Oral metoclopramide with or without diphenhydramine: potential for prevention of late nausea and vomiting induced by cisplatin.

Late nausea and vomiting lasting 2-7 days occurs in 20%-68% of patients receiving cisplatin. We therefore studied the effects of oral metoclopramide given at doses of 20, 50, or 100 mg four times a day for 25 doses (7 days) beginning after cisplatin to determine the maximum tolerated dose (MTD) for prophylactic oral antiemetic regimens. Patients were stratified into younger (less than or equal to 30 yr old) and older (greater than 30 yr old) groups. Dose escalation was performed without or with concomitant oral diphenhydramine. For the younger group, the MTD for metoclopramide without diphenhydramine was less than 20 mg, and the MTD with diphenhydramine was 20 mg. For the older group, the MTD without diphenhydramine was 20 mg, and the MTD with diphenhydramine was 50 mg. Extrapyramidal reactions in the younger group and agitated depression in the older group were the major dose-terminating toxic effects. Patient acceptance of these regimens was excellent.

Improved methods for venous access: the Port-A-Cath, a totally implanted catheter system.

We prospectively evaluated the performance and rate of long-term complications with the Port-A-Cath (PAC), a totally implanted vascular access system. Two catheter styles were evaluated, a small-bore (SB) catheter (0.51-mm diameter) and a large-bore (LB) catheter (1.02-mm diameter), in conjunction with the use of a strict catheter care protocol. The PAC performed well, and with both SB and LB systems, no significant extravasation, skin necrosis, hematoma, septum damage or leakage, or subcutaneous portal infections occurred after 7,240 days of implantation and 1,435 days of use. Complications with the PAC system consisted of catheter occlusion (seven patients, 21.5%) and one instance of possible catheter infection (3.1%). Occlusions were limited to patients implanted with the SB catheter (seven of 16, 43.8%), and five of the seven (71.4%) occurred in patients receiving continuous infusion chemotherapy and/or total parenteral nutrition. Patency of the PAC system was maintained using a regular flushing schedule once every 30 days, a significant advantage compared with the daily maintenance schedule required with externally placed venous catheters. The results of this study suggest that the PAC system can provide a safe and reliable method for venous access in patients requiring intermittent or prolonged intravenous therapy.

High-dose intravenous metoclopramide versus combination high-dose metoclopramide and intravenous dexamethasone in preventing cisplatin-induced nausea and emesis: a single-blind crossover comparison of antiemetic efficacy.

We tested the safety and antiemetic effectiveness of intravenous (IV) dexamethasone (DXM) as an adjunct to high-dose IV metoclopramide (MCP) to prevent nausea and vomiting induced by high-dose cisplatin chemotherapy. Response was determined by using objective and subjective criteria. Thirty patients were randomly assigned to receive MCP alone at a dose of 2 mg/kg IV for three doses or the same dose of MCP plus 20 mg of DXM IV for three doses. Twenty evaluable patients received a second course of cisplatin and were crossed over to the opposite arm. Study results did not show a statistically significant advantage of combination MCP plus DXM over MCP alone using strict objective criteria for antiemetic response. However, patients subjectively preferred MCP plus DXM over MCP alone by nearly a 6:1 ratio, regardless of the randomization sequence. Although the addition of DXM does not appear to objectively improve emetic protection with high-dose MCP, we recommend MCP plus DXM to prevent nausea and vomiting induced by high-dose cisplatin chemotherapy when the use of steroids is not contraindicated, in view of patient preference for the combination.

Clinical pharmacokinetics of high-dose metoclopramide in cancer patients receiving cisplatin therapy.

Using a sensitive and specific high-pressure liquid chromatographic (HPLC) assay, we measured serum levels of metoclopramide in 18 cancer patients receiving high-dose intravenous (IV) therapy to prevent cisplatin-induced emesis. Ten patients were treated with one or more courses with metoclopramide alone (1.0 to 3.0 mg/kg) in an open-label study, and eight patients were treated with a fixed 2.0 mg/kg dose of metoclopramide with or without adjunct dexamethasone (20 mg) using a randomized, crossover design. The pharmacokinetics of metoclopramide were determined, and the relationship between serum levels and clinical response was evaluated. The pharmacokinetic parameters of high-dose metoclopramide were found to be similar to those reported for standard promotility doses, and no dose dependency was demonstrated over the range of doses studied. No clear correlation between serum metoclopramide levels and prevention of cisplatin-induced emesis was observed. The addition of dexamethasone resulted in clinical improvement in two of eight patients, but had no effect on serum metoclopramide levels or kinetic parameters. Results in this study population do not show metoclopramide levels to be related to antiemetic effect following IV cisplatin therapy.

