Phase II Study of Nelarabine (compound 506U78) in Children and Young Adults With Refractory T-Cell Malignancies: A Report From the Children’
http://www.100md.com
《临床肿瘤学》
the Texas Children’s Cancer Center, Baylor College of Medicine, Houston, TX
Dana Farber Cancer Institute and Children’s Hospital, Boston, MA
Children’s Hospital of Los Angeles, and Keck School of Medicine, University of Southern California, Los Angeles, CA
Children’s National Medical Center, Washington, DC
Children’s Oncology Group, Arcadia, CA, Bethesda, MD, and Gainesville, FL
Duke University Medical Center, Durham, NC
GlaxoSmithKline, Collegeville, PA
Hospital Sainte-Justine, Montreal, Canada
National Cancer Institute, Bethesda, MD
Tomorrow’s Children’s Institute, Hackensack, NJ
Vanderbilt Children’s Hospital, Nashville, TN
ABSTRACT
PURPOSE: Nelarabine (compound 506U78), a water soluble prodrug of 9-b-d-arabinofuranosylguanine, is converted to ara-GTP in T lymphoblasts. We sought to define the response rate of nelarabine in children and young adults with refractory or recurrent T-cell disease.
PATIENTS AND METHODS: We performed a phase II study with patients stratified as follows: stratum 1: 25% bone marrow blasts in first relapse; stratum 2: 25% bone marrow blasts in second relapse; stratum 3: positive CSF; stratum 4: extramedullary (non-CNS) relapse. The initial nelarabine dose was 1.2 g/m2 daily for 5 consecutive days every 3 weeks. There were two dose de-escalations due to neurotoxicity on this or other studies. The final dose was 650 mg/m2/d for strata 1 and two patients and 400 mg/m2/d for strata 3 and four patients.
RESULTS: We enrolled 121 patients (106 assessable for response) at the final dose levels. Complete plus partial response rates at the final dose levels were: 55% in stratum 1; 27% in stratum 2; 33% in stratum 3; and 14% in stratum 4. There were 31 episodes of grade 3 neurologic adverse events in 27 patients (18% of patients).
CONCLUSION: Nelarabine is active as a single agent in recurrent T-cell leukemia, with a response rate more than 50% in first bone marrow relapse. The most significant adverse events associated with nelarabine administration are neurologic. Further studies are planned to determine whether the addition of nelarabine to front-line therapy for T-cell leukemia in children will improve survival.
INTRODUCTION
Based on the observation that patients with purine nucleoside phosphorylase (PNP) deficiency have a severe T-cell immune deficiency resulting from the abnormal accumulation of deoxyguanosine triphosphate (dGTP) in T cells,1,2 deoxyguanosine and its analogs have been investigated as antileukemic drugs. Although deoxyguanine itself is cytotoxic to T cells,3 the deoxyguanosine derivative ara-G (9-b-d-arabinofuranosylguanine) is more toxic than deoxyguanosine to T lymphoblasts and is resistant to degradation by endogenous PNP.3 Once within cells, ara-G is phosphorylated by deoxycytidine kinase and deoxyguanosine kinase,4,5 and the subsequent accumulation of intracellular ara-GTP results in inhibition of DNA synthesis.3 Ara-G is markedly more toxic to T lymphoblasts than to blasts derived from other leukemic cell types, probably due to enhanced accumulation of ara-GTP in T cells compared with B cells.4,6,7 Furthermore, deoxyguanosine inhibits proliferation to a greater extent in T-lymphoblasts than in mature T cells.3 This differential activity has created considerable interest in the possible role of ara-G as an agent for the treatment of T-cell malignancies.
Nelarabine (compound 506U78; 2-amino-9-b-d-arabinofuranosyl-6-methoxy-9H-purine), a water-soluble prodrug of ara-G,8 is rapidly demethoxylated in vivo to ara-G by adenosine deaminase.8 After intravenous (IV) administration of nelarabine, essentially all of the dose is rapidly converted to ara-G, and the area under the concentration-time curve (AUC) for ara-G is nine-fold greater than that of nelarabine. Thus, nelarabine functions not as a cytotoxic agent, but as a prodrug whose intravenous administration provides excellent systemic exposure to the active agent, ara-G.
In a phase I study of nelarabine administered as a 1-hour infusion daily for 5 days in children and adults with refractory hematologic malignancies, a striking response rate was observed in patients with T-cell malignancies. Limited activity was also seen in patients with B-lineage disease. Neurologic toxicity including ataxia, confusion, and coma was dose limiting in adults at doses 60 mg/kg/d (1.8 g/m2/d).9
The primary objective of this phase II study was to better clarify the response rate to nelarabine administered as a 1-hour infusion daily for 5 days in patients 21 years of age at initial diagnosis, with first or subsequent relapse of T-cell acute lymphoblastic leukemia (T-ALL). Other objectives were to describe the response rate in patients with T-cell non-Hodgkin’s lymphoma (T-NHL) and in patients with CNS relapse in addition to bone marrow or extramedullary relapse.
PATIENTS AND METHODS
Consent
The protocol was reviewed by the institutional review boards at participating institutions. Informed consent/assent was obtained from all subjects according to Federal and institutional guidelines.
Eligibility
Patients younger than 21 years of age at the time of initial diagnosis with refractory or recurrent T-cell malignancies (acute lymphoblastic leukemia or non-Hodgkin’s lymphoma) were eligible. Patients with isolated CNS relapse were not eligible. Other eligibility criteria included: predicted life expectancy 8 weeks, Karnofsky performance status more than 50%; no severe uncontrolled infection; bilirubin less than 1.5 mg/dL; ALT less than 5x the upper limit of normal; creatinine normal for age, or creatinine clearance or glomerular filtration rate more than 60 mL/min/1.73 m2; no concurrent anticancer therapy (including radiation therapy); at least 6 weeks since administration of nitrosoureas or craniospinal or hemipelvic radiation; and recovery from toxicity of prior therapy. Pregnant or lactating women and patients with persistent pre-existing grade 2 or worse neurotoxicity were excluded.
Treatment Assignment
Patients were assigned to one of four strata depending on the site of relapse and number of prior relapses. Stratum 1 included patients who had T-ALL with 25% bone marrow blasts, with or without concomitant extramedullary relapse other than CNS, in first relapse or refractory to initial induction therapy. Stratum 2 included patients who had T-ALL with 25% bone marrow blasts, with or without concomitant extramedullary relapse other than CNS, with a second or subsequent bone marrow recurrence or a first recurrence refractory to at least one reinduction attempt. Stratum 3 included patients who had T-ALL or T-NHL with positive bone marrow and CSF (> 5% bone marrow blasts and cranial nerve involvement or leukemic blasts present on a cytocentrifuged sample of CSF). To gain baseline experience with neurologic toxicity before patients with CNS disease enrolled, accrual to stratum 3 began after approximately 10 patients were enrolled in strata 1 and 2. Stratum 4 included patients who had T-ALL or T-NHL with extramedullary relapse and less than 25% blasts in the bone marrow (excluding isolated CNS relapse). This stratum began accruing at the same time as stratum 3. There was no random assignment.
