Pharmacology of intrathecal VP-16-213 in dogs.

VP-16-213 is an anticancer drug that is active against a number of malignancies including small cell lung cancer, lymphoma, and leukemia which are often complicated by the development of leptomeningeal carcinomatosis. To investigate the potential usefulness of VP-16-213 for intrathecal administration, the pharmacology and toxicity of intrathecal VP-16-213 was determined. VP-16-213 at varying doses (0.01-1.0 mg.kg) was instilled intrathecally in dogs. Plasma, CSF, spinal cord, and brain tissue drug concentrations were determined by radiochemical and high performance liquid chromatography technique. Drug concentrations were strikingly higher in spinal cord tissue near the injection site compared to more distal cord sites. CSF concentration of VP-16-213 is 3-4 logs higher compared to concurrent plasma levels. Severe neurotoxicity occurred at the higher doses used. Due to limited diffusion and extremely low doses which could be used without life-threatening neurotoxicity, VP-16-213 does not appear to be a useful agent for intrathecal administration.

Pharmacokinetics of potential anti-AIDS agents thiofoscarnet and foscarnet in the cat.

The pharmacokinetics of thiophosphonoformate (TPFA) and phosphonoformate (foscarnet, PFA) were studied in normal adult cats, a species susceptible to feline immunodeficiency virus (FIV) infection. Parent drugs and metabolites were quantitated by high-performance liquid chromatography (HPLC). TPFA had a mean terminal plasma half-life of 42 min, a total clearance of 4.58 ml/min/kg, and a renal clearance of 1.24 ml/min/kg (N = 4). TPFA underwent in vivo metabolism to PFA and thiophosphonic acid (TPA); the latter was inactive against HIV reverse transcriptase. The 6-h cumulative urinary excretion was 42.3% of the intravenous administered dose of TPFA, consisting of 23.5% unchanged TPFA, 13.8% PFA, and 5.0% TPA. In comparison, PFA had a mean (N = 5) terminal half-life of 172 min and a total clearance of 1.88 ml/min/kg, approximating its renal clearance. There was no evidence of PFA metabolism. Oral doses of TPFA were administered either in enteric-coated capsules or in solution by gavage. The mean oral bioavailability of encapsulated TPFA and PFA was 22 and 8%, respectively. When given by gavage, TPFA had a higher mean bioavailability (33%), but with a greater variability. Based on the 6-h cumulative urinary excretion of TPFA, the mean oral bioavailability of TPFA was 44%, similar to that based on plasma data. The TPFA appears to be superior to PFA because of its greater oral bioavailability and its ability to deliver an active metabolite, PFA, to the systemic circulation after oral dosing.

Pharmacokinetics of homoharringtonine in dogs.

We studied the pharmacokinetics and distribution of homoharringtonine (HHT), an antitumor alkaloid, in anesthetized dogs using chromatographic and radiochemical techniques. Uniformly tritiated HHT was administered i.v. to five dogs at doses of 0.05 to 0.34 mg/kg, 200 microCi per animal. Unchanged HHT disappeared in a triphasic manner from the plasma with an initial plasma t1/2 of 9.4 +/- 4.2 min, an intermediary t1/2 of 1.4 +/- 0.5 h, and a terminal t1/2 of 40.6 +/- 4.6 h. The plasma clearance was 114.0 +/- 20.1 ml/kg-1 h-1 and the steady-state volume of distribution was 6.2 +/- 0.7 1/kg. In 72 h, 40.1% +/- 4.0% of the administered radioactivity was excreted in the urine, 17.8% +/- 2.7% of which was unchanged HHT. HHT was metabolized extensively to one major and two minor metabolites. Biliary excretion of total radioactivity was 14.4% in 5 h, 2% of which was HHT. HHT concentration in the CSF was highest 4 h after drug administration, about 40% of the concentration in the concurrent plasma. At autopsy 5 h after dosing, the highest percentage of HHT was in the liver (7.4%), followed by the small intestine (2.5%), stomach (1.0%), pancreas (0.8%), kidneys (0.8%), and lungs (0.7%). The heart, spleen, large intestine, and brain each retained less than 0.5%. However, 24 h after dosing, 4% of the HHT still remained in the liver, 1% in the small intestine, and less than 1% in the other organs. HHT seems to be extensively metabolized in dogs and partially retained in the body.

Central nervous system (CNS) penetration of homoharringtonine (HHT).