Dose ranging phase I study of the serotonin antagonist GR38032F for prevention of cisplatin-induced nausea and vomiting.

GR38032F is a specific 5-HT3 (serotonin) receptor antagonist with antiemetic activity in animal and early human studies. We performed a dose-ranging phase I study of GR38032F in 43 evaluable patients receiving cisplatin 60 120 mg/m2 for the first time (38 of these patients were chemotherapy-naive). Intravenous GR38032F was administered over a dose range from 0.01 to 0.48 mg/kg given three times at four-hour intervals beginning one half hour before cisplatin, and patients were observed for 24 hours. An additional five patients were treated with 0.18 mg/kg given three times at six-hour intervals. Excellent antiemetic efficacy was noted, with 44% of patients experiencing no vomiting and 26% no nausea. Major protection from vomiting (less than or equal to 2 episodes) and from nausea (less than or equal to 2 hours) was experienced by 81% and 44%, respectively. Mild to moderate headache (40%), lightheadedness (21%), and elevated transaminase (19%) were the most common adverse events reported. One patient experienced an apparent hypersensitivity reaction that responded to conventional medications. No extrapyramidal reactions or akathisia were seen. GR38032F was effective through most of the dose range. However, efficacy decreased at the 0.01 mg/kg level and number and intensity of adverse events increased at the 0.48 mg/kg level. Analysis of those patients receiving high-dose cisplatin (100 to 120 mg/m2) revealed a positive association of GR38032F dose and antiemetic activity (Fisher's exact test, two-sided; P less than .05). The 5-HT3 receptor antagonists may provide antiemetic efficacy similar to high-dose metoclopramide without antidopaminergic toxicity. The maximum recommended dose on this schedule of GR38032F is 0.36 mg/kg.

Combination metoclopramide and dexamethasone: an effective antiemetic regimen in outpatients receiving non-cisplatin chemotherapy.

In a pilot study a combination of metoclopramide and dexamethasone was administered to 29 patients receiving emetogenic chemotherapy. Metoclopramide was given intravenously (IV) at a dose of 0.5 mg/kg one-half hour before the start of chemotherapy, and then given at the same dose orally two, five, and eight hours after chemotherapy. Dexamethasone was given at a dose of 10 mg IV immediately following the first dose of metoclopramide, then given at a dose of 8 mg orally six, 12, and 18 hours after chemotherapy. The chemotherapy regimens most commonly used were standard FAC, FAM, and BACOD regimens. Twenty-six of 29 patients received outpatient treatment. Complete protection against both nausea and vomiting was seen in 69% (20/29) patients; six others (21%) experienced mild nausea but no vomiting, resulting in 90% (26/29) of the patients having total emetic protection with combination metoclopramide and dexamethasone. Eighty-eight percent (15/17) of the patients with no prior chemotherapy had no nausea or vomiting, one (6%) had only mild nausea, and the remaining patient (6%) had one emesis. Forty-two percent (5/12) of the patients with prior chemotherapy had complete antinausea and emetic protection, five (42%) had nausea without vomiting, and the remaining two patients experienced one or two emesis. Side effects were minimal when present and included mild drowsiness (five patients), akathisia (three patients), diarrhea (one patient), and hot flashes (one patient). Combination metoclopramide and dexamethasone therapy can effectively prevent emesis in 94% of patients receiving potentially emetogenic chemotherapy and can prevent nausea and emesis in 88% of untreated patients. Studies defining the optimal dose and scheduling needed to maintain such antinausea and antiemetic protection are necessary.

Synthesis and dopaminergic activity of (+/-)-, (+)-, and (-)-2-dipropylamino-5-hydroxy-1,2,3,4-tetrahydronaphthalene.

