[CANCER RESEARCH 46, 5953-5958, November 1986] Fludarabine Phosphate (NSC 312878) Infusions for the Treatment of Acute : Phase I and Neuropathological Study1

David R. Spriggs,2 Edward Stopa, Robert J. Mayer, William Schoene, and Donald W. Kufe3 Dana Farber Cancer Institute [D. R. S., R. J. M.. D. W. K.J and Department ofPathology, Brigham and Women 's Hospital [E. S., W. SJ, Boston, Massachusetts 0211S

ABSTRACT a relationship between dose and lethargy. The schedule of drug administration in these studies was not a 5-day infusion. A Fludarabine phosphate (NSC 312878), an resist ant analogue of 9-/9-D-arabinofuranosyladenine, has entered clinical trials. single bolus was used in one study while the other used a single bolus followed by a 48-h infusion. Eleven patients with acute leukemia in relapse received 14 courses of fludarabine phosphate as a 5-day continuous infusion administered at Based on the myelosuppressive effects of fludarabine phos doses of 40 to 100 mg/m2/day. Toxicity was characterized by uniform phate in Phase I testing, we instituted a Phase I and II trial to myelosuppression, as well as occasional nausea, vomiting, and hepato- identify the toxicity and efficacy of fludarabine phosphate in toxicity. Three episodes of metabolic acidosis and lactic acidemia were the treatment of acute leukemia. The drug was administered as noted. In addition, three patients suffered neurotoxicity. Two of these a 5-day continuous infusion at a starting dose equivalent to the three patients had a severe neurotoxicity syndrome characterized by maximum tolerated dose reached in patients with solid tumors. blindness, encephalopathy, and coma. Neither patient recovered neuro This report describes 11 patients with acute leukemia treated logical function. Neuropathological findings at autopsy were character with continuous infusion fludarabine phosphate. ized by a diffuse, necrotizing leukoencephalopathy which was most severe in the occipital lobes. The medullary pyramids and posterior columns MATERIALS AND METHODS were also severely affected. This sporadic fatal neurotoxicity was ob served only at doses greater than 40 mg/m2/day. The maximum tolerated Selection of Patients. All patients entered on this trial had acute dose for a 5-day infusion of fludarabine phosphate is thus 40 mg/m2/day. leukemia in relapse as documented by bone marrow aspiration and biopsy. Ten patients had acute myelogenous leukemia and one had acute lymphocytic leukemia. All patients had already received at least INTRODUCTION one established drug regimen for acute leukemia. The characteristics of ara-C4 has become a keystone in the treatment of acute the patients are listed in Table 1. Ten of 11 patients had been treated with high dose ara-C prior to entry. All patients were followed until leukemia (1). The related purine nucleoside, ara-A, has had less death and no patient received subsequent antineoplastic therapy because clinical utility because of its rapid deamination by ADA (2). of infection, poor performance status, or patient preference. Prior to Inhibitors of ADA, such as DCF, have been used clinically to entry, patients had an Eastern Cooperative Oncology Group perform prolong the effective half-life of ara-A (3, 4). However, DCF ance status of 2 or better (Eastern Cooperative Oncology Group 2 = also increases intracellular dATP pools which compete with confined to bed less than 50% of waking hours) serum creatinine <1.5 ara-A for incorporation into DNA (5). This antagonism, as well mg/dl and serum bilirubin <1.5 mg/dl. Informed consent was obtained as the undesirable clinical side effects of DCF, has stimulated from all patients. the development of ADA-resistant ara-A analogues (6). F-ara- Treatment Plan. Fludarabine phosphate was provided by the National