Meningeal leukemia.A follow-up study

zyx zyxwv zyxw MENINGEAL LEUKEMIA A Follow-up S t u d y BOYDA. NIES,M.D., LOUIS B. THOMAS, M.D., AND EMILJ FREIREICH, M.D. T classified as having clinical meningeal leuHE CLINICAL AND PATHOLOGICAL FEATURES of meningeal leukemia have been re- kemia. Those patients with a maximum cell ported in detail in previous publications from count of from 5 to 10 cells/mm.3 were classified this institution.4r9 T h e patient group used in as having possible clinical meningeal leukemia. Histological sections of arachnoid were rethose studies consisted of the first 150 patients with acute leukemia admitted to the National examined microscopically and amount of leuCancer Institute from 1953 to 1958. Of the kemic infiltration of the arachnoid was graded 150 patients, central nervous system material quantitatively according to criteria previously for pathologic examination was available in described.* T h e brains were examined for the 117. Since that time, not only has the clinical presence of hydrocephalus, which was quanawareness of the meningeal leukemia syn- titated from 0 to 4f.4 drome increased, but the value of intrathecal RESULTS aminopterin also has been demonstrated by several favorable reports in the literature.?, 8,149 15 General characteristics: T h e 94 patients in For these reasons, follow-up clinical and paththe current study are classified by cell types, ologic study of meningeal leukemia in a more age and sex in Table 1. T h e distributions of recent group of patients with acute leukemia cell types, age and sex are comparable to the seemed indicated. T h e comparison of results 1953 to 1958 series of 150 patients of whom obtained in this later series with findings in 117 had pathologic examination of the centhe previously reported series form the basis tral nervous system. of this report. T h e median duration of survival from date of diagnosis of acute leukemia to death is MATERIALS AND METHODS shown in Table 2. Only patients in whom From December 1961 to June 1963 99 pa- central nervous system autopsies were done tients with acute leukemia had autopsies per- are included in this analysis. Both children formed. Of these patients, 94 in whom sections and adults with acute lymphocytic leukemia of the central nervous system were available in the 1961 to 1963 series had an increase of 4 for examination constitute the group under months in median survival time when comstudy. Detailed pathologic studies of these pared to similar patients in the 1953 to 1958 patients have been included in a separate re- series. A life table comparing children with port.l2 T h e medical records of these 94 pa- acute lymphocytic leukemia showed an intients were reviewed and each patient was classified by cell type, age and sex. T h e cell counts of all lumbar punctures done in this TABLE1 group were tabulated and correlated with Patients with Acute Leukemia Included clinical symptoms and hematologic findings. in 1961-1963 Studv Those patients with cerebrospinal fluid cell Lympho- Myelo- Undiffercounts of greater than 10 lymphocytes and/or cytic cytic entiated Total mononuclear cells/mm.3 in the absence of a Children* 40 9 0 49 positive bacteriologic culture or gross blood Male 23 4 0 27 Female 17 5 0 22 contamination of the cerebrospinal fluid were zyxw zyxwvuts From the Medicine and Pathologic Anatomy Branches, National Cancer Institute, National Institutes of Health, Bethesda, Maryland. Received for publication August 20, 1964. Adults Male Female TOTAL zyxw 10 50 33 8 2 - 13 42 *Fifteen years of age or less. 546 45 2 20 2 2 0 30 91 15 zyxwvutsr zyxwvutsrqpon zyxwvu MENINGEAL LEUKEMIA Nies et al. No. 5 TABLE 2 Median Survival from Date of Diagnosis of Acute Leukemia t o Death Lymphocytic ~ R/Iyelocytic 547 TABLE 3 Signs and Symptoms in 38 Episodes of Meningeal Leukemia (1961-1963) ________ Autopsy series Children Adults Children Adults 1953-58 1961-63 9 mos. 13 nios. 3 mos. 7 mos. 6 mos. 4 mos. 4 mos. 4 mos. creased survival in approximately 60% of such patients in the 1961 to 1963 group when compared