced and full model by 3.84, corresponding to a nominal p-value of 0.05 for one additional parameter).

Pharmacokinetic-Pharmacodynamic Analysis

For each pharmacodynamic timepoint assessment, inolimomab exposures were estimated from individual pharmacokinetic profiles predicted from the previously described model. They were defined as either the maximal serum concentration (Cmax), the area under the serum concentration-time curve (AUC) over the last dosing interval, the cumulated AUC, or the AUC intensity (cumulated AUC/duration) from the first to the last dosing before pharmacodynamic assessment. The cumulated AUC corresponded to the cumulated sum of all AUCs computed for each dosing interval before pharmacodynamic assessment; the duration used in AUC intensity calculation corresponded to the treatment duration before pharmacodynamic assessment. Graphical exploration of the exposure-effect relationships was performed with all pharmacodynamic assessments by plotting the estimated cumulative probabilities of ordered scores (composite and organ scores) versus distribution quantiles (25%, 50%, 75% and 100%) of the above-defined drug exposures. Apparent relationships were then quantified by proportional odds models.[26] The cumulative probabilities of the observed score were linked to pharmacokinetic exposure through logit transformation. The nature of this link was tested with different pharmacodynamic models such as the maximum effect (Emax), log-linear or linear models. Interindividual variability of some key parameters was assumed to follow either a normal or a log-normal distribution. Parameter estimation was performed using the Laplacian estimation method in NONMEM. The adequacy of the different developed models and selection of the basic model were evaluated by comparing predicted and observed probabilities.

Model Qualification

The model qualification for the pharmacokinetic and pharmacokinetic-pharmacodynamic models was conducted in two steps. With regard to the pharmacokinetic model, after inspection of the basic graphics (predictions of a typical patient versus observations, individual predictions versus observations, weighted residuals versus observations, individual predictions and observations versus time), a visual predictive check was conducted.[27] It consisted of simulating (with NONMEM) 200 new datasets with identical patients, dosage regimens and sampling times, and then graphically comparing the simulated concentrations with the observed ones. The qualification of the pharmacokinetic-pharmacodynamic model was also based on a visual predictive check. Here, the purpose was to test the model’s ability to predict the probability of observing a grade. Therefore, we compared graphically the model predicted grade-exposure relationship with the observed one. Then, a predictive check was specifically conducted to qualify the pharmacokinetic-pharmacodynamic model for its clinical purpose.[28,29] It consisted of simulating (using NONMEM) 1000 new datasets by sampling patients with their score record time and simulating the grade for each measurement. Then we compared the statistic deduced from these simulations with the observed ones. This statistic, which is a quantity that depends only on data, was chosen in order to highlight the treatment effect.


Pharmacokinetic Analysis

Among all pharmacokinetic models tested, the best results were obtained with a two-compartment model. The goodness-of-fit plot for this model showed that the mean population and individual predicted concentrations were in good agreement with the observed ones (close to identity) except for a few concentrations over 20 µg/mL which were under-predicted. Other tested models included a third compartment or a Michaelis-Menten elimination. The three-compartment model was not identifiable and a nonlinear elimination did not show a significant improvement fit. Therefore the two-compartment model was eventually retained. The correlation between all pharmacokinetic parameters was then introduced. Parameter estimates are presented in table III. The elimination half-life of the compound for a typical patient was 44.5 hours.
Table III

Population pharmacokinetic parameters obtained from the final model

The most important goodness-of-fit plots of the final pharmacokinetic model are presented in figure 1. Although the figure shows a slight under-prediction for high concentrations, a visual predictive check (figure 2) revealed that this had no impact on predicted concentrations: the proportion of concentration points outside the 80% CI band was in agreement with the expected one. The pharmacokinetic model was consequently considered to be qualified.
Fig. 1

Goodness-of-fit plots of the final pharmacokinetic model: (a) population predicted (PRED) vs observed (OBS) concentrations; (b) weighted residuals (WRES) vs PRED concentrations.
Fig. 2

Visual predictive check of the final pharmacokinetic model from 200 × 21 simulated patients. In order to normalise the scale, all concentrations (observed and predicted) were divided by the actual received doses.

