Volume=20 36, Supplement=20 2, Pages S2-S16 (August = 2009)

4 of 10

ABSTRACT

FULL=20 TEXT

FULL-TEXT PDF (646=20 KB)

Central Nervous System Disease in=20 Hematologic Malignancies: Historical Perspective and = Practical=20 Applications

Ching-Hon Puia 3D"Corresponding

, Eckhard Thielb

Acute lymphoblastic leukemia = (ALL)=20 5-year survival rates are approaching 90% in children and = 50% in=20 adults who are receiving contemporary risk-directed = treatment=20 protocols. Current efforts focus not only on further = improving cure=20 rate but also on patient quality of life. Hence, all = protocols=20 decrease or limit the use of cranial irradiation as central = nervous=20 system (CNS)-directed therapy, even in patients with = high-risk=20 presenting features, such as the presence of leukemia cells = in the=20 cerebrospinal fluid (even resulting from traumatic lumbar = puncture),=20 adverse genetic features, T-cell immunophenotype, and a = large=20 leukemia cell burden. Current strategies for CNS-directed = therapy=20 involve effective systemic chemotherapy (eg, dexamethasone,=20 high-dose methotrexate, intensive asparaginase) and early=20 intensification and optimization of intrathecal therapy. = Options=20 under investigation for the treatment of relapsed or = refractory CNS=20 leukemia in ALL patients include thiotepa and intrathecal = liposomal=20 cytarabine. CNS involvement in non-Hodgkin lymphoma (NHL) is = associated with young age, advanced stage, number of = extranodal=20 sites, elevated lactate dehydrogenase, and International = Prognostic=20 Index score. Refractory CNS lymphoma in patients with NHL = carries a=20 poor prognosis, with a median survival of 2 to 6 months; the = most=20 promising treatment, autologous stem cell transplant, can = extend=20 median survival from 10 to 26 months. CNS prophylaxis is = required=20 during the initial treatment of NHL subtypes that carry a = high risk=20 of CNS relapse, such as B-cell ALL, Burkitt lymphoma, and=20 lymphoblastic lymphoma. The use of CNS prophylaxis in the = treatment=20 of diffuse large B-cell lymphoma is controversial because of = the low=20 risk of CNS relapse (=E2=88=BC5%) in this population. In = this article, we=20 review current and past practice of intrathecal therapy in = ALL and=20 NHL and the risk models that aim to identify predictors of = CNS=20 relapse in NHL.

Article Outline

=E2=80=A2 Abstract

=E2=80=A2 Diagnosing=20 CNS Disease in Patients With ALL and NHL

=E2=80=A2 Risk=20 Factors for CNS Disease Relapse: Patients With ALL

=E2=80=A2 Treatment=20 Strategies for CNS Control During Primary ALL = Therapy

=E2=80=A2 Systemic=20 Chemotherapy

=E2=80=A2 Corticosteroids

=E2=80=A2 MTX

=E2=80=A2 Asparaginase

=E2=80=A2 Thiopurines

=E2=80=A2 Intensification=20 of CSF-Directed Therapy

=E2=80=A2 Triple=20 Intrathecal Therapy Versus Intrathecal MTX

=E2=80=A2 Triple=20 Intrathecal Therapy Versus Intrathecal Cytarabine

=E2=80=A2 Optimization=20 of CSF-Directed Therapy

=E2=80=A2 Cranial=20 Irradiation

=E2=80=A2 Prognosis=20 and Treatment of Refractory CNS Leukemia

=E2=80=A2 Risk=20 Factors for Refractory CNS Disease in NHL

=E2=80=A2 Prognosis=20 and Treatment of Refractory CNS Disease in NHL

=E2=80=A2 CNS=20 Prophylaxis in Burkitt Lymphoma and Lymphoblastic = Lymphoma

=E2=80=A2 CNS=20 Prophylaxis in Diffuse Large B-Cell Lymphoma

=E2=80=A2 Conclusions

=E2=80=A2 Acknowledgment

=E2=80=A2 References

=E2=80=A2 Copyright

In the early 1970s, presymptomatic = central=20 nervous system (CNS) therapy changed the prognosis of = pediatric=20 acute lymphoblastic leukemia (ALL). Before then, more than = half of=20 the complete remissions induced by systemic chemotherapy = ended in=20 CNS relapse.1 The = contemporary=20 use of risk-directed effective systemic chemotherapy and=20 CNS-directed treatment (intrathecal chemotherapy with or = without=20 cranial radiation) increased 5-year event-free survival = (EFS) rates=20 for childhood ALL to 80% or more in some studies.1, 2 = Still, CNS=20 relapse remains an important cause of morbidity and = mortality and=20 can occur in up to 6% of ALL patients.3 = Recent clinical=20 trials in ALL treatment focus on further reducing CNS = relapse rates=20 and improving long-term survival rates by providing more = precise=20 individualized therapy that avoids over- or under-treatment. = Studies=20 also are investigating ways of minimizing late complications = such as=20 secondary cancer, neurocognitive defects, and multiple=20 endocrinopathy by eliminating or decreasing the dose of = cranial=20 irradiation. This review presents the history and current = practice=20 of CNS-directed treatments for ALL and discusses = CNS-directed=20 therapy and prophylaxis in non-Hodgkin lymphoma (NHL). It is = hoped=20 that providing a historical perspective on the role of CNS = therapy=20 in treating ALL and NHL may serve as a paradigm for the = potential=20 benefit and techniques of CNS prophylaxis in other = malignancies.

Diagnosing CNS = Disease in=20 Patients With ALL and NHL

Accurate diagnosis of CNS leukemia = or lymphoma=20 is essential. Although several methods for detecting CNS = leukemia=20 are available, the standard method is light microscopic = examination=20 of cytospin preparations of cerebrospinal fluid (CSF) = samples for=20 the presence of leukemic cells. While CSF cytology is = regarded as=20 the gold standard for diagnosis, false-negatives or = false-positives=20 occur.4 = Staining for=20 terminal deoxynucleotidyl transferase may help distinguish = normal=20 lymphocytes from leukemic cells in cases with questionable=20 morphology. Immunocytology to detect leukemia-associated = cell=20 surface antigens also has been used to establish the = diagnosis.5

In contrast to cytology, magnetic = resonance=20 imaging (MRI) has been reported to be highly sensitive for = the=20 presence of meningeal pathology but nonspecific for the = disease=20 entity.6 For = example,=20 nonmalignant meningeal enhancement may be observed following = diagnostic or therapeutic lumbar puncture.7 = Zeiser and=20 colleagues5 = found that the=20 relative sensitivity of immunocytology compared with MRI = varies for=20 different malignancies. Although MRI was more sensitive than = immunocytology in the detection of solid tumor neoplastic = meningitis=20 (immunocytology: 46%, MRI: 100%), immunocytology was more = sensitive=20 than MRI for the diagnosis of B-lineage ALL (immunocytology: = 89%,=20 MRI: 44%) and B-cell NHL (immunocytology: 95%, MRI: 48%).5

More recent advances in detection = have focused=20 on flow cytometry and polymerase chain reaction (PCR) = methods for=20 improving sensitivity. Hegde and colleagues8 = demonstrated=20 that multicolor flow cytometry using multiple antibody = panels for=20 light chains and B- and T-cell antigens is capable of = detecting=20 neoplastic cells that constitute as little as 0.2% of total = CSF=20 lymphocytes. In a more recent evaluation of 60 patients who = had CSF=20 malignancies, the diagnostic value of flow cytometry was = found to be=20 more than twice that of cytomorphology.9 = Because of its=20 increased sensitivity, the National Comprehensive Cancer = Network=20 recommends the routine use of flow cytometry for the = diagnosis of=20 CNS lymphoma,10 = but strict=20 requirements for sample processing and the lack of = standardization=20 are problematic. Because cytomorphologic examination of the = CSF=20 provides additional diagnostic value, it should still be = performed=20 in conjunction with flow cytometry.

The development of PCR as a = diagnostic=20 technique based on the detection of clonal rearrangements of = the=20 immunoglobulin or T-cell receptor genes also has potential = as a=20 sensitive test for CSF involvement from lymphoma.11, = 12 In = a=20 retrospective PCR study that used two sets of consensus = primers from=20 the immunoglobulin V and J region, five of seven specimens = from=20 patients with CNS malignancy who were suspected of, but not=20 diagnosed with, lymphoma by conventional cytology were = positive by=20 PCR. In addition, five of 13 specimens from patients with = lymphoma=20 that showed no cytologic evidence of malignancy were = positive by=20 PCR.11 = Another study=20 that used nested PCR of the complimentarity determining = region III=20 found that PCR results and CSF cytology were discordant in = 10 of 17=20 patients with primary CNS lymphoma.12 = However,=20 patients in this study had already been treated, and further = analysis of the CSF samples obtained immediately after = diagnosis was=20 required. Nevertheless, PCR may prove to be a useful method = along=20 with conventional cytology.

Risk Factors for CNS = Disease=20 Relapse: Patients With ALL

Several factors associated with an = increased=20 risk of CNS relapse have been identified for pediatric = patients with=20 ALL.3, 13 = First among=20 these is the presence of leukemic blast cells in the CSF.=20 Traditionally, patients were considered at increased risk of = CNS=20 disease if blast cells apparent in the CSF were accompanied = by a=20 white blood cell (WBC) count that exceeded 5 cells/=CE=BCL = in the CSF.=20 However, in the 1990s, it was observed that the presence of = any=20 number of blast cells in the CSF, regardless of leukocyte = count, was=20 associated with an increased risk of CNS relapse.14, = 15 = Therefore, a=20 risk classification for CNS disease was proposed: CNS1, = denoting the=20 absence of identifiable leukemic cells in CSF; CNS2, = denoting the=20 presence of blast cells in a sample that contains <5 = WBCs/=CE=BCL; and=20 CNS3, a sample that contains =E2=89=A55 WBC/=CE=BCL with = identifiable blast=20 cells, or the presence of a cerebral mass or cranial nerve = palsy=20 with leukemic cells in CSF.14

The relevance of this CNS risk = classification=20 has been a subject of controversy, because a number of = groups did=20 not observe significant differences in outcome for patients=20 categorized as CNS1 versus CNS2.15, = 16, = 17, = 18, = 19 = However, it is=20 now known that the classification is relevant in the context = of the=20 treatment received.3 = Specifically,=20 studies that did not include early CNS-directed therapy = found that=20 patients who had CSF blast cells without pleiocytosis (CNS2) = had a=20 3.2-fold increased risk (95% confidence interval [CI], = 1.1=E2=80=939.5) for=20 CNS relapse compared with patients with no leukemic cells in = their=20 CSF (CNS1),14 = whereas=20 studies that did feature early CNS-directed therapy or = prophylactic=20 cranial irradiation identified no differences in CNS relapse = rates=20 for CNS1 versus CNS2 patients.16, = 17

In the past decade, multiple studies = have=20 reported an increased incidence of CNS relapse associated = with=20 traumatic lumbar puncture with the presence of blast = cells.16, = 18, = 20 = Several=20 explanations for these observations have been proffered. = These=20 include iatrogenic introduction of blast cells, more = advanced=20 leukemia, unrecognized CNS2 or CNS3 cases in patients with = traumatic=20 punctures, or compromise of subsequent intrathecal = treatments.3, 16 = Despite this,=20 the association between traumatic lumbar puncture and CNS = relapse=20 risk renders it imperative that diagnostic lumbar punctures = be=20 performed without trauma. Therefore, St. Jude Children's = Research=20 Hospital now routinely performs this procedure under deep = sedation=20 or general anesthesia, transfuses patients with = thrombocytopenia=20 (ie, <100 =C3=97 109/L), = and=20 administers age-adjusted intrathecal treatment immediately = after CSF=20 collection.3, 21 In = addition,=20 only the most experienced clinicians perform the procedure, = since=20 experience is one of the most important determinants of = success.22

