Optimizing radionuclide therapy with 177Lu-HA-DOTATATE in patients with neuroendocrine tumors using semi-physiological population PK models
H. Siebinga (1, 2), B.J. de Wit-van der Veen (2), T.P.C. Dorlo (3), A.D.R. Huitema (1, 4, 5), J.J.M.A. Hendrikx (1, 2)
(1) Department of Pharmacy & Pharmacology, The Netherlands Cancer Institute, Amsterdam, The Netherlands. (2) Department of Nuclear Medicine, The Netherlands Cancer Institute, Amsterdam, The Netherlands. (3) Department of Pharmacy, Uppsala University, Uppsala, Sweden (4) Department of Clinical Pharmacy, University Medical Center Utrecht, Utrecht University, Utrecht, The Netherlands. (5) Department of Pharmacology, Princess Máxima Center for Pediatric Oncology, Utrecht, The Netherlands
Objectives
Somatostatin receptor (SSTR) expression on neuroendocrine tumor (NET) cells enables targeting of NETs with radiopharmaceuticals for both diagnosis (with Gallium-68 (68Ga) HA-DOTATATE) and therapy (with Lutetium-177 (177Lu) HA-DOTATATE). Using this theranostic approach (i.e. using the same peptide for imaging and treatment), diagnostic imaging may predict individual therapeutic organ and tumor uptake and could guide precision medicine, since efficacy and toxicity are proportional to uptake in tumors and healthy tissues, respectively. However, differences in biodistribution are observed between both radiopharmaceuticals and additional knowledge on exact pharmacokinetic (PK) differences are urgently needed. Hence, semi-physiological population PK models were developed to assess PK differences between 68Ga-HA-DOTATATE and 177Lu-HA-DOTATATE. Subsequently, 177Lu-HA-DOTATATE kidney (organ at risk) and tumor uptake was individually predicted based on single time point imaging with 68Ga-HA-DOTATATE.
Methods
A previously developed PBPK model for 68Ga-HA-DOTATATE in NET patients (1) was used to inform a simplified multicompartmental model, which enabled straightforward comparison of both radiopharmaceuticals and Bayesian estimation. A semi-physiological NLMEM was developed using NONMEM (ver. 7.5) and consisted of 6 compartments, representing blood, spleen, kidney, tumor, other SSTR expressing organs and a lumped rest compartment. Due to limited 68Ga-HA-DOTATATE observations, a bottom-up model development approach was used and imaging data (n=9) was only used for model evaluation (2). Structural model parameters consisted of renal excretion (0.25 h-1), kin and kout values to compartments, compartment volumes (4, 0.21, 0.3, 4 and 50 L for compartments 1, 2, 3, 5 and 6 respectively (3), and measured tumor volume for compartment 4), fraction unbound (0.69) and maximum binding capacity (BMAX) values for compartments 2-5. IIV was added to uptake rate parameters (fixed based on assumed population variability). Factors assumed to affect PK parameters were added a priori, which were tumor volume on spleen uptake (k12) (tumor sink effect, based on PBPK simulations) and on uptake into tumors (k14) (increased uptake with increasing volume, using a power equation).
A similar model structure was used for 177Lu-HA-DOTATATE. Fraction unbound in plasma and renal excretion (k10) were set to 0.57 and 0.575 h-1, respectively (4, 5). k12, k13, k14 and k40 were sequentially estimated using patient imaging data (n=9, 4 time points), because they were expected to differ for 177Lu-HA-DOTATATE. IIV was added to receptor expressions rather than kin values. Final model performance for 177Lu-HA-DOTATATE was evaluated based on visual inspection of predictions compared to observations.
Lastly, Bayesian estimates were determined based on individual 68Ga-HA-DOTATATE data. Using these Bayesian estimates, individual 177Lu-HA-DOTATATE tumor and kidney uptake was predicted using individual parameters for 177Lu-HA-DOTATATE (based on established differences between PK parameters of both radiopharmaceuticals). Predictions were evaluated based on a relative prediction error (RPE) for area under the curve (AUC).
Results
Model predictions adequately described observed data in kidney and tumors for both peptides. Population uptake rate parameters for spleen, kidney and tumors differed by a 0.29-fold (15% RSE), 0.49-fold (15% RSE) and 1.43-fold (14% RSE), respectively, for 177Lu-HA-DOTATATE compared to 68Ga-HA-DOTATATE. 177Lu-HA-DOTATATE tumor AUCs could be predicted reasonably well based on individual 68Ga-HA-DOTATATE single time point scan data (RPEs -40-28%). Also, these individual predictions based on Bayesian estimates outperformed 177Lu-HA-DOTATATE population predictions for 7 out of 9 patients, confirming that diagnostic imaging could be used to predict therapeutic 177Lu-HA-DOTATATE tumor uptake. However, kidney predictions were less accurate (RPEs -53–41%), possibly due to complexity of interpretation of kidney scan observations.
Conclusions
Our semi-physiological NLMEM indicated lower uptake rate parameters for kidney and spleen for 177Lu-HA-DOTATATE compared to 68Ga-HA-DOTATATE, while the tumor uptake rate was 1.43-fold higher. Individual 177Lu-HA-DOTATATE kidney and tumor uptake was predicted and showed more accurate predictions for tumor compared to kidney. All in all, a framework for image-based precision medicine of radionuclide therapy was developed.
References:
[1] Siebinga H, de Wit-van der Veen BJ, Beijnen JH, Stokkel MPM, Dorlo TPC, Huitema ADR, et al. A physiologically based pharmacokinetic (PBPK) model to describe organ distribution of (68)Ga-DOTATATE in patients without neuroendocrine tumors. EJNMMI Res. 2021;11(1):73.
[2] Huizing DMV, Peters SMB, Versleijen MWJ, Martens E, Verheij M, Sinaasappel M, et al. A head-to-head comparison between two commercial software packages for hybrid dosimetry after peptide receptor radionuclide therapy. EJNMMI Phys. 2020;7(1):36.
[3] Basic anatomical and physiological data for use in radiological protection: reference values. A report of age- and gender-related differences in the anatomical and physiological characteristics of reference individuals. ICRP Publication 89. Ann ICRP. 2002;32(3-4):5-265.
[4] Puszkiel A, Bauriaud-Mallet M, Bourgeois R, Dierickx L, Courbon F, Chatelut E. Evaluation of the Interaction of Amino Acid Infusion on (177)Lu-Dotatate Pharmacokinetics in Patients with Gastroenteropancreatic Neuroendocrine Tumors. Clin Pharmacokinet. 2019;58(2):213-22.
[5] Lubberink M, Wilking H, Ost A, Ilan E, Sandstrom M, Andersson C, et al. In Vivo Instability of (177)Lu-DOTATATE During Peptide Receptor Radionuclide Therapy. J Nucl Med. 2020;61(9):1337-40.