Treatment
Nelarabine was provided by the Cancer Therapy Evaluation Program of the National Cancer Institute (NCI). Based on the phase I study and ongoing adult studies, the initial dose selected was 1.2 g/m2 administered as a 1-hour intravenous infusion daily for 5 consecutive days every 3 weeks. After the first patient treated at this dose experienced a grade 4 neurologic adverse event, the study was amended to use a dose of 900 mg/m2 daily for 5 days. Subsequently, based on adverse events observed in concurrent trials of nelarabine in adults, the dose in this study was reduced further to 650 mg/m2. In addition, because it seemed likely that the nelarabine dose might be lower in future studies in which this agent would be combined with other cytotoxic anticancer drugs, the dose of nelarabine for patients in strata 3 and 4 (patients with CNS or extramedullary relapse) was reduced to 400 mg/m2 in order to obtain pilot data on response and adverse events at a lower dose.
Patients without CNS leukemia at study entry had no intrathecal chemotherapy during the first two courses of nelarabine. Subsequent triple intrathecal therapy (methotrexate, cytarabine, hydrocortisone) was administered at the treating physician’s discretion in standard age-adjusted doses. Patients with CNS leukemia at study entry (stratum-3 patients) received triple intrathecal therapy on day 7 after the start of nelarabine, weekly for 4 weeks total, on days 1 of weeks 6 and 9, then every 6 weeks for 12 weeks, then every 9 weeks.
Patient Evaluation
Patients had physical examinations daily during drug administration, weekly during course 1, then before each course. CBCs were obtained every other day during drug administration, then weekly. Electrolytes, blood urea nitrogen, and creatinine were measured weekly, and bilirubin, ALT, total protein, albumin, calcium, phosphorus, magnesium, uric acid, and urinalysis were obtained before each course. For patients with leukemia, bone marrow aspirates (BMA) were obtained at week 3 (day 21 of cycle 1) and week 6 (day 21 of cycle 2). If the week-6 BMA was the first to document a complete response, (eg, the week-3 BMA showed partial response (PR) or was hypoplastic), a BMA was also performed at week 9. Lumbar puncture with CSF cytology was obtained prestudy and was repeated with intrathecal therapy in stratum-3 patients and as clinically indicated in all others.
Statistics
For patients with leukemia, we defined a complete response (CR) as an M1 marrow (< 5% blasts) and a PR as an M2 marrow (< 25% blasts). To be considered a CR or PR, the bone marrow had to demonstrate adequate cellularity. No criteria for recovery of peripheral blood counts were used. For patients with extramedullary disease, we used standard radiographic and physical examination response criteria (CR: no evidence of disease; PR: > 50% decrease in the sum of the products of the maximum perpendicular diameters of all measurable lesions, no evidence of progression in any lesion, and no new lesions). Although the primary goal of the study was to determine the "early response" rate (ie, the response rate by day 21 of cycle 1), we categorized responses as "confirmed" if the CR or PR were documented by two evaluations performed 3 or more weeks apart. Toxicity was graded according to the NCI Common Toxicity Criteria version 2. Patients were followed for toxicity while on study; off study criteria included death; progressive disease; development of unacceptable toxicity (which was then followed until resolution if possible); patient/family refusal of further therapy; patient lost to follow-up; change to different therapy (eg, BMT). Patients with significant neurologic toxicity attributed to nelarabine were removed from study and not rechallenged with drug.
This study had a two-stage design in which response rates were estimated after an initial accrual of 19 assessable patients each on stratum 1 (patients with leukemia in first relapse) and stratum 2 (patients with leukemia in second or later relapse). Accrual in both strata proceeded to the second stage because at least four patients achieved a CR or PR on each. In stage 2, an additional 14 patients were accrued, to bring the total accrual to 33 on both strata. Per the study design, the drug was to be declared "active" if at least eight of the 33 patients achieve an early PR or CR. With this design, the probability of concluding that the proportion of early responders is greater than 0.15 when the true proportion is at most 0.15, is equal to 0.096; and the probability of concluding that the proportion is greater than 0.15 when the true proportion is 0.35, is 0.904. When the true proportion is at most 0.15, the probability of early stopping of the trial is at least 60%, while the probability of early stopping when the true proportion is 0.35 is 6%.
Because of the dose de-escalations described above, accrual for purposes of defining the response rate was restarted from 0 for strata 1 and 2 when patients began to enter at the 650-mg/m2 dose level, and for strata 3 and 4 when patients began to enter at the 400-mg/m2 dose level. For strata 3 and 4, there were no formal statistical goals; the response rate was described based on accrual at the time the trial was closed to further enrollment.
Fisher’s exact test was used to compare response rates across strata, and neurological toxicities across dose levels.
RESULTS
This study was open to accrual from June 2, 1997 to July 19, 2002. A total of 153 patients (111 male) were registered on the study. The median age was 11.5 years (range, 0.6 to 21.7 years). Eighteen patients were not assessable for response because they were ineligible (12 patients—10 did not meet required prestudy laboratory criteria; one was improperly registered; one signed an incorrect consent form), did not receive drug (two patients), did not receive all doses due to toxicity (two patients), received the wrong dose (one patient), or had early death without disease progression (one patient). Fifteen were not assessable for toxicity because they were ineligible (12 patients), did not receive drug (two patients), or received the wrong dose (one patient). Five additional patients were partially assessable for toxicity because of early disease progression (three patients) or missed laboratory data (two patients) All patients who received at least one dose of nelarabine were considered assessable for neurologic toxicity.
Response
Responses from patients at all dose levels are shown in Table 1. For stratum 1 (leukemia in first relapse) patients at the 650-mg/m2 dose level, the response rate was 55% with 16 CRs (11 confirmed, five early) and two PRs (both confirmed). For stratum 2 (leukemia in second or greater relapse) patients at the 650-mg/m2 dose level, the response rate was 27% with seven CRs and one PR (all confirmed). For stratum 3 (leukemia in any relapse, with positive CNS) patients at the 400-mg/m2 dose level, the response rate was 33% with five CRs (three confirmed, two early) and two PRs, and for stratum 4 (extramedullary relapse) patients at the 400-mg/m2 dose level, the response rate was 14% with three PRs (two confirmed, one early). The primary reason for patients’ with an early response not being classified as a confirmed response was lack of a second bone marrow evaluation, not disease progression. There was a significant difference (P = .012) overall in response rates among the four strata. The difference in response rates at the 650-mg/m2 dose level between stratum 1 and 2 was significant, with P = .0228. There was no significant difference in response rates at the 400-mg/m2 dose level between stratum 3 and 4.
Although the study did not have an evaluation of the effect of nelarabine in reducing or clearing CNS disease as an objective, the schedule of nelarabine administration 1 week before the first intrathecal therapy provided some preliminary information. Eight of 22 patients who had positive CSF cytology before the study treatment and had lumbar punctures as scheduled, had negative CSF cytology at the time of the day-7 lumbar puncture, before the institution of intrathecal therapy.
Neurologic Toxicity
Neurologic toxicity is reported for all subjects who were enrolled and received at least one dose of drug, even if they were found to be ineligible or otherwise inevaluable at audit. Among 151 subjects who received at least one dose of nelarabine, there were 31 episodes of grade 3 neurologic adverse events in 27 patients (18% of patients; Table 2). Three patients had both grade 3 peripheral and grade 3 central neurologic adverse events. Not all neurologic adverse events were attributed to drug; for example, one patient had grade 3 hallucinations during an episode of documented sepsis; another had lower extremity weakness that was related to spinal cord metastases. The grade 3 neurologic adverse events are detailed in Table 3. There were 13 episodes of grade 3 peripheral neurologic adverse events. These events consisted of peripheral neuropathy manifested by weakness, numbness, and/or paresthesias, predominantly in the lower extremities. In at least two cases, the peripheral neuropathy appeared to be gradually reversible, though not transient. In other cases, the patients died of progressive disease before resolution of the neuropathy.
