The Effect of an Amino Acid-lowering Diet on the Rate of Melphalan Entry into Brain and Xenotransplanted Glioma1

Melphalan (i.-phenylalanine mustard, L-PAM, alkeran; molecular weight, 305,000) is transported across tumor cell membranes and the blood-brain barrier by the large neutral amino acid (LNÕA) transport system. Normally, plasma LNAA levels are high enough and the affin ity low enough that this system does not transport much melphalan into the brain. However, plasma amino acids can be reduced by fasting and protein-free diet. We used this method to reduce competition and to increase melphalan transport into brain tumors. In nude mice fasted for 12 h and then fed a protein-free diet for 2 and 6 h, mean plasma LNAA levels were 46% and 42% of control values. Nude mice with xenotransplanted D-54MG human gliomas were used to study tissue distribution and uptake kinetics of [3H]melphalan in a control group and a diet group (after a 12-h fast and 2 h of a 0% protein diet). The A, (blood-to-tissue transfer constant) of melphalan, determined by graphical analysis and by nonlinear fitting to a 2-compartment model, was higher in the diet group in all tumor regions except the necrotic center of subcutaneous tumors; the increase was significant in the tumor periphery of brain and s.c. tumors. The ratio of A ,s (diet to control) varied from 1.2 to 1.3 in brain tumors, 1.9 to 2.1 in subcutaneous tumors, and 1.8 to 3.1 in tumor-free brain. The apparent |'H)melphalan distribution space was significantly higher in the tumor periphery of both brain and subcuta neous tumors of the 15and 30-min diet group. We also measured blood-brain barrier transport of |a-14C|aminoisobutyric acid and blood flow (with [IMI]iodoantipyrine): the A, of [a-IJC]aminoisobutyric acid was 28.1 ±6.6 (SE) in brain tumors and 24.3 ±8.9 «tl/g/min in subcu taneous tumors. Blood flow was 58.2 —¿ » 3.9 in brain tumors and 5.2 ±0.4 ml/100 g/min in subcutaneous tumors. Fasting, when combined with a protein-free diet, reduces plasma amino acid levels and thereby reduces competition between melphalan and LNAAs. This may increase the amount of melphalan that can enter a brain tumor without increasing the administered drug dose and suggests a therapeutic manipulation that can be used to increase the delivery of melphalan. INTRODUCTION Melphalan6 is a chemotherapeutic drug synthesized in 1954 (1) during a period that resulted in numerous bischloroethyReceived 3/19/91; accepted 8/4/92. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accord ance with 18 U.S.C. Section 1734 solely to indicate this fact. 'Supported by NIH Grants NS12745. RR03321. DK26912. NS20023, CA44640. CAI 1898. NS00958; by American Cancer Society Grant CH-403; and by the Richard M. Lilienfeld and Mark Moritz Memorial Funds. I. F. was the recipient of an Arlene and Marshall Bennett Fellowship in Neuro-oncology from the Division of Neurosurgery, Evanston Hospital. 2 To whom requests for reprints should be addressed, at 2650 Ridge Avenue, Evanston, IL 60201. 3 Present address: Department of Neurosurgery. Universität des Saarlandes. Homburg/Saar, Germany. 4 Present address: Department of Neurology and Psychiatry. University Medical School of Debrecen, P. O. Box 31, H-4014 Debrecen, Hungary. 5 Present address: Department of Pathology. University Medical School of De brecen, P. O. Box 23, H-4014 Debrecen, Hungary. 'The abbreviations used are: melphalan, L-phenylalanine mustard. L-PAM, alkeran. molecular weight 305,000; AA, amino acid, BBB, blood-brain barrier; LNAA, large neural amino acid; AIB, n-aminoisobutyric acid; IAP. iodoantipyrine; DS, distribution space; ANOVA. one-way analysis of variance; PA, permeabilitysurface area product. lamine (nitrogen mustard) derivatives. Melphalan represented an intentional attempt to link an alkylating agent to a metabolic precursor; it is the t-isomer formed by a reaction between the AA phenylalanine and mechlorethamine (nitrogen mustard). Melphalan is active against several cancers, including medulloblastoma (2). Because of its phenylalanine component, mel phalan is transported across the cell membrane of several tumor cell lines by AA carrier transport systems (3-6). Its physicochemical characteristics make melphalan a poor candidate to pass across the normal BBB. It exists in a largely ionized form at physiological pH (7), and its rate of passage across the BBB by passive diffusion should be restricted. How ever, the BBB possesses a well-developed transport system for LNAAs, including phenylalanine. By means of an isolated brain perfusion technique, Greig et al. demonstrated that melphalan crosses the BBB by facilitated transport via the LNAA trans port system, although the affinity of the transport system for melphalan is low (7), as is the rate of brain entry (8). Cornford et al. (9) demonstrated that melphalan enters brain and brain tumor by the LNAA transporter and that, in sufficient concen tration, it can inhibit the brain uptake of phenylalanine. Nor mally, the rate of brain entry of melphalan is low because cir culating LNAAs are present in high concentration in the blood, have a higher affinity for the transport systems, and compete with melphalan for access to the transport system. Plasma levels of AAs can be influenced by dietary intake and by the body's metabolic adjustments to changing plasma AA levels. These changes apparently do not influence the kinetic behavior of the AA transport systems at the BBB but can result in variation in the species and amounts of AAs that are pre sented to the carrier systems and that enter the brain. Fernstrom and Faller (10) demonstrated that changes in brain levels of neutral AAs correlated with the amount of dietary AAs. Glaeser et al. (11) extended this relationship: plasma levels of many AAs (including the LNAAs) decreased while brain levels increased in rats fasted overnight and fed a 0% protein diet. We reasoned that lowered plasma AA levels should reduce compe tition between melphalan and circulating LNAAs, thereby pro viding a way to increase melphalan entry into brain and brain tumors. We tested this hypothesis in a model that uses xenotransplanted human glioma cell lines (12). In addition, we measured tumor blood flow and transcapillary transport of a passively distributed water-soluble compound to gain more in formation about other factors influencing the blood-to-tumor transport of melphalan. MATERIALS AND METHODS