Other major factors associated with = an=20 increased risk of relapse in the CNS include a large = leukemia cell=20 burden and T-cell immunophenotype.2, 3 = Recent trials at=20 St. Jude Children's Hospital have shown that patients who = have=20 T-lineage ALL, blasts in the CSF, or pre-B ALL with the = t(1;19)=20 chromosome translocation have an independently higher risk = of CNS=20 relapse (Pui C-H, unpublished observation). In this regard, = it is=20 well recognized that T-cell ALL cases, especially those with = a high=20 presenting leukocyte count, are at an increased risk of CNS = relapse,=20 and these patients are candidates for cranial irradiation in = many=20 clinical trials.3, 23, = 24, = 25

A number of additional high-risk = genetic=20 features have been associated with CNS relapse risk. = Leukemic cell=20 genotypes associated with an unfavorable CNS prognosis = include=20 hypodiploidy (<45 chromosomes per leukemia cell), the=20 translocation t(4;11) with the MLL-AF4 fusion gene, = the=20 translocation t(1;19) with the E2A-PBX1 fusion gene, = and the=20 Philadelphia chromosome, t(9; 22).2 Host = pharmacogenetics also can contribute to treatment = responsiveness and=20 consequently to patient outcome. Two gene polymorphisms = associated=20 with predisposition for CNS relapse have been identified. = These=20 include the vitamin D receptor start site in high-risk ALL = patients=20 and the thymidylate synthase 3/3 genotype in low-risk = patients (Figure 1).26, = 27 = The=20 polymorphism of the thymidylate synthase gene has been = linked to=20 increased expression of the enzyme, one of the major targets = of=20 methotrexate (MTX), and the vitamin D receptor regulates the = expression of proteins that are important in the disposition = of=20 vincristine, etoposide, daunorubicin, prednisone, and = dexamethasone.=20 Examination of host genotype thus may be used to optimize = therapy=20 selection and individualize chemotherapy dosing. More = recently, high=20 expression of interleukin-15 in pediatric ALL but low = expression in=20 adult ALL in diagnostic leukemic cells was associated with = CNS=20 leukemia at diagnosis and subsequent relapse, suggesting a=20 pathogenetic role of this cytokine in leukemic cell = migration into=20 the CNS.28, = 29

Figure=20 1. Pharmacogenetic determinants of central = nervous=20 system (CNS) relapse. VDR, vitamin D receptor; TYM,=20 thymidylate; TYMS, thymidylate synthetase. This = research was=20 originally published in Blood. Rocha JC, et al.=20 Pharmacogenetics of outcome in children with acute=20 lymphoblastic leukemia. Blood. 2005;105:4752-8. = =C2=A9 the=20 American Society of Hematology.

Compared with childhood ALL, = relatively little=20 is known about CNS leukemia in adults with ALL. = Approximately 5% to=20 10% of adults with ALL have CNS involvement at = presentation.30, = 31, = 32, = 33, = 34 = Factors=20 associated with CNS disease at presentation include a higher = WBC=20 count, T-cell immunophenotype, and the presence of a = mediastinal=20 mass.31 = Identified=20 risk factors for CNS relapse in adult ALL include elevated = lactate=20 dehydrogenase levels, S-phase fraction, mature B-cell = phenotype,=20 high WBC counts, and CNS leukemia at diagnosis.35, = 36, = 37 = Despite=20 intensified chemotherapy and craniospinal irradiation (24 Gy = cranial=20 and 12 Gy spinal) or hematopoietic stem cell transplantation = with=20 13.2 Gy of fractionated total body irradiation in = approximately half=20 of the patients, those with CNS leukemia at diagnosis who = had=20 achieved remission had a significantly higher risk of any = type of=20 CNS relapse, combined or isolated (11.9% v 5.6%; = P =3D=20 .04), and a poorer survival rate (29% v 38% at 5 = years;=20 P <.03) compared with other patients.31

Treatment Strategies = for CNS=20 Control During Primary ALL Therapy

The recognition that ALL is a = heterogeneous=20 disease has led to treatment directed according to = phenotype,=20 genotype, and risk of relapse. CNS-directed therapy starts = early in=20 the clinical course and varies depending on the patient's = risk of=20 relapse and the intensity of systemic treatment. Important=20 prophylactic strategies for CNS control include effective = systemic=20 chemotherapy and early intensification and optimization of=20 intrathecal therapy.

Systemic=20 Chemotherapy

Corticosteroids

Corticosteroids are an essential = component of=20 ALL treatment. Some studies have suggested potential = advantages of=20 dexamethasone over prednisone for CNS control. Two studies = comparing=20 conventional dosages of dexamethasone and prednisone = reported that=20 prednisone-treated patients had lower survival rates and = experienced=20 CNS relapse at nearly twice the rate of those treated with=20 dexamethasone.38, = 39 It = was=20 suggested that such differences in CNS relapse rates may be = due to=20 either: (1) the longer biological half-life of dexamethasone = versus=20 prednisone (>32 hours v 4 to 6 hours, = respectively),38, = 40 or = (2) the=20 low-protein binding property of dexamethasone resulting in a = higher=20 CNS bioavailability.41 = However, such=20 comparisons are difficult to make without knowledge of drug=20 equivalent doses. Interestingly, a third study that used a = 50%=20 higher dose of prednisone (60 mg; n =3D 114) did not find = significant=20 differences in CNS relapse or survival rates for = dexamethasone (8=20 mg/m2; n =3D 117) compared = with=20 prednisone,42 = suggesting=20 that when equivalent doses are matched, the two drugs are = comparable=20 in CNS outcome. Nevertheless, the use of dexamethasone is = still=20 preferred in the clinical setting, because it is believed to = yield=20 better CNS as well as systemic control.

MTX

High-dose intravenous (IV) MTX was = introduced=20 as a treatment for ALL that might be expected to prevent CNS = relapse=20 because of its ability to cross the blood-brain barrier = (BBB). The=20 question of whether this is the case in practice was = addressed in a=20 meta-analysis of eight studies that compared cranial = radiation plus=20 CSF-directed therapy with high-dose MTX plus intrathecal = therapy.43 = This study=20 found that high-dose MTX reduced hematologic relapse and = improved=20 EFS but had only a marginal effect on the control of CNS = relapse=20 (odds ratio [OR], 0.81 [0.63=E2=80=931.03]; P =3D = .08). It should be=20 noted that this meta-analysis included studies using a wide = range of=20 MTX doses (0.5 g/m2 to 8 = g/m2) and dose schedules, and it is = plausible=20 that significant CNS control may be observed only with the = higher=20 MTX doses.43 = Furthermore,=20 even a MTX dose as high as 33.6 g/m2=20 would not improve outcome if the MTX treatments were given = too late=20 in the treatment course or rescued with leucovorin too early = or at=20 too high a dose.44

Asparaginase

Another drug used to treat ALL, = asparaginase,=20 also can provide effective CNS control. However, its = efficacy for=20 preventing CNS relapse depends on the type and dose of = asparaginase=20 used. Moghrabi and colleagues45 = demonstrated=20 that patients treated with asparaginase derived from = Erwinia=20 (n =3D 139; 5-year CNS relapse: 6%) experienced CNS relapse = at nearly=20 six times the rate of patients treated with Escherichia=20 coli=E2=80=93derived asparaginase (n =3D 147; 5-year CNS = relapse rate: 1%;=20 P <.01) when the two preparations were = administered at the=20 same dose (25,000 IU/m2 = weekly =C3=97 20=20 doses). However, it is well recognized that the = Erwinia=20 preparation had a much shorter half-life than the E = coli=20 preparation. Thus, one must carefully select the optimal = dose=20 depending on the type of asparaginase preparation = used.

Thiopurines

Of the thiopurines, mercaptopurine = has=20 conventionally been used to treat ALL, whereas thioguanine = has been=20 used to treat acute myeloid leukemia (AML). The preference = for=20 mercaptopurine over thioguanine in ALL was historically = based and=20 had not been clinically evaluated until recently. Several = recent=20 studies have reported superior CNS control associated with=20 thioguanine treatment of childhood ALL. One study comparing=20 mercaptopurine (75 mg/m2; n = =3D 1,017)=20 and thioguanine (50=E2=80=9360 mg/m2; n =3D=20 1,010) for the treatment of childhood ALL found that CNS = relapse=20 rates were significantly lower (P =3D .01) for=20 thioguanine-treated patients (3.5%) than for = mercaptopurine-treated=20 patients (5.5%).46 = The=20 thioguanine group also had a higher (P =3D .02) = 5-year EFS=20 (85.1%) than the mercaptopurine group (77.1%).46 A = second=20 study, which used slightly lower doses of thioguanine (40 = mg), also=20 reported a significantly lower risk (OR, 0.53; 95% CI, = 0.30=E2=80=930.92;=20 P =3D .02) of isolated CNS relapse (2.5%; n =3D 748) = than for=20 mercaptopurine (4.6%; n =3D 744).47 = Most recently,=20 a smaller pilot study (n =3D 51) reported an 8-year CNS = relapse rate=20 of zero for patients given IV thioguanine (480 mg/m2) during consolidation and = maintenance,=20 and oral thioguanine (60 mg/m2/d)=20 during maintenance.48 = However, all=20 three studies reported high rates of liver toxicity = associated with=20 thioguanine treatment.

Stork et al46 = reported that=20 20% of thioguanine-treated patients developed veno-occlusive = disease=20 (VOD) of the liver, and that 5% developed portal = hypertension. In=20 the study by Jacobs et al,48 = six patients=20 (12%) experienced reversible VOD while receiving oral = thioguanine,=20 and the study was amended to discontinue thioguanine. = Although a=20 lower thioguanine dose (40 mg) was associated with a = slightly lower=20 occurrence of VOD (11%), associated liver toxicity renders = it an=20 unacceptable alterna4tive to mercaptopurine. Therefore,=20 mercaptopurine remains the thiopurine of choice for = childhood ALL.=20 Whether or not short courses of thioguanine could improve = outcome=20 without adding undue toxic effects is = unknown.

Intensification of=20 CSF-Directed Therapy

Triple Intrathecal = Therapy=20 Versus Intrathecal MTX

MTX has been the standard drug used = for=20 intrathecal therapy. It was thought that the addition of = cytarabine=20 and a corticosteroid (ie, triple intrathecal therapy) may = have=20 additive or synergistic benefits, and the added = corticosteroid may=20 reduce arachnoiditis associated with MTX therapy. Until = recently,=20 this had never been evaluated in a clinical trial. The = Children's=20 Cancer Group 1952 clinical trial compared intrathecal MTX (n = =3D=20 1,018) with triple intrathecal therapy (MTX, cytarabine, and = hydrocortisone sodium succinate; n =3D 1,009) for = pre-symptomatic CNS=20 treatment of childhood ALL.49 = Compared with=20 intrathecal MTX, triple intrathecal therapy reduced the risk = of CNS=20 relapse (3.4% =C2=B1 1.0% v 5.9% =C2=B1 1.2%, = P =3D .004) but was=20 linked to significantly worse overall survival (6-year = survival:=20 triple intrathecal therapy 90.3% v intrathecal MTX = 94.4%;=20 P =3D .01). The reduction in overall survival was due = to a=20 significantly greater number of testicular and bone marrow = relapses=20 that have lower salvage rates than CNS relapse.49 = One=20 explanation for this result is that a so-called isolated CNS = relapse=20 could be early manifestation of systemic relapse, and the = improved=20 CNS control secured with triple intrathecal therapy favors = overt=20 leukemic relapse in other sites at a later time. Therefore, = if=20 triple intrathecal therapy is used, it is imperative that = concurrent=20 intense systemic treatment is also provided to prevent bone = marrow=20 and testicular relapse.