A was synthesized as one of these analogues (7). This compound Cancer Institute. The sterile lyophilized powder containing sodium hydroxide to adjust the pH to 6.5-8.5 was reconstituted with sterile is resistant to deamination by ADA and has antitumor activity water. The drug was administered i.v. in 500 ml of 5% dextrose and in vitro and in vivo (8). Because it is relatively insoluble, the water over a 24-h period. monophosphate derivative, fludarabine phosphate (NSC The drug was administered as a continuous infusion for 5 consecutive 312878), has been formulated for clinical use. days. Patients with decreased numbers of leukemic cells on day 14 bone Fludarabine phosphate has activity in a number of animal marrow aspirates or biopsies were eligible for a second course of tumor systems including LI210 leukemia, CD8F mammary treatment beginning on day 15. The starting dose of fludarabine phos adenocarcinoma, P388 leukemia, and human LX-1 lung tumor phate was 40 mg/m2/day and at least three courses were administered xenograft (9). The preclinical toxicology of fludarabine phos at each dose level. Doses of 40, 60, and 80 mg/m2/day were tested. Two patients each received a single course of 100 mg/m2/day. phate has demonstrated myelosuppression of all hematopoietic lineages, as well as vomiting, diarrhea, liver enzyme abnormal All treatments were administered on the inpatient service at the Dana ities, and occasional azotemia. The inital Phase I trials of Farber Cancer Institute. Patients received standard supportive care including antibiotics, antiemetics. and blood products. fludarabine phosphate demonstrated that myelosuppression is the dose-limiting toxicity (10, 11). Mild lethargy was reported Study Parameters. During therapy, patients had daily complete blood counts and twice weekly measurements of liver function and renal following bolus administration, while there was no other re function. Electrolytes, blood gases, and other clinical studies were ported neurotoxicity. These studies, however, did not establish performed as indicated. During the 4 weeks after therapy, complete blood counts were obtained at least twice weekly and measures of liver Received 3/7/86; revised 6/24/86; accepted 7/31/86. and renal function were monitored weekly. All patients were considered The costs of publication of this article were defrayed in part by the payment évaluablefortoxicity and 10 of 11 patients were évaluableforresponse. of page charges. This article must therefore be hereby marked advertisement in A complete remission in this study was defined as the achievement of accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1This investigation was supported by Grant CH-199 awarded by the American an Ml bone marrow as well as the recovery of normal peripheral blood counts (WBC >3000/mm3, >IOO,000/mm3, and hemoglobin Cancer Society. 'Supported by USPHS Grant 1K08 CA00994-01 awarded by the National >10 g/dl.) A remission of less than 30 days in duration was not Cancer Institute, Department of Health and Human Services. 'Supported by an American Cancer Society Faculty Research Award. To considered to be complete. Patients who did not achieve complete whom requests for reprints should be addressed, at the Dana Farber Cancer remission were analyzed for the mechanism of failure ( 12). Institute, 44 Binney Street, Boston. MA 02115. 4The abbreviations used are: ara-C, 9-/J-D-arabinofuranosylcytosine; F-ara-A, RESULTS 9-|8-D-arabinofuranosyl-2-iluoroadenine; ara-A, 9-/3-r>arabinofuranosyladenine; DCF, 2'-deoxycoformycin; ADA, adenosine deaminase; CSF, cercbrospinal fluid; The first objective of this study was to determine the maxi CNS, central nervous system; PAS, periodic acid-Schiff. mum tolerated dose of fludarabine phosphate in the treatment 5953