In an effort to identify further the structural requirements for central dopamine receptor agonists, some monohydroxyl analogs of the known agonist 5,6-dihydroxy-2-dipropylamino-1,2,3,4-tetrahydronaphthalene were synthesized. They were examined for production of emesis in dogs and stereotyped behavior in rats. The most potent was 5-hydroxy-2-dipropylamino-1,2,3,4-tetrahydronaphthalene, which was more potent than apomorphine but less so than the dihydroxyl analog. The two enantiomers of the monohydroxyl analog were synthesized by conventional methods from an optically active intermediate, 2-benzylamino-5-methoxy-1,2,3,4-tetrahydronaphthalene. The resolution of this amine was performed with the aid of mandelic acid. Dopaminergic activity was found to be confined to the levo enantiomer. Requirements for both substitution and chirality in the tetralines were found to correspond closely to those known for the dopaminergic aporphines.

Synthesis and pharmacology of some 2-aminotetralins. Dopamine receptor agonists.

A series of 2-amino-1,2,3,4-tetrahydronaphthalene compounds bearing substituents on the nitrogen and in the aromatic ring was synthesized from beta-tetralone intermediates. Compounds were screened in vivo for dopaminergic activity using tests in which apomorphine was especially active. It was found that apparent dopaminergic activity is inherent in 2-dialkylaminotetralins, the dipropylamine substitution being the most consistently productive amine group studies. Activity was greatly enhanced by proper substitution in the aromatic ring. The 5,6-dihydroxy group was the best potentiating group found. These data support the idea that the extended conformation for the phenylethylamine moiety of ampmorphine and dopamine is favorable for dopaminergic agonist activity. They also suggest that an unetherified catechol group may not be essential for such activity.

Potent gastrin-releasing peptide (GRP) antagonists derived from GRP (19-27) with a C-terminal DPro psi [CH2NH]Phe-NH2 and N-terminal aromatic residues.

We have previously reported that octapeptides with a -DPro psi[CH2NH]Phe- NH2 C-terminus are potent GRP antagonists and have greatly enhanced in vivo stability. Now we report the detailed syntheses of such peptides and additional attempts to further increase metabolic stability. Replacement of the -DPro psi[CH2NH]Phe-NH2 with a "-DPro-statine"-Phe-NH2 led to less potent antagonistic activity. The introduction of ThiAla and BzthAla, to replace His and Trp, respectively, did not increase activity. A series of analogs having different aromatic residues at the N-terminal, other than 3-phenylpropionic acid, are equally potent. These residues show increased activity when hydrophilic substitutions are added to the aromatic ring. Replacement of the C-terminal Phe by DPhe and D2Nal is tolerated. Even though none of these peptides have higher activity than the original lead peptide, they are potentially more metabolically stable.

BW 1023U90: a new GRP receptor antagonist for small-cell lung cancer cells.

Gastrin-releasing peptide (GRP) receptor antagonists were synthesized and their ability to interact with small-cell lung cancer (SCLC) cells determined. [125I] BW1023U90, bound with high affinity (Kd = 2 nM) to a single class of sites (Bmax = 55 fmol/mg protein) using SCLC cell line NCI-H345. [125I] BW1023U90 binding was time dependent and reversible even at 37 degrees C as the ligand was minimally internalized. Specific [125I] BW1023U90 binding was inhibited with high affinity by GRP as well as bombesin (BB) but not neuromedin B (NMB). BW1023U90 inhibited the ability of BB to elevate cytosolic Ca2+ and increase the growth of SCLC cells. A BW1023U90 analogue, BW2258U89 (10 micrograms/day, SC) slowed SCLC xenograft format on in nude mice and [125I] BW 1023U90 localized to SCLC tumors 1 h after injection into nude mice. BW2258U89 (4% by weight) was placed in microspheres and slowly released over a 3-week period in nude mice bearing SCLC xenografts. The microspheres containing BW2258U89 strongly inhibited SCLC growth in vivo. A radioimmunoassay was developed for the GRP receptor antagonists and the rabbit antiserum cross-reacted totally with BW2258U89 or BW1023U90. BW2258U89 immunoreactivity (5 nM) was detected in the plasma of nude mice containing the microspheres after 1 week. These data suggest that GRP receptor antagonists bind to receptors on SCLC tumors.

Examination of the correlation of serum metoclopramide levels with antiemetic efficacy in patients receiving cisplatin.

