William R. Crom

Conventional compared with individualized chemotherapy for childhood acute lymphoblastic leukemia

BACKGROUND The rate of clearance of antileukemic agents differs by a factor of 3 to 10 among children with acute lymphoblastic leukemia. We hypothesized that the outcome of treatment would be improved if doses were individualized to prevent low systemic exposure to the drugs in patients with fast drug clearance. METHODS We stratified and randomly assigned 182 children with newly diagnosed acute lymphoblastic leukemia to postremission regimens that included high-dose methotrexate and teniposide plus cytarabine. The doses of these drugs were based on body-surface area (in the conventional-therapy group) or the rates of clearance of the three medications in each patient (in the individualized-treatment group). In the individualized-treatment group, doses were increased in patients with rapid clearance and decreased in patients with very slow clearance. RESULTS Patients who received individualized doses had significantly fewer courses of treatment with systemic exposures below the target range than did patients who received conventional doses (P<0.001 for each medication). Among the patients with B-lineage leukemia, those who received individualized therapy had a significantly better outcome than those given conventional therapy (P=0.02); the mean (+/-SE) rates of continuous complete remission at five years were 76+/-6 percent and 66+/-7 percent, respectively. There was no significant difference between treatments for patients with T-lineage leukemia (P=0.54). In a proportional-hazards model, the time-dependent systemic exposure to methotrexate, but not to teniposide or cytarabine, was significantly related to the risk of early relapse in children with B-lineage leukemia. CONCLUSIONS Adjusting the dose of methotrexate to account for the patients ability to clear the drug can improve the outcome in children with B-lineage acute lymphoblastic leukemia.

Patient characteristics associated with high-risk methotrexate concentrations and toxicity.

PURPOSE Following high-dose methotrexate (HD-MTX) treatment, delayed MTX elimination is an important problem because it necessitates increased leucovorin rescue and additional hospitalization for hydration and urinary alkalinization. Our purpose was to identify factors associated with high-risk MTX plasma concentrations (defined by plasma concentration > or = 1.0 mumol/L at 42 hours from the start of MTX) and with toxicity. PATIENTS AND METHODS Variables associated with MTX concentrations and toxicity were assessed in 134 children treated with one to five courses of HD-MTX (900 to 3,700 mg/m2 intravenously [i.v.] over 24 hours for a total of 481 courses) for acute lymphoblastic leukemia (ALL). RESULTS High-risk MTX concentrations, toxicity (usually mild mucositis), and delay in resuming continuation chemotherapy occurred in 106 (22%), 123 (26%), and 66 (14%) of 481 courses, respectively. Using a mixed effects model for repeated measures, high-risk MTX concentrations were significantly associated with a higher MTX area-under-the-concentration-time curve (AUC), low urine pH, emesis, low MTX clearance, low urine output relative to intake, use of antiemetics during the MTX infusion, and concurrent intrathecal therapy (all p values < .01). Clinical toxicities and delay in resumption of continuation chemotherapy due to myelosuppression were more common in those with high 42-hour MTX concentrations, despite increased leucovorin rescue for all patients with high-risk MTX concentrations. However, with individualized rescue, no patient developed life-threatening toxicity. A more aggressive hydration and alkalinization regimen for subsequent courses reduced the frequency of high-risk MTX concentrations to 7% of courses (13 of 183) (P = .0001), and the frequency of toxicity decreased to 11% of courses (P = .0074). CONCLUSION This study identified several clinical variables that influence MTX disposition that, when modified, can reduce the frequency of high-risk MTX concentrations and toxicity.

Cerebrospinal fluid pharmacokinetics and penetration of continuous infusion topotecan in children with central nervous system tumors.