Triple Intrathecal = Therapy=20 Versus Intrathecal Cytarabine

In an ongoing multi-institutional = AML trial,=20 three of the first 33 patients who received intrathecal = cytarabine=20 had isolated CNS relaps. When triple intrathecal therapy was = used=20 instead of intrathecal cytarabine, none of the subsequent 79 = patients experienced CNS relapse. These observations suggest = that=20 triple intrathecal therapy is more effective than = intrathecal=20 cytarabine alone at reducing CNS relapse rates.50

Optimization of = CSF-Directed=20 Therapy

Optimal drug distribution throughout = the CNS=20 is essential for effective CNS-directed therapy. Several = factors=20 associated with intrathecal administration contribute to = optimal=20 drug distribution. For example, patient body position = following=20 intralumbar dosing has important consequences for = intraventricular=20 drug perfusion. A more than 10-fold decrease in ventricular = dose can=20 occur if patients do not remain prone for 1 hour after = dosing.51 = The use of=20 atraumatic noncutting spinal needles (22 gauge or lower) can = reduce=20 the likelihood of CSF leaks and decrease the risk of = postlumbar=20 puncture headaches3, 52 = and possible=20 cerebral bleeding or thrombosis.53 = Another=20 technique purported to improve lumbar puncture success rates = involves removal of the stylet immediately after passage = through the=20 epidermal and subcutaneous tissues (early stylet removal). = Baxter=20 and colleagues54 = demonstrated=20 that early stylet removal can significantly improve lumbar = puncture=20 success in infants <12 weeks of age (OR, 2.4; 95% CI, = 1.1=E2=80=935.2).=20 The use of good technique for diagnostic and therapeutic = lumbar=20 puncture is imperative not only for ensuring treatment = efficacy but=20 also for preventing traumatic lumbar puncture at diagnosis = leading=20 to increased CNS relapse risk.16, = 18, = 20

Cranial=20 Irradiation

Cranial irradiation has played a = central role=20 in the successful treatment of CNS leukemia since the 1960s. = However, the use of cranial irradiation in children is = complicated=20 by a very high risk of secondary neoplasms, neurocognitive = defects,=20 and multiple endocrinopathies, including lifelong growth = hormone=20 deficiency.55 = Therefore, at=20 St. Jude Children's Research Hospital, cranial radiation is = no=20 longer used as a preventive measure; it is used only as a = salvage=20 measure in patients who develop CNS relapse. In most = collaborative=20 groups, cranial irradiation is still recommended for 2% to = 20% of=20 patients at very high risk of CNS relapse, especially those = with=20 T-cell lineage or CNS leukemia (CNS3 status) at diagnosis.3 = Investigators=20 participating in the Berlin-Frankfurt-M=C3=BCnster trials = reduced the=20 dose of cranial irradiation to 12 Gy for high-risk patients = without=20 CNS leukemia at diagnosis. However, even with the lower = dose, there=20 was still a cumulative incidence of 1.7% (95% CI, = 0.1%=E2=80=933.4%) of=20 secondary neoplasms at 15 years.25

The St. Jude Total Therapy Study XV = was=20 designed to determine if prophylactic CNS irradiation could = be=20 omitted totally from the treatment of childhood ALL.21 = This study=20 adopted effective treatment components of successful = clinical trials=20 (eg, reinduction therapy, intensive asparaginase, and = intensive=20 triple intrathecal therapy) and tailored treatments = according to=20 risk and prognosis. The intensity of postremission = consolidation,=20 continuation, and reinduction therapy was based on the level = of=20 residual disease at the end of induction. The dosage of MTX = was=20 targeted to achieve a steady-state concentration of 65 = =CE=BCmol/L in=20 high-risk patients and 33 =CE=BCmol/L in low-risk patients. = The=20 mercaptopurine dose was adjusted according to thiopurine=20 methyltransferase genotype, thioguanine levels in red blood = cells,=20 and absolute neutrophil counts to ensure adequate drug = exposure but=20 to avoid excessive hematopoietic toxicities and to decrease = the risk=20 of therapy-related cancers.21 = The results=20 are encouraging, with a 5-year EFS of 86% =C2=B1 3%, an = overall survival=20 rate of 94% =C2=B1 2% (n =3D 498),2 and = a cumulative=20 risk of CNS relapse of approximately 3%.21

Prognosis and = Treatment of=20 Refractory CNS Leukemia

The prognosis for patients who = experience CNS=20 relapse varies with National Cancer Institute/Rome risk = criteria at=20 diagnosis of ALL, time from diagnosis to relapse, and = whether or not=20 cranial irradiation was used during initial treatment.3 In a = study of 74=20 children with an isolated CNS relapse, patients who had an = initial=20 remission duration exceeding 18 months had a 4-year EFS rate = of=20 77.7% =C2=B1 6.4%, but those with a first remission shorter = than 19=20 months had a 4-year EFS of only 51% =C2=B1 11.3%.56 = Adult ALL=20 patients who develop CNS recurrence have a poor prognosis. = In a=20 study of 32 patients who had CNS recurrence following = complete=20 remission, the median survival duration was 6 months; 28% of = patients were alive at 1 year, 6% were alive at 4 years, and = only=20 two patients were alive and in complete remission at 10+ = years.57 In = a second=20 study, only three of 22 patients (14%) with isolated CNS = relapse=20 were alive 2 years postrelapse, and the 5-year survival rate = was=20 estimated to be zero.58

Currently, there are a number of = treatment=20 options for patients who have refractory CNS leukemia. = Frequent=20 early triple intrathecal therapy (ie, days 1, 3, and 5) is = one=20 option that has been used in patients with Burkitt-type (L3) = leukemia.59 = Intrathecal=20 therapy with MTX plus prednisone (days 2, 3, 4, and 5) and=20 cytarabine (day 6) also has been used in patients with = B-cell=20 neoplasms.60 = However,=20 neither of these approaches has been evaluated in clinical = trials=20 that included patients with CNS disease that was refractory = to=20 conventional treatment.

Another option is intrathecal = administration=20 of liposomal cytarabine (Table 1)13; = compared with=20 unencapsulated cytarabine, it has a longer half-life (141 = hours=20 v 3.4 hours, respectively) and improved CSF = distribution.37, = 61, = 62, = 63, = 64, = 65, = 66, = 67, = 68, = 69, = 70 = Treatment with=20 liposomal cytarabine (25, 35, and 50 mg) was evaluated in a = phase I=20 trial of 18 children with overt leptomeningeal disease that = was=20 refractory to conventional treatment.61 Of = the seven=20 leukemia (ALL or AML) patients who received liposomal = cytarabine,=20 four achieved complete remission and three had a partial=20 remission.61 = This study=20 also found that the maximum tolerated dose of liposomal = cytarabine=20 is lower in children (35 mg) than in adults (50 mg), and = that=20 intrathecal liposomal cytarabine must be concomitantly = administered=20 with dexamethasone to prevent chemical arachnoiditis.61 = Additional=20 precaution should be taken with the simultaneous or closely = timed=20 administration of intrathecal liposomal cytarabine with = systemic=20 agents that penetrate the BBB (eg, high-dose MTX and = high-dose=20 cytarabine), because severe neurotoxicity can occur, as = shown in a=20 study conducted at the M.D. Anderson Cancer Center of adults = with=20 ALL.67 In = a=20 subsequent study, using a modified M.D. Anderson regimen in = which=20 liposomal cytarabine and high-dose chemotherapy were given 3 = weeks=20 apart, two of 14 adults with leukemia or lymphoma still = developed=20 significant neurologic events.69 = However, in a=20 recent pediatric study, six heavily pretreated children with = ALL and=20 CNS relapse responded well to liposomal cytarabine at doses = of 25 mg=20 to 50 mg (median, 35 mg) without neurologic complications, = despite=20 concomitant systemic administration of high-dose cytarabine = (2=20 g/m2) in five patients.70 = Whether the=20 use of age-adjusted (hence reduced) dose, concomitant = administration=20 of intrathecal prednisone, or younger age contributed to = this=20 finding is uncertain. It also could reflect simply the small = number=20 of patients studied. Additional studies are needed to learn = the=20 optimal use of this highly effective agent.

Table 1.

Summary of Studies = Using=20 Intrathecal Liposomal=20 Cytarabine


Type of Study	First Author, Year (n, age)	Liposomal Cytarabine Dose, Total Doses, = and=20 Frequency	Measures to Prevent Neurotoxicity	Concurrent Cranial RT or High-Dose=20 Chemotherapy	Neurotoxicity (=E2=89=A5grade 3)	Proportion of Patients With = =E2=89=A5Grade III=20 Neurotoxicity (95% CI)	Cytologic Response
Phase I	Kim, 199363 (n=20 =3D 12, 6-73 yr)	12.5-125 mg; 47 doses; = every=20 2-3 wk	NR	Cranial RT: 2	Encephalopathy: 3 Headache: = 2	5=20 of 47 cycles 10.6%=20 (1.8%-19.5%)	7 of 9 patients
	Chamberlain, 199564 (n=20 =3D 9, 23-67 yr)	75 mg; 18 doses; every = 2 wk	DXM oral	NR	None	0=20 of 9 cycles 0% = (0%-28%)	6 of 8 patients
	Bomgaars, 200461 (n=20 =3D 18, 4-19 yr)	25-50 mg; 78 doses; = every 2 wk=20 =C3=97 1 mo, every 4 wk =C3=97 2 mo, every 8 wk =C3=97 = 12 mo	DXM oral/IV	None	Headache: 3	3=20 of 18 cycles 16.6% = (0%-33.9%)	8 of 14 = patients
Phase II	Glantz, 199965 (n=20 =3D 14, 35-86 yr)	50 mg; 74 doses; every = 2 wk =C3=97=20 2 mo, every 4 wk =C3=97 8 mo	DXM oral/IV	None	Headache: 4 Meningismus: 2 Confusion: 2 Somnolence: = 2	10=20 of 74 cycles 13.5%=20 (5.7%-21.3%)	10 of 14 = patients
	Glantz, 199965 (n=20 =3D 31, 19-74 yr)	50 mg; 102 doses; = every 2 wk =C3=97=20 2 mo, every 4 wk =C3=97 2 mo	DXM oral/IV	Cranial or spinal RT: = 4	Headache: 4 Altered mental status: 5 Seizures: 1 Sensory/motor: 1 Drug-related meningitis: 3 CNS infection: = 3	17=20 of 31 cycles 54.8%=20 (37.3%-72.3%)	8 of 31 = patients
	Jaeckle, 200166 (n=20 =3D 53, 28-74 yr)	50 mg; 177 doses; = every 2 wk =C3=97=20 2 mo, every 4 wk =C3=97 2 mo	DXM oral/IV	Cranial RT: 13	Headache: = 2 Arachnoiditis:=20 4	6=20 of 177 cycles 3.4% = (0.7%-6%)	12 of 43 = patients
	Jabbour, 200767 (n=20 =3D 31, >18 yr)	50 mg on days 2 and 15 = of=20 hyperCVAD cycle, day 10 of MA cycle	DXM = oral/IV Liposomal cytarabine given >7 = days after=20 high-dose = cytarabine	High-dose methotrexate = and=20 cytarabine: 31	Papilloedema and blindness:=20 1 Increased intracranial pressure:=20 2 Cauda equina syndrome:=20 2	5=20 of 31 cycles16.1%=20 (3.2%-29.1%)	Prophylactic = use 1=20 combined marrow and CNS=20 relapse
Retrospective = series	Sancho, 200637 (n=20 =3D 6, 5-50 yr)	50 mg; 29 doses; every = 2 wk=20 (25 mg in 5-yr-old)	DXM oral/IV	NR	None	0=20 of 6 cycles 0% = (0%-39%)	2 of 3 patients
	Benesch, 200768 (n=20 =3D 5, 5-18 yr)	15-50 mg; 33 doses = every 2-4=20 wk	DXM oral/IV	High-dose methotrexate (8 = g/m2): = 1 TBI: = 1	Encephalopathy: = 1 Seizures:=20 1	2=20 of 5 cycles 40% = (0%-82.9%)	3 of 5 patients
	Sancho, 200737 (n=20 =3D 10, 18-57 yr)	50 mg; 39 doses every = 2 or 4=20 wk (35 mg in 18-yr-old)	DXM oral/IV	High-dose cytarabine: = 6	Headache: 3	3=20 of 10 cycles 30% = (1.6%-58.4%)	9 of 9 patients
	McClune, 200769 (n=20 =3D 14, 23-72 yr)	50 mg; 40 doses; every = 3 wk=20 (25 mg if IVT)	DXM = oral/IV Liposomal cytarabine every 3=20 wk	High-dose methotrexate = and=20 cytarabine: 14	Hyponatremia and somnolence:=20 1 Headache:=20 1	2=20 of 14 cycles 14.3% = (0%-32.6%)	Prophylactic = use No=20 CNS relapses
	Parasole, 200870 (n=20 =3D 6, 2-26 yr)	35 mg = (median=20 dose); 33 doses; every 2 wk	DXM = oral/IV Intrathecal prednisone (2=20 patients) Age-adjusted=20 dose	High-dose=20 cytarabine (2 g/m2): = 5	None	0=20 of 6 cycles 0% = (0%-39%)	6 of 6 = patients

Abbreviations: CNS, = central=20 nervous system; hyperCVAD, hyperfractionated = cyclophosphamide,=20 vincristine, doxorubicin, and dexamethasone; DXM,=20 dexamethasone; IVT, intraventricular; LP, lumbar = puncture; MA,=20 maintenance; CI, confidence interval; NR: not = reported; RT:=20 radiation therapy.