The existence of a threshold serum metoclopramide level above which total protection from cisplatin-induced vomiting is more likely to occur has been proposed. We monitored serum metoclopramide levels prior to the third metoclopramide dose in the first cisplatin treatment cycle in patients receiving metoclopramide 2 mg/kg x 4 as part of a randomized double-blind cross-over study comparing single-agent metoclopramide with combination metoclopramide and dexamethasone. Serum samples from 35 patients (17 receiving single-agent metoclopramide and 18 receiving the combination) were analyzed using reverse-phase high-pressure liquid chromatography (HPLC). A wide variation in metoclopramide levels was observed (range 273-3380 ng/ml). Serum levels obtained from the same patient on multiple treatment cycles were well correlated, and the addition of dexamethasone did not alter serum metoclopramide levels. No threshold level could be identified for the two groups (single-agent or combination antiemetic therapy) considered individually or considered together. However, significantly more vomiting episodes and a lower incidence of total protection were noted in patients with metoclopramide levels above 1469 ng/ml receiving metoclopramide alone. This effect was nullified in the combination antiemetic group. Our data do not support a directly proportional relationship between serum metoclopramide level and antiemetic protection. However, a non-linear relationship with a possible agonist/antagonist effect is suggested.

Modification of cardiovascular actions of 2-amino-5,6-dihydroxytetralin by N,N-di-n-propyl substitution.

Effects of dopamine (DA) and the N,N-di-n-propyl derivative of 2-amino-5,6-dihydroxytetralin (dipropyl-A-5,6-DTN) were compared on renal blood flow in phenoxybenzamine pretreated dogs. Equivalent increments in renal blood flow were observed in the dose range of 12-190 nmol (ED50 of DA, 1.4 x 10(-8) mol; ED50 of dipropyl-A-5,6-DTN, 1.2 x 10(-8) mol, n = 7). Effects of 47 nmol of DA and dipropyl-A-5,6-DTN were reduced equally by simultaneous injections of 0.5 mg sulpiride (76 +/- 4 and 59 +/- 6%, respectively, n = 5). Intravenous injections of dipropyl-A-5,6-DTN in doses of 4-16 micrograms/kg elevated arterial blood pressure and decreased heart rate and cardiac contractility in open chest, vagotomized dogs, confirming its neuronal inhibitory effect mediated by DA2-presynaptic receptors. Dipropyl-A-5,6-DTN also caused dose-related femoral artery vasoconstriction in the hexamethonium pretreated dog. The hypertensive effect of intravenous dipropyl-A-5,6-DTN and its femoral vasoconstrictor effect were abolished by phenoxybenzamine. These results show that dipropyl-A-5,6-DTN possesses DA1-, DA2- and alpha-adrenergic activities. It is significant that by introducing di-n-propyl substitution on the nitrogen of 5,6-dihydroxy-2-aminotetralin, potent DA1-agonist activity can be conferred upon a molecule which is inactive in that respect. Further, beta-adrenergic activity of the primary amine is replaced by potent alpha-adrenergic activity by this substitution.

BW2258U89: a GRP receptor antagonist which inhibits small cell lung cancer growth.

The ability of reduced peptide bond analogues of gastrin releasing peptide (GRP) to antagonize small cell lung cancer (SCLC) GRP receptors was investigated. BW462U89, BW1023U90, BW2123U89 and BW2258U89 inhibited binding of (125I-Tyr4) BN to NCI-H345 cells with IC50 values of 5, 6, 140 and 10 nM respectively. The GRP analogues had no effect on basal cytosolic Ca2+ but inhibited the increase caused by 10 nM BN. BW462U89 reversibly blocked the increase in cytosolic Ca2+ caused by BN. The GRP analogues (1 microM) inhibited NCI-H345 colony formation in the absence or presence of 10 nM BN. Also, BW2258U89 (0.4 mg/kg, s.c. daily) inhibited xenograft growth in nude mice. These data indicate that BW2258U89 inhibits SCLC growth in vitro and in vivo.

A novel bombesin receptor antagonist selectively blocks the satiety action of peripherally administered bombesin.

To investigate the effect of a new, specific antagonist for bombesin receptors on the satiating action of exogenous bombesin, adult male rats were adapted to a nondeprivation test regimen with daily access to a palatable liquid food. In a prefeeding paradigm, rats received intraperitoneal injections of the bombesin receptor antagonist, BW2258U89 (6.25, 25, 50, or 100 micrograms kg-1) or vehicle 20 min before, and then a second injection of either bombesin (4 micrograms kg-1), gastrin-releasing peptide (GRP; 16 micrograms kg-1), the C-terminal octapeptide of cholecystokinin (CCK-8; 4 micrograms kg-1), or vehicle 5 min before a 2-h feeding test. BW2258U89 pretreatment antagonized the satiating actions of bombesin and GRP18-27 in a very potent, dose-related manner, but did not antagonize the satiating action of CCK-8. These differential results with BW2258U89 are consistent with prior results showing the potency of this antagonist for bombesin receptor-mediated effects in visceral systems; in addition, they demonstrate the selectivity of the compound for the satiating actions of peripherally administered bombesin and bombesin-like peptides.