Downloaded from cancerres.aacrjournals.org on October 1, 2021. © 1986 American Association for Cancer Research. FLUDARABINE PHOSPHATE INFUSIONS FOR ACUTE LEUKEMIA of acute leukemia. AH 14 courses of therapy were évaluablefor toxicity. The potential for late toxicity (defined as toxicity first 30 appearing more than 28 days after the initiation of treatment) was évaluablein8 of 11 patients. The clinical toxicity of fludarabine phosphate is summarized 25 in Table 2. Severe myelosuppression was uniformly seen, af fecting WBC, platelets, and RBC. Intensive support was re quired for all courses including transfusions, RBC transfusions, and broad spectrum antibiotics for fever and neu- 20 tropenia. The duration of leukopenia, defined as WBC <1000/ mM/L mm3, was more than 2 weeks in 10 of 11 patients. One patient with acute lymphocytic leukemia in relapse received a dose of 15 40 mg/m2/day but his peripheral lymphoblast count never fell below 1000/dl. Both patients receiving a dose of 100 mg/m2/ day for 5 days had leukopenia for more than 30 days (32 and 10 45 days). Three patients received a second dose of fludarabine phosphate 15-17 days after the initiation of treatment because of persistent leukemic cells in the day 14 bone marrow aspirate. In these patients, the leukopenic phase persisted for 14-30 days after the second treatment. Nausea and vomiting were seen in 6 of 14 courses but were generally mild and readily controlled with antiemetics at all dose levels. Hepatotoxicity, as manifested by 2-3-fold eleva tions of the serum transaminases, was observed in 6 of 14 579 11 courses. This toxicity was reversible within 7 days of treatment DAY OF TREATMENT and not clearly related to dose. One patient had a significant Fig. 1. Fludarabine phosphate-related metabolic acidosis. An infusion of flu but reversible elevation of serum creatinine (1.5-3.3 mg/dl) darabine phosphate was administered over 5 days. Measurement of the plasma while receiving fludarabine phosphate concurrently with neph- bicarbonate (•)and anióngap (columns) demonstrated a progressive acidosis. A serum lactate (arrow) drawn on the seventh day of treatment was elevated (23 rotoxic antibiotics. Another patient had mild, reversible mg/dl). F-ARA-AMP, 5-fluoro-9-/3-D-arabinofuranosyladenine 5'-phosphate. changes in serum creatinine occurring 2 weeks after fludarabine phosphate treatment. Two patients developed erythematous several days after the completion of the fludarabine phosphate maculopapular rashes while receiving fludarabine phosphate infusion. The three patients who developed metabolic acidosis and combination antibiotic therapy. Both episodes were mild were otherwise clinically similar to others in the study. Concur and resolved after altering the antibiotic regimen. rent medications, hepatic function, renal function, and prior During fludarabine phosphate infusions, three patients de treatment were similar for both groups. Furthermore, no alter veloped significant metabolic acidosis associated with lactic native explanation for metabolic acidosis (such as sepsis, renal acidemia. The course of one patient is illustrated in Fig. 1. In disease, ketosis, or salicylate administration) could be identi each case, a rising anión gap, decreasing serum bicarbonate fied. Episodes of metabolic acidosis occurred in patients receiv levels, and increasing serum lactate became apparent 2 to 3 ing doses of 40, 60, and 100 mg/m2/day. days after beginning the fludarabine phosphate infusion. Each The most serious toxicity seen in this study was late neuro- patient had an appropriate respiratory compensation and urine toxicity. Two patients had severe neurotoxicity associated with pH fell to <6. The metabolic acidosis resolved spontaneously blindness and coma. These cases will be presented in detail. Case 1. A 41-year-old woman was found to have acute my- Table 1 Patient characteristics logenous leukemia (M4) in August 1983. Remission was in Patients 11 duced with two cycles of (45 mg/m2/day, days 1- Courses 14 Age 19-59 yr (median age, 32 yr) 3) and ara-C (200 mg/m2/day, days 1-7). She received four Sex 8 male, 3 female cycles of high dose ara-C consolidation (3 g/m2 every 12 h for Performance status 0-2 (median, 1) Prior therapy ara-C and daunorubicin, II; high four doses). No neurotoxicity was noted. She relapsed 21 days dose ara-C, 10 following her fourth cycle of high dose ara-C. Reinduction with No. of prior therapies 0 = 0 patients 1 =0 two cycles of daunorubicin and ara-C was unsuccessful and she 2 = 5 was enrolled on the fludarabine phosphate trial. 3 or more = 6 The patient received fludarabine phosphate (60 mg/m2/day) No. of prior complete remissions 0 = 1 patients 1 =5 for 5 days with a prompt clearing of peripheral blasts. However, 2 = 5 a bone marrow aspiration 14 days after the start of treatment showed the persistence of blasts and a second course of fludar Table 2 Toxicity of fludarabine phosphate infusion abine phosphate (60 mg/m2/day for 5 days) was begun 3 days No. of courses/total later. Fever and persistent pancytopenia were managed with a variety of antibiotics including ticarcillin, gentamicin, trime- Toxicity Grade 1 Grade 2 Grade 3 Grade 4 thaprim-sulfamethoxazole, cefaperazone, vancomycin, and am- Leukopenia 14/14 Thrombocytopenia 14/14 photericin B. Another bone marrow biopsy 40 days after the Nausea and vomiting 3/14 3/14 start of fludarabine phosphate treatment showed 70% blasts. Hepatotoxicity 3/14 3/14 Nephrotoxicity 2/14 The patient remained hospitalized for the next 30 days, receiv Neurotoxicity 1/14 2/14 ing supportive care for infection and pancytopenia. Skin rash 2/14 The patient was found to be lethargic and confused on the 5954

