Pergamon

Int. J. Radiation Or~cology Biol. Phys., Vol. 29, No. 4, pp. 771-776, 1994 Copyright 8 1994 Elsevier Science Ltd Printed in the USA. All rights reserved

0360-3016/94 $6.00 + .OO

0360-3016(94)EOOOl-Z

??Biology Original Contribution

PROTECTIVE EFFECTS OF GLUTATHIONE ON CISPLATIN NEUROTOXICITY IN RATS

GUIDO CAVALETTI, M.D.,*+, CLAUDIO MINOIA, M.D.,” MARCO SCHIEPPATI, M.D.§ AND

GIOVANNI TREDICI, M.D.+

*Institute of Human Anatomy, University of Milan, Milan; +Department of Neurology, University of Milan, S. Gerard0 Institute for Biomedical Sciences, Monza; *Clinica de1 Lavoro, Pavia; and PInstitute of Clinical Neurology,

University of Genoa, Genoa, Italy

Purpose: Different attempts have been made to minimize the neurotoxicity of cisplatin (DDP) and the use of “nemoprotective” drugs seems to be a promising strategy. In rats we compared the effects on the dorsal root ganglia neurons and peripheral nerves of the administration of DDP alone or in combination with glutathione (GSH), a putative “neuroprotective” drug. Methods and Materials: Twenty-four Wistar rats were treated with DDP alone (2 mg/kg/week) or with the same dose of DDP plus GSH (300 mg/week) for nine cycles and they were compared to 12 untreated age-matched rats. All the animals underwent either neurophysiological examination of the tail nerve or pathologic examination of the dorsal root ganglia. Analytical determination of the platinum concentration in dorsal root ganglia was also performed. Results: Morphologic and morphometric evaluations demonstrated a reduced incidence of pathologic changes in DDP plus GSH-treated rats with respect to DDP-treated ones. In agreement with the morphological findings, the platinum concentration in the dorsal root ganglia was lower and sensory nerve conduction velocity in the tail nerve less markedly decreased in the animals treated with DDP plus GSH with respect to those treated with DDP alone. Conclusion: We conclude that the administration of GSH is effective in reducing the neurotoxic effects of DDP, thus supporting the preliminary results obtained in clinical trials in humans.

Cisplatin, Glutathione, Nervous system, Toxicity, Morphology, Neurophysiology, Toxicology.

INTRODUCTION

The use of high total doses of cisplatin (DDP) (cis-di- amine-dichloroplatinum II) in humans is frequently pre- vented by the onset of a disabling peripheral neuropathy (15), clinically characterized by sensory impairment (mostly involving vibration sense), deep reflex disappear- ance, and ataxia ( 18, 19). This sensory neuropathy is sec- ondary to dorsal root ganglia (DRG) neuronal involve- ment (4, 8, 23). Different attempts have been made to minimize the neurotoxic effects of DDP by modifying the schedules of its administration or by adding to it different “neuroprotective” drugs (2, 14, 22). Among the latter, glutathione (GSH) has been shown to be effective in re- ducing DDP neurotoxicity in different trials in humans which were evaluated on clinical grounds alone (9, 13). Moreover, in an experimental model in rats, GSH has been reported to be effective in ameliorating some neu- rophysiological parameters in sensory nerve conduction (11) after DDP treatment. However, the involvement of

DRG neurons can be demonstrated only with a morpho- logical approach, since electrophysiological parameters can only give information regarding conduction in the peripheral and, less easily, in the central branches of DRG neurons. Therefore, we have developed a model of DDP- induced neurotoxicity, in rats, in which DRG neurons undergo clear-cut pathologic changes (5, 6). Furthermore, it is possible to determine the tissue platinum concentra- tion (7) comparing the levels in DRG with respect to the nervous and nonnervous organs. Using this model, we performed the current study with the aim of objectively comparing the effects on DRG neurons when DDP was administered alone or in combination with GSH.

METHODS AND MATERIALS

Experimental model Thirty-six female Wistar rats, weighing 180 g at the

beginning of the experiment, were used. They were in-

This work was presented in part at the Fourth European Meeting of Neuropathology, Berlin, 14- 18 July 1992.

Reprint requests to: Prof. Giovanni Tredici, Istituto di Ana- tomia Umana, v. Mangiagalli 3 1, I-20133 Milano, Italy. Acknowledgements-The valuable help of D. Fabbrica, M. Ga- man and 0. De Negri during specimen preparation is gratefully

acknowledged. The dutathione used in this study was a gift from Boehringer Mannheim Italia. This study was supported by a CNR (Italy) grant-Targeted Project “Prevention and control disease factors PF4.”