Generally tritiated homoharringtonine ([3H]HHT, 150 microCi, 430 micrograms) was administered intravenously to seven patients at varying times before surgical resection of malignant brain tumor. Plasma, urine, cerebrospinal fluid (CSF), and tumor specimens were obtained during surgery, and the concentrations of HHT, its major metabolite, and [3H]HHT equivalent were determined chromatographically and radiochemically. For [3H]HHT equivalent, the concentration in tumor ranged from 0.6 to 4.3 ng/g and the ratio of tumor to plasma concentration from 0.5 to 1.8. In one patient who had CSF available for drug determination, the CSF to plasma ratio of total [3H]HHT was 0.3 at 45 minutes after drug administration and less than 0.2 ng/ml was unchanged HHT. For unchanged HHT, drug concentration in tumor ranged from undetectable (4 patients) to 1.8 ng/g. A major metabolite of HHT was detectable in the tumor specimens of all the patients. These results indicate that homoharringtonine can penetrate into brain tumors; in 3 patients with brain tumors, the ratios of HHT concentration in the tumor to that in the concurrent plasma were greater than one.

Comparison of CNS penetration, tissue distribution, and pharmacology of VP 16-213 by intracarotid and intravenous administration in dogs.

Eight beagle dogs received [3H]VP 16-213 at 2 mg/kg administered intravenously (IV) or intra-arterially (IC) through a catheter inserted into the internal carotid artery. Blood, urine, bile, and cerebrospinal fluid (CSF) samples were collected at intervals. At 1, 6, 24 hr, and 2 weeks after drug administration the dogs were sacrificed and the major organs analyzed for drug concentration. VP 16-213 concentration was determined by radiochemical assay and high pressure liquid chromatography. The plasma t1/2 in the IC group of dogs was 1.0 hr, the volume of distribution was 1.7 L/kg and the clearance was 1.5 ml/hr/kg. In the IV group the values were 1.7, 3.9, and 1.6, respectively. The CSF concentration peaked at 1 hr by both routes, but was higher at all time points in the IC group. At 24 hr and 2 weeks after IC VP 16-213, drug concentration in brain tissue was at least four times higher in the IC group compared with the IV group. In extracranial organs the reverse was true, with the bone marrow cell concentration 1.6 times higher by IV compared to IC (267.2 ng/g and 164.5 ng/g, respectively). Two major and one minor metabolites were found in plasma, urine, bile, and tissue by both routes, however, not all metabolites were found in all organs and body fluids. No acute neurologic toxicity was noted in the IC group and no histopathologic changes by light microscopy were found in the brain or other organs. IC VP 16-213 produced higher drug concentration in the brain of dogs compared with IV administration and was well tolerated at the dosage used.

Clinical pharmacology of homoharringtonine.

Clinical pharmacokinetics of homoharringtonine (HHT) were studied in eight patients who received uniformly labeled HHT at 3-4 mg/m2 (150 mu Ci) by continuous 6-hour infusion. The drug and metabolites were quantified by radiochemical and high-performance liquid chromatographic techniques. Computerized nonlinear least-square regression and curve stripping were used to characterize HHT and total [3H]HHT equivalent pharmacokinetics. Unchanged HHT in the plasma declined biphasically, with an alpha-half-life of 0.5 +/- 0.1 hours and a beta-half-life of 9.3 +/- 1.4 hours. The total clearance of HHT was 177.4 +/- 27.7 ml X hour-1 X kg-1, and the apparent volume of distribution, estimated from the area under the drug concentration versus time curve, was 2.4 +/- 0.4 L X kg-1. Correspondingly, the total [3H]HHT equivalent disappeared from the plasma in a triphasic manner. Compared with the pharmacokinetic parameters of unchanged HHT, the terminal half-life of total 3H was 67.5 +/- 7.5 hours, 7.4 times longer; the total clearance was 30.9 +/- 3.1 ml X hour-1 X kg-1, 5.5 times slower; but the volume of distribution by area was 2.7 +/- 0.1 L X kg-1, nearly the same. The 72-hour cumulative urinary excretion of total tritium was 28.2% of the administered dose and only 38.3% of this resided in unchanged HHT. Thus, urinary excretion was not a major route of elimination of HHT. Moreover, HHT underwent extensive metabolism; one major and two minor unidentified products were detected in both plasma and urine.

Clinical pharmacology of intracarotid etoposide.

