Olivier Le Tilly (1,2), Nicolas Azzopardi (1), Claire Bonneau (3), Céline Desvignes (1,2), David Ternant (1,2), Isabelle Turbiez (3), Maya Gutierrez (3), Gilles Paintaud (1,2)
(1) EA7501 GICC, France, (2) Tours Hospital, France, (3) Institut Curie, France
Introduction: Leptomeningeal carcinomatosis is a complication with a poor prognosis found in at least 5% of breast cancer patients [1] and corresponds to a metastatic invasion of the subarachnoid space. While trastuzumab is one of the first-line therapies for the management of breast cancer overexpressing HER2, intravenous administration of monoclonal antibodies leads to low concentrations in the central nervous system. Case reports of intrathecal (IT) administrations of trastuzumab have shown promising results in patients with leptomeningeal carcinomatosis of HER2+ breast cancer but dosing regimens are empirical in absence of pharmacokinetic study.
Objectives: Our primary objective was to study the pharmacokinetics of trastuzumab after its IT administration in patients with leptomeningeal carcinomatosisof HER2+ breast cancer by population compartmental approach. Secondary objective was to estimate individual trastuzumab exposure and its relationship with disease markers.
Methods: With a population pharmacokinetic approach (Monolix software), we described the fate of trastuzumab after IT administration in 21 women included in a phase I-II clinical trial [2]. Trastuzumab was administered by IT route every week for 8 weeks and both cerebrospinal fluid (CSF) and serum were sampled to measure trough concentrations. Disappearance of tumor cells from CSF was used as biomarker of efficacy. Areas under curve (AUC) in serum and CSF, over both the first and last week of treatment, were used as markers of exposure, as well as time spent with serum or CSF concentrations above 10 mg/L [3] and 60 mg/L [4].
Results: Four dose regimen were tested: 30 mg (n = 4), 60 mg (n = 4), 100 mg (n = 4) and 150 mg (n = 9). Thirteen patients happened to also receive intravenous trastuzumab (every 3 weeks) prior to the study and its administrations were continued. Some patients showed noticeable CSF concentration fluctuations, predicted using a target-mediated drug disposition (TMDD) (AIC = 1769 and 1859, with and without TMDD, respectively). This target was latent and produced with four transit compartments and a delayed feedback.
We built the model using the following ordinary differential equations:
dSERUM/dt = – k12+ k21.CSF – k10.SERUM
dCSF/dt = k12– k21.CSF – kdeg.CSF.TARGET4
dTARGET0/dt = kin– ktr.TARGET0; kin(t) = kin.TARGET4(t = 0)/TARGET4(t)
dTARGETn/dt = ktr.TARGETn-1– kout.TARGETn; kout= ktr
dTARGET4/dt = ktr.TARGET3– kout.TARGET4– kdeg.CSF.TARGET4; TARGET4(t = 0) = kin/kout
Apparent volumes of distribution were close to physiological volumes (V1= 3.25 L, V2= 0.644 L, for serum and CSF, respectively). Estimated (zero-order) transfer from serum to CSF was very slow (k12= 0.264 mg.d-1) while estimated half-life of transfer from CSF to serum was rapid (ln(2)/k21= 2,2 days). From the individual parameters of patients, a single IT administration of 150 mg of trastuzumab corresponded to average mean residence times of 6.0 days and 23.6 days in CSF and serum, respectively.
Mean patients’ trastuzumab AUC over the first week of treatment were 425 mg.d/L and 511 mg.d/L in CSF and serum, respectively. Mean patients’ AUC over the eighth week of treatment were 381 mg.d/L and 817 mg.d/L in CSF and serum, respectively. Over 7 days, mean time spent with trastuzumab CSF concentrations above 10 mg/L and 60 mg/L were 5.2 days and 2.8 days, respectively. Over 56 days, these values were 36.3 days and 19.9 days, respectively. Cytological response to treatment in CSF was not related to trastuzumab exposure.
Conclusions: This study confirms that transfer of trastuzumab from serum to CSF is very limited and that this monoclonal antibody, when administered by IT route, is rapidly transferred to the serum.
References:
[1] Gauthier H et al. Ann Oncol Off J Eur Soc Med Oncol (2010) 21, 2183–2187
[2] Bonneau C et al. Eur J Cancer Oxf Engl 1990 (2018) 95, 75–84.
[3] Grossman SA and Krabak MJ. Cancer Treat Rev (1999) 25, 103–119.
[4] Baselga J. Ann Oncol (2001) 12, S49–55.
Reference: PAGE () Abstr 9336 [www.page-meeting.org/?abstract=9336]
Poster: Drug/Disease Modelling - Oncology