Accepted for publication 22 December 1993.

772 1. J. Radiation Oncology 0 Biology 0 Physics Volume 29, Number 4, 1994

dividually housed in plastic cages and given commercial rat cubes and tap water ad lib. Twelve animals were used as controls and 24 were treated. According to a previously reported method (56) commercially available DDP’ was diluted in 2 ml sterile 0.9% NaCl solution to administer a dose of 2 mg/kg of the drug. The solution was intra- peritoneally (IP) injected into the treated rats once a week. The drug administration was immediately followed by 3 ml of sterile 0.9% NaCl solution SC to prevent renal dam- age by means of hyperhydration. Twelve rats were treated with the solution containing only DDP (group A), while 12 animals were IP injected with 300 mg of GSH* 15 min before each DDP injection (group B). The controls were treated in exactly the same way, except that the IP injected solution did not contain DDP. The animals received the treatment for a total of nine cycles and the scheduled total dose of DDP in all treated animals was 18 mg/kg. The drug was always administered between 12:00 am and 2:00 pm to minimize possible changes in its activity due to the circadian cycle.

the case of multinucleolated cells) was in the outer half of the radius of the nucleus (21). The statistical compar- ison between mean values was performed with the two- tailed Student’s t-test for unpaired data and the SE of the mean was calculated for each value.

The animals which had completed the study (12 con- trols, ten rats in group A, and 11 in group B) were divided into two groups which were evaluated separately as fol- lows.

Morphological examination. Six controls, four rats be- longing to group A and five belonging to group B were killed under chloral anesthesia by intracardiac perfusion with 3% cold glutaraldehyde in 0.12 M phosphate buffer solution. The lumbar spinal ganglia L4-L5-L6 were dis- sected carefully and removed. All the animals were sac- rificed between 12:00 am and 2:00 pm to prevent nu- cleolar circadian modifications. Specimens were then washed in 0.18 M phosphate buffer solution, postfixed in 1% buffered 0s04, dehydrated in ethanol series and finally embedded in epoxy resin. Light-microscope observations were performed on semi-thin (1 pm) sections stained with toluidine blue. Ultrathin sections were stained with uranyl acetate and lead citrate and observed with an electron microscope.3 Morphometric determinations were per- formed in all specimens on toluidine blue stained semi- thin sections with an automatic image analyzer.4 Somatic, nuclear and nucleolar cross-sectional areas were measured in spinal ganglia neurons. The percentages of multinu- cleolated neurons and of nuclear eccentricity were also calculated. For the analysis three DRG sections, inter- spaced at more than 50 pm, were randomly chosen from each of the three different ganglia for a total of nine sec- tions in each animal. All those cell profiles (ranging from 14-5 1 in each section) in which at least one nucleolus was visible in the section plane were included in the eval- uation (21). Nucleolar eccentricity was taken as being when the nucleolar center (or that of the largest one in

Nerve conduction study. The remaining 18 rats (six controls, six in group A, six in group B) underwent neu- rophysiological evaluation of the sensory conduction ve- locity (SNCV) in the tail. Each rat was inserted into a metal container, its tail protruding through a hole in the lid. The container and the tail were firmly kept in place by one investigator, while another fixed the electrodes to the rubbed skin of the tail. The container, both investi- gators, and the tail were connected to a common ground. Both the stimulating and the recording electrodes were made of a pair of conductive wires twisted around the tail, the interwire distance being 5 mm. The distance be- tween the stimulating and recoding electrodes was 95 mm, the stimulating electrodes being placed as proximally as possible. The signal from the recording electrodes was differentially amplified (X 100,000) and fed to an aver- aging computer, with a sampling frequency of 4 kHz. Each epoch lasted 50 ms, with a pretrigger time of 10 ms. Fifty consecutive epochs were averaged. The stimulus strength was the minimum possible one that yielded the maximal amplitude of the first peak. In most cases it corresponded to three to five times the sensory threshold, as identified both by inspection of the oscilloscope display and of the averaged signal. Stimulation frequency was 2 Hz. At this intensity and frequency of stimulation, rats did not show obvious signs of discomfort (movement or vocalization), except at the very beginning of the train. The latency of antidromic sensory discharge was measured off-line by a visual inspection of the averaged traces. In all animals, the latency was set at the first negative peak of the evoked potential. Mean values of SNCV were compared with the two-tailed Student’s t-test for unpaired data.