Downloaded from cancerres.aacrjournals.org on October 1, 2021. © 1986 American Association for Cancer Research. FLUDARABINE PHOSPHATE INFUSIONS FOR ACUTE LEUKEMIA

70th day after the start of fludarabine phosphate therapy. Her tiple abscesses were found in liver, spleen, marrow, and kidney mental status continued to decline and she became incontinent consistent with disseminated fungal infection. Bone marrow of urine. Three days later, her pupils were unreactive to light biopsy sections demonstrated myelodysplasia with 30% blast and she was responsive only to painful stimuli. She expired on forms. the following day. The fresh brain weighed 1630 g, was mildly swollen on A complete autopsy was performed. The systemic findings external examination, and showed diffuse softening of the white included extensive leukemic infiltration of lungs, spleen, mar matter with focal areas of mottling. Selected portions including row, breast, and esophagus. There was necrotizing broncho- the occipital lobe of the left cerebral hemisphere were frozen. pneumonia. The fresh brain weighed 1330 g. Scant subdural All white matter appeared to be affected diffusely with the most hematomas were present over both posterior convexities. Serial severe involvement posteriorly in the occipital lobe. The brain- coronal sections through fixed brain revealed no gross abnor stem, cerebellum, and spinal cord were grossly unremarkable. malities. Transverse sections of brainstem and cerebellum were Various nerve roots, nerves, and muscles were sampled and also unremarkable. Microscopic sections were carried out on 9 portions of nerve and muscle were also frozen for future study. formalin-fixed, paraffin-embedded sections (7 brain, 2 dura). Microscopic examination was carried out on 41 formalin-fixed, Hematoxylin and eosin, Luxol fast blue, periodic acid-Schiff, paraffin-embedded sections (17 brain, 1 pituitary, 7 spinal cord, and Bielschowsky staining techniques were used. The sections 2 spinal roots, 1 dorsal root ganglia, 5 nerves, 1 sympathetic of dura showed mild subdural leukemic infiltration and hem chain, and 7 muscle). Hematoxylin and eosin, Luxol fast blue- orrhage. Slight vacuolization and myelin loss, scattered periodic PAS, Bodian, Bielschowsky, Gomori trichrome. Gram, methe- acid-Schiff-positive macrophages, and rare axonal spheroids namine silver, and acid-fast (Ziell-Nielson) staining techniques were seen focally within one occipital lobe. These changes were were used. minimal and primarily involved the optic radiations and adja The most striking findings were within the cerebral white cent white matter. There was no evidence of intracerebral matter (Fig. 2A). These areas showed a diffuse, necrotizing hemorrhage, leukostasis, or leukemic meningitis. leukoencephalopathy, characterized by vacuolization, myelin Case 2. A 32-year-old man was found to have acute undiffer- loss with numerous PAS-positive macrophages, and axonal entiated leukemia in November 1983. He was treated with swelling with spheroid formation. (Fig. 2, B and C). The necro daunorubicin (45 mg/nr/day for 3 days) and ara-C (200/mg/ sis was most severe within the occipital and parietal lobes. The m2/day for 7 days). A partial remission was achieved. A second optic radiations were particularly affected. The frontal and cycle of daunorubicin and ara-C was administered in combina temporal lobes were less extensively involved, but patchy areas tion with (2 mg/m2/\veek for 3 weeks) and predni- of necrosis were evident in the deep and subcortical white sone. A complicated hospital course ensued with Escherichia matter. Resolving embolie microinfarctions were seen in the coli and Candida tropicalis septicemia, respiratory failure, renal basal ganglia. failure, and C. tropicalis hepatitis. A complete remission was Examination of the optic nerves revealed severe necrosis with achieved, but he was left with neurological deficits including a myelin loss, numerous PAS-positive macrophages, and early severe peripheral neuropathy, vertical nystagmus, and tituba- reactive astrocytosis. Sections of cerebellum and pons showed tion. The latter findings were thought to be related to hypoxic patchy foci of vacuolization. The medullary pyramids showed insults sustained during sepsis. The patient was hospitalized vacuolar change (Fig. ID). Multiple sections of spinal cord for the next 6 months for rehabilitation. He relapsed 10 months showed marked necrosis of the dorsal columns. Slight to mod after his initial therapy without receiving maintenance or con erate vacuolization was present in spinal cord white matter (Fig. solidation therapy. 2E). The cervical and thoracic portions of the cord were more Fludarabine phosphate (100 mg/m2/day for 5 days) was involved than the lumbar region. Nerve roots, the sympathetic administered in combination with amphotericin (because of the chain, and dorsal root ganglia were all normal. The sural nerve history of fungal hepatitis). The patient had prolonged pancy- revealed a moderate to severe axonal neuropathy. Neurogenic topenia but achieved a complete bone marrow remission after muscular atrophy was seen in all muscle groups sampled with 35 days. He complained of a slight decrease in visual acuity and the most severe changes present in the muscles of the lower photophobia 44 days after starting fludarabine phosphate. Eval extremity. uation included a normal computer-assisted tomographic scan Based on the toxicity of 100-mg/m2/day infusions seen at and a lumbar puncture with CSF protein of 104 mg/dl, CSF our institution and elsewhere, the dose of this agent was de glucose of 92 mg/dl, and a CSF cell count of 1 WBC/mm3. creased to 60 mg/m2/day and the duration of infusion was The myelin basic protein concentration in the CSF was elevated extended to 7 days. A single patient was treated at this dose at 8.8 ng/ml (normal value, <4 ng/ml). The patient's vision level. The patient received a dose of 60 mg/m2/day for 7 days. was completely lost 3 days later. The patient's optic discs were Three weeks after initiating fludarabine phosphate, he was pale, visual evoked response testing elicited no cortical poten noted to have new resting tremor. A detailed neurological tials, and his mental status began to deteriorate. By the 57th evaluation was otherwise negative and no clinical evidence of day after the onset of fludarabine phosphate therapy, the patient visual impairment was found. Visual evoked potentials were was confused, lethargic, and incontinent. He had roving eye not performed. He was followed until his death 50 days after movements and increased motor tone with little spontaneous the start of fludarabine phosphate treatment, without any movement. An electroencephalogram showed poor organiza change in his neurological status. He died from progessive tion with bilateral theta and delta slowing. A repeat computer- leukemia. No autopsy was performed. assisted tomographic scan was negative. The patient remained The response to therapy was assessed by bone marrow aspi in a vegetative state for the next 30 days without improvement. rate 14 days after initiation of therapy. Patients treated with A relapse of his leukemia was documented 81 days after the doses of 40-80 mg/m2/day had persistent leukemic cells in start of fludarabine phosphate treatment and he expired 8 days their bone marrow aspirates. Three patients received a second later. course of fludarabine phosphate, beginning on the 15th day A complete postmortem examination was performed. Mul following initiation of treatment. These patients failed to 5955