The purpose of this study was to describe the cerebrospinal fluid (CSF) penetration of topotecan in humans, to generate a pharmacokinetic model to simultaneously describe topotecan lactone and total concentrations in the plasma and CSF, and to characterize the CSF and plasma pharmacokinetics of topotecan administered as a continuous infusion (CI). Plasma and CSF samples were collected from 17 patients receiving 5.5 or 7.5 mg/m2 per day as a 24-h CI (5 patients, 7 courses), or 0.5 to 1.25 mg/m2 per day as a 72-h CI (12 patients, 12 courses). CSF samples were obtained from either a ventricular reservoir (VR) or a lumbar puncture (LP). Topotecan lactone and total (lactone plus hydroxy acid) concentrations were determined by HPLC and fluorescence detection. Using MAP-Bayesian modelling, a three-compartment model was fitted simultaneously to topotecan lactone and total concentrations in the plasma and CSF. The penetration of topotecan into the CSF was determined from the ratio of the CSF to the plasma area under the concentration-time curve. The median CSF ventricular lactone concentrations, obtained prior to the end of infusion (EOI), were 0.86, 1.4, 0.73, 5.3 and 4.6 ng/ml for patients receiving 0.5, 1.0, 1.25, 5.5, and 7.5 mg/m2 per day, respectively. EOI CSF lumbar lactone concentrations measured in three patients were 0.44, 1.1, and 1.7 ng/ml for topotecan doses of 1.0, 5.5, and 7.5 mg/m2 per day, respectively. In two patients receiving 1.25 mg/m2 per day, EOI CSF concentrations were obtained simultaneously from a VR and LP; the lumbar lactone concentrations were 30% and 49% lower than the ventricular concentrations. During a 24-h and a 72-h CI, the median CSF penetration of topotecan lactone was 0.29 (range 0.10 to 0.59) and 0.42 (range 0.11 to 0.86), respectively. A three-compartment model adequately described topotecan lactone and total concentrations in the plasma and CSF. Topotecan was therefore found to significantly penetrate into the CSF in humans. The pharmacokinetic model presented may be useful in the design of clinical studies of topotecan to treat CNS tumors.

2-Chlorodeoxyadenosine produces a high rate of complete hematologic remission in relapsed acute myeloid leukemia.

PURPOSE The purine analog 2-chlorodeoxyadenosine (2-CDA) was well tolerated and showed promising anti-leukemic activity in a phase I trial conducted at St Jude Childrens Research Hospital. To substantiate and extend this result, we performed a phase II trial in a representative group of children and young adults with relapsed acute leukemia. PATIENTS AND METHODS Twenty-four patients (median age, 11 years) with acute myeloid or lymphoid leukemia in first or later relapse (acute myeloid leukemia [AML], 17; acute lymphoid leukemia [ALL], seven) were given continuous infusion 2-CDA for 5 days at 8.9 mg/m2/d. Patients with residual blast cells 10 days after treatment received a second course of 2-CDA that was identical to the first. Detailed pharmacokinetic studies were performed on plasma collected from five patients. RESULTS Eight (47%) of the 17 patients with AML had complete hematologic remissions (four after the initial course of 2-CDA), and two (12%) had partial remissions, for a total response rate of 59%. Only one child with ALL achieved remission. Seven of the responding patients underwent allogeneic or autologous bone marrow transplantation, with six remaining free of leukemia for a median of 7 months (range, 1 to 11 months). The major form of drug-induced toxicity was hematologic, with severe neutropenia and thrombocytopenia (National Cancer Institute [NCI] grade 3 or 4) developing in 34 of the 36 courses of 2-CDA. In responding patients, the median times to recovery of neutrophil counts greater than 0.5 x 10(9)/L and platelet counts greater than 50 x 10(9)/L were 18 and 21 days, respectively. There were no deaths due to toxicity. The mean steady-state plasma concentration of 2-CDA was 34.6 nmol/L (range, 20 to 54 nmol/L). CONCLUSION 2-CDA given by prolonged continuous infusion has clinically significant activity against AML and merits further testing in multidrug regimens for this disease.

Clinical pharmacodynamics of continuous infusion topotecan in children: systemic exposure predicts hematologic toxicity.