Pharmacokinetic-Pharmacodynamic Analysis

Pharmacodynamic assessments included composite scores, with a median number of score of 13 (range 3–23) per subject in the IBMTR classification, a median number of score of 12 (range 3–23) per subject in the Glucksberg classification, and a median number of score of 18 (range 7–25) per subject in the Karnofsky classification. Organ scores (skin, intestinal tract or liver) were recorded, with a median number of score of 18 (range 7–25) per subject. The pharmacokinetic-pharmacodynamic exploratory graphical analysis revealed that the composite scores (IBMTR, Glucksberg and Karnofsky) were apparently not related to drug exposure (see figure 3 showing the cumulated AUC; other graphics concerning the Cmax, AUC and AUC intensity are not shown). On the contrary, organ scores (skin, intestinal tract and liver) revealed patient improvements (i.e. the lowest-grade probability increase) when plotted against drug exposure expressed as the cumulated AUC or AUC intensity (see cumulated AUC in figure 4). When the AUC or Cmax measured over the last dosing interval was used, this relationship was less clear. Consequently, it was decided to develop the pharmacokinetic-pharmacodynamic model for the three-organ score cumulated probabilities, conditional on the cumulated AUC and AUC intensity. Based on goodness-of-fit and parameter uncertainty, only the relationship between the cumulated AUC and observed scores was finally considered. With regard to the skin, after merging grades 3 and 4 (since only four grades were observed), an Emax model with interindividual variability on the logit and cumulated AUC producing 50% of the maximum effect (EA50) gave the best results according to equation 1:
$${\rm{logit}}[{\rm{P}}({\rm{Y}} \le {\rm{j}})] = {{\rm{\alpha }}_{\rm{j}}} + \left( {{{{{\rm{E}}_{{\rm{max}}}} \bullet {\rm{cumulated\;AU}}{{\rm{C}}^{\rm{\gamma }}}} \over {{\rm{E}}{{\rm{A}}_{50{\rm{i}}}}^{\rm{\gamma }} + {\rm{cumulated\;AU}}{{\rm{C}}^{\rm{\gamma }}}}}} \right)$$
(Eq. 1)
where P(Y ≤ j) represents the probability of obtaining a score Y for skin inferior or equal to the grade j, αj represents the intercept of the logit for the grade j, and γ represents the coefficient of sigmoidicity. The individual EA50i distribution across the population revealed two groups of patients (i = 1 or 2), one with a high EA50 and the other with a low EA50. A mixture probability was then added to the EA50i random effect in order to estimate the proportion of those two subpopulations and their respective EA50.[30] For the intestinal tract and liver scores, the best results were obtained with a linear model (equation 2) and interindividual variability on the logit. All parameter estimates are presented in table IV.
$${\rm{logit}}[{\rm{P}}({\rm{Y}} \le {\rm{j}})] = {{\rm{\alpha }}_{\rm{j}}} + {\rm{slope}} \bullet {\rm{cumulated\;AUC}}$$
(Eq. 2)
where the Emax model of equation 1 is replaced by a linear model defined by a slope.
Fig. 3

Observed cumulated probabilities of composite scores: (a) International Bone Marrow Transplant Registry (IBMTR); (b) Glucksberg; and (c) Karnofsky grades as a function of the predicted cumulated area under the serum concentration-time curve (AUC). The data are split into four intervals according to the quantiles (Qs) 25%, 50% and 75%. This allows for a sufficient number of data (at least n = 64) to represent the evolution of probability of observing each grade versus different values of the predicted cumulated AUC. The red lines indicate the link between observed cumulated probabilities of the same grade. The horizontal black bars indicate the observed cumulated probabilities.
Fig. 4

Observed cumulated probabilities of organ scores: (a) skin, (b) liver and (c) intestinal tract as a function of the predicted cumulated area under the serum concentration-time curve (AUC). The data are split into four intervals according to the quantiles (Qs) 25%, 50% and 75%. This allows for a sufficient number of data (at least n = 86) to represent the evolution of probability of observing each grade versus different values of the predicted cumulated AUC. The red lines indicate the link between observed cumulated probabilities of the same grade. The horizontal black bars indicate the observed cumulated probabilities.
Table IV