Analytical determination. After the SNCV determina- tion the animals were killed with supramaximal anes- thesia, carefully washed with saline via intracardiac per- fusion, and lumbar DRG L4-L5-L6 were freshly dissected. Analytical determination of tissue platinum concentra- tions was performed in each animal of the three groups on pooled specimens of lumbar DRG. Inductively Cou- pled Plasma Mass Spectrometry (ICP-MS)’ was used to determine platinum concentration. Sample pretreatment consisted of acid digestion in a microwave oven followed by appropriate dilution. Mean values of platinum con- centration were compared with the two-tailed Student’s t-test for unpaired data.

In all morphologic, morphometric, neurophysiological and analytical evaluations, the examiner was always un-

’ Platamine, Farmitalia Carlo Erba, Italy. 2 Tationil, Boehringer Mannheim, Italy. 3 Philips CM 10.

4 Leitz Tas Plus. 5 Perkin Elmer Sciex Elan 5000.

Glutathione and cisplatin neuroprotection 0 G. CAVALETTI et al. 773

Fi ig. 1. Light micrograps of dorsal root ganglia of a control rat nucleolar eccentricity are also both present in controls their in- (a ) and of a DDP-treated one (b). Cell size reduction is evid’ ent cidence is increased after DDP administration (toluidine blue, in the treated animal. Although multinucleolated neurons a Ind bar = 10 pm).

aware of whether the specimen under examination was from a control or a treated animal.

RESULTS

Three of 36 animals died during the experiment. How- ever, the administration of DDP at a dose similar to that used in the treatment of solid tumors in humans was gen- erally well tolerated in the surviving animals, although, at the end of the experiment, both groups of treated an- imals showed reduced weight gain compared with controls (mean weight and range: controls = 316 g, 295-345 g; group A, DDP alone = 231 g, 185-255 g; group B, DDP plus GSH = 282 g, 250-310 g).

Cisplatin administration induced clear-cut signs of pathologic involvement of DRG neurons. The most striking features were an increased number of multinu- cleolated neurons and nucleolar eccentricity (Fig. 1). At the electron microscope the nucleoli often showed seg- regation of the fibrillar and granular components (Fig. 2) and, occasionally, focal clearing. Although nucleolar ec- centricity and multiple nucleoli occurred also in normal rats, their increased incidence is a sign of cell reaction to damage. Morphometric determinations revealed a signif- icant reduction of the somatic, nuclear, and nucleolar area after chronic DDP administration. The nucleolar area was the most affected of all the morphometric parameters measured in the present study. In group A rats, treated with DDP alone, the reduction in nucleolar cross-sectional area was 48%. In group B rats, treated with DDP plus GSH, the decrease was again quite marked (-34%), but it was less evident than in the previous group. The differ- ence between the two groups was significant. Also the difference in the nuclear and somatic area reduction in DDP treated rats compared to DDP plus GSH treated ones was significant when evaluated statistically. The nu- clear area was reduced by 25% in group A as opposed to 11% in group B, and the somatic area by 2 1% as opposed to 12%.

The incidence of multinucleolated DRG neurons was nearly five times higher in the rats treated with DDP alone in comparison with the controls (p < O.Ol), while in the rats treated with DDP plus GSH it was only about twice as high (p < 0.01 vs. controls and vs. group A). A re- markable increase in nucleolar eccentricity was observed in DDP treated rats, but it was more evident in group A rats. In fact, this difference was significant only when DDP-treated rats were compared with controls, while the comparison between controls and GSH-treated rats was not significant. The complete morphometric results are reported in Table 1.

The results of the SNCV determinations were consistent with the pathologic data. In fact the mean SNCV was

I. J. Radiation Oncology 0 Biology 0 Physics Volume 29, Number 4, 1994

Table 1. Morphometric results obtained in control and treated rats

Group A Group B Controls (DDP alone) (DDP + GSH)

Multiple nucleoli (%) ;t?57,

15.3* 8.2*+ (0.6 1) (0.30)

Eccentric nucleoli (%) 32.3 53.6; 40.7 (2.15) (2.80) (2.76)

Nucleolar area (Mm2) 11.5 :;)?4)

8.8* (0.34)

Nuclear area (rm2) 150 113’ ‘p;,‘:+) (2.90) (2.91)

Somatic area (rm2) 953 753* ‘s24;3 (28.22) (2 1.49) (21.77)

Nuclear/somatic ratio 0.157 0.150 0.159 Nucleolar/nuclear ratio 0.076 0.064 0.066

DDP = Cisplatin; GSH = Glutathione. Note: Values are expressed as mean (SE), except for the ratios, which

are calculated on the mean values of nucleolar, nuclear and somatic area.