PURPOSE Topotecan pharmacokinetics and pharmacodynamics were studied following a 72-hour continuous infusion in 20 children with cancer (median age, 8 years; range, 3.5 to 18). METHODS Serial plasma and urine samples were collected during the infusion and for up to 6 hours following the end of infusion. Topotecan (lactone) and total (lactone plus hydroxy acid) concentrations were determined by a sensitive and specific high-performance liquid chromatography (HPLC) assay with fluorescence detection. Using maximum a posteriori-Bayesian modeling, lactone and total plasma concentrations were described separately by a two-compartment model. Hematologic toxicity was expressed as the percent decrease in absolute neutrophil count (ANC) and platelet count. The relation between systemic exposure (SE) and hematologic toxicity was modeled using a sigmoid maximum-effect model. RESULTS Systemic clearance rates for lactone and total topotecan were (mean +/- SD) 18.5 +/- 7.0 and 6.5 +/- 2.4 L/h/m2, respectively. Urinary recovery of total topotecan was (mean +/- SD) 67.5% +/- 25.2% (n = 12 patients). SE (area under the concentration-time curve from zero to infinity [AUC] or steady-state plasma concentration [Cpss]) to either topotecan lactone or total topotecan was significantly correlated to hematologic toxicity (P < .05). Overall, patients with a higher SE to topotecan experienced greater hematologic toxicity. CONCLUSION These data demonstrate a relation between systemic exposure to topotecan and clinical effect (myelosuppression). Moreover, these data provide the basis for development of individualized topotecan administration schedules.

Pharmacokinetics of teniposide (VM26) and etoposide (VP16-213) in children with cancer

SummaryThe clinical pharmacokinetics of VM26 and VP16-213 were assessed in 15 children (median age 10 years) with acute leukemia, using a new high-performance liquid chromatography—electrochemical assay. Pharmacokinetic parameters were calculated by both model-dependent and compartment model-independent methods. These studies demonstrated substantial differences in the central volumes of distribution (VDc), steady-state volumes of distribution (VDss) and systemic clearances (Cls) of VM26 and VP16-213; with the VDc, VDss, and Cls all being smaller for VM26. Systemic clearances determined by model-independent methods were 5.2±1.0 ml/min/m2 (mean±SD) for VM26 and 17.8±11.2 ml/min/m2 for VP16-213. The major metabolites detected in serum and urine were the hydroxy acids. Low levels of the picro-lactone isomers were detected in some patients while the aglycones were not detected in the serum or urine of any patients.

Pharmacokinetics and toxicity of methotrexate in children with Down syndrome and acute lymphocytic leukemia

Children with Down syndrome and acute lymphocytic leukemia (ALL) have poor tolerance to antineoplastic drugs, including methotrexate (MTX). We evaluated MTX pharmacokinetics and toxicity in five patients with Down syndrome and ALL who had received multiple high doses of MTX (1 g/m2). Three control patients without Down syndrome were matched to each case according to sex, race, age, and initial leukocyte count. Median MTX plasma concentrations, measured 42 hours after infusion, were significantly higher in patients with Down syndrome versus control patients (average 0.47 vs 0.24 mumol/L, respectively, P = 0.03). When a 42-hour MTX concentration of 0.5 mumol/L was used to identify patients at risk for toxicity, more courses were considered at high risk for toxicity among patients with Down syndrome (31 of 62, 50%) than in control patients (13 of 214, 6.1%, P less than 0.0001). The average MTX clearance was 64.1 mL/min/m2 in Down syndrome vs an average control value of 80.6 mL/min/m2 (P = 0.13). Toxicity after each high-dose MTX course was graded according to standardized criteria. Grades 2 through 4 gastrointestinal toxicity and grades 3 and 4 hematologic toxicity occurred more frequently in the patients with Down syndrome (36% and 13.4% of courses, respectively) vs the control patients (3.6% and 0.9% respectively, P less than 0.0001 for both). This higher frequency of toxicity occurred despite higher doses and prolonged duration of leucovorin given to all patients with Down syndrome. We conclude that altered MTX pharmacokinetics may contribute to the higher incidence of MTX-induced toxicity seen in patients with Down syndrome.

Etoposide pharmacokinetics in patients with normal and abnormal organ function.