Dopamine receptors in rat striatum and nucleus accumbens; conformational studies using rigid analogues of dopamine.

A study was made of the actions of dopamine and of some 2-amino-1,2,3,4-tetrahydronaphthalenes on dopamine-sensitive adenylate cyclase in homogenates of rat striatum and nucleus accumbens. The compounds were also tested for their ability to stimulate motor activity following bilateral injection into the nucleus accumbens of conscious rats. The most active compounds on adenylate cyclase from both striatum and nucleus accumbens were dopamine and 2-amino-6,7-dihydroxy-1,2,3,4-tetrahydronaphthalene(6,7-diOHATN). The 5,6-dihydroxy analogue (5,6-diOHATN) was 50 times less active than 6,7-diOHATN in striatal homogenates and 350 times less active in homogenates of nucleus accumbens. All dihydroxy compounds tested were active in causing stimulation of motor activity, the most active compounds being 6,7-and5,6diOHATN. Both dimethoxy derivatives tested were inactive on the adenylate cyclase and as locomotor stimulants.

Impact of NADiA ProsVue PSA slope on secondary treatment decisions after radical prostatectomy.

BACKGROUND: Selecting appropriate candidates for postprostatectomy radiotherapy is challenging, because adverse pathological features cannot accurately predict clinical recurrence. Biomarkers that identify residual disease activity may assist clinicians when counseling patients on the risks, benefits and costs of secondary treatment. NADiA ProsVue PSA slope results ≤2.0 pg ml(-1) month(-1) are predictive of a reduced risk of clinical recurrence; however, its clinical utility has not yet been studied. METHODS: We prospectively enrolled men treated by radical prostatectomy in a multicenter, institutional review board-approved clinical trial. At postsurgical follow-up, investigators (N=17) stratified men into low-, intermediate- or high-risk groups for prostate cancer recurrence based on clinicopathological findings and other factors. Investigators documented their initial treatment plan for each subject and serially collected three serum samples for ProsVue testing. After the ProsVue result was reported, investigators recorded whether or not the initial treatment plan was changed. The proportion of cases referred for secondary treatment before and after ProsVue was reported, and the significance of the difference determined. RESULTS: Complete assessments were reported for 225 men, 128 (56.9%) of whom were stratified into intermediate- and high-risk groups. Investigators reported that they would have referred 41/128 (32.0%) at-risk men for secondary treatment. However, after results were known, they referred only 15/128 (11.7%) men. The difference in proportions (-20.3%, 95% confidence interval (CI) -29.9 to -10.3%) is significant (P<0.0001). Odds of a referral was significantly reduced after results were reported (odds ratio 0.28, 95% CI 0.15-0.54, P<0.0001). CONCLUSIONS: Knowledge of a ProsVue result had significant impact on the final treatment plan. A ProsVue result ⩽2.0 pg ml(-1) month(-1) significantly reduced the proportion of men at risk of recurrence who otherwise would have been referred for secondary treatment.

Intermittent androgen deprivation in prostate cancer patients: factors predictive of prolonged time off therapy.