Adapted with = permission.13 =C2=A9 2008=20 Informa UK Ltd.

Neurotoxic events = reported per=20 cycle of intrathecal liposomal cytarabine (not per=20 patient).

A fourth option under investigation = for the=20 treatment of refractory CNS disease is the alkylating agent=20 thiotepa, which has excellent CNS penetration.71 = Barredo and=20 colleagues56 = demonstrated=20 that a single systemic IV dose of thiotepa (50 mg/m2 and 65 mg/m2) is effective at clearing blast = cells in=20 the CSF of patients in first isolated CNS relapse (n =3D = 19); seven of=20 the nine patients who received the 65 mg/m2 dose achieved partial or = complete=20 remission. However, intrathecal administration of 6.9 mg and = 10 mg=20 of thiotepa showed no additional benefit to combination = therapy for=20 the treatment of neoplastic meningitis in two children with = relapsed=20 ALL.72

Risk Factors for = Refractory=20 CNS Disease in NHL

The incidence of CNS relapse in = lymphoma=20 patients varies greatly for different NHL subtypes (Table 2).1, 30, = 73, = 74, = 75, = 76, = 77, = 78, = 79, = 80, = 81 = Patients with=20 primary CNS lymphoma and primary ocular lymphoma are at = highest=20 risk, and as many as 90% of these patients will develop = refractory=20 CNS disease.79, = 82 = Also at high=20 risk are adult patients with ALL, lymphoblastic lymphoma, = and=20 Burkitt lymphoma; up to 50% of these patients develop CNS = disease.1, 30, = 73 At = lowest risk=20 of CNS relapse are patients with indolent NHL (<5%) and = diffuse=20 large B-cell lymphoma (=E2=88=BC5%).75, = 83, = 84 = Several risk=20 factors for CNS involvement in NHL have been identified. = These=20 include young age, advanced stage, bone marrow involvement, = number=20 of extranodal sites, elevated lactate dehydrogenase, = B-symptoms, and=20 poor performance status.75, = 84, = 85, = 86, = 87

Table 2.

Central Nervous System = Relapse=20 Rates for Patients With Different Subtypes of = Non-Hodgkin=20 Lymphoma Who Have Not Received CNS Prophylaxis1, 30, 73, 74, 75, 76, 77, 78, 79, 80, 81


	NHL Subtype	CNS Relapse (w/o prophylaxis)
	ALL, LBL, and BL1, 30, 73	30%-50%
	DLBCL74, 75, 76	=E2=88=BC5%(2%-27%)
	PMLCL74	17%
	Breast NHL77	11%-21%
	Testicular NHL78	7%-31%
	Intraocular NHL79	50%-90%
	Indolent NHL75	=E2=88=BC 0%-4%
	Mantle cell NHL80, 81	4%-23%

Abbreviations: CNS, = central=20 nervous system; NHL, non-Hodgkin lymphoma; ALL, acute=20 lymphoblastic leukemia; LBL, lymphoblastic leukemia; = BL,=20 Burkitt lymphoma; DLBCL, diffuse large B-cell = lymphoma; PMLCL,=20 primary mediastinal thymic large B-cell=20 lymphoma.

Refractory CNS disease in lymphoma = patients=20 appears relatively early in a large majority of patients, = usually=20 occurring 5 to 12 months (range, 0=E2=80=93146 months) after = initial=20 diagnosis.75, = 83, = 84, = 86, = 87, = 88, = 89, = 90 = Intracranial=20 disease can be parenchymal or leptomeningeal. Generally, the = majority of cases of CNS lymphoma are isolated to the = meninges=20 (=E2=88=BC65%; range, 33%=E2=80=93100%), but it is not = uncommon for CNS disease to=20 be isolated to the parenchyma (=E2=88=BC30%; range, = 10%=E2=80=9352%) or present in=20 both the meninges and the parenchyma (=E2=88=BC10%; range, = 0%=E2=80=9318%).90, = 91, = 92, = 93 = CNS disease is=20 the sole manifestation of relapse in roughly 50% of patients = and=20 occurs in the context of progressive systemic disease in the = other=20 50%.93

Prognosis and = Treatment of=20 Refractory CNS Disease in NHL

The prognosis of patients with = secondary CNS=20 lymphoma is poor. Larger studies report a median survival of = 2 to 6=20 months, with a 1-year survival rate of 2% to 25% with = conventional=20 treatment.75, = 82, = 84, = 85, = 86 = Favorable=20 prognostic factors for survival in patients with refractory = CNS=20 lymphoma include normal lactate dehydrogenase levels at CNS = disease=20 manifestation, young age, CNS disease at first diagnosis, = and=20 high-dose chemotherapy for CNS disease.94

With conventional high-dose MTX and=20 intrathecal cytarabine (=C2=B1 procarbazine), the longest = median survival=20 duration achieved is 6 months.92 = More=20 substantial increases in survival time and survival rate = have been=20 achieved by combining high-dose chemotherapy with autologous = stem=20 cell transplant (ASCT) to treat secondary CNS lymphoma. One = study of=20 patients with CNS involvement (n =3D 37/750), in which 59% = of patients=20 in remission from CNS involvement received ASCT and 41% = received an=20 allogenic transplant, reported an overall 5-year survival = rate of=20 39% and a median overall survival of 10 months.95 A = longer=20 median survival duration of 26 months and a 5-year overall = survival=20 rate of 41% =C2=B1 28% was observed in a smaller study that = included=20 patients with CNS involvement (n =3D 15/48) who received = ASCT.96 In = a third=20 study of ASCT that included 62 lymphoma patients with CNS = disease,=20 the 5-year progression-free survival was 42% if CNS disease = was=20 absent at the time of transplant and 9% in patients who had = active=20 CNS disease at the time of ASCT.97 = However,=20 procedure-related deaths occurred in 8.9% of patients with = CNS=20 involvement who achieved CNS remission at the time of ASCT = and in=20 29.4% of those with active CNS disease at the time of = ASCT.97 = Other major=20 procedure-related complications of ASCT included = interstitial=20 pneumonitis, fungal and viral infections, and cardiac = failure.97 = Thus, although=20 ASCT is one of the more promising treatments for refractory = CNS=20 lymphoma, the treatment itself carries a high degree of = morbidity=20 and mortality.

It is also interesting that the = majority of=20 patients who relapsed following ASCT did so in the CNS.97 = For patients=20 with CNS disease before transplant, 13% had an isolated CNS = relapse=20 and 15% had relapse to the CNS and one other site. Not = surprisingly,=20 the CNS relapse rate was even higher in patients who had CNS = involvement at the time of transplant; 35% relapsed in the = CNS only,=20 and 11% relapsed in both CNS and bone marrow.97 = These=20 observations suggest that malignant cells in the CSF may not = have=20 been effectively cleared by pretransplant induction = chemotherapy.=20 Thus, combining more effective CNS-directed therapy with = ASCT may=20 help reduce CNS relapse rates and improve the overall = prognosis of=20 CNS lymphoma.

This approach to the treatment of = CNS relapse=20 in patients with aggressive NHL is being evaluated in the = Response=20 Adapted Therapy study (Figure 2). In this trial, all patients = receive an=20 initial treatment of MTX (4 g/m2),=20 ifosfamide (3 =C3=97 2 g/m2), and=20 liposomal cytarabine (50 mg). If patients respond well to = this=20 initial treatment (ie, complete/partial response or no = change), they=20 can continue with induction therapy for two further cycles, = during=20 which stem cells are collected. Patients who continue to = have=20 progressive disease (ie, nonresponders) following 22 days of = treatment are switched to IV cytarabine (2 =C3=97 3 g/m2), thiotepa (40 mg/m2), and liposomal cytarabine (50 = mg) and=20 maintained on this treatment until day 43. Patients are then = staged=20 and treated with high-dose chemotherapy, including = carmustine and=20 thiotepa followed by ASCT.98, = 99

Figure=20 2. Response Adapted Therapy study for the=20 treatment of central nervous system (CNS) relapse of=20 aggressive lymphoma.98, 99 AraC,=20 cytarabine; ASCT, autologous stem cell transplant; CR, = complete response; ECOG, Eastern Cooperative Oncology = Group;=20 ETO, etoposide; HD-BCNU, high-dose carmustine; IFO,=20 ifosfamide; LC, liposomal cytarabine; NC, no change; = MTX,=20 methotrexate; NHL, non-Hodgkin lymphoma; OS, overall = survival;=20 PR, partial response; RR, response rate; TTF, time to=20 treatment failure; TT, thiotepa; SC, = subcutaneous.

CNS Prophylaxis in = Burkitt=20 Lymphoma and Lymphoblastic Lymphoma

Like ALL, certain subtypes of = high-grade=20 lymphoma, such as Burkitt lymphoma and lymphoblastic = lymphoma, are=20 associated with relatively high rates of CNS relapse (Table 2).1, 30, = 73, = 74, = 75, = 76, = 77, = 78, = 79, = 80, = 81 As = CNS-directed therapy for ALL has become more aggressive over = the=20 past 3 decades, CNS relapse rates and survival rates for = adult ALL=20 have greatly improved (Table 3).30 = Similarly, the=20 adaptation of protocols from childhood B-cell ALL has = greatly=20 improved survival rates for adults with B-cell ALL.

Table 3.

Central Nervous System = Relapse=20 Rates in Adult Acute Lymphoblastic Leukemia With = Different=20 CNS-Directed Prophylaxis30


Years	Type of CNS Prophylaxis	5-Year CNS Relapse Rate (%) for Overall=20 Population
1979-1982	None	58
1982-1988	High-dose Ara-C and = MTX	25
1988-1992	High-dose Ara-C and = MTX + IT Ara-C=E2=80=A0 after CR for = high-risk patients	27
1992-1995	High-dose Ara-C=20 and high-dose MTX + early IT MTX and Ara-C=E2=80=A1 for all patients	8=C2=B6

Abbreviations: CNS, = central=20 nervous system; Ara-C, unencapsulated cytarabine; MTX, = methotrexate; IT, intrathecal; CR, complete=20 response.

Ara-C: 3 g/m2 every 12 hours for 6 = doses; MTX:=20 400-1,600 mg/m2.

=E2=80=A0

Ara-C at 100 mg weekly = (22 IT doses=20 in 12 months).

=E2=80=A1

MTX at 12 mg on day 2 and = Ara-C at=20 100 mg on day 8.

=C2=B6

Three-year CNS relapse=20 rate.