Downloaded from cancerres.aacrjournals.org on October 1, 2021. © 1986 American Association for Cancer Research. FLUDARABINE PHOSPHATE INFUSIONS FOR ACUTE LEUKEMIA

Fig. 2. A, Luxol fast blue-PAS (myelin) stain of a coronal section of the occipital lobe with marked pallor within the optic radiations demonstrating profound white matter destruction. B, Luxol fast blue- PAS (myelin) stain of occipital lobe white matter seen in A. showing diffuse necrosis with numerous perivascular PAS-positive macrophages, x 10. C, Bielschowsky (silver) stain of right occipital lobe white matter showing numerous axonal spheriods. x 40. D, Luxol fast blue-PAS (myelin) stain of medullary pyramid showing diffuse vacuolization and pallor consistent with extensive injury- x 4. E, Luxol fast blue-PAS (myelin) stain of cervical spinal cord demonstrating pallor within the dorsal columns, x 1.

- achieve remission. Two patients died of infection while pancy- 60 days. This patient (case 2) had not previously received high topenic 17 and 23 days after treatment. Two patients received dose ara-C. 1UOmg/m2/day for 5 days and both patients had hypoplastic DISCUSSION bone marrow biopsies on day 14 without evidence of persistent leukemic cells. One of these patients, a man with an undiffer- The dose-limiting toxicity of fludarabine phosphate in Phase entiated leukemia, had a complete remission which lasted for I clinical trials has been myelosuppression (10, 11). This find- 5956

Downloaded from cancerres.aacrjournals.org on October 1, 2021. © 1986 American Association for Cancer Research. FLUDARABINE PHOSPHATE INFUSIONS FOR ACUTE LEUKEMIA ing, as well as its similarity to ara-C, made fludarabine phos meningitis, and intrathecal drug administration are thought to phate an obvious choice for testing in acute leukemia. Like ara- predispose to the development of this syndrome. C, fludarabine phosphate, when administered as a continuous The two patients described here demonstrated late and dev infusion over a 5-day period, resulted in a consistent cytore- astating neurotoxicity. In both patients, the onset of neurolog duction of circulating myeloblasts, as well as severe myelo- ical systems began 6-8 weeks after treatment, differing from suppression at all dose levels tested in this trial. The duration the patterns of other antimetabolities. There was considerable of myelosuppression at all dose levels tested in this trial. The variation in the neuropathological findings. These differences duration of myelosuppression ranged from 11 to 46 days. The may be related to the interval between death and the onset of one patient who acheived a complete remission was not re symptoms. In case 1, a very short period elapsed between the treated and thus no inference can be made regarding cumulative onset of neurological sequelae and death (4 days). The neuro toxicity or the effects of this agent on a remission marrow. pathological findings were limited to subtle demyelination and The nonmyelosuppressive toxicity of fludarabine phosphate axonal damage which predominantly involved the visual sys proved to be a critical obstacle in the further use of this agent tem. In case 2, a much longer survival was seen following the in leukemia treatment. Mild renal toxicity was seen in 3 of 11 onset of symptoms (45 days) and the leukoencephalopathy patients and 6 of 11 patients had transient elevations of serum observed was more extensive and severe. These observations transaminases. Metabolic acidosis, characterized by an in suggest that the mechanism of toxicity may be related to some creased anión gap, an elevated plasma lactate level, and com impairment of oligodendroglial or axonal function, which is pensatory respiratory alkalosis, was seen in three patients who most apparent in those brain areas having the greatest metabolic were free of overt infection, hepatic disease, or hypoxemia. In activity (i.e., visual system). Although neither patient survived each case, the acidosis had its onset after 3 days of therapy and long enough to establish the reversibility of these lesions, the resolved spontaneously after stopping the fludarabine phos extensive necrosis observed would make complete recovery phate infusion. The mechanism of this acidosis is not known. unlikely. In this regard, a recent Phase I and II study of A "spontaneous" lactic acidosis has been described in patients fludarabine phosphate has similarly reported delayed CNS tox with acute leukemia but is uncommon and poorly understood icity (20). This report, however, did not characterize the neu (13). We have excluded other etiologies of lactic acidosis, ropathological changes. including hypoxia, seizures, diabetes mellitus, and other drugs. These two patients with severe neurotoxicity did not differ Concomitant hepatic dysfunction is well known to increase the clinically from the other patients in the study. Both patients severity of lactic acidosis, although clinical evidence of hepatic were receiving broad spectrum antibiotics including amphoter- disease was not present in these patients. However, transami- icin B, as were most of the other patients under study. Pretreat nase elevations were seen in some patients receiving fludarabine ment histories were not unusual and it is of note that one of phosphate, suggesting that hepatocyte damage may occur with these patients was not previously treated with high dose ara-C, fludarabine phosphate therapy. a regimen known to be neurotoxic. No early signs of neurotox The neurotoxicity of antineoplastic agents has been recently icity were observed prior to the onset of visual impairment. The reviewed (14). can cause a variety of clinico- late onset of toxicity in patient 2 was particularly ominous. Few pathological syndromes including arachnoiditis (, leukemia patients survive for more than 60 days after unsuc ara-C), cerebellar dysfunction (5-, ara-C), somno cessful treatment and it is possible that additional episodes of lence (DCF), and necrotizing leukoencephalopathy (methotrex neurotoxicity would have been seen if more patients had sur ate, ara-C). Table 3 summarizes these toxicities as compared to vived for this period. that seen with fludarabine phosphate. The precise mechanism of action of fludarabine phosphate is Treatment with either 5-fluorouracil or ara-C can lead to an not known. Fludarabine phosphate is rapidly dephosphorylated acute cerebellar syndrome which is often reversible. In the case in vivo to form the active compound F-ara-A. Like ara-C and of ara-C, the loss of neurons from the cerebellar cortex has ara-A, F-ara-A is a potent inhibitor of DNA synthesis (21, 22). been reported, particularly in the Purkinje cell layer (15, 16). Each of these agents is metabolized to the triphosphate form Leukoencephalopathy has also been well documented in pa and is incorporated into DNA (23, 24). The extent of F-ara-A tients receiving methotrexate, often in combination with radia incorporation into DNA is highly correlated with the loss of tion or other chemotherapeutic agents. Necrosis and demyeli- clonogenic survival in vitro (25). The incorporated ara-C and nation have been described in the centrum ovale, periventricular ara-A residues alter reactivity of the 3' terminus and act as region, and pons (17-19). Whole brain radiation, leukemic relative chain terminators (23). Thus, chain elongation is slowed