There were 18 episodes of grade 3 CNS adverse events. One patient with a history of a previous seizure died without regaining consciousness 1 day after entering status epilepticus on day 4 of nelarabine; the relationship of the seizures to nelarabine is unknown. Three other patients had seizures, one 22 days after nelarabine during a period of electrolyte abnormalities, one after receiving triple intrathecal therapy and with evidence of progressive CNS leukemia, and one 5 days after nelarabine and 2 days after triple intrathecal therapy. This patient was given anticonvulsants and did not have further seizures with nelarabine. The remaining episodes of grade 3 or 4 CNS adverse events consisted of somnolence, hallucinations, headache, and one episode of retrobulbar neuritis with onset 9 weeks after nelarabine. The hallucinations and retrobulbar neuritis were attributed to intercurrent illness or medications, and not to nelarabine.
At doses of 900 mg/m2, five (28%) of 18 patients had grade 3 or 4 neurologic adverse events. One patient at the 900 mg/m2 dose level developed grade 3 somnolence and grade 4 peripheral neuropathy with a Guillan-Barre–like ascending weakness approximately 2 weeks after the start of nelarabine; she had progressive leukemia and died after the family elected to discontinue mechanical ventilation. At doses of 650 mg/m2, 22 (17%) of 133 patients had grade 3 neurologic adverse events, including the patient who died after status epilepticus. The incidence of neurologic events did not appear to vary by stratum, prior CNS radiation therapy, or current CNS status, or to be related to concurrent medications, concurrent intrathecal therapy, or particular clinical or laboratory features, though the power of the study to detect such relationships is very low. Overall, neurologic adverse events did not vary significantly across the dose ranges used in the study. Although grade 3 CNS adverse events appeared to be less frequent at doses 650 mg/m2 than at 900 mg/m2 (13 of 133 v four of 18), the difference was not statistically significant (P = .1235)
Non-Neurologic Toxicity
Of 29 assessable patients on stratum 4 (patients with extramedullary relapse and < 25% blasts in the bone marrow) two had grade 4 neutropenia and two had grade 4 thrombocytopenia during course 1 of nelarabine. No other grade 4 hematologic toxicity was reported in this stratum. Hematologic toxicity is difficult to assess in the other strata because patients in those strata had leukemia with bone marrow involvement. Nonhematologic, non-neurologic adverse events were not a major problem. Table 4 summarizes grade 3 nonhematologic toxicities for all doses and strata.
DISCUSSION
This multi-institutional phase II study confirms the impressive activity of nelarabine as a single agent in T-cell lymphoblastic malignancies. In the previous phase I study, of 28 assessable patients with T-ALL, there were 14 CR and nine PR, with responses observed at all dose levels (5 mg/kg to 75 mg/kg).9 In the present study, the overall response rate for all patients at all dose levels was 33%, with an objective response (CR + PR) rate of 55% in patients with first relapse of T-ALL and 27% in patients with second or subsequent relapse at the 650-mg/m2 dose level. The difference in response rate between these two groups is statistically significant and may be related to the amount of prior therapy. The mechanism of acquired resistance to nelarabine is believed to be decreased expression of deoxyguanosine kinase, and to a lesser extent, of deoxycytidine kinase.10 Since decreased expression of deoxycytidine kinase may also be a factor in cytarabine resistance,11 it is reasonable to speculate that one reason for the better response rate in less heavily pretreated patients is a decreased likelihood of having leukemia clones selected for this characteristic. The lower response rate in stratum 2 is also not unexpected given that these patients were selected for very refractory disease.
Nelarabine appears to have at least some activity in clearing CSF cytology when administered as a single systemic agent. This is analogous to other antimetabolites like methotrexate and cytarabine, which can be given at high-doses systemically to treat CNS leukemia.12 The design of this study does not permit accurate estimation of a CNS response rate for nelarabine as a single agent. However, the cytologic conversion noted in eight patients suggests that nelarabine may play a role in the treatment or prevention of CNS leukemia.
Patients in stratum 4, who had less than 25% bone marrow involvement by tumor at the time of study entry, offer some insight into the relative lack of myelosuppression of nelarabine as a single agent. These patients experienced little serious hematologic toxicity, with only four of 29 assessable patients having grade 4 thrombocytopenia or neutropenia. This suggests that combining nelarabine with other cytotoxic agents in future studies may be possible without having to reduce drug doses to avoid excessive myelosuppression.
It is difficult to compare the response rate of single-agent nelarabine to that of other cytotoxic agents because there is little published data specific to T-ALL. Most studies of multiagent regimens in children with first relapse of ALL report overall second CR rates in the range of 75% to 90%.13-16 In studies that identify subsets of patients with T-ALL, Buchanan et al reported CRs in 81% of 27 children, and Giona et al reported CRs in 60% of 15 children.13,15 Thus, it seems that the response rate to nelarabine administered as a single agent in this study approaches that of multiagent chemotherapy in patients with T-ALL in first bone marrow relapse.
We observed significant neurologic adverse events during nelarabine therapy, which was not unexpected. In dose-finding studies in nonhuman primates, the predominant toxicities were neurologic, with tremors, weakness, and/or ataxia observed especially when nelarabine was administered while animals were under ketamine anesthesia.17 In the phase I study, dose-limiting neurotoxicities consisting of weakness, ataxia, confusion and coma were observed in three of four adults and one of 11 children treated at the 60-mg/kg (approximately 1.8 g/m2) dose level, and in two of 31 adults treated at 40 mg/kg (1.2 g/m2). Thirty percent of the adults treated at a dose of 40 mg/kg experienced reversible somnolence on day 6 to 7 after starting therapy with nelarabine. No neurotoxicity was observed in children treated at the 40-mg/kg level. The only child treated at the 75-mg/kg (2.25-g/m2) dose level experienced severe somnolence that resolved by day 10 and was followed by a generalized seizure on day 11 and subsequent ascending paralysis and coma from which the child had not recovered 10 weeks later when he died of progressive leukemia.9
Eighteen percent of patients in this study experienced serious ( grade 3) neurologic adverse events. Although it is difficult to know how many of these events were definitely related to nelarabine, the most common nelarabine-related neurologic adverse events seem to be CNS toxicities, primarily reversible somnolence; a peripheral neuropathy that may be severe and may occur even with the first course of nelarabine therapy; and a peripheral neuropathy that may be related to the cumulative drug dose and may be gradually reversible, though not transient. Because of the relatively small numbers of patients who had neurologic adverse events, we cannot draw definitive conclusions about whether these toxicities are dose-related or reversible.