Hoelzer and colleagues36 = adapted=20 pediatric treatment regimens that consist of six short = intensive=20 5-day treatment cycles. Compared with conventional ALL = treatment=20 (ALL 01/81 study; n =3D 9), following which all patients = died,=20 survival rates obtained with the pediatric protocols B-NHL = 83 and=20 B-NHL 86 were greatly improved (overall survival: ALL 01/81: = 0%,=20 B-NHL 83: 49% and B-NHL 86: 51%; P =3D .001).36 = CNS relapse=20 rates also varied with the different combinations of = CNS-directed=20 prophylaxis used. The highest CNS relapse rate occurred with = the=20 B-NHL 83 protocol. Patients who achieved complete remission = with=20 low-dose IV MTX (0.5 g/m2) = and=20 high-dose intrathecal MTX (12 g/m2)=20 were given prophylactic cranial irradiation (24 Gy), = resulting in a=20 27% CNS relapse rate.36 A=20 substantially lower CNS relapse rate of 4% was observed in = the B-NHL=20 86 study that used systemic high-dose MTX (1.5 g/m2), triple intrathecal therapy, = and=20 subsequent cranial irradiation as prophylaxis in patients = who=20 achieved complete remission.36 A = slightly=20 higher CNS relapse rate of 12% was observed when cranial = irradiation=20 was omitted in the B-NHL 90 study (3 g/m2 MTX and triple intrathecal = therapy)=20 (Thiel E, unpublished data). A fourth ongoing trial, B-NHL = 2002, is=20 evaluating the effect of replacing cranial irradiation in = the B-NHL=20 86 protocol with a consolidation therapy of 1.5 g/m2 MTX and cytarabine (2 g/m2 for 2 days). Although optimal = CNS=20 prophylactic regimens are still being investigated, there is = universal agreement that ALL, Burkitt lymphoma, and = lymphoblastic=20 lymphoma require CNS-directed therapy to prevent CNS = relapse.76, = 100

CNS Prophylaxis in = Diffuse=20 Large B-Cell Lymphoma

Diffuse large B-cell lymphoma is an = aggressive=20 subset of NHL. There is no clear consensus as to which = patients may=20 benefit from CNS prophylaxis, and the true rate of CNS = relapse=20 following primary treatment of diffuse large B-cell lymphoma = without=20 CNS prophylaxis is unclear. Estimates range from 4% to 27%,=20 depending on many disease and patient variables, but most = studies=20 report an incidence of around 5%.74, = 75, = 76 = Although this=20 incidence is relatively low, the outcome of those patients = suffering=20 CNS relapse is extremely poor regardless of histology,101, = 102 = and=20 therefore prevention of relapse by CNS-directed therapy = needs to be=20 carefully considered.

Some studies have found that CNS = prophylaxis=20 is effective in preventing CNS relapse in diffuse large = B-cell=20 lymphoma patients. For example, CNS relapse rates were = reduced to=20 between 2% and 3% in patients who received CNS-directed = prophylaxis=20 (intrathecal MTX/high-dose MTX) following ACVBP = (doxorubicin,=20 cyclophosphamide, vindesine, bleomycin, prednisone).89, = 91 = This was about=20 half the 5% to 8% rate observed in patients treated with = standard=20 CHOP (cyclophosphamide, doxorubicin, vincristine, = prednisone) or=20 CHOP with rituximab who did not receive CNS prophylaxis (Table 4).83, = 89, = 90, = 91 = However,=20 intrathecal therapy carries a high risk of complications = (eg, CSF=20 leaks from lumbar puncture, infections or surgical = complications=20 from Ommaya reservoir placement); exposing the potentially = >90%=20 of unselected diffuse large B-cell lymphoma patients who = will not=20 develop CNS disease to these risks is unacceptable. Thus, = there is a=20 need for well-defined risk models to identify patients who = would=20 benefit most from CNS prophylaxis.

Table 4.

Central Nervous System = Relapse=20 Rates for Studies That Did or Did Not Include CNS=20 Prophylaxis83, 89, 90, 91

CHOP/CHOEP14/21


Author, Year	n	Chemotherapy	CNS Prophylaxis	CNS Relapse Rate (%)
Haioun, 200089	974	ACVBP	IT=20 MTX (15 mg) High-dose MTX (3 g/m2)	2.2
Tilly, = 200391	635	ACVBP	IT=20 MTX (15 mg) High-dose MTX (3 g/m2)	2.8
CHOP	None	8.3
Feugier, = 200483	399	CHOP	None	4.6
CHOP + rituximab	None	5.4
Boehme, = 200790	1693	Etoposide reduced CNS relapse=20 rate (P =3D=20 .017)	2.2

Abbreviations: CNS, = central=20 nervous system; ACVBP, doxorubicin, cyclophosphamide,=20 vindesine, bleomycin, prednisone; CHOP, = cyclophosphamide,=20 doxorubicin, vincristine, prednisone; CHOEP, CHOP + = etoposide;=20 IT, intrathecal.

Combined CNS relapse rate = includes=20 both isolated CNS relapses and CNS relapses that = occurred=20 concurrent with other systemic=20 relapse.

Several risk factors that = distinguish=20 subgroups with a greater risk of CNS disease have been = consistently=20 identified. Clinical risk factors include advanced stage,86, = 88, = 102=20 International Prognostic Index (IPI),83, = 89 = elevated=20 lactate dehydrogenase,75, = 84, = 87, = 89 = involvement in=20 more than one extranodal site,75, = 84, = 89 = young age,75, = 103 = low=20 albumin,75 = B-symptoms,88 = and=20 retroperitoneal lymph node involvement.75 = Many studies=20 have identified specific sites of extranodal involvement as = risk=20 factors for CNS involvement, including bone marrow,75, = 87, = 88, = 103 = breast,77 = testes,86, = 88, = 104 = epidural=20 space,100, = 101 = orbit,104 = and=20 nasal/paranasal sinuses.104

There is little consensus among = clinicians=20 regarding which diffuse large B-cell lymphoma patients = should=20 receive CNS prophylaxis. A survey of practice in Canada = reported=20 that CNS prophylaxis was given to 29% to 45% of patients = with=20 lymphoma presenting in the sinuses, testes, epidural space, = or=20 orbit.105 = Less=20 frequently, patients with advanced stage (12%) or elevated = lactate=20 dehydrogenase and extranodal disease (<10%) received CNS=20 prophylaxis. A second 2005 British survey found slightly = more=20 aggressive practices.106 = This survey=20 reported that 65% to 88% of clinicians used CNS prophylaxis = in cases=20 that involved sinuses, testes, orbit, and bone marrow. About = one=20 third (34%) treated diffuse large B-cell lymphoma with stage = IV=20 disease, and fewer than 21% treated patients with a high = IPI, more=20 than one extranodal site, or raised lactate dehydrogenase.106 = That only a=20 minority of clinicians responding to these surveys used = factors such=20 as stage IV disease, IPI score, number of extranodal sites, = and=20 elevated lactate dehydrogenase as indications for CNS = prophylaxis is=20 of concern, especially as these factors have been identified = in=20 several large series as being the most important risk = factors for=20 CNS relapse.75, = 84, = 89, = 102, = 107

Still, risk-model studies suggest = that these=20 factors may not adequately define a subgroup of patients = suitable=20 for prophylactic therapy. For example, van Besien and = colleagues84 = identified two=20 factors (elevated lactate dehydrogenase [relative risk (RR), = 7.0;=20 95% CI, 2.0=E2=80=9338.0] and >1 extranodal site [RR, = 5.5; 95% CI,=20 2.1=E2=80=9314.9]) as independent predictors of CNS = recurrence by=20 multivariate logistic regression. Yet, if only patients with = both=20 risk factors were treated with CNS prophylaxis, 15.4% of all = patients would be treated (a subgroup with a 17.4% risk of = CNS=20 recurrence), which would only capture 46% of those predicted = to=20 develop CNS disease.84 = Hollender and=20 colleagues75 = identified=20 five risk factors as significant in multivariate analysis: = elevated=20 lactate dehydrogenase (RR, 2.1; 95% CI, 1.0=E2=80=934.4), = serum albumin=20 <35 g/L (RR, 2.5; 95% CI, 1.3=E2=80=934.6), age <60 = years (RR, 2.8;=20 95% CI, 1.5=E2=80=935.4), retroperitoneal disease (RR, 1.9; = 95% CI,=20 1.0=E2=80=933.5), and >1 extranodal site (RR, 3.0; 95% = CI, 1.7=E2=80=935.4). If=20 four or more of these risk factors were present, the = probability of=20 a CNS recurrence within 5 years would be at least 25% (Table 5).75 By = this model,=20 if patients with at least four risk factors were selected = for CNS=20 prophylaxis, 12% of the overall population would be treated = (a=20 subgroup with a >25% risk of CNS relapse). However, this = would=20 result in treatment of only 54% of all patients predicted to = develop=20 CNS relapse. Thus, deciding who to treat based on known risk = factors=20 alone fails to identify about half of all patients who may = benefit=20 from prophylaxis, leaving the debate surrounding CNS = prophylaxis for=20 diffuse large B-cell lymphoma still controversial.76, = 93

Table 5.

Probability for Central = Nervous=20 System Relapse According to the Number of Risk Factors = Present=20 for Non-Hodgkin Lymphoma=20 Patients


No. of Risk Factors	No. of Patients	No. of Patients With CNS Involvement	Probability of a CNS Recurrence (%) = Within 5=20 Years (CI, 95%)
0	155	2	1.9(0%-4.6%)
1	356	6	2.0(0.4%-3.6%)
2	312	6	2.8(0.5%-5.2%)
3	202	8	6.2(1.9%-10.5%)
4	115	19	25.3(14.8%-35.8%)
5	29	7	32.7(11.6%-53.8%)

Abbreviations: CNS, = central=20 nervous system; CI, confidence interval.

Reprinted with = permission.75 =C2=A9 2002=20 European Society for Medical=20 Oncology.

Conclusions

Optimal CNS control in ALL requires = effective=20 systemic and intrathecal treatment. Isolated CNS relapse of = ALL is=20 highly curable in children who have a long initial remission = (>18=20 months) and who have not received prior cranial irradiation = but is=20 associated with a poor prognosis in adults. Liposomal = cytarabine is=20 effective in treating refractory CNS leukemia; however, = additional=20 studies are needed to optimize the use of this agent. CNS=20 prophylaxis is not required in indolent NHL, but it is = required for=20 the treatment of adult B-cell ALL, Burkitt lymphoma, and=20 lymphoblastic lymphoma. The use of CNS prophylaxis in the = treatment=20 of diffuse large B-cell lymphoma is controversial. Improved = risk=20 models that more accurately identify predictors of CNS = relapse are=20 needed so that prophylactic treatments can be targeted to=20 highest-risk NHL patients. To date, the most effective = treatment for=20 CNS relapse in NHL is ASCT. The devastating effects of CNS = relapse=20 necessitate additional research, focusing on more precise=20 risk-directed therapy to avoid over- or under-treatment, and = on=20 optimal CNS-directed treatment.

Acknowledgments

Professor Thiel wishes to thank the = GMALL=20 Study and PCNSL-G1 Study Grants of the Deutsche Krebshilfe.=20 Professor Pui is supported by Grant CA21765 from the = National=20 Institutes of Health and by the American Lebanese Syrian = Associated=20 Charities (ALSAC).

References

1.=20 1Simone JV. History of the treatment of childhood = ALL: a=20 paradigm for cancer cure. Best Pract = Res Clin=20 Haematol. 2006;19:353=E2=80=93359. Abstract=20 | Full=20 Text | Full-Text=20 PDF (77 KB) | MEDLINE<= /A>=20 | CrossRef=20

2.=20 2Pui CH, Evans WE. = Treatment of acute lymphoblastic=20 leukemia. N Engl J Med. 2006;354:166=E2=80=93178. CrossRef=20

3.=20 3Pui CH, Howard SC. = Current management and challenges of = malignant=20 disease in the CNS in paediatric leukaemia. = Lancet = Oncol.=20 2008;9:257=E2=80=93268. Ab= stract=20 | Fu= ll=20 Text | Full-Text=20 PDF (285 KB) | CrossRef=20

4.=20 4Glantz MJ, Cole BF, = Glantz LK,=20 Cobb J, Mills P, Lekos A,=20 et al. Cerebrospinal=20 fluid cytology in patients with cancer: minimizing = false-negative=20 results. Cancer. 1998;82:733=E2=80=93739.