to the 1953 to 1958 series (Fig. 1). O n the other hand, no significant increase in median survival was demonstrated for patients with acute myelocytic leukemia. A life table of adults with acute myelocytic leukemia showed nearly identical curves when the 2 groups were compared. Clinical meningeal leukemia: I n the 1953 to 1958 series the diagnosis of meningeal leukemia was made on the basis of a neurological syndrome which most frequently consisted of signs and symptoms of increased intracranial pressure. Of the few patients in the earlier clinical series who had lumbar punctures done during asymptomatic periods, none had elevated cerebrospinal fluid cell counts. Subsequent pathologic and cytologic studies in patients with acute leukemia, however, have No signs or symptoms Lethargy, irritability, or confusion Headache Vomiting Seizures Papilledema or spread sutures on X-ray VII nerve palsy VI nerve palsy 37 11 6 5 3 28 16 13 8 7 18 11 3 4 1 demonstrated that an elevated cerebrospinal fluid cell count in the absence of a positive bacteriologic culture or gross blood contamination is a more sensitive method of diagnosing meningeal leukemia than are neurologic signs or symptom~.~.5 Elevation of the cerebrospinal fluid cell count, as previously described, therefore became the sole criterion for clinical meningeal leukemia in the 1961 to 1963 series. T h e frequency of clinical meningeal leukemia in both groups is demonstrated in Figure 2. Clinical meningeal leukemia was recognized in 42% of patients with acute lymphocytic leukemia and in 12% of patients with acute myelocytic leukemia in the 1961 zyxwvuts zyxwvuts zyxwvutsr v) I- z FIG. 1. Survival % No. 14 of children with acute lymphocytic leukemia. T h e 9.1 and 13.2-month figures are the 50% survival times for each group. 2 W ILL. 0 L z W 0 u W a MONTHS FROM DIAGNOSIS OF ACUTE LEUKEMIA 548 zyxwvutsr zyxwvu zy zyxw CANCER M a y 1965 VOl. 18 TABLE 4 zyxwv Relationship of Meningeal Leukemia t o Systemic Leukemia episodes No. in marrow relapse No. in partial marrow remission Median time t o relapse No. in complete marrow remission Median time to relapse 25 38 17 24 1 3 2 wks. 8 wks.** 7* 11 3 wks. 15 wks.** No. 1953-1958 1961-1963 *Includes 2 patients with complete clinical and peripheral remission but in whom no bone marrow examination was done. ** Does not include the second of 2 episodes of meningeal leukemia occurring in same patient during same bone marrow remission. to 1963 series. O n the other hand, in the 1953 to 1958 series, clinical meningeal leukemia was recognized in 25y0 of patients with acute lymphocytic leukemia and in 4% of patients with acute myelocytic leukemia. However, if only those patients in whom major central nervous system signs or symptoms were present are included, the prevalence of meningeal leukemia in both series was approximately the same. Patients were considered to have had more than one episode of clinical meningeal leukemia when pleocytosis reappeared after at least a one-month period of normal cerebrospinal fluid cell counts. T h e 27 patients with clinical meningeal leukemia in the 1961 to 1963 series had 38 episodes of clinical meningeal leukemia. T h e principal clinical findings in these 38 episodes are shown in Table 3. There were n o central nervous system signs or symptoms in more than a third of the episodes. I n contrast, as previously noted, central nervous system signs or symptoms occurred in all patients with clinical meningeal leukemia in the 1953 to 1958 series. When central nervous signs and/or symptoms were present in patients in the 1961 to 1963 series, lethargy, headache, vomiting, papilledema and cranial nerve palsies were the most common. I n the 1953 to 1958 series, nausea and anorexia, lethargy, vomiting, headache, convulsions and papilledema each occurred in more than 5001, of the patients with meningeal leukemia. T h e relationship of clinical meningeal leukemia to the bone marrow status of the patient is shown in Table 4. T h e criteria for bone marrow remission and relapse are those of Cancer Chemotherapy