The mechanism of nelarabine-induced neurologic toxicity is unknown, but numerous other anticancer agents, especially antimetabolites, are associated with similar toxicities. Intrathecal chemotherapy is well known to produce neurologic toxicity, including paralysis.18 Of note, cytarabine, especially when administered in high doses, has also been associated with severe peripheral neuropathy, including paralysis.19-21 Interestingly, one patient who received nelarabine on this study without significant neurologic adverse events experienced an ascending peripheral neuropathy approximately 6 weeks later after receiving high-dose cytarabine in a bone marrow transplant preparative regimen. (For purposes of completeness, this patient is included in the analysis of neurologic adverse events associated with the nelarabine study.) Another patient who received high-dose cytarabine in the same transplant regimen but who had never received nelarabine experienced similar severe neurologic adverse events.22 Thus this toxicity is not unique to nelarabine, but may be associated with purine nucleoside toxicity in general. Furthermore, purine nucleoside phosphorylase deficiency is associated with spasticity and other neurologic abnormalities in some patients, perhaps due to CNS consequences of abnormal purine metabolism.23,24 In addition, brain and nerve tissue express high levels of deoxyguanosine kinase activity, potentially leading to high concentrations of cytotoxic ara-GTP in these tissues.5 These mechanisms are speculative, and further research should be performed to elucidate the risks factors, cause, and management of nelarabine-related neurologic toxicity.
The results of our study indicate that nelarabine is remarkably active as a single agent in recurrent T-cell leukemia in children. The most significant adverse events associated with nelarabine administration are neurologic, which occur occasionally and can be severe. Further studies are planned to determine whether the addition of nelarabine to front-line therapy for T-cell leukemia in children will improve outcome.
Authors' Disclosures of Potential Conflicts of Interest
The following authors or their immediate family members have indicated a financial interest. No conflict exists for drugs or devices used in a study if they are not being evaluated as part of the investigation. Employment: Tom A. Lampkin, GlaxoSmithKline. Consultant/Advisory Role: Michael B. Harris, Advanced Viral. Stock Ownership: Tom A. Lampkin, GlaxoSmithKline, Michael B. Harris, Amgen, Pfizer. Honoraria: Joanne Kurtzberg, GlaxoSmithKline. Research Funding: Joanne Kurtzberg, BW/GlaxoSmithKline. For a detailed description of these categories, or for more information about ASCO’s conflict of interest policy, please refer to the Author Disclosure Declaration and Disclosures of Potential Conflicts of Interest found in Information for Contributors in the front of each issue.
Acknowledgment
We thank Ms Susan Milligan for her excellent data management support.
NOTES
This study originated as a Pediatric Oncology Group/Children’s Cancer Group intergroup study.
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
Giblett E, Ammann A, Wara D, et al: Nucleoside-phosphorylase deficiency in a child with severely defective T-cell immunity and normal B-cell immunity. Lancet 1 : 1010 -1013, 1975
Cohen A, Gudas L, Ammann A, et al: Deoxyguanosine triphosphate, a possible toxic metabolite in immunodeficiency associated with purine nuclotide phosphorylase deficiency. J Clin Invest 61 : 1405 -1410, 1978
Cohen A, Lee J, Gelfand E: Selective cytotoxicity of deoxyguanosine and arabinosyl guanine for T-leukemic cells. Blood 61 : 660 -666, 1983
Ullman B, Martin Jr: D: Specific cytotoxicity of arabinosylguanine toward cultured T lymphoblasts. J Clin Invest 74 : 951 -955, 1984
Rodriguez CO Jr, Mitchell BS, Ayres M, et al: Arabinosylguanine is phosphorylated by both cytoplasmic deoxycytidine kinase and mitochondrial deoxyguanosine kinase. Cancer Res 62 : 3100 -3105, 2002
Shewach D, Daddona P, Ashcraft E, et al: Metabolism and selective cytotoxicity of 9-b-D-arabinofuranosylguanine in human lymphoblasts. Cancer Res 45 : 1008 -1014, 1985
Shewach D, Mitchell B: Differential metabolism of 9-b-D-arabinofuranosylguanine in human leukemic cells. Cancer Res 49 : 6498 -6502, 1989
Lambe C, Averett D, Paff M, et al: 2-Amino-6-methoxypurine arabinoside: An agent for T-cell malignancies. Cancer Res 55 : 3352 -3356, 1995
Kurtzberg J, Keating M, Plunkett W, et al: Compound 506 (2-amino-6-methoxypurine arabinoside) is active against T-cell malignancies: Preliminary results of an ongoing phase I trial. Proc Am Soc Clin Oncol 15 : 27 , 1996 (abstr 2022)
Lotfi K, Mansson E, Peterson C, et al: Low level of mitochondrial deoxyguanosine kinase is the dominant factor in acquired resistance to 9-beta-D-arabinofuranosylguanine cytotoxicity. Biochem Biophys Res Commun 293 : 1489 -1496, 2002
Kakihara T, Fukuda T, Tanaka A, et al: Expression of deoxycytidine kinase (dCK) gene in leukemic cells in childhood: Decreased expression of dCK gene in relapsed leukemia. Leuk Lymphoma 31 : 405 -409, 1998
Blaney S, Poplack D: Neoplastic meningitis: Diagnosis and treatment considerations. Med Oncol 17 : 151 -162, 2000
Buchanan GR, Rivera GK, Boyett JM, et al: Reinduction therapy in 297 children with acute lymphoblastic leukemia in first bone marrow relapse: A Pediatric Oncology Group Study. Blood 72 : 1286 -1292, 1988
Henze G, Fengler R, Hartmann R, et al: Six-year experience with a comprehensive approach to the treatment of recurrent childhood acute lymphoblastic leukemia (ALL-REZ BFM 85): A relapse study of the BFM group. Blood 78 : 1166 -1172, 1991
Giona F, Testi AM, Rondelli R, et al: ALL R-87 protocol in the treatment of children with acute lymphoblastic leukaemia in early bone marrow relapse. Br J Haematol 99 : 671 -677, 1997
Lawson SE, Harrison G, Richards S, et al: The UK experience in treating relapsed childhood acute lymphoblastic leukaemia: A report on the medical research council UKALLR1 study. Br J Haematol 108 : 531 -543, 2000
Nelarabine Investigator's Brochure. 1996
Hirschfeld S, Flowers C, Smith M: Toxicity from intrathecal therapy in pediatric hematologic neoplasms: Is the incidence increasing. Proc Am Soc Clin Oncol 17 : 539A , 1998 (abstr 2069)
Baker W, Royer G, Weiss R: Cytarabine and neurological toxicity. J Clin Oncol 9 : 679 -693, 1991
Kornblau SM, Cortes-Franco J, Estey E: Neurotoxicity associated with fludarabine and cytosine arabinoside chemotherapy for acute leukemia and myelodysplasia. Leukemia 7 : 378 -383, 1993
Scherokman B, Filling-Katz M, Tell D: Brachial plexus neuropathy following high-dose cytarabine in acute monoblastic leukemia. Cancer Treat Rep 69 : 1005 -1006, 1985
Rodriguez V, Kuehnle I, Heslop H, et al: Guillain-Barrre syndrome ofter allogeneic hematopoietic stem cell transplantation. Bone Marrow Transpl 29 : 515 -517, 2002
Rijksen G, Kuis W, Wadman SK, et al: A new case of purine nucleoside phosphorylase deficiency: Enzymologic, clinical, and immunologic characteristics. Pediatr Res 21 : 137 -141, 1987
Stoop JW, Zegers BJ, Hendrickx GF, et al: Purine nucleoside phosphorylase deficiency associated with selective cellular immunodeficiency. N Engl J Med 296 : 651 -655, 1977(Stacey L. Berg, Susan M. )
Dana Farber Cancer Institute and Children’s Hospital, Boston, MA
Children’s Hospital of Los Angeles, and Keck School of Medicine, University of Southern California, Los Angeles, CA
Children’s National Medical Center, Washington, DC
Children’s Oncology Group, Arcadia, CA, Bethesda, MD, and Gainesville, FL
Duke University Medical Center, Durham, NC
GlaxoSmithKline, Collegeville, PA
Hospital Sainte-Justine, Montreal, Canada
National Cancer Institute, Bethesda, MD
Tomorrow’s Children’s Institute, Hackensack, NJ
Vanderbilt Children’s Hospital, Nashville, TN
ABSTRACT
PURPOSE: Nelarabine (compound 506U78), a water soluble prodrug of 9-b-d-arabinofuranosylguanine, is converted to ara-GTP in T lymphoblasts. We sought to define the response rate of nelarabine in children and young adults with refractory or recurrent T-cell disease.