5.=20 5Zeiser R, Burger JA, = Bley TA,=20 Windfuhr-Blum M, Schulte-Monting J,=20 Behringer DM. Clinical=20 follow-up indicates differential accuracy of magnetic = resonance=20 imaging and immunocytology of the cerebral spinal fluid for = the=20 diagnosis of neoplastic meningitis=E2=80=94a single centre=20 experience. Br J Haematol. 2004;124:762=E2=80=93768. ME= DLINE=20 | CrossRef=20

6.=20 6Phillips ME, Ryals TJ,=20 Kambhu SA, Yuh WT. Neoplastic vs inflammatory meningeal = enhancement=20 with Gd-DTPA. J Comput Assist Tomogr. 1990;14:536=E2=80=93541. MED= LINE

7.=20 7Hoelzer D, = Gokbuget N. New approaches to acute lymphoblastic = leukemia=20 in adults: where do we go?. Semin = Oncol.=20 2000;27:540=E2=80=93559. MED= LINE

8.=20 8Hegde U, Filie A, = Little RF,=20 Janik JE, Grant N, Steinberg SM,=20 et al. High = incidence of=20 occult leptomeningeal disease detected by flow cytometry in = newly=20 diagnosed aggressive B-cell lymphomas at risk for central = nervous=20 system involvement: the role of flow cytometry versus=20 cytology. Blood. 2005;105:496=E2=80=93502. ME= DLINE=20 | CrossRef=20

9.=20 9Bromberg JEC, Breems DA,=20 Kraan J, Bikker G, van der Holt B, = Smitt PS,=20 et al. CSF flow = cytometry=20 greatly improves diagnostic accuracy in CNS hematologic=20 malignancies. Neurology. 2007;68:1674=E2=80=931679. CrossRef=20

10.=20 10Brem SS, Bierman PJ, = Black P,=20 Brem H, Chamberlain MC, Chiocca EA,=20 et al. Central = nervous=20 system cancers. J Natl Compr Canc Netw. 2008;6:456=E2=80=93504.

11.=20 11Rhodes CH, Glantz MJ, = Glantz L,=20 Lekos A, Sorenson GD, Honsinger C,=20 et al. A = comparison of=20 polymerase chain reaction examination of cerebrospinal fluid = and=20 conventional cytology in the diagnosis of lymphomatous=20 meningitis. Cancer. 1996;77:543=E2=80=93548.

12.=20 12Gleissner B, Siehl J, = Korfel A,=20 Reinhardt R, Thiel E. CSF=20 evaluation in primary CNS lymphoma patients by PCR of the = CDR III=20 IgH genes. Neurology. 2002;58:390=E2=80=93396. MED= LINE

13.=20 13Bhojwani D, Pui CH. = Intrathecal liposomal cytarabine: more = friend=20 than foe?. Leuk Lymphoma. 2008;49:1427=E2=80=931430. CrossRef=20

14.=20 14Mahmoud HH, Rivera GK,=20 Hancock ML, Krance RA, Kun LE, Behm FG,=20 et al. Low = leukocyte=20 counts with blast cells in cerebrospinal fluid of children = with=20 newly diagnosed acute lymphoblastic leukemia. = N Engl J = Med.=20 1993;329:314=E2=80=93319. ME= DLINE=20 | CrossRef=20

15.=20 15Smith M, Arthur D, = Camitta B,=20 Carroll AJ, Crist W, Gaynon P,=20 et al. Uniform = approach=20 to risk classification and treatment assignment for children = with=20 acute lymphoblastic leukemia. J Clin = Oncol.=20 1996;14:18=E2=80=9324.

16.=20 16Burger B, Zimmermann M, = Mann G,=20 K=C3=BChl J, L=C3=B6ning L, Riehm H,=20 et al. Diagnostic=20 cerebrospinal fluid examination in children with acute = lymphoblastic=20 leukemia: significance of low leukocyte counts with blasts = or=20 traumatic lumbar puncture. J Clin = Oncol.=20 2003;21:184=E2=80=93188. CrossRef=20

17.=20 17Gilchrist GS, Tubergen DG, = Sather HN, et al. Low numbers of CSF blasts at diagnosis = do not=20 predict for the development of CNS leukemia in children with = intermediate-risk acute lymphoblastic leukemia: a Childrens = Cancer=20 Group report. J Clin Oncol. 1994;12:2594=E2=80=932600.

18.=20 18te Loo DM, Kamps WA, van = der=20 Does-van den Berg A, van Wering ER, de=20 Graaf SS. Prognostic = significance of blasts in the cerebrospinal fluid without=20 pleiocytosis or a traumatic lumbar puncture in children with = acute=20 lymphoblastic leukemia: experience of the Dutch Childhood = Oncology=20 Group. J Clin Oncol. 2006;24:2332=E2=80=932336. CrossRef=20

19.=20 19van den Berg H, Vet R, den = Ouden E, Behrendt H. Significance of lymphoblasts in = cerebrospinal=20 fluid in newly diagnosed pediatric acute lymphoblastic = malignancies=20 with bone marrow involvement: possible benefit of=20 dexamethasone. Med Pediatr Oncol. 1995;25:22=E2=80=9327. MEDL= INE=20 | CrossRef=20

20.=20 20Gajjar A, Harrison PL,=20 Sandlund JT, Rivera GK, Ribeiro RC, = Rubnitz JE,=20 et al. Traumatic lumbar=20 puncture at diagnosis adversely affects outcome in childhood = acute=20 lymphoblastic leukemia. Blood. 2000;96:3381=E2=80=933384. ME= DLINE

21.=20 21Pui CH, Campana D, = Pei D,=20 et al. Treating = childhood=20 acute lymphoblastic leukemia without cranial = irradiation.=20 N Engl J=20 Med. 2009;360:(in press).

22.=20 22Howard SC, Gajjar AJ, = Cheng C,=20 Kritchevsky SB, Somes GW, Harrison PL,=20 et al. Risk = factors for=20 traumatic and bloody lumbar puncture in children with acute=20 lymphoblastic leukemia. JAMA. 2002;288:2001=E2=80=932007. M= EDLINE=20 | CrossRef=20

23.=20 23Conter V, Schrappe M, = Aric=C3=BC M,=20 Reiter A, Rizzari C, D=C3=B6rdelmann M,=20 et al. Role of = cranial=20 radiotherapy for childhood T-cell acute lymphoblastic = leukemia with=20 high WBC and good response to prednisone. = J Clin = Oncol.=20 1997;15:2786=E2=80=932791.

24.=20 24Goldberg JM, Silverman LB, = Levy DE, Dalton VK, Gelber RD, = Lehmann L,=20 et al. Childhood T-cell=20 acute lymphoblastic leukemia: the Dana-Farber Cancer = Institute acute=20 lymphoblastic leukemia consortium experience. = J Clin = Oncol.=20 2003;21:3616=E2=80=933622. CrossRef=20

25.=20 25Schrappe M, Reiter A,=20 Ludwig WD, Harbott J, Zimmermann M, = Hiddemann W,=20 et al. Improved = outcome=20 in childhood acute lymphoblastic leukemia despite reduced = use of=20 anthracyclines and cranial radiotherapy: results of trial = ALL-BFM 90=20 (German-Austrian-Swiss ALL-BFM Study Group). = Blood. 2000;95:3310=E2=80=933322. ME= DLINE

26.=20 26Rocha JC, Cheng C, = Liu W,=20 Kishi S, Das S, Cook EH,=20 et al. Pharmacogenetics=20 of outcome in children with acute lymphoblastic = leukemia.=20 Blood. 2005;105:4752=E2=80=934758. M= EDLINE=20 | CrossRef=20

27.=20 27Krajinovic M, Costea I,=20 Chiasson S. Polymorphism of=20 the thymidylate synthase gene and outcome of acute = lymphoblastic=20 leukaemia. Lancet. 2002;359:1033=E2=80=931034. Ab= stract=20 | Fu= ll=20 Text | Full-Text=20 PDF (58 KB) | MEDLINE=20 | CrossRef=20

28.=20 28Cario G, Izraeli S, = Teichert A,=20 Rhein P, Skokowa J, M=C3=B6ricke A,=20 et al. High=20 interleukin-15 expression characterizes childhood acute=20 lymphoblastic leukemia with involvement of the CNS.=20 J Clin=20 Oncol. 2007;25:4813=E2=80=934820. CrossRef=20

29.=20 29Wu S, Burmeister T, = Baldus C,=20 Schwartz S, Notter M, Goekbuget N,=20 et al. Low = expression of=20 IL-15 in adult B-lineage acute lymphoblastic leukemia is = associated=20 with central nervous system involvement at initial = presentation=20 [abstract 2389]. Blood. 2007;110.

30.=20 30Cortes J, O'Brien SM, = Pierce S,=20 Keating MJ, Freireich EJ, = Kantarjian HM. The value of high-dose systemic = chemotherapy and=20 intrathecal therapy for central nervous system prophylaxis = in=20 different risk groups of adult acute lymphoblastic = leukemia.=20 Blood. 1995;86:2091=E2=80=932097. ME= DLINE

31.=20 31Lazarus HM, Richards SM,=20 Chopra R, Litzow MR, Burnett AK, = Wiernik PH,=20 et al. Central = nervous=20 system involvement in adult acute lymphoblastic leukemia at=20 diagnosis: results from the international ALL trial MRC = UKALL=20 XII/ECOG E2993. Blood. 2006;108:465=E2=80=93472. ME= DLINE=20 | CrossRef=20

32.=20 32Fi=C3=A8re D, Lepage E, = Sebban C,=20 Boucheix C, Gisselbrecht C, Vernant JP,=20 et al. Adult = acute=20 lymphoblastic leukemia: a multicentric randomized trial = testing bone=20 marrow transplantation as postremission therapy (The French = Group on=20 Therapy for Adult Acute Lymphoblastic Leukemia). = J Clin = Oncol.=20 1993;11:1990=E2=80=932001.

33.=20 33Kantarjian HM, O'Brien S,=20 Smith TL, Cortes J, Giles FJ, Beran M,=20 et al. Results = of=20 treatment with hyper-CVAD, a dose-intensive regimen, in = adult acute=20 lymphocytic leukemia. J Clin Oncol. 2000;18:547=E2=80=93561.

34.=20 34Petersdorf SH, Kopecky KJ, = Head DR, Boldt DH, Balcerzak SP, Wun T,=20 et al. Comparison of the=20 L10M consolidation regimen to an alternative regimen = including=20 escalating methotrexate/L-asparaginase for adult acute = lymphoblastic=20 leukemia: a Southwest Oncology Group study. = Leukemia. 2001;15:208=E2=80=93216. MED= LINE=20 | CrossRef=20

35.=20 35Kantarjian HM, Walters RS, = Smith TL, Keating MJ, Barlogie B, = McCredie KB,=20 et al. Identification of=20 risk groups for development of central nervous system = leukemia in=20 adults with acute lymphocytic leukemia. Blood. 1998;72:1784=E2=80=931789. ME= DLINE

36.=20 36Hoelzer D, Ludwig WD, = Thiel E,=20 Gassmann W, L=C3=B6ffler H, Fonatsch C,=20 et al. Improved = outcome=20 in adult B-cell acute lymphoblastic leukemia. = Blood. 1996;87:495=E2=80=93508. MED= LINE

37.=20 37Sancho JM, Ribera JM, = Oriol A,=20 Hernandez-Rivas JM, Rivas C, Bethencourt C,=20 et al. Central = nervous=20 system recurrence in adult patients with acute lymphoblastic = leukemia: frequency and prognosis in 467 patients without = cranial=20 irradiation for prophylaxis. Cancer. 2006;106:2540=E2=80=932546.

38.=20 38Bostrom BC, Sensel MR,=20 Sather HN, Gaynon PS, La MK, Johnston K, = et al. Dexamethasone=20 versus prednisone and daily oral versus weekly intravenous=20 mercaptopurine for patients with standard-risk acute = lymphoblastic=20 leukemia: a report from the Children's Cancer Group.=20 Blood. 2003;101:3809=E2=80=933817. M= EDLINE=20 | CrossRef=20

39.=20 39Mitchell CD, Richards SM,=20 Kinsey SE, Lilleyman J, Vora A, = Eden TO.=20 Benefit of dexamethasone = compared with=20 prednisolone for childhood acute lymphoblastic leukaemia: = results of=20 the UK Medical Research Council ALL97 randomized = trial.=20 Br J=20 Haematol. 2005;129:734=E2=80=93745. ME= DLINE=20 | CrossRef=20

40.=20 40Meikle AW, = Tyler FH. Potency and duration of action of=20 glucocorticoids (Effects of hydrocortisone, prednisone and=20 dexamethasone on human pituitary-adrenal function).=20 Am J=20 Med. 1997;63:200=E2=80=93207. Abstract=20 | Full-Text=20 PDF (975 KB) | MEDLINE= =20 | CrossRef=20

41.=20 41Balis FM, Lester CM,=20 Chrousos GP, Heideman RL, Poplack DG. = Differences in cerebrospinal fluid = penetration=20 of corticosteroids: possible relationship to the prevention = of=20 meningeal leukemia. J Clin Oncol. 1987;5:202=E2=80=93207.