Pharmacokinetics studies were performed in ten patients who received VP-16 by intracarotid infusion at 100-300 mg/m2. VP-16 was analyzed by high-pressure liquid chromatography. ESTRIP and NONLIN were used to characterize VP-16 pharmacokinetics. VP-16 disappeared biphasically, with a t1/2 beta of 6.1 +/- 1.4 h; the total clearance was 26.8 +/- 2.8 ml/min/m2, and the Vss was 8.8 +/- 1.6 l/m2. The pharmacokinetics was not significantly different after administration by the IV route. However, at a lower dosage, less than 140 mg/m2, the half-life appears to be shorter. This may or may not be significant, since VP-16 pharmacokinetics is quite variable and the number of patients studied is relatively small. Overall, the brain and brain tumor do not appear to have any first-pass effect on VP-16 pharmacokinetics.

Clinical pharmacokinetics of 9, 10-anthracenedicarboxaldehyde-bis [(4,5-dihydro-1 H-imidazol-2-yl)hydrazone]dihydrochloride.

We studied the clinical pharmacokinetics of the anthracene derivative bisantrene using high-performance liquid chromatographic analysis. We administered the drug to ten patients at 120-250 mg/m2 IV; one of these patients also received a second dose of 120 mg/m2 6 weeks later, and another received 150 mg/m2 weekly for three doses. Bisantrene disappeared from the plasma biphasically, with an initial t1/2 of 0.6 +/- 0.3 h and a terminal t1/2 of 24.7 +/- 6.9 h after single doses. The apparent volume of distribution according to the area under the curve was 42.1 +/- 5.9 l/kg, and the total clearance was 1045.5 +/- 51.0 ml/kg/h. The 96-h cumulative urinary excretion was 3.4% +/- 1.1% of dose; thus, renal excretion was a minor route of elimination for this agent. Bisantrene pharmacokinetics in the patient who received a second dose after 6 weeks showed insignificant changes. However, in the patient who was given this drug weekly for 3 weeks, the plasma t1/2 of the drug during the terminal phase became increasingly longer, while the total clearance was significantly reduced. These results suggest that bisantrene may accumulate in the body and that caution is essential in the event of frequent administration.

Clinical pharmacology of 4-demethoxydaunorubicin (DMDR).

DMDR, a daunorubicin derivative with a higher therapeutic index and lower cardiotoxicity than either the parent drug or doxorubicin, is active when given PO in experimental animals. We studied its pharmacokinetics in ten patients receiving DMDR IV or PO or IV and PO sequentially at 10-12.5 mg/m2. DMDR and its metabolites were quantified by high-performance liquid chromatography and fluorometry. In nine patients who received DMDR IV the unchanged drug disappeared from the plasma biphasically with a mean terminal half-life of 27.0 +/- 5.5 h, an apparent volume of distribution of 63.9 +/- 12.61 kg-1, and a total clearance of 1.9 +/- 0.41 kg-1 h-1. In 24 h only 5.1% +/- 1.1% of the dose was excreted in the urine. In comparison, in 19 studies the plasma half-life of DMDR given PO was 34.8 +/- 6.7 h, 2.3% +/- 1.3% was excreted in the urine in 24 h, and the maximum plasma drug concentration was reached in about 1 h. The bioavailability of DMDR given PO was about 39% according to comparison of the areas under the plasma DMDR concentration versus time curves for the two routes, but 45% according to comparison of the 24-h cumulative urinary excretion rates. In one patient with severe liver dysfunction following oral administration, the plasma DMDR half-life was 56.8 h, more than twice the average length. By either route, the drug was quickly metabolized to one major metabolite, DMDR-ol. The plasma half-life of DMDR-ol was 72.5 +/- 24.7 h, or 35.7 +/- 7.4 when DMDR was administered IV or PO. In the plasma, DMDR-ol always exceeded DMDR in concentration. Moreover, the 24 h cumulative urinary excretion of DMDR-ol as a percentage of the dose of DMDR administered was 7.8 following IV and 7.4 following PO administration.

Sequential administration of thymidine, 5-fluorouracil, and PALA. A phase I-II study.