* = p i 0.0 1 when compared with controls. + = p < 0.0 I when compared with controls and group A.

lower in both groups of treated rats (group A = 36.7 m/ s, range 33.3-39.1; group B = 41.2 m/s, range 39.1-42.0) than in controls (44.4 m/s, range 42%50.0) but the dif- ference was more evident and statistically significant only in the animals treated with DDP alone (group A, p < 0.0 1 vs. controls).

Platinum concentration was different in the two groups of treated rats, with higher values being found in the an- imals treated with DDP alone. The mean platinum con- centration was 2 1 nanograms/g of tissue in group A (range 18-23) and 13 nanograms/g in group B (range lo- 14, p < 0.01 vs. group A), while the concentration in the con- trols was always lower than 1 nanogram/g (p < 0.00 1 vs. both groups of treated rats).

DISCUSSION

The use of DDP in the treatment of different solid tu- mors has become more and more widespread since its effectiveness has been demonstrated by clinical trials and its nephrotoxicity has been overcome by means of hy- perhydration. These achievements have prompted clini- cians to schedule higher doses of the drug to obtain the best tumor response ratio, but these schedules are often severely neurotoxic, and this may cause withdrawal from DDP treatment ( 15).

It is believed that DRG neurons are the main targets of DDP neurotoxicity in humans, since neurophysiologic studies performed by different groups have unequivocally demonstrated that both the peripheral and central sensory pathways are affected during DDP treatment, with a spar-

Fig. 2. Electron micrograph obtained from DDP-treated rats. The typical features found in DDP-intoxicated rats (a), the neu- ron is multinucleolated and one of the nucleoli (on the left)

shows initial signs of segregation of its components. (b) Segre- gation of the nucleolar components and a coiled body close to the nucleolus are shown at an higher magnification (uranyl ac- etate and lead citrate, bar = 1 pm).

Glutathione and cisplatin neuroprotection 0 G. CAVALETTI et a/. 775

ing of motor nerves (4, 8). This pattern of involvement is well explained by the selective involvement of the DRG neurons to which DDP has easy access due to the less effective blood-nerve barrier at this site with respect to both the central nervous system and peripheral nerves (17).

At present, the only available and reliable method to demonstrate DRG neuron involvement is morphological examination. In humans, the definite demonstration of DRG neuron degeneration induced by DDP treatment has only recently been reported in a small series of nec- roscopies (10) because of the difficulty in obtaining ade- quate specimens. In rats however, it has been demon- strated experimentally that evident pathologic changes occur in the DRG neurons following acute high-dose DDP administration (2 l), but no nerve damage was observed, probably due to the short survival time of the animals in that experiment. Also the chronic administration of low doses of DDP can induce the same clear changes in the DRG neurons. In this case however, it is important to observe that an associated, secondary, mild damage of peripheral nerves has been demonstrated pathologically (5, 6) and neurophysiologically (11). The pathologic in- volvement of both DRG neurons and peripheral nerves is treatment-dependent and it is not related to the reduced weight gain which treated rats undergo, as fully discussed in a previous study (6). The demonstration of primary DRG neuron damage and subsequent peripheral nerve involvement closely reproduces the clinical features of DDP neurotoxicity in humans.

In this experimental model of chronic DDP neurotox- icity, we tested the protective effectiveness of GSH, a drug that in humans seems to reduce DDP toxicity without interfering with its antineoplastic activity (9, 13). In the experiment, the animals were treated with DDP, at a dose comparable to that used in clinical practice, alone or in combination with high doses of GSH. The morphological examination of DRG neurons, the electrophysiological evaluation of SNCV, and the analytical determination of DDP tissue concentrations all showed that GSH reduces the changes induced by DDP. Our experimental data therefore, lend support to the results of the clinical trials which have indicated that GSH has a protective effect on the neurotoxicity induced by DDP after prolonged treat- ment in humans and are in agreement with the neuro- physiological results reported by Hamers et al. (11) in the peripheral nerves of the rat.