National Service Center as modified by Acute Leukemia Cooperative Group B.l Of the 38 episodes of meningeal leukemia in the 1961 to 1963 series 11 (29y0) occurred during complete bone marrow remission while in the earlier group 7 of 25 episodes (28y0) occurred while the patients were in complete bone marrow remission. Treatment: Treatment for meningeal leukemia in the 1961 to 1963 series consisted primarily of intrathecal aminopterin in a usual dose of 2.5 mg. per square meter of body surface area when given alone or in a dose of 6 mg. per square meter when given with citrovorum factor. Treatment ordinarily was continued at weekly intervals until the cerebrospinal fluid cell count dropped to less than 5 cells/mm.3 Table 5 shows the response to intrathecal aminopterin in patients with both acute lymphocytic and acute myelocytic leukemia. I n 24 of 27 episodes of meningeal leukemia occurring in patients with acute lymphocytic leukemia and in all 4 episodes TABLE 5 Response of Meningeal Leukemia t o Therapy (1961-1963) Type of treatment Decrease t o 0-4 cells or negative arachnoid Decrease t o 5-10 cells Death before adequate treatment zyxw zyxwvutsr zyx Episodes No response - ~ Lymphocytic Intrathecal aminopterin BCNU Other M yelocytic Intrathecal aminopterin BCNU 32 27 2 3 5 4 1 2 1 0 0 0 1 0 0 0 4 0 0 0 0 0 0 24 2 2 1 zyxwvutsrqpon MENINGEAL LEUKEMIA Nies et al. No. 5 54 9 TABLE 6 Major Complications of Meningeal Leukemia Series VI Nerve Total patients Papilledema Seizures Coma palsy VII Nerve Palsy termine its effect on clinical meningeal leukemia. I n 2 children with acute lymphocytic leukemia, BCNU was successful in reducing the cerebrospinal fluid cell count to normal. Both patients, who died within 2 months after starting this therapy, had no demonstrable leukemic infiltration of the arachnoid at autopsy. Two 25-mg. doses of BCNU at weekly intervals were not successful in eliminating leukemic cells from the arachnoid in a 20year-old male with acute myelogenous leukemia who had 23 mononuclear cells/mm.3 in the cerebrospinal fluid 2 weeks before death. No follow-up spinal fluid cell counts were obtained prior to death. Treatment of meningeal leukemia in the 1961 to 1963 series was associated with a decrease i n morbidity from serious complications of meningeal leukemia when compared to the 1953 to 1958 group. Table 6 shows the frequency of the major objective sequelae of meningeal leukemia in the 2 series. Papilledema, seizures, coma and cranial nerve palsies each occurred less frequently in the 1961 to 1963 series. T h e length of time from the onset of me- in patients with acute myelocytic leukemia, treatment was successful either in reducing the cerebrospinal fluid cell count to less than 5 cells/mm.3 or in producing a n arachnoid free of demonstrable leukemic cells at autopsy. Two additional patients with acute lymphocytic leukemia died before a complete course of intrathecal aminopterin could be given. I n the remaining patient treatment with intrathecal aminopterin reduced the cerebrospinal fluid cell count to between 5 and 10 cells/mm.3 Clinical symptoms were relieved in all patients who received a complete course of treatment. T h e median duration of remission of clinical meningeal leukemia was 9 weeks in patients treated with intrathecal aminopterin who subsequently either had a relapse of clinical meningeal leukemia or died and had leukemic infiltration of the arachnoid at autopsy. An additional 6 patients died and had no demonstrable leukemic-cell infiltration of their arachnoid at autopsy, a median of 5 weeks after the onset of remission of their previous meningeal leukemia. In 3 far advanced patients, BCNU (1,3 bis (2-chloroethy1)-1-nitrosourea) was given to de- zyxwvu zyxw zyxw zyxwvuts TABLE 7 Frequency of Leukemic Arachnoid Infiltration in Patients with Acute Leukemia (1961-1963) Total Lymphocytic No. of patients No. with arachnoid infiltration a t autopsy “Cured” Myelocytic No. of patients No. with arachnoid infiltration a t autopsy “Cured” No L.P. done mm.3 11 or more cells/mrn.3 5-10 