PATIENTS AND METHODS: We performed a phase II study with patients stratified as follows: stratum 1: 25% bone marrow blasts in first relapse; stratum 2: 25% bone marrow blasts in second relapse; stratum 3: positive CSF; stratum 4: extramedullary (non-CNS) relapse. The initial nelarabine dose was 1.2 g/m2 daily for 5 consecutive days every 3 weeks. There were two dose de-escalations due to neurotoxicity on this or other studies. The final dose was 650 mg/m2/d for strata 1 and two patients and 400 mg/m2/d for strata 3 and four patients.
RESULTS: We enrolled 121 patients (106 assessable for response) at the final dose levels. Complete plus partial response rates at the final dose levels were: 55% in stratum 1; 27% in stratum 2; 33% in stratum 3; and 14% in stratum 4. There were 31 episodes of grade 3 neurologic adverse events in 27 patients (18% of patients).
CONCLUSION: Nelarabine is active as a single agent in recurrent T-cell leukemia, with a response rate more than 50% in first bone marrow relapse. The most significant adverse events associated with nelarabine administration are neurologic. Further studies are planned to determine whether the addition of nelarabine to front-line therapy for T-cell leukemia in children will improve survival.
INTRODUCTION
Based on the observation that patients with purine nucleoside phosphorylase (PNP) deficiency have a severe T-cell immune deficiency resulting from the abnormal accumulation of deoxyguanosine triphosphate (dGTP) in T cells,1,2 deoxyguanosine and its analogs have been investigated as antileukemic drugs. Although deoxyguanine itself is cytotoxic to T cells,3 the deoxyguanosine derivative ara-G (9-b-d-arabinofuranosylguanine) is more toxic than deoxyguanosine to T lymphoblasts and is resistant to degradation by endogenous PNP.3 Once within cells, ara-G is phosphorylated by deoxycytidine kinase and deoxyguanosine kinase,4,5 and the subsequent accumulation of intracellular ara-GTP results in inhibition of DNA synthesis.3 Ara-G is markedly more toxic to T lymphoblasts than to blasts derived from other leukemic cell types, probably due to enhanced accumulation of ara-GTP in T cells compared with B cells.4,6,7 Furthermore, deoxyguanosine inhibits proliferation to a greater extent in T-lymphoblasts than in mature T cells.3 This differential activity has created considerable interest in the possible role of ara-G as an agent for the treatment of T-cell malignancies.
Nelarabine (compound 506U78; 2-amino-9-b-d-arabinofuranosyl-6-methoxy-9H-purine), a water-soluble prodrug of ara-G,8 is rapidly demethoxylated in vivo to ara-G by adenosine deaminase.8 After intravenous (IV) administration of nelarabine, essentially all of the dose is rapidly converted to ara-G, and the area under the concentration-time curve (AUC) for ara-G is nine-fold greater than that of nelarabine. Thus, nelarabine functions not as a cytotoxic agent, but as a prodrug whose intravenous administration provides excellent systemic exposure to the active agent, ara-G.
In a phase I study of nelarabine administered as a 1-hour infusion daily for 5 days in children and adults with refractory hematologic malignancies, a striking response rate was observed in patients with T-cell malignancies. Limited activity was also seen in patients with B-lineage disease. Neurologic toxicity including ataxia, confusion, and coma was dose limiting in adults at doses 60 mg/kg/d (1.8 g/m2/d).9
The primary objective of this phase II study was to better clarify the response rate to nelarabine administered as a 1-hour infusion daily for 5 days in patients 21 years of age at initial diagnosis, with first or subsequent relapse of T-cell acute lymphoblastic leukemia (T-ALL). Other objectives were to describe the response rate in patients with T-cell non-Hodgkin’s lymphoma (T-NHL) and in patients with CNS relapse in addition to bone marrow or extramedullary relapse.
PATIENTS AND METHODS
Consent
The protocol was reviewed by the institutional review boards at participating institutions. Informed consent/assent was obtained from all subjects according to Federal and institutional guidelines.
Eligibility
Patients younger than 21 years of age at the time of initial diagnosis with refractory or recurrent T-cell malignancies (acute lymphoblastic leukemia or non-Hodgkin’s lymphoma) were eligible. Patients with isolated CNS relapse were not eligible. Other eligibility criteria included: predicted life expectancy 8 weeks, Karnofsky performance status more than 50%; no severe uncontrolled infection; bilirubin less than 1.5 mg/dL; ALT less than 5x the upper limit of normal; creatinine normal for age, or creatinine clearance or glomerular filtration rate more than 60 mL/min/1.73 m2; no concurrent anticancer therapy (including radiation therapy); at least 6 weeks since administration of nitrosoureas or craniospinal or hemipelvic radiation; and recovery from toxicity of prior therapy. Pregnant or lactating women and patients with persistent pre-existing grade 2 or worse neurotoxicity were excluded.
Treatment Assignment
Patients were assigned to one of four strata depending on the site of relapse and number of prior relapses. Stratum 1 included patients who had T-ALL with 25% bone marrow blasts, with or without concomitant extramedullary relapse other than CNS, in first relapse or refractory to initial induction therapy. Stratum 2 included patients who had T-ALL with 25% bone marrow blasts, with or without concomitant extramedullary relapse other than CNS, with a second or subsequent bone marrow recurrence or a first recurrence refractory to at least one reinduction attempt. Stratum 3 included patients who had T-ALL or T-NHL with positive bone marrow and CSF (> 5% bone marrow blasts and cranial nerve involvement or leukemic blasts present on a cytocentrifuged sample of CSF). To gain baseline experience with neurologic toxicity before patients with CNS disease enrolled, accrual to stratum 3 began after approximately 10 patients were enrolled in strata 1 and 2. Stratum 4 included patients who had T-ALL or T-NHL with extramedullary relapse and less than 25% blasts in the bone marrow (excluding isolated CNS relapse). This stratum began accruing at the same time as stratum 3. There was no random assignment.