42.=20 42Igarashi S, Manabe A, = Ohara A,=20 Kumagai M, Saito T, Okimoto Y,=20 et al. No = advantage of=20 dexamethasone over prednisolone for the outcome of standard- = and=20 intermediate-risk childhood acute lymphoblastic leukemia in = the=20 Tokyo Children's Cancer Study Group L95-14 protocol.=20 J Clin=20 Oncol. 2005;23:6489=E2=80=936498. CrossRef=20

43.=20 43Clarke M, Gaynon P, = Hann I,=20 Harrison G, Masera G, Peto R,=20 et al. CNS-directed=20 therapy for childhood acute lymphoblastic leukemia: = Childhood ALL=20 Collaborative Group overview of 43 randomized trials. = J Clin=20 Oncol. 2003;21:1798=E2=80=931809. CrossRef=20

44.=20 44Pui CH, Relling MV,=20 Evans WE. Is mega = dose of=20 methotrexate beneficial to patients with acute lymphoblastic = leukemia?. Leuk Lymphoma. 2006;47:2431=E2=80=932432. ME= DLINE=20 | CrossRef=20

45.=20 45Moghrabi A, Levy DE, = Asselin B,=20 Barr R, Clavell L, Hurwitz C,=20 et al. Results = of the=20 Dana-Farber Cancer Institute ALL Consortium Protocol 95-01 = for=20 children with acute lymphoblastic leukemia. = Blood. 2007;109:896=E2=80=93904. ME= DLINE=20 | CrossRef=20

46.=20 46Stork LC, Sather H,=20 Hutchinson R. Comparison of=20 mercaptopurine (mp) with thioguanine (tg) and it = methotrexate with=20 it =E2=80=9Ctriples=E2=80=9D (itt) in children with sr-all: = results of ccg1952=20 [abstract 585]. Blood. 2002;100:156a.

47.=20 47Vora A, Mitchell CD, = Lennard L,=20 Eden TO, Kinsey SE, Lilleyman J,=20 et al. Toxicity = and=20 efficacy of 6-thioguanine versus 6-mercaptopurine in = childhood=20 lymphoblastic leukaemia: a randomised trial. = Lancet. 2006;368:1339=E2=80=931348. Ab= stract=20 | Fu= ll=20 Text | Full-Text=20 PDF (202 KB) | CrossRef=20

48.=20 48Jacobs SS, Stork LC,=20 Bostrom BC, Hutchinson R, Holcenberg J,=20 Reaman GH, et al.Children's Oncology Group; = National=20 Cancer Institute Substitution=20 of oral and intravenous thioguanine for mercaptopurine in a=20 treatment regimen for children with standard risk acute=20 lymphoblastic leukemia: a collaborative Children's Oncology=20 Group/National Cancer Institute pilot trial = (CCG-1942).=20 Pediatr=20 Blood Cancer. 2007;49:250=E2=80=93255. CrossRef=20

49.=20 49Matloub Y, Lindemulder S,=20 Gaynon PS, Sather H, La M, Broxson E,=20 et al.Children's Oncology Group Intrathecal triple therapy decreases = central=20 nervous system relapse but fails to improve event-free = survival when=20 compared with intrathecal methotrexate: results of the = Children's=20 Cancer Group (CCG) 1952 study for standard-risk acute = lymphoblastic=20 leukemia, reported by the Children's Oncology Group.=20 Blood. 2006;108:1165=E2=80=931173. M= EDLINE=20 | CrossRef=20

50.=20 50Rubnitz J, Razzouk B, = Bowman P,=20 Dahl G, Lacayo N, Taub J,=20 et al. Improved = remission=20 induction rate of childhood AML: preliminary results of the = AML02=20 trial [abstract 275]. Blood. 2005;106(part=20 1 of 2):84a.=20

51.=20 51Blaney SM, Poplack DG,=20 Godwin K, McCully CL, Murphy R, = Balis FM.=20 Effect of body position on = ventricular CSF=20 methotrexate concentration following intralumbar=20 administration. J Clin Oncol. 1995;13:177=E2=80=93179.

52.=20 52Armon C, Evans RW. = Addendum to assessment: prevention of=20 post-lumbar puncture headaches: report of the Therapeutics = and=20 Technology Assessment Subcommittee of the American Academy = of=20 Neurology. Neurology. 2005;65:510=E2=80=93512. CrossRef=20

53.=20 53Stam J. Thrombosis of the cerebral veins and=20 sinuses. N Engl J Med. 2005;352:1791=E2=80=931798. CrossRef=20

54.=20 54Baxter AL, Fisher RG, = Burke BL,=20 Goldblatt SS, Isaacman DJ, Lawson ML. = Local anesthetic and stylet styles: = factors=20 associated with resident lumbar puncture success.=20 Pediatrics. 2006;117:876=E2=80=93881.

55.=20 55Pui CH, Cheng C, = Leung W,=20 Rai SN, Rivera GK, Sandlund JT,=20 et al. Extended = follow-up=20 of long-term survivors of childhood acute lymphoblastic=20 leukemia. N Engl J Med. 2003;349:640=E2=80=93649. CrossRef=20

56.=20 56Barredo JC, Devidas M,=20 Lauer SJ, Billett A, Marymont M, = Pullen J,=20 et al. Isolated = CNS=20 relapse of acute lymphoblastic leukemia treated with = intensive=20 systemic chemotherapy and delayed CNS radiation: a Pediatric = Oncology Group study. J Clin Oncol. 2006;24:3142=E2=80=933149. CrossRef=20

57.=20 57Surapaneni UR, Cortes JE,=20 Thomas D, O'Brien S, Giles FJ, Koller C, = et al. Central = nervous=20 system relapse in adults with acute lymphoblastic = leukemia.=20 Cancer. 2002;94:773=E2=80=93779.

58.=20 58Fielding AK, Richards SM,=20 Chopra R, Lazarus HM, Litzow MR, Buck G, = et al.Medical Research Council of the United Kingdom = Adult ALL=20 Working Party; Eastern Cooperative Oncology = Group Outcome of 609 adults after relapse of = acute=20 lymphoblastic leukemia (ALL); an MRC UKALL12/ECOG 2993 = study.=20 Blood. 2007;109:944=E2=80=93950. ME= DLINE=20 | CrossRef=20

59.=20 59Patte C, Auperin A, = Michon J,=20 Behrendt H, Leverger G, Frappaz D, = et al.Soci=C3=A9t=C3=A9=20 Fran=C3=A7aise d'Oncologie = P=C3=A9diatrique The Societe Francaise d'Oncologie = Pediatrique=20 LMB89 protocol: highly effective multiagent chemotherapy = tailored to=20 the tumor burden and initial response in 561 unselected = children=20 with B-cell lymphomas and L3 leukemia. Blood. 2001;97:3370=E2=80=933379. ME= DLINE=20 | CrossRef=20

60.=20 60Woessmann W, Seidemann K,=20 Mann G, Zimmermann M, Burkhardt B, = Oschlies I,=20 et al.BFM Group The=20 impact of the methotrexate administration schedule and dose = in the=20 treatment of children and adolescents with B-cell neoplasms: = a=20 report of the BFM Group Study NHL-BFM95. Blood. 2005;105:948=E2=80=93958. ME= DLINE=20 | CrossRef=20

61.=20 61Bomgaars L, Geyer JR,=20 Franklin J, Dahl G, Park J, Winick NJ,=20 et al. Phase I = trial of=20 intrathecal liposomal cytarabine in children with neoplastic = meningitis. J Clin Oncol. 2004;22:3916=E2=80=933921. CrossRef=20

62.=20 62Rueda Dominguez A, Olmos = Hidalgo D,=20 Viciana Garrido R, Torres Sanchez E. Liposomal cytarabine (DepoCyte) for = the=20 treatment of neoplastic meningitis. Clin = Transl=20 Oncol. 2005;7:232=E2=80=93238. CrossRef=20

63.=20 63Kim S, Khatibi S, = Howell SB,=20 McCully C, Balis FM, Poplack DG. Prolongation of drug exposure in = cerebrospinal=20 fluid by encapsulation into DepoFoam. Cancer = Res.=20 1993;53:1596=E2=80=931598. ME= DLINE

64.=20 64Chamberlain MC, = Kormanik P,=20 Howell SB, Kim S. Pharmacokinetics of intralumbar = DTC-101 for the=20 treatment of leptomeningeal metastases. Arch = Neurol.=20 1995;52:912=E2=80=93917. MED= LINE

65.=20 65Glantz MJ, LaFollette S,=20 Jaeckle KA, Shapiro W, Swinnen L, = Rozental JR,=20 et al. Randomized trial=20 of a slow release versus a standard formulation of = cytarabine for=20 the intrathecal treatment of lymphomatous meningitis. = J Clin=20 Oncol. 1999;17:3110=E2=80=933116.

66.=20 66Jaeckle KA, Phuphanich S,=20 Bent MJ, Aiken R, Batchelor T, = Campbell T,=20 et al. Intrathecal=20 treatment of neoplastic meningitis due to breast cancer with = a=20 slow-release formulation of cytarabine. Br J = Cancer.=20 2001;84:157=E2=80=93163. MED= LINE=20 | CrossRef=20

67.=20 67Jabbour E, O'Brien S,=20 Kantarjian H, Garcia-Manero G, Ferrajoli A,=20 Ravandi F, et al. Neurologic complications associated = with=20 intrathecal liposomal cytarabine given prophylactically in=20 combination with high-dose methotrexate and cytarabine to = patients=20 with acute lymphocytic leukemia. Blood. 2007;109:3214=E2=80=933218. M= EDLINE=20 | CrossRef=20

68.=20 68Benesch M, Sovinz P, = Krammer B,=20 Lackner H, Mann G, Schwinger W,=20 et al. Feasibility and=20 toxicity of intrathecal liposomal cytarabine in 5 children = and young=20 adults with refractory neoplastic meningitis. = J Pediatr = Hematol=20 Oncol. 2007;29:222=E2=80=93226. MED= LINE=20 | CrossRef=20

69.=20 69McClune B, Buadi FK, = Aslam N,=20 Przepiorka D. Intrathecal=20 liposomal cytarabine for prevention of meningeal disease in = patients=20 with acute lymphocytic leukemia and high-grade = lymphoma.=20 Leuk=20 Lymphoma. 2007;48:1849=E2=80=931851. CrossRef=20

70.=20 70Parasole R, Menna G, = Marra N,=20 Petruzziello F, Locatelli F, Mangione A,=20 et al. Efficacy = and=20 safety of intrathecal liposomal cytarabine for the treatment = of=20 meningeal relapse in acute lymphoblastic leukemia: = experience of two=20 pediatric institutions. Leuk = Lymphoma.=20 2008;49:1553=E2=80=931559. CrossRef=20

71.=20 71Heideman RL, Cole DE, = Balis F,=20 Sato J, Reaman GH, Packer RJ,=20 et al. Phase I = and=20 pharmacokinetic evaluation of thiotepa in the cerebrospinal = fluid=20 and plasma of pediatric patients: evidence for = dose-dependent plasma=20 clearance of thiotepa. Cancer Res. 1989;49:736=E2=80=93741. MED= LINE

72.=20 72Fisher PG, = Kadan-Lottick NS,=20 Korones DN. Intrathecal=20 thiotepa: reappraisal of an established therapy. = J Pediatr = Hematol=20 Oncol. 2002;24:274=E2=80=93278. MED= LINE=20 | CrossRef=20

73.=20 73Hoelzer D, = Gokbuget N. Treatment of lymphoblastic lymphoma in = adults. Best Pract Res Clin Haematol. = 2002;15:713=E2=80=93728. Abstract=20 | Full-Text=20 PDF (508 KB) | MEDLINE<= /A>=20 | CrossRef=20

74.=20 74Bishop PC, Wilson WH,=20 Pearson D, Janik J, Jaffe ES, = Elwood PC.=20 CNS involvement in primary = mediastinal=20 large B-cell lymphoma. J Clin Oncol. 1999;17:2479=E2=80=932485.