OBJECTIVES: We hypothesize that prostate cancer (PC) patients who achieve and maintain an undetectable prostate-specific antigen (UD-PSA) on androgen deprivation therapy (ADT) have a predominantly androgen-dependent cancer cell population sensitive to apoptosis that allows for a prolonged time off ADT. This study summarizes patient- and treatment-related factors associated with a prolonged time off ADT in patients electing intermittent androgen deprivation (IAD). METHODS: Hormone-naïve patients with PC were treated with ADT using an antiandrogen and a luteinizing-hormone-releasing hormone-agonist. Of 255 consecutive patients, 216 (85%) achieved a UD-PSA (< 0.05 ng/ml). Ninety-three (43%) of 216 elected to stop ADT after maintaining a UD-PSA for a median of one year. Patients were followed off therapy and advised to restart ADT if the PSA level reached > or = 5.0 ng/ml. Forty-one patients received finasteride as part of IAD induction and as maintenance off therapy; these patients are excluded from the current study and are the focus of another publication. The remaining 52 patients are assessable for response being either in the off-phase of IAD > or = 1 year or having restarted IAD. RESULTS: In the first IAD cycle, the median duration of the on-phase of IAD was 16 months (mean 19.0 months, range 3.6-71 months), and the median off-phase duration was 15.5 months (mean 24.1 months, range 3.2-87+ months). In 28 patients who maintained a UD-PSA for > or = 1 year, their median off-phase duration was 29 months (mean 35.8 months, range 7.8-87+ months), with nine (32%) still off IAD after a median follow-up of 62 months. Significant (p < 0.05) independent factors associated with prolonged off-phase duration by multivariate analysis included UD-PSA on ADT > or = 1 year (p = 0.010), PSA-only recurrence after local therapy (p = 0.039), and reaching a testosterone level > or = 150 ng/dl in > or = 4 months off ADT (p = 0.041). After a median of 66 months of follow-up, only one (2%) patient developed androgen-independent PC. CONCLUSIONS: Hormone-naïve patients who achieve and maintain a UD-PSA for at least one year during ADT may initiate IAD and anticipate a prolonged off-phase duration. Attainment of a UD-PSA on ADT may serve as an in vivo sensitivity test of a patient's tumor cell population, and allow for better selection of those best suited for IAD.

Antibody specific to the alpha subunit of the guanine nucleotide-binding regulatory protein Go: developmental appearance and immunocytochemical localization in brain.

A polyclonal rabbit antibody (9120) against the alpha subunit of the "other" guanine nucleotide-binding protein Go (alpha o) was raised against a synthetic alpha o peptide fragment (Asp-Gly-Ile-Ser-Ala-Ala-Lys-Asp-Val) attached to a branched core system. Antiserum 9120, at a final dilution of 1:400, can detect alpha o in as little as 0.2 microgram of Go on immunoblots, and, at a final dilution of 1:20,000, can detect alpha o in 1 microgram of Go on immunoblots. Antiserum and affinity-purified antibody are specific to alpha o. No cross-reactivity was detected to the alpha subunits of the stimulatory or inhibitory guanine nucleotide-binding regulatory proteins or of transducin (alpha s, alpha i, or alpha T) or to the beta or gamma subunits. Immunoblots revealed a high density of alpha o in rat brain and lung membrane preparations, but other tissues (such as adipose tissue, heart, erythrocytes, and liver) have no detectable alpha o. Developmental studies showed that alpha o in rat brain was low before birth, increased after birth, and reached the full adult level at 4 weeks of age. In contrast, ADP-ribosylation of 40-kDa proteins increased for up to 1 week and then decreased. Immunocytochemistry revealed that alpha o was localized to somatic and synaptic membranes in rat brain, whereas little or no alpha o was detected in the cytoplasm of neuronal cell bodies. Our observations suggest that Go in brain might have a role in membrane signal transduction at synaptic and extrasynaptic sites.

Treatment of astrocytoma with a 5-day cisplatin infusion.

Although cisplatin may have intrinsic activity against astrocytoma, limited penetration into the central nervous system may severely curtail delivery of adequate amounts of drug to the tumor. In an attempt to increase the dose of delivered drug without markedly increasing toxicity, cisplatin was administered by 5-day continuous intravenous infusion at a dose of 28 mg/m2/day (total dose 140 mg/m2) to 15 evaluable patients (5 with astrocytoma Grade III and 10 with astrocytoma Grade IV). Median age was 39 years, median Karnofsky Performance Status was 80%, and all patients had been previously treated with other modalities. One patient (7%) achieved Partial Response as demonstrated by increased strength of paretic extremities, increased Karnofsky Performance Status, and decrease of enhancing tumor mass on CT scan. Although nephrotoxicity was minimal, nausea and vomiting (usually mild) was seen in 14 patients. Myelosuppression was also common (anemia in 7 patients, leukopenia in 4 patients and thrombocytopenia in 3 patients). Ototoxicity was seen in 5 patients and may represent a combined modality toxicity with prior cranial radiotherapy. Routes and schedules which allow a higher peak serum concentration of cisplatin (such as subselective intracerebral artery infusion) may be necessary to achieve greater central nervous system drug penetration and to maximally exploit cisplatin in the treatment of astrocytoma.