Population pharmacodynamic parameters obtained from the final models

A visual predictive check of the pharmacokinetic-pharmacodynamic models for the three organs (figure 5) revealed an overall good agreement between the 80% CI band and the observed grade probabilities. The next step consisted of global qualification of the pharmacokinetic-pharmacodynamic analysis. For this purpose, we evaluated the model by a predictive check. We evaluated whether the combination of the pharmacokinetic model and the three-organ score models (for the skin, intestinal tract and liver) correctly predicted the global therapy effect expressed by the IBMTR score. In this way, our model could be used to verify a treatment effect. The chosen test statistic was the number of the observed grade at a given time, either calculated from the observed data or predicted depending on the model. Simulation was performed for the overall treatment duration, but the statistic is presented only at the start of treatment (figure 6 a–d; n = 21) and at the end of the treatment period on day 28 (figure 6 e–h; n = 11). This graph revealed good agreement between the 90% CI band and the observed IBMTR score. The combination of the three models to obtain the IBMTR score therefore qualified the model for use in defining the effect of treatment without modelling the IBMTR score itself. The comparison of figures 6ad and figures 6eh revealed that IBMTR scores decreased with treatment time: the IBMTR score at day 1 had greater probability of being observed at a grade B or C and eventually D, whereas at day 28 the highest probability of being observed and predicted was at grade A and then B or C. This observation confirmed the global therapy effect as had already been demonstrated by the pharmacokinetic-pharmacodynamic profiles of each organ.
Fig. 5

Visual predictive check of final pharmacokinetic-pharmacodynamic models from 200 × 21 simulated patients. The observations plotted in figure 4 are compared with the model predictions. The thin and thick black lines indicate the prediction and 80% CI, respectively, of the predicted cumulated probabilities.
Fig. 6

Predictive check of final pharmacokinetic-pharmacodynamic models: histogram of International Bone Marrow Transplant Registry (IBMTR) scores from 1000 × 21 simulated patients at the start of treatment (a–d) and from 1000 × 11 simulated patients still in the study at day 28 (e–h). The dotted lines indicate the 90% CI of the simulated IBMTR scores. The black lines indicate the observed IBMTR scores of 21 patients at the start of treatment (a–d) and 11 patients still in the study at day 28 (e–h).


Following promising results observed in 32 corticosteroid-resistant patients who presented with acute GvHD[31] and, more recently, in a retrospective analysis of 85 corticosteroid-resistant patients with grade II–IV acute GvHD,[32] inolimomab was proposed as an upfront therapy. As the investigators in these clinical trials had already observed heterogeneity in the organ response with, for instance, a better and more prolonged response occurring in cutaneous acute GvHD,[32] it therefore clearly appeared that the pharmacodynamics of inolimomab needed to be investigated. Concerning inolimomab and concomitant treatments, patients have received very different exposures in terms of duration as well as dosage. Some investigators carried out a multivariate analysis, suggesting that a higher total dose of inolimomab might be predictive of a better response.[32] In this context, it also appeared important to identify the pharmacokinetics of this drug[32] in order for analyses to take into account the real inolimomab treatment exposure, and to understand more precisely the pharmacodynamics and their relationship to the pharmacokinetics. This type of approach in acute GvHD is not very widespread and has not previously involved any type of monoclonal antibody. Only a few investigators have previously tried to link the pharmacokinetics and pharmacodynamics for prophylactic treatment. For instance, some investigators noticed a correlation between ciclosporin trough blood concentrations in the early post-transplantation period and the probability of observing an acute GvHD.[33] By splitting the population into four groups (no acute GvHD, mild acute GvHD, moderate acute GvHD and severe acute GvHD) they observed a decrease in mean ciclosporin trough blood concentrations for all time periods. Other investigators defined binary criteria as the probability of observing at least grade II acute GvHD and linked this criteria to the busulfan AUC at steady state by a logistic function.[34] Finally, other investigators used a threshold value of the AUC of unbound mycophenolate mofetil in week 1 after transplant to define two groups of patients and observed different cumulative proportions of observing a grade II–IV acute GvHD as a function of time.[35]

Our study modelled the pharmacokinetics of inolimomab and succeeded in modelling its exposure-effect relationship. For pharmacokinetic modelling, the population approach, taking into account design heterogeneity and the individual treatment history, allowed identification of a two-compartment model. Despite some under-predictions at the higher concentrations, the observations were, on the whole, well predicted and the model was qualified for calculating individual treatment exposure, which could not be directly computed from the observed data.

To highlight pharmacokinetic-pharmacodynamic relationships, we initially considered acute GvHD according to the Glucksberg and IBMTR reference scores and the Karnofsky classification. Since those combined scores are not arranged in order (i.e. grade 0 < grade I < grade II…), the pharmacokinetic-pharmacodynamic relationship is not easy to reveal. For instance, grade B in the IBMTR classification can correspond to a grade 2 of skin involvement and two grades 0 of liver and intestinal involvement or a grade 0 of skin involvement, a grade 1 of liver involvement and a grade 2 of intestinal involvement. We found out that the calculation of the IBMTR and Glucksberg grades, as well as of the global performance provided by the Karnofsky score, are not adapted to highlighting an exposure-effect relationship of inolimomab. In fact, some investigators have already identified better efficacy of inolimomab in a cutaneous form of acute GvHD.[32] In the context of targeted efficacy in one organ, one can easily understand that a composite measure of the effect is not relevant to the pharmacokinetic-pharmacodynamic analysis.