Twenty-seven patients with colorectal adenocarcinoma, (12) non-small cell bronchogenic carcinoma, (11) gastric adenocarcinoma (3), and adenocarcinoma of unknown primary lesion (1) were treated with the combination of thymidine (TdR), 5-fluorouracil (FU), and N-phosphonacetyl-L-aspartic acid (PALA). PALA 1 g/m2 was given over 1 hour on day 1, followed on day 2 by 30 g of TdR given over 3 hours. FU, 150-300 mg/m2, was administered sequentially over 1 hour immediately following TdR infusion. There were no responses seen using this dose schedule. Gastrointestinal and central nervous system toxicities were dose-limiting. Myelosuppression was seen at all dose levels and was not dose related. Fever and infection occurred in 16% and 3% of the courses. The maximum tolerated dosages on this schedule were PALA, 1 g/m2; TdR, 30 g; and FU, 250 mg/m2. Pharmacologic studies done revealed the following half-lives: TdR, 1.6 hours; thymine, 5.0 hours; FU, 6.8 hours; and FUDR, 3.7 hours. The significant prolongation of the half-life of FU with this drug combination implies that the tumor tissues may be exposed longer to the anticancer action of FU.

Induction of DNA strand breaks and cross-links by 2,5-diaziridinyl-3,6-bis(carboethoxyamino)-1,4-benzoquinone in Chinese hamster ovary cells.

The DNA-damaging effects of 2,5-diaziridinyl-3,6-bis(carboethoxyamino)-1,4-benzoquinone (AZQ) in Chinese hamster ovary cells were investigated. As determined by alkaline elution, DNA strand breaks were observed in cells treated with 50 microM AZQ for 2 hr. The single-strand break frequency was 31.3 +/- 5.3 (S.D.) rad equivalents. Strand breaks could also be detected at lower drug concentration if proteinase K treatment was included before DNA elution. In comparison, DNA cross-links were apparent in cells treated with as low as 6.25 microM AZQ. The cross-linking frequencies were 39.7, 124.3, 230.3, and 625.1 rad-equivalents for 6.25, 12.5, 25, and 50 microM AZQ, respectively. Both DNA-DNA and DNA-protein cross-links in AZQ-treated cells were revealed by the proteinase K assay. The DNA strand breaks induced by AZQ were rapidly rejoined within 1 hr after drug removal. DNA interstrand cross-links increased within the first 6 hr of postincubation and then slightly decreased by 12 hr, and most of the cross-links disappeared after cells were allowed to recover for 24 hr. DNA-protein cross-links were immediately formed during the drug treatment period and were gradually decreased after drug removal.

Inhibition of ribonucleotide reductase by caracemide.

Caracemide, a new antitumor agent now in clinical trial, was tested against partially purified ribonucleotide reductase from rat Novikoff tumor. It was found to be an active inhibitor, about one-ninth as effective as hydroxyurea.

Alkylation of DNA by the new anticancer agent 3,6-diaziridinyl-2,5-bis(carboethoxyamino)-1,4-benzoquinone (AZQ).

The bifunctional cross-linking activity of 3,6-diaziridinyl-2,5-bis(carboethoxyamino)-1,4-benzoquinone (AZQ, NSC 182986) on isolated calf thymus DNA was studied, using ethidium bromide fluorescence assay. Between 1 and 350 microM AZQ produced a dose-dependent cross-linking effect in the presence of a 10-fold excess of sodium borohydride. No cross-linking was observed in the absence of the reducing agent. AZQ can also be activated by NADH and NADPH at pH 4. The AZQ cross-linking activity exhibited a strong pH dependency, highest at acidic pH, lower at alkaline pH and not seen under neutral conditions. It was also significantly inhibited under anaerobic conditions. At pH 5 the binding ratio was 1 molecule of AZQ per 191 bases at an AZQ dose of 300 microM. Our results suggest that reduced AZQ behaved like a bifunctional alkylating agent.

Central nervous system pharmacology of Baker's antifolate (NSC139105) in man.

Radiolabelled Baker's Antifolate (BAF) was administered to 6 patients undergoing surgical resection of intracerebral tumors. Levels of radioactivity in resected tumor and edematous brain adjacent to tumor were generally higher than levels in concurrent plasma samples and were generally comparable to levels in temporalis muscle. Levels in tumor cyst fluid were far lower than concurrent plasma levels and levels in surrounding tumor. Chromatography was performed on tumor from 2 patients and revealed that only a small proportion of the radioactivity represented unchanged BAF. The major metabolite present in tissues was 1 000 times less potent as an inhibitor of dihydrofolate reductase than was BAF. Five patients had cerebrospinal fluid (CSF) sampled following administration of tracer doses of radiolabelled BAF. Radioactivity levels were far lower in CSF than in plasma. Levels of radioactivity in the CSF were also far lower than levels in tumor and brain samples from other patients and were slightly lower than tumor cyst fluid levels. Two patients had CSF collected after they received therapeutic doses of BAF. In these patients, both CSF and plasma were assayed using a dihydrofolate reductase inhibition assay. As with tracer dose studies, CSF concentrations of BAF were substantially lower than were concurrent plasma concentrations. Thus it appears that only very low concentrations of BAF are attainable in human CSF and intracerebral tumor, although a metabolite which is a very weak inhibitor of dihydrofolate reductase attains high concentrations in tumor.