So far no conclusive explanations for the GSH neu- roprotection have been proposed. First of all, a direct in- teraction between GSH and DDP in the blood seems un- likely. The pharmacokinetics of the two drugs are com- pletely different since GSH is rapidly cleared from the blood (within 15 min after a single intravenous admin- istration) ( l), while DDP has a biphasic blood distribution, with a first peak 25-49 min and a second one 58-73 h after IV administration (3). Moreover, the administration of GSH does not impair DDP anti-cancer activity in clin- ical trials in humans (9) and in experimental models in rats (11, 24), while a decrease in its activity would be induced if it were directly inactivated by GSH in the blood. On the other hand. GSH can enter a cell only if y-glutamyl transpeptidase (T-GT) is expressed on its plasma mem- brane. To our knowledge, this has been demonstrated in the peripheral nerves of the rat (20) while no information about the situation in DRG neurons is available. There- fore, it can be hypothesized that either GSH acts directly on DRG neurons if -r-GT is expressed by these neurons, or there is a peripheral uptake if y-GT is only expressed peripherally in the nerves. In the latter case, retrograde transport of GSH along the axons might occur, thus al- lowing the drug to reach the perikarion of the DRG neu- rons after peripheral uptake, and finally to act on the target structures of DDP toxicity. A third hypothesis might be that GSH neuroprotection is secondary to the well-dem- onstrated protective effect of the drug on the kidney, where y-GT is expressed and GSH accumulates (16, 24). Ac- cording to this hypothesis, the use of GSH allows a better metabolism of cisplatin so that blood DDP concentration might be less elevated and, therefore, nonneurotoxic, al- though still effective on the tumor. This hypothesis is sup- ported by the fact that increased DDP clearance has re- cently been demonstrated after GSH administration (12).

The results obtained in our study demonstrate that GSH interferes at least in part with the mechanism(s) of DDP- induced DRG neuronopathy in this experimental model of chronic DDP administration, although it is unclear whether the neuroprotection is due to direct action on the DRG neurons, secondary to nephroprotection, or both. This interference might be at the basis of the prom- ising but preliminary results thus far reported in clinical trials in humans. On the basis of our experimental study we conclude that the use of GSH, in humans, as a pro- tective drug during DDP treatment is worth investigating further in future studies.

REFERENCES

I. Ammon, H. P. T.; Melien, M. C. M.; Verspohl, E. J. Phar- macokinetics of intravenously administered glutathione in the rat. J. Pharm. Pharmacol. 38:721-725; 1986.

2. Apfel, S. C.; Arezzo, J. C.; Lipson, L. A.; Kessle, J. A. Nerve growth factor prevents experimental cisplatin neuropathy. Ann. Neurol. 3 1:76-80; 1992.

3. Calabresi, P.; Chabner, B. A. Antineoplastic agents. In: Goodman, G. A., Rall, T. W., Nies, A. S., Taylor P., eds.

The pharmacological basis of therapeutics. New York: Per- gamon; 1985:1249-1251.

4. Cavaletti, G.; Marzorati, L.; Bogliun, G.; Colombo, N.; Marzola, M.; Pittelli, M. R.; Tredici, G. Cisplatin-induced peripheral neurotoxicity is dependent on total-dose intensity and single-dose intensity. Cancer 69:203-207; 1990.

5. Cavaletti, G.; Petruccioli, M. G.; Tredici, G.; Marmiroli, P.; Barajon, I.; Fabbrica, D.; Di Francesco, A. Effects of

716 1. J. Radiation Oncology 0 Biology 0 Physics Volume 29, Number 4, 1994

6.

7.

8.

9.

10.

11.

12.

13.

14.