3MethotrexateLateConfusion, neurotoxicity5-FluorouracilAcuteAtaxia, phosphateLate OnsetClinical (months)Blindness, manifesta nausea,vomiting, somnolence.spasticity, dys-metriaHighnystagmus, nystag incontinencequadraparesis, tionsDose, fever,meningismusIntrathecaltremors,dementiaHigh musdysme-triaHigh de mentiaDepends schedule de route4-72 dosemethotrexateor dose1daily or weekly doseara-C, >60mg/m2on dose pendenceDurationCofactorsNeuropathologicalfindingMethotrexateAcuteHeadache,routeMonths-yearsRadiationintrathecal >4-6dosesWeeks and pro tractedscheduleMonths hUnknownCSF -6 wk (may be irreversi beirreversible)UnknownLoss(may (reversibilityunknown)UnknownOccipitalhUnknown ble)UnknownLoss therapyDemyelination, pleiocytosisTable coagu aswellof Purkinje cells of Pur mattermostwhite withastrocyticlative necrosis ininferioras neuronal loss kinje cellsFludarabine involved; ex reaction.Periventricular olivary and den tensive demyelina whitematter tate nucleiara-CAcuteAtaxia, tion with axonal in involved.graymost juryDCFAcuteSomnolenceProlongedinfusions24-72 matter sparedAntimetabolile