cells/m m.3 Other** CSF pleocytosis 21 5 3 0-4 cells/ 50 6 15 22 3 2 11 1o* 3 s 42 21 15 5 1 0 8 4 2 2 0 0 2*t *Patients with clinical meningeal leukemia but no meningeal leukemia a t autopsy. All patients with pleocytosis of 11 cells/mm.3 or more who subsequently had no leukemic infiltration of the arachnoid a t autopsy received treatment with either intrathecal aminopterin or BCNU. **Pleocytosis secondary t o bacterial meningitis or peripheral blood contamination of the CSF. tone patient had a negative arachnoid a t autopsy but had leukemic infiltration of non-neural areas. z zyxwvz CANCER May 1965 550 cytic leukemia in the 1953 to 1958 series (15%) and in 6 patients with acute lymphocytic leukemia in the 1961 to 1963 series (12%). I n patients with acute myelocytic leukemia 7 of 43 in the 1953 to 1958 series (16y0) and 8 of 42 in the 1961 to 1963 series (19yo) had some degree of leukemiccell infiltration of the arachnoid. Infiltration classified as 3-4+ was seen in 3 patients with acute myelocytic leukemia in the 1953 to 1958 series (7%) and i n one patient with acute myelocytic leukemia in the 1961 to 1963 series zyxwvutsrq Patients with major neurologic symptoms Patients with CSF pleocytosis only W Vol. 18 I953 -1958 I-T 1961-1963 -zyxwvutsrq 1 25 % LYMPH. MYELOG. (93) (57) (2%)- T h e relation of leukemic infiltration of the arachnoid to the lumbar puncture data (1961 to 1963) is shown i n Table 7. Of 50 patients with acute lymphocytic leukemia, 6 did not have a lumbar puncture at any time during their course. Of these patients 3 had arachnoid infiltration by leukemic cells at autopsy. Of 15 patients with acute lymphocytic leukemia in whom a t least one spinal fluid was examined but in whom n o cerebrospinal fluid pleo- a zyxwvuts zyxw LYMPH. MYELOG. (50) (42) ...... ....... ...... ...... 0 - 5 ...... ....... FIG. 2. Distribution of meningeal leukemia by cell type. 1-2+ 1953-1958 4 3-4+ 1961-1963 I00 ningeal leukemia occurring during a period of bone marrow remission to subsequent bone marrow relapse also was increased in the 1961 to 1963 series. Table 4 shows that the median time to bone marrow relapse when clinical meningeal leukemia occurred during complete bone marrow remission was 15 weeks in the 1961 to 1963 series compared to only 3 weeks in the 1953 to 1958 series. Similar figures for meningeal leukemia occurring in partial bone marrow remission were 8 weeks for the 1961 to 1963 series and 2 weeks for the 1953 to 1958 group. PuthoEogy: Figure 3 shows the relation of the frequency and degree of leukemic infiltration of the arachnoid to the type of acute leukemia in both series. T h e percentage of patients with arachnoid infiltrations in the 2 groups is nearly identical. Some degree of leukemic-cell infiltration of the arachnoid was noted in 28 of 74 patients with acute lymphocytic leukemia in the 1953 to 1958 series (38%) and in 22 of 50 patients in the 1961 to z 0 I- a a 80 5 LL z I 60 I3 v) I- =w I2 LL O s 40 20 0 LYMPH. MYELOG. (74) (43) LYMPH. MYELOG. (50) (42) zyxwvutsrqpo zyxwvutsrqpon zyxwv zyxwvut MENINGEAL LEUKEMIA No. 5 551 Nies et al. TABLE 8 Frequency of Hydrocephalus Pts. with Clinical Meningeal Leukemia Pts. Without Clinic; Meningeal Leukemia Hydrocephalus Series No. of brains 1953-1958 1961-1963 20 22 1-2+ 10 4 cytosis was present 2 nevertheless had leukemic cell infiltration of the arachnoid at autopsy. I n each of these 2 patients the last lumbar puncture was done more than a month prior to death. I n 21 patients with acute lymphocytic leukemia who previously had presented with clinical meningeal leukemia, 10 had no evidence of leukemic infiltrations of the arachnoid at autopsy. Each of these 10 patients was treated with either intrathecal aminopterin or BCNU. Of the 11 patients with previously recognized clinical meningeal leukemia who had leukemic-cell infiltration of the arachnoid at autopsy, 7 had received adequate treatment for their last episode of clinical meningeal leukemia. I n each of these 7 patients, the last lumbar puncture, including 3 done in the week prior to