Treatment
Nelarabine was provided by the Cancer Therapy Evaluation Program of the National Cancer Institute (NCI). Based on the phase I study and ongoing adult studies, the initial dose selected was 1.2 g/m2 administered as a 1-hour intravenous infusion daily for 5 consecutive days every 3 weeks. After the first patient treated at this dose experienced a grade 4 neurologic adverse event, the study was amended to use a dose of 900 mg/m2 daily for 5 days. Subsequently, based on adverse events observed in concurrent trials of nelarabine in adults, the dose in this study was reduced further to 650 mg/m2. In addition, because it seemed likely that the nelarabine dose might be lower in future studies in which this agent would be combined with other cytotoxic anticancer drugs, the dose of nelarabine for patients in strata 3 and 4 (patients with CNS or extramedullary relapse) was reduced to 400 mg/m2 in order to obtain pilot data on response and adverse events at a lower dose.
Patients without CNS leukemia at study entry had no intrathecal chemotherapy during the first two courses of nelarabine. Subsequent triple intrathecal therapy (methotrexate, cytarabine, hydrocortisone) was administered at the treating physician’s discretion in standard age-adjusted doses. Patients with CNS leukemia at study entry (stratum-3 patients) received triple intrathecal therapy on day 7 after the start of nelarabine, weekly for 4 weeks total, on days 1 of weeks 6 and 9, then every 6 weeks for 12 weeks, then every 9 weeks.
Patient Evaluation
Patients had physical examinations daily during drug administration, weekly during course 1, then before each course. CBCs were obtained every other day during drug administration, then weekly. Electrolytes, blood urea nitrogen, and creatinine were measured weekly, and bilirubin, ALT, total protein, albumin, calcium, phosphorus, magnesium, uric acid, and urinalysis were obtained before each course. For patients with leukemia, bone marrow aspirates (BMA) were obtained at week 3 (day 21 of cycle 1) and week 6 (day 21 of cycle 2). If the week-6 BMA was the first to document a complete response, (eg, the week-3 BMA showed partial response (PR) or was hypoplastic), a BMA was also performed at week 9. Lumbar puncture with CSF cytology was obtained prestudy and was repeated with intrathecal therapy in stratum-3 patients and as clinically indicated in all others.
Statistics
For patients with leukemia, we defined a complete response (CR) as an M1 marrow (< 5% blasts) and a PR as an M2 marrow (< 25% blasts). To be considered a CR or PR, the bone marrow had to demonstrate adequate cellularity. No criteria for recovery of peripheral blood counts were used. For patients with extramedullary disease, we used standard radiographic and physical examination response criteria (CR: no evidence of disease; PR: > 50% decrease in the sum of the products of the maximum perpendicular diameters of all measurable lesions, no evidence of progression in any lesion, and no new lesions). Although the primary goal of the study was to determine the "early response" rate (ie, the response rate by day 21 of cycle 1), we categorized responses as "confirmed" if the CR or PR were documented by two evaluations performed 3 or more weeks apart. Toxicity was graded according to the NCI Common Toxicity Criteria version 2. Patients were followed for toxicity while on study; off study criteria included death; progressive disease; development of unacceptable toxicity (which was then followed until resolution if possible); patient/family refusal of further therapy; patient lost to follow-up; change to different therapy (eg, BMT). Patients with significant neurologic toxicity attributed to nelarabine were removed from study and not rechallenged with drug.
This study had a two-stage design in which response rates were estimated after an initial accrual of 19 assessable patients each on stratum 1 (patients with leukemia in first relapse) and stratum 2 (patients with leukemia in second or later relapse). Accrual in both strata proceeded to the second stage because at least four patients achieved a CR or PR on each. In stage 2, an additional 14 patients were accrued, to bring the total accrual to 33 on both strata. Per the study design, the drug was to be declared "active" if at least eight of the 33 patients achieve an early PR or CR. With this design, the probability of concluding that the proportion of early responders is greater than 0.15 when the true proportion is at most 0.15, is equal to 0.096; and the probability of concluding that the proportion is greater than 0.15 when the true proportion is 0.35, is 0.904. When the true proportion is at most 0.15, the probability of early stopping of the trial is at least 60%, while the probability of early stopping when the true proportion is 0.35 is 6%.
Because of the dose de-escalations described above, accrual for purposes of defining the response rate was restarted from 0 for strata 1 and 2 when patients began to enter at the 650-mg/m2 dose level, and for strata 3 and 4 when patients began to enter at the 400-mg/m2 dose level. For strata 3 and 4, there were no formal statistical goals; the response rate was described based on accrual at the time the trial was closed to further enrollment.
Fisher’s exact test was used to compare response rates across strata, and neurological toxicities across dose levels.
RESULTS
This study was open to accrual from June 2, 1997 to July 19, 2002. A total of 153 patients (111 male) were registered on the study. The median age was 11.5 years (range, 0.6 to 21.7 years). Eighteen patients were not assessable for response because they were ineligible (12 patients—10 did not meet required prestudy laboratory criteria; one was improperly registered; one signed an incorrect consent form), did not receive drug (two patients), did not receive all doses due to toxicity (two patients), received the wrong dose (one patient), or had early death without disease progression (one patient). Fifteen were not assessable for toxicity because they were ineligible (12 patients), did not receive drug (two patients), or received the wrong dose (one patient). Five additional patients were partially assessable for toxicity because of early disease progression (three patients) or missed laboratory data (two patients) All patients who received at least one dose of nelarabine were considered assessable for neurologic toxicity.
Response
Responses from patients at all dose levels are shown in Table 1. For stratum 1 (leukemia in first relapse) patients at the 650-mg/m2 dose level, the response rate was 55% with 16 CRs (11 confirmed, five early) and two PRs (both confirmed). For stratum 2 (leukemia in second or greater relapse) patients at the 650-mg/m2 dose level, the response rate was 27% with seven CRs and one PR (all confirmed). For stratum 3 (leukemia in any relapse, with positive CNS) patients at the 400-mg/m2 dose level, the response rate was 33% with five CRs (three confirmed, two early) and two PRs, and for stratum 4 (extramedullary relapse) patients at the 400-mg/m2 dose level, the response rate was 14% with three PRs (two confirmed, one early). The primary reason for patients’ with an early response not being classified as a confirmed response was lack of a second bone marrow evaluation, not disease progression. There was a significant difference (P = .012) overall in response rates among the four strata. The difference in response rates at the 650-mg/m2 dose level between stratum 1 and 2 was significant, with P = .0228. There was no significant difference in response rates at the 400-mg/m2 dose level between stratum 3 and 4.
Although the study did not have an evaluation of the effect of nelarabine in reducing or clearing CNS disease as an objective, the schedule of nelarabine administration 1 week before the first intrathecal therapy provided some preliminary information. Eight of 22 patients who had positive CSF cytology before the study treatment and had lumbar punctures as scheduled, had negative CSF cytology at the time of the day-7 lumbar puncture, before the institution of intrathecal therapy.
Neurologic Toxicity
Neurologic toxicity is reported for all subjects who were enrolled and received at least one dose of drug, even if they were found to be ineligible or otherwise inevaluable at audit. Among 151 subjects who received at least one dose of nelarabine, there were 31 episodes of grade 3 neurologic adverse events in 27 patients (18% of patients; Table 2). Three patients had both grade 3 peripheral and grade 3 central neurologic adverse events. Not all neurologic adverse events were attributed to drug; for example, one patient had grade 3 hallucinations during an episode of documented sepsis; another had lower extremity weakness that was related to spinal cord metastases. The grade 3 neurologic adverse events are detailed in Table 3. There were 13 episodes of grade 3 peripheral neurologic adverse events. These events consisted of peripheral neuropathy manifested by weakness, numbness, and/or paresthesias, predominantly in the lower extremities. In at least two cases, the peripheral neuropathy appeared to be gradually reversible, though not transient. In other cases, the patients died of progressive disease before resolution of the neuropathy.
