Leukemia & Lymphoma, 2014; Early Online: 1–3 © 2014 Informa UK, Ltd. ISSN: 1042-8194 print / 1029-2403 online DOI: 10.3109/10428194.2013.852667
ORIGINAL ARTICLE: CLINICAL
Sensitivity and specificity of cerebrospinal fluid flow cytometry for the diagnosis of leukemic meningitis in acute lymphoblastic leukemia/lymphoma
Leuk Lymphoma Downloaded from informahealthcare.com by University of Maastricht on 06/18/14 For personal use only.
Zahi Mitri1, Momin T. Siddiqui2, Fuad El Rassi3, Jeannine T. Holden2, Leonard T. Heffner3, Amelia Langston3, Edmund K. Waller3, Elliott Winton3, Morgan McLemore3, Leon Bernal-Mizrachi3, David Jaye2, Martha Arellano3 & Hanna Jean Khoury3 1Department of Medicine, 2Department of Pathology and Laboratory Medicine and 3Department of Hematology and
Medical Oncology, the Winship Cancer Institute of Emory University, Emory University School of Medicine, Atlanta, GA, USA
which dictates the number of prophylactic intrathecal (IT) chemotherapy injections [5]. The diagnosis of leukemic meningitis is based on the presence of leukemic cells in the CSF by light microscopy [6–8]. Given the limitations associated with cytology, flow cytometry immunophenotyping (FCI) is often applied to CSF samples [9]. The sensitivity and specificity of CSF FCI for the diagnosis of leukemic meningitis in ALL are unknown. We therefore conducted this retrospective analysis.
Abstract The presence of leukemic blasts detected by light microscopy in cerebrospinal fluid (CSF) establishes the diagnosis of leukemic meningitis in acute lymphoblastic leukemia/lymphoma (ALL). Flow cytometry immunophenotyping (FCI) is a very sensitive method that detects a minute number of aberrant cells, and is increasingly performed on CSF samples. We sought to determine the sensitivity and specificity of CSF FCI for the diagnosis of leukemic meningitis in ALL. Between November 2007 and August 2011, 800 CSF samples from 80 patients with ALL were available from diagnostic lumbar punctures (LPs; n ⫽ 80), followup LPs (n ⫽ 687) and at the time of relapse (n ⫽ 33). FCI was performed on 267 samples, and only identified aberrant cells in cytologically confirmed cases of leukemic meningitis. A blinded review of all cases with detectable CSF nucleated cells confirmed these findings. We conclude that CSF FCI has a 100% sensitivity and specificity for the detection of lymphoblasts. However, additional studies are needed to define the role this procedure plays in the diagnosis of leukemic meningitis.
Materials and methods Basic demographics, disease characteristics, CSF and bone marrow findings, systemic and IT chemotherapy treatments, complications, remission, relapse and survival of patients diagnosed with ALL between November 2007 and August 2011 at the Winship Cancer Institute of Emory University were captured onto study-specific case report forms. CSF status at diagnosis was classified according to well-accepted criteria: number of red blood cells, number of white blood cells (WBCs) and presence of blasts. Briefly, central nervous system (CNS) 1 was defined by the presence of fewer than 10 red blood cells per microliter and no identifiable leukemic blast cells after cytocentrifugation, and CNS 2 by the presence of ⱕ 5 WBCs per microliter with leukemic blast cells after cytocentrifugation. A traumatic lumbar puncture (TLP) was defined as TLP (⫺) if it had 10 red blood cells per microliter and no leukemic blast cells after cytocentrifugation [8]. Leukemic meningitis was defined by the presence of ⬎ 5 blasts per microliter [10]. Patients were classified into high or intermediate risk for CSF leukemic involvement at diagnosis based on LDH levels (⬎ or ⬍ 600 U/L) only [5], given that mitotic indices to calculate S ⫹ G2M were not available. Patients
Keywords: Lymphoblastic lymphoma, cerebrospinal fluid, flow cytometry, sensitivity, specificity
Introduction Cerebrospinal fluid (CSF) involvement by leukemic blasts occurs in less than 10% of adult patients with newly diagnosed acute lymphoblastic leukemia/lymphoblastic lymphoma (ALL) [1,2], and in 5–15% at the time of relapse [3,4]. Risk factors for CSF involvement include a high serum lactate dehydrogenase (LDH) level, and a high percentage of proliferative cells as measured by the S ⫹ G2M phases. Based on these findings, patients are classified into low, intermediate and high risk for CSF involvement,
Correspondence: Hanna Jean Khoury, MD, FACP, Emory University School of Medicine, 1165 C Clifton Road, Room C1152, Atlanta, GA 30322, USA. E-mail: [email protected] There is an accompanying commentary that discusses this paper. Please refer to the issue Table of Contents. Received 26 March 2013; revised 26 September 2013; accepted 29 September 2013