5957

Downloaded from cancerres.aacrjournals.org on October 1, 2021. © 1986 American Association for Cancer Research. FLUDARABINE PHOSPHATE INFUSIONS FOR ACUTE LEUKEMIA and reduplication of certain DN A segments may result during nosyladenine in combination with deoxycoformycin. Cancer Chemother. Pharamacol., 10: 125-128, 1983. a single (26). Unlike ara-C and ara-A, F-ara-A is also 5. Plunkett, W., Benjamin, R., Keating, M.. and Freireich, E. Modulation of 9- incorporated into RNA (25). The extent of F-ara-A incorpora /3-o-arabinofuranosyladenine 5'-triphosphate and deoxyadenosine triphos phate in leukemic cells by 2'-deoxycoformycin during therapy with 9-/3-D- tion into RNA also correlates with loss of clonogenic survival. arabinofuranosyladenine. Cancer Res., 42: 2092-2096, 1982. Additionally, F-ara-A has been shown to inhibit S-adenosyl- 6. Major, P., Agarwal, R., and Kufe, D. Clinical pharmacology of deoxycofor homocysteine hydrolase and ribonucleotide reducíase(21). The mycin. Blood, SS: 91-96. 1981. 7. Montgomery, J. A. Studies on the biological activity of purine and pyrimidine relative importance of these actions remains unclear. analogs. Med. Res. Rev., 2: 211-308. 1982. The mechanism for fludarabine phosphate neurotoxicity is 8. Brockman, R. W., Schabel, E. M.. and Montgomery, J. A. Biological activity uncertain. Adenosine and its analogues are potent presynaptic of 9-0-o-arabinofuranosyl-2-fluoroadenine. Biochem. Pharmacol., 26:2193- depressants in a variety of animal systems (27, 28). The intra- 2196. 1977. 9. Montgomery, J. A. Has the well gone dry? The first Cain Memorial Award ventricular administration of adenosine to dogs results in CNS lecture. Cancer Res., 42: 3911-3917, 1982. 10. Hutton, J. J., Van Hoff, D. D., Kühn,J., Phillips, J., Heish, M., and Clark, depression, as evidenced by drowsiness (27). Rat studies have G. Phase l clinical investigation of 9-f(-D-arabinofuranosyl-2-adenine 5'- also demonstrated that deoxyadenosine has CNS depressant monophosphate (NSC 213887), a new purine . Cancer Res., effects (28). Furthermore, rat brain studies have suggested the 4-/.-4183-4186, 1984. presence of an adenosine receptor (29). Finally, structure-activ 11. Leiby, J. M., Grever, M. R., Staubus, A. E., Neidhart, J. A., and Malspeis, L. A Phase I clinical investigation of fludarabine phosphate by loading dose/ ity studies suggest that halogenation of position 2 of the purine continuous infusion. Proc. Am. Assoc. Cancer Res., 26:669, 1985. ring will result in a significant enhancement of CNS depressant 12. Preisler, H. D. Failure of remission induction in acute myelocytic leukemia. Med. Pediatr. Oncol., 4: 275-276, 1978. activity, as well as resistance to ADA activity. 13. Walters, W. C. Ill, Hall. J. D., and Schwartz, W. B. Spontaneous lactic DCF, a pure ADA inhibitor, has been used as an antineo- acidosis: the nature of the acid-base disturbance and considerations in diag plastic agent, both alone and in combination with ara-A (30, nosis and management. Am. J. Med., 35: 781-786, 1963. 14. Kaplan, R. S., and Wiernik, P. H. Neurotoxicity of antineoplastic drugs. 31). As a single agent, neurotoxicity was an important dose- Semin. Oncol., 9: 103-130, 1982. limiting toxicity. Sixty % of patients had central nervous system 15. Reil)I. J. L., and Brown, W. J. Acute cerebellar syndrome secondary to 5- fluorouracil therapy. Neurology, 14: 961-964, 1964. toxicity, ranging from lethargy to coma (30). The severity of 16. Lazarus, H. M., Herzig, G. P., Phillips, G., Roessman, U., and Fishman, D. neurotoxicity for DCF is related to dose. Although the penetra Central nervous system toxicity of high dose systemic cytosine arabinoside. tion of DCF into the cerebrospinal fluid is limited, elevated Cancer (Phila.). 48: 2577-2582, 1981. 17. Rubinstein, L. J., Herman, M. M., Long, T. T., and Wilbur, J. R. Dissemi levels of deoxyadenosine have been noted in the cerebrospinal nated necrotizing leukoencephalopathy: a complication of treated central fluid following DCF treatment (32). No anatomic changes have nervous system leukemia and . Cancer (Phila.), 35:291-305,1975. been reported in conjunction with DCF treatment. 18. Shapiro, W. R., Chernik, N. T., and Posner, J. B. Necrotizing encephalopathy following intraventricular instillation of methotrexate. Arch. Neural., 28: Whether F-ara-A or another metabolite of fludarabine phos 96-102. 