death, revealed a normal cerebrospinal fluid cell count. Of the 4 remaining patients, 3 died before a n adequate course of intrathecal aminopterin could be given. T h e final patient in this group had 6 lymphocytes per 111111.3 i n the cerebrospinal fluid 2 weeks prior to death and was not subsequently treated. Of 5 patients with acute lymphocytic leukemia and possible clinical meningeal leukemia, 2 had no evidence of leukemic cell infiltration of the arachnoid at autopsy. Each of these patients had an episode of possible clinical meningeal leukemia more than 2 months prior to death. Neither received effective treatment for meningeal leukemia. Of the 3 patients with acute lymphocytic leukemia and possible meningeal leukemia who had leukemic cell arachnoid infiltration at autopsy, one had received intrathecal aminopterin and had a normal cerebrospinal fluid cell count 2 days prior to death. T h e other 2 patients in this group had revealed 6 to 10 cells per mm.3 i n the cerebrospinal fluid in the month prior to death and subsequently received no intrathecal treatment. Three patients with acute lymphocytic leukemia had elevated cerebrospinal fluid white ceII counts associated with either bac- Hydrocephalus 34+ No. of brains 5 1 42 72 1-2+ 9 16 3-4+ 0 0 terial meningitis or intracranial hemorrhage. Each of these patients had leukemic-cell infiltration of the arachnoid at autopsy. Lumbar punctures were not performed in 21 of the 42 patients with acute myelocytic leukemia. Only 4 of these 21 had arachnoid infiltration by leukemic cells at autopsy. Three of these 4 were in blastic crisis and had intracerebral hemorrhage associated with intracerebral leukostasis and leukemic nodules. Both patients with negative spinal fluid examinations during life but with leukemic infiltration of the arachnoid also had intracerebra1 leukostasis at autopsy. Of the 5 patients with acute myelocytic leukemia and clinical meningeal leukemia, 3 showed no evidence of arachnoid infiltration at autopsy following treatment. Leukemic cell infiltration was noted, however, in the non-neural areas of the brain in one of these 3 patients. Of the 2 patients with acute myelocytic leukemia in whom an arachnoid free of leukemic cells was not achieved, one was the previously described patient treated with BCNU. T h e other patient had been treated with intrathecal aminopterin and had a normal cerebrospinal fluid cell count one month prior to death but did not have subsequent Cerebrospinal fluid examinations. I n the previous study, a high degree of correlation was found between hydrocephalus and the meningeal leukemia syndrome.4 Hydrocephalus was thought to be due probably to leukemic infiltrations within the arachnoid which interfered with cerebrospinal fluid flow and subsequently resulted in increased intracranial pressure. I n the 1953 to 1958 series 15 of 20 patients (75%) with previous clinical meningeal leukemia had hydrocephalus. Of these, 5 (25%) had 3-4+ ventricular dilatation. T h e brains of 42 other patients without the meningeal leukemia syndrome also were examined in the previous study and 9 (21%) showed mild or moderate hydrocephalus; none showed severe ventricular dilatation. I n the zy zyxwvu zyxwv zyxwv CANCER M a y 1965 present study only 5 of the 22 brains available for examination in patients with clinical meningeal leukemia (2301,) had evidence of hydrocephalus. Hydrocephalus was noted in only 3 of the 13 brains (2301,) of patients with major neurologic signs or symptoms. T h e only patient with marked (3-4+) ventricular dilation was a female child with acute lymphocytic leukemia who had a prolonged episode of meningeal leukemia in 1960 which was not treated with effective therapy for more than 4 months. Of the 72 patients without clinical meningeal leukemia, 16 (22%) had evidence of hydrocephalus at autopsy. I n none of these patients was the hydrocephalus severe. These results are summarized in Table 8. DISCUSSION T h e recognition of meningeal leukemia during life has increased, primarily due to the increase in the number of lumbar