1
2
Z. Mitri et al.
diagnosed with leukemic meningitis received twice-weekly IT chemotherapy until CSF blasts cleared, followed by consolidation with 4-weekly IT injections [10]. Patients’ risk for relapse was determined according to cytogenetic results [11,12].
Leuk Lymphoma Downloaded from informahealthcare.com by University of Maastricht on 06/18/14 For personal use only.
Flow cytometry immunophenotyping CSF samples were not obtained at night or weekends. Upon completion of the procedure at the bedside, samples were hand delivered to the laboratory by the person performing the procedure. No stabilizers were added, and shortly after receipt, samples underwent standard brief hypotonic red cell lysis and were resuspended in RPMI medium with 10% fetal calf serum before staining with antibodies. Antibodies were used according to the manufacturer’s instructions and purchased from Becton Dickinson (San Jose, CA). The target number of events per tube was 10 000 using an open gate with threshold set on forward scatter just below lymphocytes. Data acquisition on the cytometer ceased once 10 000 events were acquired or when the tube contents had been depleted. For final analyses, gates were placed on lymphocytes and blasts on the forward by side scatter plot. The volume of CSF for the flow tube was 2–4 mL. Median WBCs in CSF were 9/μL, with range 1–29. When available, blood or marrow blast immunophenotype was used to customize an assay that incorporated antibodies of interest; otherwise, the antibody combination included for B-cell ALL: CD10– fluorescein isothiocyanate (FITC)/CD19–phycoerythrin (PE)/CD34–allophycocyanin (APC)/CD45–peridinin chlorophyll protein (PerCP), and for T-ALL: CD2, CD5, CD7, CD45 and/or CD3, CD4, CD8 and CD45. After centrifugation, the CSF supernatant was discarded, and residual cells were mixed with lineage-specific color combinations of four antibodies conjugated to the following fluorochromes: FITC, PE, PerCP and APC. Samples were acquired on a FACSCanto™ cytometer and analyzed for phenotypic aberrancies using Cellquest™ or Diva™ software (BD Biosciences, San Jose, CA) [13]. Statistics were descriptive, and performed using SPSS 16.0. The results were reported as a proportion. This retrospective study was approved by Emory University Institutional Review Board.
Results Patients Table I summarizes the characteristics of the 80 newly diagnosed patients with ALL. The median number of IT injections was 8 (range, 1–28) for intermediate risk and 12
Table I. Patient characteristics (n ⫽ 80). Age (years) Gender: M/F (%) Cytogenetics, n (%) Normal Poor prognosis Philadelphia chromosome Failed or not available WBC count, median (range) Lineage: B-cell/T-cell (%) LDH ⬎ 600, n (%) LDH ⬎ ULN, n (%) Treatment, n (%) HCVAD [14] COG AALL 0232 [15]
46 (16–77) 64/36 37 (46) 6 (8) 27 (34) 10 (12) 8.3 ⫻ 109/L (0.4–138) 81/19 29 (36%) 65 (81%) 63 (79%) 17 (21%)
M, male; F, female; LDH, lactate dehydrogenase; ULN, upper limit of normal; HCVAD, hyper-fractionated cyclophosphamide, vincristine, adriamycin and dexamethasone regimen; COG AALL 0232, Children’s Oncology Group Protocol AALL 0232.