repeated administration of low doses of cisplatin on the rat nervous system. Int. J. Tiss. Reac. 18: 15 I- 157; 199 1. Cavaletti, G.; Tredici, G.; Marmiroli, P.; Petruccioli, M. G.; Barajon, I.; Fabbrica, D. Morphometric study of the sensory neuron and peripheral nerve changes induced by chronic cisplatin (DDP) administration in rats, Acta Neuropathol. (Berl.) 84:364-371; 1992. Cavaletti, G.; Tredici, G.; Pizzini, G.; Minoia, A. Tissue platinum concentrations and cisplatin schedules. Lancet ii: 1003; 1990. Daugaard, G. K.; Petrera, J.; Trojaborg, W. Electrophysi- ological study of the peripheral and central neurotoxic effect of cis-platin. Acta. Neurol. Stand. 76:86-93; 1987. Di Re, F.; Bohm, S.; Oriana, S.; Spatti, G. B.; Zunino, F. Efficay and safety of high-dose cisplatin and cyclophospha- mide with glutathione protection in the treatment of bulky advanced epithelial ovarian cancer. Cancer Chemother. Pharmacol. 25:355-360; 1990. Gregg, R. W.; Molepo, J. M.; Monpetit, V. J. A.; Mikael, N. Z.; Redmond, D.; Gadia, M.; Stewart, D. J. Cisplatin neurotoxicity: The relationship between dosage, time and platinum concentration in tissues, and morphological evi- dence of toxicity. J. Clin. Oncol. 10:795-803; 1992. Hamers, F. P. T.; Brakkee, J. H.; Cavalletti, E.; Tedeschi, M.; Marmonti, L.; Pezzoni, G.; Neijt, J. P.; Gispen, W. H. Reduced glutathione protects against cisplatin-induced neurotoxicity in rats. Cancer Res. 53:544-549; 1993. Leone, R.; Fracasso, M. E.; Soresi, E.; Cimino, G.; Tedeschi, M.; Castoldi, D.; Monzani, V.; Colombi, L.; Usari, T.; Ber- nareggi, A. Influence of glutathione administration on the disposition of free and total platinum in patients after ad- ministration of cisplatin. Cancer Chemother. Pharmacol. 29:385-390; 1992. Marzorati, L.; Bogliun, G.; Cavaletti, G.; Frattola, L. Effect on the peripheral nerves of the combined administration of cisplatin (DDP) and glutathione (GSH) (Abstr). It. J. Neurol. Sci. 12(Suppl.): 14; 199 1. Mollman, J. E.; Glover, D. J.; Hogan, W. M.; Furman, R. E. Cisplatin neuropathy. Risk factors, prognosis and pro- tection by WR-2721. Cancer 61:2192-2195; 1988.

15.

16.

17.

18.

19.

20.

21.

22.

23

24.

Mollman, J. E. Cisplatin neurotoxicity. N. Engl. J. Med. 322:126-127; 1990. Ohno, Y.; Jones, T. W.; Ormstad, K. Ally1 alcohol toxicity in isolated renal epithelial cells: Protective effects of low molecular weight thiols. Chem-Biol. Interact. 52:289-299; 1985. Petterson, C. A. V.; Salahndalin, T. S.; Olsson, Y. Presence of plasma proteins in spinal nerve roots. An immunohis- tochemical study in the rat. Acta Neurol. Stand. 82:21-27; 1990. Riggs, J. E.; Ashraf, M.; Snyder, R.; Gutmann, L. Prospective nerve conduction studies in cisplatin therapy. Ann. Neurol. 23:92-94: 1988. Roelofs, R. I.; Hrushesky, W.; Rogin, J.; Rosenberg, L. Pe- ripheral sensory neuropathy and cisplatin chemotherapy. Neurol. 34:934-938; 1984.

Romero, F.; Segura-Aguilar, J.; Monsalve, E.; Hermenegildo, C.; Nies, E.; Puertas, F. J.; Roma, J. Antioxidant and glu- tathione-related enzymatic activities in rat sciatic nerve. Neurotoxicol. Teratol. 12:603-605; 1990. Tomiwa, K.; Nolan, C.; Cavanagh, J. B. The effects of cis- platin on rat spinal ganglia: A study by light and electron microscopy and by morphometry. Acta Neuropathol. (Berl.) 69:295-308; 1986.

van der Hoop, R. G.; Vecht, C. J.; van den Burg, M. E. L.; Elderson, A.; Boogerd, W.; Heimans, J. J.; Vries, E. P.; van Houwelingen, J. C.; Jennekens, F. G. I.; Gispen, W. H.; Neijt, J. P. Prevention of cisplatin neurotoxicity with an ACTH(4-9) analogue in patients with ovarian cancer. N. Engl. J. Med. 322:89-94; 1990. Walsh, T. J.; Clark, A. W.; Parhad, I. M.; Green, W. R. Neurotoxic effects of cisplatin therapy. Arch. Neurol. 39: 719-720; 1982. Zunino, F.; Pratesi, G.; Micheloni, A.; Cavalletti, E.; Sala, F.; Tofanetti, 0. Protective effect of reduced glutathione against cisplatin-induced renal and systemic toxicity and its influence on the therapeutic activity of the antitumor drug. Chem-Biol. Interact. 70:89-101; 1989.