Precise guidelines for dose modification of etoposide in patients with hepatic dysfunction have not been determined. Etoposide pharmacokinetics were determined in 17 patients. Nine patients had bilirubin less than or equal to 1 mg/dL and eight had bilirubin ranging from 1.9 to 23 mg/dL. Twelve patients received etoposide 100 mg/m2 days 1, 3, and 5, in combination with cisplatin 70 mg/m2 or iproplatin 225 mg/m2 on day 1. Five patients received only one dose of etoposide. Etoposide was measured using a published high pressure liquid chromatography (HPLC) method which also quantitates picro etoposide and its hydroxy acid. Systemic clearance, Vdss and t1/2 beta averaged (+/- SD) 21.4 (+/- 7.4) mL/min/m2, 10.7 (+/- 4.1) L/m2, and 8.1 (+/- 2.8) hours in the nine patients with bilirubin less than or equal to 1 mg/dL, and 22.4 (+/- 9.6) mL/min/m2, 13.6 (+/- 11.3) L/m2, and 8.4 (+/- 3.9) hours in the eight patients with bilirubin 1.9 to 23.0 mg/dL. Stepwise multiple linear regression analysis of liver and renal function tests and other patient-specific variables identified creatinine clearance as the strongest predictor of etoposide systemic clearance (r2 = 40.8). Serum albumin was identified as the next strongest predictor, improving the r2 to 57.3%. Cumulative biliary excretion of unchanged etoposide and glucuronide or sulfate conjugates over 48 hours accounted for less than 3% of the dose in six patients studied. Toxicity occurred in patients with normal and abnormal bilirubin and was unrelated to etoposide clearance. Patients with total bilirubin 1.9 to 23 mg/dL, but creatinine clearance greater than 30 mL/min/m2 had etoposide clearance within the range for patients with normal liver function (16.8 to 35 mL/min/m2). Although these patients did not have reduced etoposide clearance, the major routes of etoposide non-renal elimination remain to be clearly defined. Additional patients should be evaluated to establish more precise guidelines for dosing etoposide in patients with abnormal liver function.

Dextromethorphan and caffeine as probes for simultaneous determination of debrisoquin-oxidation and N-acetylation phenotypes in children

The feasibility and reliability of simultaneously determining debrisoquin oxidation and N‐acetylation phenotypes was assessed in children with use of two innocuous substrate probes given by mouth, 30 mg dextromethorphan (Pertussin ES) and 25 to 46 mg caffeine (Coca‐Cola beverage). Twenty‐six children and adolescents (aged 3 to 21 years) were studied three times, once with each substrate given alone and once with the two substrates given together. Urine was collected for 4 hours, and the molar urinary metabolic ratios for dextromethorphan:dextrorphan and for two caffeine metabolites (AFMU: 1X) were detennined by HPLC ultraviolet assays. The urinary metabolic ratios for both substrates were not significantly different when the substrates were given alone compared with when they were given together. There also was no difference in either the oxidation or acetylation phenotype assignments when the two substrates were given alone and when they were given together. No adverse effects were observed. We conclude that dextromethorphan and caffeine can be given together to simultaneously determine oxidation and acetylation phenotypes and can thereby provide an innocuous, noninvasive method for the assessment of polymorphic drug metabolism in various pediatric populations.

Pharmacokinetics of vincristine in children and adolescents with acute lymphocytic leukemia

We studied the pharmacokinetics of vincristine in children with acute lymphocytic leukemia by means of a specific high-performance liquid chromatographic assay with ultraviolet and electrochemical detection and a limited sampling strategy. Our objectives were to characterize the disposition of vincristine in pediatric patients, to determine clinical, demographic, or biochemical variables related to variability in vincristine pharmacokinetic parameters, and to assess the relationship between pharmacokinetic parameters and vincristine neurotoxicity. Plasma samples were collected at 5 and 30 minutes, and 1, 3, and 24 hours after a rapid intravenous injection during 3 minutes. Vincristine-induced neurotoxicity was retrospectively evaluated by chart review. Pharmacokinetic studies were completed for 64 doses in 54 children between 2 months and 18 years of age (median, 4.3 years), including 2-month-old monozygous twin girls. Vincristine clearance, estimated by Bayesian methods, was highly variable, with a mean (SD) clearance of 19.9 (14.9) ml/min per kilogram or 482 (342) ml/min per square meter. Mean clearance for all subjects was faster than in published studies of adults, which may be related in part to the greater specificity of the assay used in our study, as well as to age-related differences in drug disposition. Vincristine-associated neurotoxicity was frequent but mild and was not predicted by vincristine systemic exposure; however, neurotoxicity may have been underestimated. Clearance in one patient who received concomitant treatment with pentobarbital exceeded the 75th percentile for all patients, and four of five patients receiving concomitant histamine2 antagonists had clearances below the 25th percentile for all subjects, suggesting that drugs that induce or inhibit hepatic cytochrome P-450 enzymes may affect vincristine disposition. Further studies are needed to identify the factors responsible for interpatient variability in vincristine disposition and to develop improved dosing guidelines.