Our pharmacokinetic-pharmacodynamic analysis logically focused on each of the three organs and on treatment exposure. It clearly appeared that the relationship was significant regardless of which treatment exposure measurement was chosen. The treatment effect reached a maximum for the skin and was modelled with an Emax model. A mixture model revealed two populations of patients: sensitive and non-sensitive.[3638] For the liver as well as for the intestinal tract, the treatment effect appeared only above a threshold of cumulated AUC. It was illustrated by a large decrease in patients with severe symptoms (grade 4) and a significant increase in patients without symptoms (grade 0). This means that a larger exposure than that needed for the skin is required to reach the same efficacy in those organs. It also explains why clinically, the skin is the first organ to be cured whereas it seems to be more difficult to treat the liver and the intestinal tract. This is illustrated by the pharmacokinetic-pharmacodynamic graphs at day 8: from 15% to 20% of patients still present with the highest grade for these two organs. Some investigators have attributed this phenomenon to the difference in the bioavailability of inolimomab or the pathophysiology of acute GvHD depending on organ involvement.[32]

Although the modelling of the IBMTR score could not be performed, this approach allowed us to easily simulate this score by the combination of the three organs on which it is based. It was also possible to use it to define the overall response at the end of treatment and therefore to verify the treatment effect. However, the benefit of this type of approach is that it takes into account the time dependency of the data. In this way, our models can predict the effect for a given patient and a treatment schedule, as a function of time throughout the study (see figure 7) and also give a more precise idea of the disease evolution under treatment. It is also possible to know if one organ presents a rapid or slow remission. However, these predictions must be used with caution for the time being. As in our models, the treatment effect is related to the cumulated AUC, and we modelled the pharmacokinetic-pharmacodynamic relationship by positive monotonous functions, predicted effects automatically increase with time. Therefore, no acute GvHD relapse could be predicted by the model. Moreover, based on the cumulated AUC, the same response will be obtained for concentration y during x hours or concentration x during y hours, regardless of the values of x and y.[39] In these conditions, dosage or regimen optimisation based on modelling is not possible. In fact, we used the cumulated AUC because we assumed that there would be a delay between concentrations and effect. In the present study, the nature of the pharmacodynamic data did not allow us to model this delay in another way (with an effect compartment, for instance). However, the perspective of this work, along with the future data, will be to link organ scores to serum concentrations using a more physiological model that takes into account the drug and effect accumulation. On the basis of this, dosage and regimen optimisation for future trials, as well as individual therapeutic monitoring of patients, could be performed.
Fig. 7

Observations and individual predictions versus time of final pharmacokinetic and pharmacokinetic-pharmacodynamic models for three patients (IDs 6, 10 and 18) enrolled in the clinical trial: (a) shows the observed concentrations and pharmacokinetic predictions as well as the predicted corresponding cumulated area under the serum concentration-time curve (AUC); (b) shows the observed pharmacodynamic grades; (c) shows the predicted probability of a grade observation. For patient ID 6, pharmacodynamic results for the skin, for patient ID 10, pharmacodynamic results for the liver, and for patient ID 18, pharmacodynamic results for the intestinal tract.


With this analysis, we highlighted and modelled a pharmacokinetic-pharmacodynamic relationship between the cumulated AUC of inolimomab and the skin, liver and intestinal tract scores. The modelling of the data allowed us to describe observations as well as to predict an overall response at the end of treatment for this population using IBMTR scores. This approach, which was validated for its objective, allowed us to better understand the treatment effect over time and represents the first step towards optimising the dosage for future patients. However, it does still present some limitations due, in particular, to the limited number of patients. Further trials are needed to improve clinical use of these models.


The use of trade names is for product identification purposes only and does not imply endorsement.



The authors wish to thank OPi SA, Limonest, France, for providing the pharmacokinetic and pharmacodynamic samples, and for reviewing and approving the manuscript. The funding of this study was provided by OPi SA and the Faculté de Médecine Lyon Sud at the Université de Lyon, Lyon, France. I. Darlavoix and C. Vermot-Desroches are employees of OPi SA. C. Dartois is supported by the Institut de Recherches Internationales Servier. P. Girard is supported by INSERM, France. The other authors have no conflicts of interest that are directly relevant to the content of this study.

Copyright information

© Adis Data Information BV 2007