(range, 0–22) for high risk patients. Median follow-up for all patients was 16 months (range, 2–45).
CSF status at diagnosis Among the 80 CSF samples obtained at the time of diagnosis, one was diagnostic for leukemic meningitis, 40 (51%) were classified as CNS 1, one (1%) as CNS 2 and 38 (48%) were TLP (⫺) [8]. FCI was performed on 66/80 (83%) diagnostic CSF samples, and confirmed the presence of aberrant cells in the one positive sample. No aberrant cells were identified in the remaining 65 samples.
Follow-up CSF analyses A total of 720 CSF samples were collected beyond the initial diagnostic LP. The median number of CSF samples available per patient was 10 (range 1–29). Of these 720 samples, 687 were collected during treatment that included: induction, consolidation, intensification, maintenance and in the post-transplant setting. Additionally, 33 CSF samples were collected from 15 patients at the time of relapse. FCI was performed on 186/687 follow-up samples (27%) and 15/33 (45%) samples at the time of relapse. During ALL treatment, analysis of two successive CSF samples from the patient who had leukemic meningitis at diagnosis showed a decreasing number of leukemic blasts by both cytology and FCI, which was followed by clearance of these leukemic cells after the third CSF analysis. One sample was suspicious for leukemic meningitis by light microscopy, and led to a repeat LP that failed to show leukemic cells by both cytology and FCI. The other 683 samples were negative by cytology, including the 183 samples with available FCI that also did not detect aberrant cells.
Table II. Results of CSF analyses in 80 adult patients with ALL. CSF samples from diagnostic LP (n ⫽ 80) Diagnostic test result Negative Positive
CSF samples during treatment and post-transplant (n ⫽ 687)
CSF samples at relapse (n ⫽ 33)
Cytology (n ⫽ 80)
FCI (n ⫽ 66)
Cytology (n ⫽ 687)
FCI (n ⫽ 186)
Cytology (n ⫽ 33)
FCI (n ⫽ 15)
79/80 1/80
65/66 1*/66
685/687 2/687
184/186 2*/186
23/33 10/33
8/15 7†/15
CSF, cerebrospinal fluid; ALL, acute lymphoblastic leukemia/lymphoma; LP, lumbar puncture; FCI, flow cytometry immunophenotyping. *CSF samples with leukemic cells detected by FCI are the same samples positive by cytology. †CSF FCI was not done in 3/10 leukemic meningitis CSF samples diagnosed by cytology.
CSF flow cytometry in adult ALL
Leuk Lymphoma Downloaded from informahealthcare.com by University of Maastricht on 06/18/14 For personal use only.
Fifteen patients experienced ALL relapse at a median of 12 months from diagnosis (range, 7–41): one had isolated CSF relapse, and 14 had systemic relapse that included CSF leukemic involvement in 5/14. Thirty-three CSF samples were collected from these 15 patients, with FCI available in 15/33 samples. FCI was performed on 7/10 samples collected from the five patients who relapsed with leukemic meningitis, and confirmed the presence of leukemic cells in all seven samples. Of the 23 CSF samples collected from the nine patients with systemic relapse only, FCI was performed on eight samples and failed to detect aberrant cells. Table II summarizes results of CSF analyses in these 80 adult patients with ALL.
Discussion In the absence of published reports, this retrospective study was designed to determine the sensitivity and specificity of CSF FCI for the detection of leukemic meningitis in adult patients with ALL. A large number of CSF samples collected at various stages in the course of ALL treatment were analyzed (n ⫽ 800), with FCI performed in 83% of diagnostic LPs, and 45% of relapsed samples. The sensitivity and specificity of FCI was 100%, as FCI as performed at our institution confirmed the presence of leukemic blasts in all CSF samples diagnosed with leukemic meningitis, and failed to detect aberrant cells in any of the samples that were negative by cytology. Of interest, in two patients who relapsed with CSF involvement, the absence of aberrant cells by FCI in the diagnostic LP was not predictive of leukemic meningitis-free survival. FCI was not performed on diagnostic samples in the other three patients with CSF involvement at relapse. In this study, we considered cytology as the gold standard against which FCI was evaluated. Given that it is not possible to ascertain whether the cytopathologist’s microscopic evaluation was influenced by the readily available FCI results, all cases with detectable CSF nucleated cells at diagnosis (n ⫽ 28) or follow-up/relapse (n ⫽ 14) were reviewed by a blinded cytologist (M.T.S.). None of these cases were considered positive for leukemic meningitis by this microscopic review. In an era when cost-effective medicine will likely reflect quality of care and influence reimbursement, this study provides an insight into a technology that is very sensitive, widely available and commonly used. Our analysis suggests that FCI has a very high sensitivity and specificity for the diagnosis of leukemic meningitis in adults with ALL.