75.=20 75Hollender A, Kvaloy S, = Nome O,=20 Skovlund E, Lote K, Holte H. Central nervous system involvement = following=20 diagnosis of non-Hodgkin's lymphoma: a risk model.=20 Ann=20 Oncol. 2002;13:1099=E2=80=931107. ME= DLINE=20 | CrossRef=20

76.=20 76McMillan A. Central nervous system-directed = preventative=20 therapy in adults with lymphoma. Br J = Haematol.=20 2005;131:13=E2=80=9321. MED= LINE=20 | CrossRef=20

77.=20 77Aviles A, Delgado S, = Nambo MJ,=20 Neri N, Murillo E, Cleto S. Primary breast lymphoma: results of a = controlled=20 clinical trial. Oncology. 2005;69:256=E2=80=93260.

78.=20 78Zucca E, Conconi A, = Mughal TI,=20 Sarris AH, Seymour JF, Vitolo U,=20 et al.International Extranodal Lymphoma Study=20 Group Patterns of = outcome and=20 prognostic factors in primary large-cell lymphoma of the = testis in a=20 survey by the International Extranodal Lymphoma Study = Group.=20 J Clin=20 Oncol. 2003;21:20=E2=80=9327. CrossRef=20

79.=20 79Hormigo A, Abrey L,=20 Heinemann MH, DeAngelis LM. Ocular presentation of primary central = nervous=20 system lymphoma: diagnosis and treatment. = Br J = Haematol.=20 2004;126:202=E2=80=93208. ME= DLINE=20 | CrossRef=20

80.=20 80Montserrat E, Bosch F,=20 Lopez-Guillermo A, et al. CNS involvement in mantle-cell = lymphoma.=20 J Clin=20 Oncol. 1996;14:941=E2=80=93944.

81.=20 81Oinonen R, Franssila K,=20 Elonen E. Central = nervous=20 system involvement in patients with mantle cell = lymphoma.=20 Ann=20 Hematol. 1999;78:145=E2=80=93149. MED= LINE=20 | CrossRef=20

82.=20 82Jahnke K, Thiel E, = Martus P,=20 Herrlinger U, Weller M, Fischer L, = et al.German=20 Primary Central Nervous System Lymphoma Study=20 Group Relapse of = primary=20 central nervous system lymphoma: clinical features, outcome = and=20 prognostic factors. J Neurooncol. 2006;80:159=E2=80=93165. CrossRef=20

83.=20 83Feugier P, Virion JM, = Tilly H,=20 Haioun C, Marit G, Macro M,=20 et al. Incidence and risk=20 factors for central nervous system occurrence in elderly = patients=20 with diffuse large-B-cell lymphoma: influence of = rituximab.=20 Ann=20 Oncol. 2004;15:129=E2=80=93133. MED= LINE=20 | CrossRef=20

84.=20 84van Besien K, Ha CS, = Murphy S,=20 McLaughlin P, Rodriguez A, Amin K,=20 et al. Risk = factors,=20 treatment, and outcome of central nervous system recurrence = in=20 adults with intermediate-grade and immunoblastic = lymphoma.=20 Blood. 1998;91:1178=E2=80=931184. ME= DLINE

85.=20 85Colocci N, Glantz M,=20 Recht L. Prevention = and=20 treatment of central nervous system involvement by = non-Hodgkin's=20 lymphoma: a review of the literature. Semin = Neurol.=20 2004;24:395=E2=80=93404. MED= LINE=20 | CrossRef=20

86.=20 86Bollen EL, Brouwer RE,=20 Hamers S, Hermans J, Kluin M, = Sankatsing SU,=20 et al. Central = nervous=20 system relapse in non-Hodgkin lymphoma (A single-center = study of 532=20 patients). Arch Neurol. 1997;54:854=E2=80=93859. MED= LINE

87.=20 87Tomita N, Kodama F, = Sakai R,=20 Koharasawa H, Hattori M, Taguchi J,=20 et al. Predictive factors=20 for central nervous system involvement in non-Hodgkin's = lymphoma:=20 significance of very high serum LDH concentrations.=20 Leuk=20 Lymphoma. 2000;38:335=E2=80=93343. MED= LINE

88.=20 88Keldsen N, Michalski W,=20 Bentzen SM, Hansen KB, Thorling K. = Risk factors for central nervous = system=20 involvement in non-Hodgkins-lymphoma=E2=80=94a multivariate = analysis.=20 Acta=20 Oncol. 1996;35:703=E2=80=93708. CrossRef=20

89.=20 89Haioun C, Besson C, = Lepage E,=20 Thieblemont C, Simon D, Rose C,=20 et al. Incidence and risk=20 factors of central nervous system relapse in histologically=20 aggressive non-Hodgkin's lymphoma uniformly treated and = receiving=20 intrathecal central nervous system prophylaxis: a GELA study = on 974=20 patients (Groupe d'Etudes des Lymphomes de l'Adulte). = Ann=20 Oncol. 2000;11:685=E2=80=93690. MED= LINE=20 | CrossRef=20

90.=20 90Boehme V, Zeynalova S,=20 Kloess M, Loeffler M, Kaiser U, = Pfreundschuh M,=20 et al.German High-Grade Non-Hodgkin's Lymphoma Study = Group=20 (DSHNHL) Incidence = and risk=20 factors of central nervous system recurrence in aggressive=20 lymphoma=E2=80=94a survey of 1693 patients treated in = protocols of the=20 German High-Grade Non-Hodgkin's Lymphoma Study Group=20 (DSHNHL). Ann Oncol. 2007;18:149=E2=80=93157. MED= LINE=20 | CrossRef=20

91.=20 91Tilly H, Lepage E, = Coiffier B,=20 Blanc M, Herbrecht R, Bosly A, = et al.Groupe=20 d'Etude des Lymphomes de l'Adulte Intensive conventional chemotherapy = (ACVBP=20 regimen) compared with standard CHOP for poor-prognosis = aggressive=20 non-Hodgkin lymphoma. Blood. 2003;102:4284=E2=80=934289. M= EDLINE=20 | CrossRef=20

92.=20 92Bokstein F, Lossos A,=20 Lossos IS, Siegal T. Central nervous system relapse of = systemic=20 non-Hodgkin's lymphoma: results of treatment based on = high-dose=20 methotrexate combination chemotherapy. Leuk = Lymphoma.=20 2002;43:587=E2=80=93593. MED= LINE=20 | CrossRef=20

93.=20 93Hill QA, Owen RG. = CNS prophylaxis in lymphoma: who to = target and=20 what therapy to use. Blood Rev. 2006;20:319=E2=80=93332. | CrossRef=20

94.=20 94Jahnke K, Thiel E, = Martus P,=20 Schwartz S, Korfel A. Retrospective study of prognostic = factors in=20 non-Hodgkin lymphoma secondarily involving the central = nervous=20 system. Ann Hematol. 2006;85:45=E2=80=9350. MEDL= INE=20 | CrossRef=20

95.=20 95Kasamon YL, Jones RJ,=20 Piantadosi S, Ambinder RF, Abrams RA,=20 Borowitz MJ, et al. High-dose therapy and blood or marrow=20 transplantation for non-Hodgkin lymphoma with central = nervous system=20 involvement. Biol Blood Marrow Transplant. = 2005;11:93=E2=80=93100. Abstract=20 | Full=20 Text | Full-Text=20 PDF (155 KB) | MEDLINE=20 | CrossRef=20

96.=20 96Alvarnas JC, Negrin RS,=20 Horning SJ, Hu WW, Long GD, Schriber JR, = et al. High-dose therapy=20 with hematopoietic cell transplantation for patients with = central=20 nervous system involvement by non-Hodgkin's lymphoma. = Biol=20 Blood Marrow Transplant. 2000;6:352=E2=80=93358. Abstract=20 | Full-Text=20 PDF (95 KB) | MEDLINE=20 | CrossRef=20

97.=20 97Williams CD, Pearce R,=20 Taghipour G, Green ES, Philip T,=20 Goldstone AH. Autologous bone=20 marrow transplantation for patients with non-Hodgkin's = lymphoma and=20 CNS involvement: those transplanted with active CNS disease = have a=20 poor outcome=E2=80=94a report by the European Bone Marrow = Transplant=20 Lymphoma Registry. J Clin Oncol. 1994;12:2415=E2=80=932422.

98.=20 98Fischer L, Korfel A,=20 Lupp V. Kombinierte=20 systemische und intrathekale Chemotherapie mit nachfolgender = Hochdosischemotherapie und autologer = Stammzelltransplantation bei=20 ZNS-Rezidiven aggressiver Lymphome. http://www.lymphome.de/Gruppen/G-PCNSL-SG/Protokolle/NHL-Z= NS-Rezidiv/index.jsp(home page of the German Competence = Network=20 Malignant Lymphomas); 2007.

99.=20 99Fischer L, Korfel A, = Jahnke K,=20 Kiewe P, Thiel E. High-dose methotrexate and ifosfamide = for CNS=20 relapse of aggressive lymphoma [abstract 12524]. = J Clin = Oncol.=20 2007;25:606s.

100.=20 100MacKintosh FR, Colby TV,=20 Podolsky WJ, Burke JS, Hoppe RT, = Rosenfelt FP,=20 et al. Central = nervous=20 system involvement in non-Hodgkin's lymphoma: an analysis of = 105=20 cases. Cancer. 1982;49:586=E2=80=93595.

101.=20 101Bashir RM, Bierman PJ, = Vose JM,=20 Weisenburger DD, Armitage JO. Central nervous system involvement in = patients=20 with diffuse aggressive non-Hodgkin's lymphoma. = Am J Clin=20 Oncol. 1991;14:478=E2=80=93482. MED= LINE

102.=20 102Zinzani PL, Magagnoli M,=20 Frezza G, Prologo G, Gherlinzoni F, = Bendandi M,=20 et al. Isolated = central=20 nervous system relapse in aggressive non-Hodgkin's lymphoma: = the=20 Bologna experience. Leuk Lymphoma. 1999;32:571=E2=80=93576. MED= LINE

103.=20 103Ersboll J, Schultz HB,=20 Thomsen BL, Keiding N, Nissen NI. = Meningeal involvement in non-Hodgkin's = lymphoma:=20 symptoms, incidence, risk factors and treatment. = Scand J=20 Haematol. 1985;35:487=E2=80=93496.

104.=20 104Liang R, Chiu E,=20 Loke SL. Secondary = central=20 nervous system involvement by non-Hodgkin's lymphoma: the = risk=20 factors. Hematol Oncol. 1990;8:141=E2=80=93145. MEDL= INE=20 | CrossRef=20

105.=20 105Buckstein R, Lim W, = Franssen E,=20 Imrie KL. CNS = prophylaxis and=20 treatment in non-Hodgkin's lymphoma: variation in practice = and=20 lessons from the literature. Leuk = Lymphoma.=20 2003;44:955=E2=80=93962. MED= LINE=20 | CrossRef=20

106.=20 106Cheung CW, Burton C, = Smith P,=20 Linch DC, Hoskin PJ, Ardeshna KM. = Central nervous system = chemoprophylaxis in=20 non-Hodgkin lymphoma: current practice in the UK.=20 Br J=20 Haematol. 2005;131:193=E2=80=93200. ME= DLINE=20 | CrossRef=20

107.=20 107Bos GM, van Putten WL, van = der=20 Holt B, van den Bent M, Verdonck LF,=20 Hagenbeek A. For = which=20 patients with aggressive non-Hodgkin's lymphoma is = prophylaxis for=20 central nervous system disease mandatory? (Dutch HOVON=20 Group). Ann Oncol. 1998;9:191=E2=80=93194. MEDL= INE=20 | CrossRef=20

a St. Jude's Children's Research = Hospital,=20 Memphis, TN

b Charit=C3=A9 University Medicine = Berlin, Berlin,=20 Germany

Address = correspondence=20 to Ching-Hon Pui, MD, St. Jude's Children's Research = Hospital, 262=20 Danny Thomas Place, Memphis, TN 38105

STATEMENT OF CONFLICT OF = INTEREST:=20 The authors received honoraria for lectures from=20 Mundipharma.

PII: S0093-7754(09)00092-X

doi:10.1053/j.seminoncol.2009.05.002

=C2=A9 2009 = Published by=20 Elsevier Inc.

3D""=20
4 of 10