Received: 26 April 2016

Revised: 30 May 2016

Accepted: 8 June 2016

DOI: 10.1002/pbc.26128

Pediatric Blood & Cancer

RESEARCH ARTICLE

The American Society of Pediatric Hematology/Oncology

Leukemic blasts are present at low levels in spinal fluid in one-third of childhood acute lymphoblastic leukemia cases Mette Levinsen1

Hanne V. Marquart2

Birgitte K. Albertsen4 Cornelis Pronk7 Riitta Niinimäki11

Line Groth-Pedersen1

Mette K. Andersen5

Aina Ulvmoen8

Thomas L. Frandsen1

Goda Vaitkeviˇciene˙ 9

Mervi Taskinen12

Jonas Abrahamsson3

Maria Jeppesen1

Arja Harila-Saari6

Päivi M. Lähteenmäki10 Kjeld Schmiegelow1,13,14

for the Nordic Society of Pediatric Hematology and Oncology (NOPHO) 1 Department of Pediatrics and Adolescent

Medicine, University Hospital Rigshospitalet, Copenhagen, Denmark 2 Department of Clinical Immunology, University

Abstract Background: Central nervous system (CNS) involvement is associated with relapse in childhood acute lymphoblastic leukemia (ALL) and is a diagnostic challenge.

Hospital Rigshospitalet, Copenhagen, Denmark 3 Department of Pediatrics, Institution of Clinical

Sciences, Sahlgrenska University Hospital, Gothenburg, Sweden

Procedure: In a Nordic/Baltic prospective study, we assessed centralized flow cytometry (FCM) of locally fixed cerebrospinal fluid (CSF) samples versus local conventional cytospin-based cytology (CC) for detecting leukemic cells and evaluating kinetics of elimination of leukemic cells in CSF.

4 Department of Pediatrics, Aarhus University

Hospital, Aarhus, Denmark

Results: Among 300 patients with newly diagnosed ALL, 87 (29%) had CSF involvement by FCM,

5 Department of Clinical Genetics, University

while CC was positive in 30 (10%) of 299 patients with available CC data (P < 0.001). Patients

Hospital Rigshospitalet, Copenhagen, Denmark

with FCM+/CC+ had higher CSF leukemic blast counts compared to patients positive by FCM only

6 Department of Pediatrics, Astrid Lindgrens

(medians: 0.10 vs. 0.017 leukemic blasts/𝜇l, P = 0.006). Patients positive by FCM had higher white

Hospital, Stockholm, Sweden

blood cell counts in peripheral blood than patients negative by FCM (medians: 45 × 109 /l vs. 10

7 Department of Pediatrics, Skåne University

Hospital, Lund, Sweden 8 Department of Pediatrics, Oslo University

Hospital, Norway 9 Centre for Pediatric Oncology and Hematology,

University Children’s Hospital, Vilnius, Lithuania 10 Department of Pediatrics, Turku University

Hospital, Turku, Finland

× 109 /l, P < 0.001), were younger (medians: 3 years vs. 4 years, P = 0.03), and more frequently had T-cell ALL (18/87 vs. 16/213, P = 0.001). At treatment day 15, five of 52 patients (10%) who had CSF positive by FCM at diagnosis remained so despite at least two doses of weekly intrathecal chemotherapy. Conclusions: Longer follow-up is needed to clarify whether FCM positivity has prognostic significance and is an indicator for intensified CNS-directed therapy.

11 Department of Pediatrics, Oulu University

Hospital, Oulu, Finland 12 Children and Adolescents, Helsinki University

KEYWORDS

ALL, acute, leukemias, CSF leukemia, minimal residual disease

Hospital, Helsinki, Finland 13 Institute of Clinical Medicine, Faculty of

Health and Medical Sciences, University of Copenhagen, Denmark 14 Division of Pediatric Hematology/Oncology,

Perlmutter Cancer Center, NYU Langone Medical Center, New York Correspondence Kjeld Schmiegelow, Department of Pediatrics and Adolescent Medicine, JMC-5704, Juliane Marie Center, University Hospital Rigshospitalet, Blegdamsvej 9, DK-2100 Copenhagen, Denmark Email: [email protected]

Pediatr Blood Cancer 2016;00: 1–8

wileyonlinelibrary.com/journal/pbc

c 2016 Wiley Periodicals, Inc. 

1

2

LEVINSEN ET AL .

1

INTRODUCTION

Copenhagen, Denmark. Only results obtained by CC were used for treatment allocation, and FCM results were blinded to the treating

Risk-adapted systemic and central nervous system (CNS)-directed

physician. Some centers only had ethical approval to send fixed CSF

chemotherapy with or without cranial irradiation has increased

samples until treatment day 15 for patients with blasts at CC, while

5-year overall survival rates for childhood acute lymphoblastic

other centers sent CSF samples from all patients regardless of CC-

leukemia (ALL) to 90% or more.1 The CNS is a therapeutic sanc-

status until treatment day 15. Patients who were FCM− at diagnosis

tuary, since the blood–brain barrier and blood–cerebrospinal fluid

did not have their subsequent CSF samples analyzed by FCM. All CSF

(CSF) barrier prevent adequate cytotoxic CSF levels of most system-

samples classified as CC+ (CNS2 and CNS3) underwent central review

ically administered antileukemic agents.2 On the other hand, CNS-

at the Department of Pathology, University Hospital Rigshospitalet.

directed therapy may cause permanent neurological and neuroen-

The study was approved by local ethical committees and parents gave

docrine

damage.3–5

Thus, improved strategies for reliable capture

of patients with or without CNS involvement are needed to allow personalized intensified or reduced CNS-directed treatment.6

informed consent to study participation according to the Helsinki Declaration. Patients with newly diagnosed ALL were treated according to

CNS relapse occurs in 3–8% of patients with ALL and may account

Interfant-06 (patients below 1.00 year, EudraCT number 2005-

for 30–40% of all relapses.7–9 Furthermore, seeding from the CNS to

004599-19), EsPhALL (Ph+ ALL, NCT00287105), or NOPHO

the bone marrow has been proposed as a potential mechanism for

ALL2008 (all other patients, EudraCT number 2008-003235-20)

systemic relapse.10 The standard method for detecting CNS involve-

protocols. (For details of CNS status, treatment regimens, and MRD

ment is light microscopic examination of CSF cytospin preparations for

response, see Supplementary Appendix S1.)

the presence of leukemic cells.11,12 Patients without leukemic blasts in cytospin preparations are classified as CNS1.13 Two to three percent of patients have CNS leukemia at diagnosis (CNS3), defined as ≥5 leukocytes/𝜇l CSF with blasts in cytospin preparations, cra-

2.2 Flow cytometric immunophenotyping and fluorescence in situ hybridization (FISH)

nial nerve palsy, or cerebral mass by retinal inspection or CNS

CSF samples were collected into tubes containing TransFix

imaging.7,9,14–16

R 

(Inter-

Patients with CNS3 are at increased risk of CNS

medico Ltd., Hellerup, Denmark) and shipped to University Hospi-

relapse, but CNS3 status is associated with other risk factors, for exam-

tal Rigshospitalet. The CSF sample volume was measured and 2 ml

ple, T-cell disease or hyperleukocytosis.7,17 CNS involvement at lower

of FACSflow (BD Biosciences [BD], San Jose, CA) was added to the

levels (CNS2, defined as