There were 18 episodes of grade 3 CNS adverse events. One patient with a history of a previous seizure died without regaining consciousness 1 day after entering status epilepticus on day 4 of nelarabine; the relationship of the seizures to nelarabine is unknown. Three other patients had seizures, one 22 days after nelarabine during a period of electrolyte abnormalities, one after receiving triple intrathecal therapy and with evidence of progressive CNS leukemia, and one 5 days after nelarabine and 2 days after triple intrathecal therapy. This patient was given anticonvulsants and did not have further seizures with nelarabine. The remaining episodes of grade 3 or 4 CNS adverse events consisted of somnolence, hallucinations, headache, and one episode of retrobulbar neuritis with onset 9 weeks after nelarabine. The hallucinations and retrobulbar neuritis were attributed to intercurrent illness or medications, and not to nelarabine.
At doses of 900 mg/m2, five (28%) of 18 patients had grade 3 or 4 neurologic adverse events. One patient at the 900 mg/m2 dose level developed grade 3 somnolence and grade 4 peripheral neuropathy with a Guillan-Barre–like ascending weakness approximately 2 weeks after the start of nelarabine; she had progressive leukemia and died after the family elected to discontinue mechanical ventilation. At doses of 650 mg/m2, 22 (17%) of 133 patients had grade 3 neurologic adverse events, including the patient who died after status epilepticus. The incidence of neurologic events did not appear to vary by stratum, prior CNS radiation therapy, or current CNS status, or to be related to concurrent medications, concurrent intrathecal therapy, or particular clinical or laboratory features, though the power of the study to detect such relationships is very low. Overall, neurologic adverse events did not vary significantly across the dose ranges used in the study. Although grade 3 CNS adverse events appeared to be less frequent at doses 650 mg/m2 than at 900 mg/m2 (13 of 133 v four of 18), the difference was not statistically significant (P = .1235)
Non-Neurologic Toxicity
Of 29 assessable patients on stratum 4 (patients with extramedullary relapse and < 25% blasts in the bone marrow) two had grade 4 neutropenia and two had grade 4 thrombocytopenia during course 1 of nelarabine. No other grade 4 hematologic toxicity was reported in this stratum. Hematologic toxicity is difficult to assess in the other strata because patients in those strata had leukemia with bone marrow involvement. Nonhematologic, non-neurologic adverse events were not a major problem. Table 4 summarizes grade 3 nonhematologic toxicities for all doses and strata.
DISCUSSION
This multi-institutional phase II study confirms the impressive activity of nelarabine as a single agent in T-cell lymphoblastic malignancies. In the previous phase I study, of 28 assessable patients with T-ALL, there were 14 CR and nine PR, with responses observed at all dose levels (5 mg/kg to 75 mg/kg).9 In the present study, the overall response rate for all patients at all dose levels was 33%, with an objective response (CR + PR) rate of 55% in patients with first relapse of T-ALL and 27% in patients with second or subsequent relapse at the 650-mg/m2 dose level. The difference in response rate between these two groups is statistically significant and may be related to the amount of prior therapy. The mechanism of acquired resistance to nelarabine is believed to be decreased expression of deoxyguanosine kinase, and to a lesser extent, of deoxycytidine kinase.10 Since decreased expression of deoxycytidine kinase may also be a factor in cytarabine resistance,11 it is reasonable to speculate that one reason for the better response rate in less heavily pretreated patients is a decreased likelihood of having leukemia clones selected for this characteristic. The lower response rate in stratum 2 is also not unexpected given that these patients were selected for very refractory disease.
Nelarabine appears to have at least some activity in clearing CSF cytology when administered as a single systemic agent. This is analogous to other antimetabolites like methotrexate and cytarabine, which can be given at high-doses systemically to treat CNS leukemia.12 The design of this study does not permit accurate estimation of a CNS response rate for nelarabine as a single agent. However, the cytologic conversion noted in eight patients suggests that nelarabine may play a role in the treatment or prevention of CNS leukemia.
Patients in stratum 4, who had less than 25% bone marrow involvement by tumor at the time of study entry, offer some insight into the relative lack of myelosuppression of nelarabine as a single agent. These patients experienced little serious hematologic toxicity, with only four of 29 assessable patients having grade 4 thrombocytopenia or neutropenia. This suggests that combining nelarabine with other cytotoxic agents in future studies may be possible without having to reduce drug doses to avoid excessive myelosuppression.
It is difficult to compare the response rate of single-agent nelarabine to that of other cytotoxic agents because there is little published data specific to T-ALL. Most studies of multiagent regimens in children with first relapse of ALL report overall second CR rates in the range of 75% to 90%.13-16 In studies that identify subsets of patients with T-ALL, Buchanan et al reported CRs in 81% of 27 children, and Giona et al reported CRs in 60% of 15 children.13,15 Thus, it seems that the response rate to nelarabine administered as a single agent in this study approaches that of multiagent chemotherapy in patients with T-ALL in first bone marrow relapse.
We observed significant neurologic adverse events during nelarabine therapy, which was not unexpected. In dose-finding studies in nonhuman primates, the predominant toxicities were neurologic, with tremors, weakness, and/or ataxia observed especially when nelarabine was administered while animals were under ketamine anesthesia.17 In the phase I study, dose-limiting neurotoxicities consisting of weakness, ataxia, confusion and coma were observed in three of four adults and one of 11 children treated at the 60-mg/kg (approximately 1.8 g/m2) dose level, and in two of 31 adults treated at 40 mg/kg (1.2 g/m2). Thirty percent of the adults treated at a dose of 40 mg/kg experienced reversible somnolence on day 6 to 7 after starting therapy with nelarabine. No neurotoxicity was observed in children treated at the 40-mg/kg level. The only child treated at the 75-mg/kg (2.25-g/m2) dose level experienced severe somnolence that resolved by day 10 and was followed by a generalized seizure on day 11 and subsequent ascending paralysis and coma from which the child had not recovered 10 weeks later when he died of progressive leukemia.9
Eighteen percent of patients in this study experienced serious ( grade 3) neurologic adverse events. Although it is difficult to know how many of these events were definitely related to nelarabine, the most common nelarabine-related neurologic adverse events seem to be CNS toxicities, primarily reversible somnolence; a peripheral neuropathy that may be severe and may occur even with the first course of nelarabine therapy; and a peripheral neuropathy that may be related to the cumulative drug dose and may be gradually reversible, though not transient. Because of the relatively small numbers of patients who had neurologic adverse events, we cannot draw definitive conclusions about whether these toxicities are dose-related or reversible.