Effect of thymidine on the toxicity and antitumor activity of cis-diamminedichloroplatinum (II).

The effect of thymidine (TdR) on the preclinical toxicity of cis-diamminedichloroplatinum (II) (DDP) was investigated in the BDF1 mouse and the Sprague-Dawley rat. The effect of TdR on the antitumor activity of DDP was investigated using the ascites P388 murine leukemia model. TdR at 500 mg/kg consistently decreased the recovery of body weight after DDP treatment IP, but did not affect the lethal toxicity of DDP to non-tumor-bearing mice or those with the P388 murine leukemia. This effect was greatest when TdR was injected 30 min prior to DDP and at higher doses of DDP. A 500-mg/kg dose of TdR did not affect the antitumor activity of DDP 5 mg/kg administered on days 1, 5, and 9. Treatment of rats with TdR 500 mg/kg according to various schedules of timing relative to a 5-mg/kg dose of DDP did not consistently affect the DDP-related loss in body weight or nephrotoxicity at day 3. Pretreatment of mice with TdR 1,500 mg/kg 30 min prior to DDP 5 mg/kg (every 4 days X 3) resulted in a slower recovery of body weight, which became more pronounced with increasing doses of DDP. Pretreatment of ascites P388-bearing mice with TdR 1,500 mg/kg increased the number of early deaths when mice were treated with DDP 5 mg/kg (days 1, 5, and 9). These data suggest that the cytotoxicity of DDP is increased by TdR only at higher doses of either drug, but that the antitumor activity against P388 murine leukemia is not affected.

Human tissue distribution of 4'-(9-acridinylamino)-methanesulfon-m-anisidide (NSC 141549, AMSA).

Concentrations of AMSA were determined by HPLC in autopsy tissue samples from five patients who had received the drug antemortem. Relative organ concentrations of AMSA varied from patient to patient; however, concentrations were generally highest in gallbladder, liver, and kidney, while low levels were generally but not invariably found in lung, testicle, muscle, fat, spleen, bladder, pancreas, colon, prostate, and brain. One patient with ventricular fibrillation and seizures had high tissue AMSA concentrations in myocardium, but low concentrations in brain. Another patient with seizures during treatment had high brain concentrations of AMSA. Relative organ concentrations were similar to those found in mice, except that mice have high AMSA concentration in their spleens whereas our patients did not, even when the spleen was infiltrated with leukemic cells. High tissue concentrations of AMSA were still present 2 weeks after treatment. AMSA concentration was lower in a renal oncocytoma (1.1 micrograms/g) than in surrounding kidney (2.4 micrograms/g).

Pharmacological disposition of 1,4-dihydroxy-5-8-bis[[2 [(2-hydroxyethyl)amino]ethyl]amino]-9,10-anthracenedione dihydrochloride in the dog.

DHAQ, a new antitumor agent, has been selected for clinical trial on the basis of its activity against a number of transplantable rodent tumors. In anticipation of the clinical trial of this agent, the pharmacology of DHAQ was studied in beagles by high-pressure liquid chromatographic and radiochemical techniques that are specific for the unchanged drug. 14C-DHAQ was administered IV to beagles at a dose of 5 mg/kg, 100-125 microCi total. With a maximal plasma concentration of 75 +/- 2.7 ng/ml, DHAQ was eliminated from the plasma with a half-life of 28.1 h during the terminal phase. The total clearance of DHAQ was 10.1 +/- 0.4 mg/kg/min, while the apparent volume of distribution was 26.6 +/- 4.9 l/kg. In 48 h, 2.4% +/- 0.6% of the dose was excreted in the urine and 3.0% +/- 0.1% in the bile as the unchanged drug. At autopsy performed 5 h after dosing, the highest percentage of the administered DHAQ was in the liver (49.7% +/- 2.7%), followed by the small intestine (7.1% +/- 0.7%), kidneys (2.7% +/- 0.1%), lung (1.9% +/- 0.3%), spleen (1.6% +/- 0.3%), and stomach (1.3% +/- 0.1%). The heart, large intestine, pancreas, gallbladder, urinary bladder, and brain each retained less than 1% of the dose. However, 24 h after dosing 10.6% of the drug was detected in the liver and 2.9% in the small intestine. In terms of the percentage of the dose, the distribution of DHAQ in the other organs either remained unchanged or increased slightly. In concentrations varying from 10 ng/ml to 10 micrograms/ml the drug was 70%-80% bound to plasma protein. DHAQ was metabolized to two unidentified metabolites. Thus, this drug appeared to be cleared from the plasma of beagle dogs chiefly by tissue binding, leading to possible persistence of the drug in certain body compartments.