1973. phate is responsible for the observed neurotoxicity remains 19. Breuer, A. C., Blank, N. K., and Schoene, W. C. Multifocal pontine lesions in cancer patients treated with and CNS radiotherapy. Cancer unclear. The central nervous system has a high level of ADA (Phila.), 41: 2112-2120, 1978. and high cerebrospinal fluid concentrations of 2-fluoro-9-/3-D- 20. Warrell, R., and Herman, E. Phase I and II study of fludarabine phosphate arabinofuranosylhypoxanthine may result from prolonged flu in leukemia: therapeutic efficacy with delayed central nervous system toxicity. J. Clin. Oncol., 4: 74-79, 1986. darabine phosphate infusions (32). Recent studies have shown 21. White, E. L., Shaddix, S. C., Brockman. R. W.. and Bennett. L. L. Compar that up to 10% of F-ara-A is metabolized to 2-fluoroadenosine ison of the actions of 9-/3-D-arabinofuranosyl-2-fluoroadenine and 9-/3-D- triphosphate via the toxic intermediate compound 2-fluoroad- arabinofuranosyladenine on target enzymes from mouse tumor cells. Cancer Res., 42: 2260-2264, 1982. enine (33). Further animal studies would be required to identify 22. Dow, L. W., Bell, D. E.. Poulakes, T., and Fridland, A. Differences in the metabolic pathway responsible for fludarabine phosphate- metabolism and cytotoxicity between 9-0-D-arabinofuranosyladenine and 9- /3-D-arabinofuranosyl-2-fluoroadenine in human leukemic lymphoblasts. induced neurotoxicity. Cancer Res., 40: 1405-1410, 1980. Only one complete remission was noted in this study of 23. Kufe, D. W., and Spriggs, D. R. Biochemical and cellular pharmacology of patients with refractory leukemia. This patient received a dose cytosine arabinoside. Si-min. Oncol., 12 (Suppl. 3): 34-48, 1985. of 100 mg/m2/day. Patients receiving less than 100 mg/m2/day 24. Kufe, D. W., Major, P., Munroe, D., Egan, E. M., and Herrick, D. Relation ship between incorporation of 9-/i-i>-arabinofuranosyladenine in 1.1210 and for 5 days did not achieve marrow hypoplasia on day 14 and cytotoxicity. Cancer Res., 43: 2000-2004. 1983. 25. Spriggs, D., Robbins. G., Mitchell, T., and Kufe, D. W. Incorporation of 9- no responses were seen at these doses. However, doses as low /3-r>arabinofuranosyl-2-fluoroadenine into HL-60 cellular RNA and DNA. as 60 mg/m2/day have been associated with neurotoxicity. Biochem. Pharmacol., 35: 247-252, 1986. Thus, no safe and effective dose could be identified. Although 26. Crowther, P. J., Cooper, I. A., and Woodcock, D. M. Biology of cell killing 1-0-D-arabinofuranosylcytosine and its relevance to molecular mechanisms fludarabine phosphate in other schedules of administration of cytotoxicity. Cancer Res., 45:4291-4300, 1985. might be useful in reducing leukemic cell burden, the sporadic 27. Haulica, I., Ababei. L., Branisteanu, D.. and Topoliceanu, F. Preliminary and devastating neurotoxicity must be defined more clearly data on the possible hypnogenic role of adenosine. Neurochemistry, 21: 1019-1020, 1973. before the clinical utility of this agent can be determined. Until 28. Phyllis, J. W., Kostopoulos, J. K., Edstrom, J. P., and Ellis, S. W. Role of studies elucidate the mechanism of the neurotoxicity or provide adenosine and adenine nucleotides in central nervous function. /;/.•H.P. Bauer and G. I I¡ruminomi(eds). Physiological and Regulatory Functions of methods to predict and avoid this complication, fludarabine Adenosine and Adenine Nucleotides, pp. 343-360. New York: Raven Press, phosphate would not be recommended as a S-day continuous 1979. infusion for the treatment of acute leukemia. 29. Premont, J.. Tassin, J. P.. Blanc, G., and Bockaert, J. Effects of adenosine on adenylate cyclase in broken cell preparations of central nervous system. In: H. P. Baer and G. I. »ruminomi (eds). Physiological and Regulatory REFERENCES Functions of Adenosine and Adenine Nucleotides, pp. 259-270. New York: Raven Press, 1979. 1. Gale, R. P. Advances in the treatment of acute myelogenous leukemia. N. 30. Major, P. P., Agarwal, R. P.. and Kufe, D. W. Deoxycoformycin: neurological Engl. J. Med.. 300: ! 189-1199, 1979. toxicity. Cancer Chemother. Pharmacol.. 5: 193-196. 1981. 2. Choo, I >., and Kimball. A. Deamination of arabinofuranosyladenine by 31. Major, P. P.. Agarwal, R. P.. and Kufe, D. W. Clinical pharmacology of adenosine deaminase and inhibition by arabinosyl-6-. Cancer arabinofuranosyladenine in combination with deoxycoformycin. Cancer Res., 32: 1721-1724. 1972. Chemother. Pharmacol., 10: 125-128, 1983. 3. Woo. P.. Dion, H., Lang. S.. Daini, L., and Durkhar, L. A novel adenosine 32. William, B. B., and Lerner, A. M. Adenosine deaminase in human tissues and ara-A deaminase inhibitor, (A)-3-(2-deoxy-£)-D-fr}YAro-pento-furanosyl)- and fluids. Clin. Res., 23:418, 1975. 3.6,7.8-tetrahydroimidaz(4.5-

5958

Downloaded from cancerres.aacrjournals.org on October 1, 2021. © 1986 American Association for Cancer Research. Fludarabine Phosphate (NSC 312878) Infusions for the Treatment of Acute Leukemia: Phase I and Neuropathological Study

David R. Spriggs, Edward Stopa, Robert J. Mayer, et al.

Cancer Res 1986;46:5953-5958.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/46/11/5953

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/46/11/5953. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.

Downloaded from cancerres.aacrjournals.org on October 1, 2021. © 1986 American Association for Cancer Research.