The mechanism of nelarabine-induced neurologic toxicity is unknown, but numerous other anticancer agents, especially antimetabolites, are associated with similar toxicities. Intrathecal chemotherapy is well known to produce neurologic toxicity, including paralysis.18 Of note, cytarabine, especially when administered in high doses, has also been associated with severe peripheral neuropathy, including paralysis.19-21 Interestingly, one patient who received nelarabine on this study without significant neurologic adverse events experienced an ascending peripheral neuropathy approximately 6 weeks later after receiving high-dose cytarabine in a bone marrow transplant preparative regimen. (For purposes of completeness, this patient is included in the analysis of neurologic adverse events associated with the nelarabine study.) Another patient who received high-dose cytarabine in the same transplant regimen but who had never received nelarabine experienced similar severe neurologic adverse events.22 Thus this toxicity is not unique to nelarabine, but may be associated with purine nucleoside toxicity in general. Furthermore, purine nucleoside phosphorylase deficiency is associated with spasticity and other neurologic abnormalities in some patients, perhaps due to CNS consequences of abnormal purine metabolism.23,24 In addition, brain and nerve tissue express high levels of deoxyguanosine kinase activity, potentially leading to high concentrations of cytotoxic ara-GTP in these tissues.5 These mechanisms are speculative, and further research should be performed to elucidate the risks factors, cause, and management of nelarabine-related neurologic toxicity.
The results of our study indicate that nelarabine is remarkably active as a single agent in recurrent T-cell leukemia in children. The most significant adverse events associated with nelarabine administration are neurologic, which occur occasionally and can be severe. Further studies are planned to determine whether the addition of nelarabine to front-line therapy for T-cell leukemia in children will improve outcome.
Authors' Disclosures of Potential Conflicts of Interest
The following authors or their immediate family members have indicated a financial interest. No conflict exists for drugs or devices used in a study if they are not being evaluated as part of the investigation. Employment: Tom A. Lampkin, GlaxoSmithKline. Consultant/Advisory Role: Michael B. Harris, Advanced Viral. Stock Ownership: Tom A. Lampkin, GlaxoSmithKline, Michael B. Harris, Amgen, Pfizer. Honoraria: Joanne Kurtzberg, GlaxoSmithKline. Research Funding: Joanne Kurtzberg, BW/GlaxoSmithKline. For a detailed description of these categories, or for more information about ASCO’s conflict of interest policy, please refer to the Author Disclosure Declaration and Disclosures of Potential Conflicts of Interest found in Information for Contributors in the front of each issue.
Acknowledgment
We thank Ms Susan Milligan for her excellent data management support.
NOTES
This study originated as a Pediatric Oncology Group/Children’s Cancer Group intergroup study.
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
Giblett E, Ammann A, Wara D, et al: Nucleoside-phosphorylase deficiency in a child with severely defective T-cell immunity and normal B-cell immunity. Lancet 1 : 1010 -1013, 1975
Cohen A, Gudas L, Ammann A, et al: Deoxyguanosine triphosphate, a possible toxic metabolite in immunodeficiency associated with purine nuclotide phosphorylase deficiency. J Clin Invest 61 : 1405 -1410, 1978
Cohen A, Lee J, Gelfand E: Selective cytotoxicity of deoxyguanosine and arabinosyl guanine for T-leukemic cells. Blood 61 : 660 -666, 1983
Ullman B, Martin Jr: D: Specific cytotoxicity of arabinosylguanine toward cultured T lymphoblasts. J Clin Invest 74 : 951 -955, 1984
Rodriguez CO Jr, Mitchell BS, Ayres M, et al: Arabinosylguanine is phosphorylated by both cytoplasmic deoxycytidine kinase and mitochondrial deoxyguanosine kinase. Cancer Res 62 : 3100 -3105, 2002
Shewach D, Daddona P, Ashcraft E, et al: Metabolism and selective cytotoxicity of 9-b-D-arabinofuranosylguanine in human lymphoblasts. Cancer Res 45 : 1008 -1014, 1985
Shewach D, Mitchell B: Differential metabolism of 9-b-D-arabinofuranosylguanine in human leukemic cells. Cancer Res 49 : 6498 -6502, 1989
Lambe C, Averett D, Paff M, et al: 2-Amino-6-methoxypurine arabinoside: An agent for T-cell malignancies. Cancer Res 55 : 3352 -3356, 1995
Kurtzberg J, Keating M, Plunkett W, et al: Compound 506 (2-amino-6-methoxypurine arabinoside) is active against T-cell malignancies: Preliminary results of an ongoing phase I trial. Proc Am Soc Clin Oncol 15 : 27 , 1996 (abstr 2022)
Lotfi K, Mansson E, Peterson C, et al: Low level of mitochondrial deoxyguanosine kinase is the dominant factor in acquired resistance to 9-beta-D-arabinofuranosylguanine cytotoxicity. Biochem Biophys Res Commun 293 : 1489 -1496, 2002
Kakihara T, Fukuda T, Tanaka A, et al: Expression of deoxycytidine kinase (dCK) gene in leukemic cells in childhood: Decreased expression of dCK gene in relapsed leukemia. Leuk Lymphoma 31 : 405 -409, 1998
Blaney S, Poplack D: Neoplastic meningitis: Diagnosis and treatment considerations. Med Oncol 17 : 151 -162, 2000
Buchanan GR, Rivera GK, Boyett JM, et al: Reinduction therapy in 297 children with acute lymphoblastic leukemia in first bone marrow relapse: A Pediatric Oncology Group Study. Blood 72 : 1286 -1292, 1988
Henze G, Fengler R, Hartmann R, et al: Six-year experience with a comprehensive approach to the treatment of recurrent childhood acute lymphoblastic leukemia (ALL-REZ BFM 85): A relapse study of the BFM group. Blood 78 : 1166 -1172, 1991
Giona F, Testi AM, Rondelli R, et al: ALL R-87 protocol in the treatment of children with acute lymphoblastic leukaemia in early bone marrow relapse. Br J Haematol 99 : 671 -677, 1997
Lawson SE, Harrison G, Richards S, et al: The UK experience in treating relapsed childhood acute lymphoblastic leukaemia: A report on the medical research council UKALLR1 study. Br J Haematol 108 : 531 -543, 2000
Nelarabine Investigator's Brochure. 1996
Hirschfeld S, Flowers C, Smith M: Toxicity from intrathecal therapy in pediatric hematologic neoplasms: Is the incidence increasing. Proc Am Soc Clin Oncol 17 : 539A , 1998 (abstr 2069)
Baker W, Royer G, Weiss R: Cytarabine and neurological toxicity. J Clin Oncol 9 : 679 -693, 1991
Kornblau SM, Cortes-Franco J, Estey E: Neurotoxicity associated with fludarabine and cytosine arabinoside chemotherapy for acute leukemia and myelodysplasia. Leukemia 7 : 378 -383, 1993
Scherokman B, Filling-Katz M, Tell D: Brachial plexus neuropathy following high-dose cytarabine in acute monoblastic leukemia. Cancer Treat Rep 69 : 1005 -1006, 1985
Rodriguez V, Kuehnle I, Heslop H, et al: Guillain-Barrre syndrome ofter allogeneic hematopoietic stem cell transplantation. Bone Marrow Transpl 29 : 515 -517, 2002
Rijksen G, Kuis W, Wadman SK, et al: A new case of purine nucleoside phosphorylase deficiency: Enzymologic, clinical, and immunologic characteristics. Pediatr Res 21 : 137 -141, 1987
Stoop JW, Zegers BJ, Hendrickx GF, et al: Purine nucleoside phosphorylase deficiency associated with selective cellular immunodeficiency. N Engl J Med 296 : 651 -655, 1977(Stacey L. Berg, Susan M. )