punctures performed in patients with acute leukemia. I n contrast to the 1953 to 1958 series, lumbar punctures frequently were performed in patients who did not have major central nervous system symptoms in the 1961 to 1963 series. This was an attempt to diagnose meningeal leukemia before it became symptomatic. I n addition, intrathecal chemotherapy was incorporated as part of the treatment of a number of the patients during the 1961 to 1963 period.2,3 I n those patients with acute lymphocytic leukemia in the 1961 to 1963 series an actual increase in the frequency of meningeal leukemia may have occurred because of their longer life span when compared to the 1953 to 1958 group. Previous reports have stressed that meningeal leukemia was recognized only infrequently in the pre-chemotherapy era and have postulated that the subsequent increased life span in patients with acute leukemia has allowed time for leukemic cells to infiltrate and proliferate in the central nervous ~ystem.~-ll T h e studies of Thomas, et al. have provided support for this hypothesis. I n mice inoculated subcutaneously with L1210 leukemia, only those mice whose lives were lengthened by treatment with methotrexate had diffuse leukemic infiltration in the arachnoid at autopsy.13 T h e value of doing routine lumbar punctures in patients with acute lymphocytic leukemia is demonstrated both by the relatively frequent finding of pleocytosis in the absence of neurologic symptoms and also by the pres- Val. 18 ence of leukemic arachnoid infiltration at autopsy in 3 of 6 patients with acute lymphocytic leukemia who never had a lumbar puncture. I n patients with acute myelocytic leukemia, however, at least minimal central nervous system symptoms accompanied each episode of clinical meningeal leukemia. Furthermore, only 4 of 21 patients with acute myelocytic leukemia who had no lumbar puncture done during life had leukemic infiltration of the arachnoid at autopsy. In 3 of the 4, meningeal leukemia was not an important clinical problem because of an associated fatal intracerebral hemorrhage due to intracerebral leukostasis and leukemic nodule formation. Treatment, which consisted primarily of intrathecal aminopterin in 1961 to 1963, was clearly effective in controlling clinical symptoms and cerebrospinal pleocytosis due to leukemic infiltration of the arachnoid. T h e complication of hydrocephalus largely was prevented also in the 1961 to 1963 series. T h e lengthening of time to subsequent bone marrow relapse when cerebrospinal fluid pleocytosis occurred during bone marrow remission, which was observed in the 1961 to 1963 series, may have been due either to earlier recognition of meningeal leukemia during bone marrow remission or to treatment with intrathecal aminopterin. T h e former possibility seems more likely since a recent study has shown no prolongation of bone marrow remissions in patients receiving routine monthly injections of intrathecal aminopterin, when compared to patients receiving oral therapy 0nly.3 Although intrathecal aminopterin afforded good paliation, it was, however, not successful in reducing the degree or frequency of microscopic infiltration of the arachnoid by leukemic cells. Therefore, new therapeutic techniques are needed if all leukemic cells are to be eliminated from the central nervous system in patients with acute leukemia. T h e intraventricular infusion of massive doses of antifolk agents as described by RalP may be more successful in eradicating all leukemic cells. This procedure is a formidable one, however, and cannot be used in every patient with cerebrospinal fluid pleocytosis. BCNU, which is given orally or intravenously, also may improve the treatment of meningeal leukemia. I n mice, i t is the only current drug capable of eradicating L1210 leukemia from the arachnoid and central nervous system.lZ I n humans this drug has shown promise and, despite its severe delayed toxic effects on the marrow, No. 3 zyxwvutsrqp MENINGEAL LEUKEMIA Nies et al. liver and lung, may prove to be a useful drug in attempting to eliminate