Determination of 1-(2-chloroethyl)-3-(2,6-dioxo-3-piperidyl)-1-nitrosourea in plasma by negative chemical ionization mass spectrometry.

A sensitive and specific method for the quantitative determination of a new antitumor agent, 1-(2-chloroethyl)-3-(2,6-dioxo-3-piperidyl)-1-nitrosourea, (PCNU) has been developed for the analysis of plasma. This assay involves extraction of the plasma sample, separation of the drug by thin-layer chromatography and mass spectrometric detection of the negative ions derived from the unchanged drug and its 2H4-labeled analog. Ion current profile peaks are obtained when drug samples are introduced into the heated source using a desorption chemical ionization probe. Detection limits are below 1.0 ng ml-1 of plasma. This method has been applied to determine the in vitro rate of decomposition of PCNU in plasma and the plasma clearance of this drug from a cancer patient.

Immune restoration and/or augmentation of local graft versus host reaction by traditional Chinese medicinal herbs.

The in vitro restorative effect of aqueous extracts from two traditional Chinese medicinal herbs were studied in 19 cancer patients and in 15 normal healthy donors. Using the local graft versus host (GVH) reaction as a test assay for T-cell function, the extract from astragalus membranaceus (10 microgram/ml) induced a restored reaction in nine of ten patients with an increase in local GVH reaction from 18.2 plus/minus 15.8 mm3 to 112.9 plus/minus 94.2 mm3 (P less than 0.01). The extract from ligustrum lucidum, likewise effected an immune restoration in nine of 13 cancer patients with an increase in local GVH reaction from 32.3 plus/minus 36.1 mm3 to 118 plus/minus 104.9 mm3 (P less than 0.01). This degree of immune restoration appears to be complete as it equals the local GVH reaction observed among untreated mononuclear cells from normal healthy donors (82.8 plus/minus 41.1 mm3, P greater than 0.1). These results suggest that both extracts of the traditional Chinese medicinal herbs contain potent immune stimulants which may provide the rational basis for their therapeutic use as biological response modifiers.

Human pharmacokinetics of a new acridine derivative, 4'-(9-acridinylamino)methanesulfon-m-anisidide (NSC 249992).

The clinical pharmacology of 4'-(9-acridinylamino)methanesulfon-m-anisidide (amsacrine) was studied, utilizing [9-14C]amsacrine i.v. in 19 patients with disseminated neoplasms. The mean terminal plasma half-life for total 14C ranged from 34 hr in patients with normal organ function to 46 hr in patients with severe liver disease. For unchanged amsacrine, the mean values of plasma half-life were 7.4 and 17.2 hr for patients with normal and abnormal liver function, respectively. The plasma half-lives of 14C were prolonged, while those for unchanged amsacrine appeared to be normal in patients with renal dysfunction. The mean 72-hr cumulative urinary excretion of total 14C varied from 35% in normal patients to 49% in patients with severe liver disease, while patients with renal disease excreted only 2 to 16%. In comparison, the urinary excretion of unchanged amsacrine was 12, 20 and 2% of the administered dose, respectively, in these same patients. Amsacrine biliary excretion studied in two patients showed about 8 and 36% of the administered radioactivity excreted in the bile in 72 hr, with less than 2% as unchanged amsacrine. Cerebrospinal fluid concentrations of amsacrine were below 2% of the simultaneous plasma levels in three patients. Impaired amsacrine drug clearance was frequently associated with liver dysfunction. Patients with impaired amsacrine drug clearance experienced the most severe clinical toxicity. Hepatic metabolism and biliary excretion appear the most important routes for amsacrine elimination. Renal elimination, although less important, is significant in patients with severe kidney dysfunction. To avoid excessive clinical toxicity, initial dose reductions of 30 to 40% are recommended for patients with severe liver or renal disease or for those who have pharmacologically documented impaired drug clearance.