those leukemic cells remaining after intensive aminopterin treatment. SUMMARY T h e clinical and pathological features of meningeal leukemia in 94 patients with acute leukemia dying between December 1961 and June 1963 were studied and compared to a similar series of patients with acute leukemia who died between 1953 and 1958. Meningeal leukemia was recognized more frequently during life in the 1961 to 1963 series than in the 1953 to 1958 group, primarily because of the increased number of lumbar punctures done in patients without major central nervous system symptoms in the later group. I n addition, an actual increase in the frequency of leukemic infiltration of the arachnoid may 553 have occurred in patients with acute lymphocytic leukemia in the 1961 to 1963 series due to their longer life span. T h e importance of routine lumbar punctures in patients with acute lymphocytic leukemia was demonstrated by a significant incidence of cerebrospinal fluid pleocytosis and leukemic cell infiltration of the arachnoid in asymptomatic patients. Treatment with intrathecal aminopterin was successful in controlling the clinical manifestations of meningeal leukemia and in preventing hydrocephalus. Despite the clinical effectiveness of therapy, however, the overaIl frequency and degree of leukemic infiltration of the arachnoid at autopsy was approximately the same in the 1961 to 1963 series as in the 1953 to 1958 series. I t is concluded that new therapeutic approaches are necessary to eradicate leukemic cells from the central nervous system of patients with acute leukemia. zyxwvuts zyx zyxwvu zyxwvutsrq REFERENCES 1. BISEL,H. F.: Letter to the Editor. Blood 11: 676, 1956. 2. FREI, E., 111, and TAYLOR,R. J.: T h e effect of continuous alternating intrathecal chemotherapy on the duration of remission in acute leukemia (abstr.). Proc. Am. Assn. Cancer Res. 4: 20, 1963. 3. FKEIREICH, E. J., KARON,M., and FREI, E., 111: Quadruple combination therapy (VAMP) for acute lymphocytic leukemia of childhood (abstr.). Proc. Am. Assn. Cancer Res. 5: 20, 1964. 4. MOORE,E. W., THOMAS, L. B., SHAW,R. K., and FREIREICH, E. J.: T h e central nervous system in acute leukemia-A postmortem study of 117 consecutive cases with particular reference to hemorrhage, leukemic infiltrations and the syndrome of meningeal leukemia. Arch. Intern. M e d . 103: 451, 1960. 5. NIES, B. A., MALMGREN, R., CHU, E., DEL VECCHIO,P., THOMAS, L., and FREIREICH, E. J.: Cytological Examination of the Central Nervous System in Acute Leukemia. In preparation. 6. RALL,D.: Experimental studies of the blood brain barrier. Cancer Res. Sup@ In press. 7. RIESELBACH, R. E., MORSE,E. E., RALL, D. P., FREI,E., 111, and FREIREICH, E. J.: Intrathecal aminopterin therapy of meningeal leukemia. Arch. Intern. Med. 111: 620, 1963. 8. SHANBRUM, E., MILLER,S., and FAIRBANKS, V. F.: Intrathccal administration of amethopterin in leukemic encephalopathy of young adults. New Eng. J. M e d . 265: 169, 1961. 9. SHAW,R. K., MOORE,E. W., FREIREICH, E. J., and THOMAS, L. B.: Meningeal leukemia-A syndrome resulting from increased intracranial pressure in patients with acute leukemia. Neurology 1 0 823, 1960. 10. STEFFEY,J. M.: T h e central nervous system manifestations of leukemia. J. Pediat. 60: 183, 1962. 11. SULLIVAN, M. P.: Intracranial complications of leukemia in children. Pediatrics 20: 757, 1957. 12. THOMAS, L. B.: Pathology of leukemia in the brain and meninges-Postmortem studies of patients with acute leukemia and of mice inoculated with LIZ10 leukemia. Cancer Res. Sup@ In press. 13. THOMAS,L. B., CHIRIGOS,. M. P., HUMPHREYS, S. R., and GOLDIN,A.: Development of meningeal leukemia (L1210) during treatment of subcutaneously inoculated mice with methotrexate. Cancer 17: 352, 1964. 14. WHITESIDE, J. A,, PHILIPS,F. S., DARGEON, H. W., and BURCHENAL, J. H.: Intrathecal amethopterin in neurologic manifestations of leukemia. Arch. Intern. Med. 101: 279, 1958. 15. WOLLNER, N., MURPHEY, M. L., and GORDON, C. S.: A study of intrathecal methotrexate (abstr.). Proc. Am. A n n . Cancer Res. 3: 74, 1959.