3
However, with the availability of a cytologist with expertise in hematological malignancies, FCI did not appear to provide additional clinically relevant information as compared to cytology. Potential conflict of interest: Disclosure forms provided by the authors are available with the full text of this article at www.informahealthcare.com/lal.
References [1] Thomas X, Le QH. Central nervous system involvement in adult acute lymphoblastic leukemia. Hematology 2008;13:293–302. [2] Lazarus HM, Richards SM, Chopra R,et al. Central nervous system involvement in adult acute lymphoblastic leukemia at diagnosis: results from the international ALL trial MRC UKALL XII/ECOG E2993. Blood 2006;108:465–472. [3] Surapaneni UR, Cortes JE, Thomas D, et al. Central nervous system relapse in adults with acute lymphoblastic leukemia. Cancer 2002;94:773–779. [4] Reman O, Pigneux A , Huguet F, et al. Central nervous system involvement in adult acute lymphoblastic leukemia at diagnosis and/or at fi rst relapse: results from the GET-LALA group. Leuk Res 2008;32:1741–50. [5] Kantarjian HM, Walters RS, Smith TL, et al. Identification of risk groups for development of central nervous system leukemia in adults with acute lymphocytic leukemia. Blood 1988;72:1784–1789. [6] Mastrangelo R, Poplack D, Bleyer A , et al. Report and recommendations of the Rome workshop concerning poor-prognosis acute lymphoblastic leukemia in children: biologic bases for staging stratifi cation and treatment.Med Pediatr Oncol 1986;14:191–194. [7] Mahmoud HH, Rivera GK, Hancock ML, et al. L ow leukocyte counts with blast cells in cerebrospinal fluid of children with newly diagnosed acute lymphoblastic leukemia. N Engl J Med 1993;329:314– 319. [8] Gajjar A , Harrison PL, Sandlund JT, et al. Traumatic lumbar puncture at diagnosis adversely aff ects outcome in childhood acute lymphoblastic leukemia. Blood 2000;96:3381–3384. [9] Subirá D, Castañón S, Román A , et al. Flow cytometry and the study of central nervous disease in patients with acute leukaemia. Br J Haematol 2001;112:381–384. [10] Cortes J. Central nervous system involvement in adult acute lymphocytic leukemia. Hematol Oncol Clin North Am 2001;15:145–162. [11] Rowe JM. Prognostic factors in adult acute lymphoblastic leukaemia. Br J Haematol 2010;150:389–405. [12] Hoelzer D, Gökbuget N, Ottmann O, et al. Acute lymphoblastic leukemia. Hematology Am Soc Hematol Educ Program 2002: 162–192. [13] Li S, Juco J, Mann KP, Holden JT. Flow cytometry in the differential diagnosis of lymphocyte-rich thymoma from precursor T-cell acute lymphoblastic leukemia/lymphoblastic lymphoma. Am J Clin Pathol 2004;121:268–274. [14] Garcia-Manero G, Kantarjian HM. The hyper-CVAD regimen in adult acute lymphocytic leukemia. Hematol Oncol Clin North Am 2000;14:1381–1396, x–xi. [15] Al-Khabori M, Minden MD, Yee KW, et al. Improved survival using an intensive pediatric-based chemotherapy regimen in adults with T-cell acute lymphoblastic leukemia. Leuk Lymphoma 2010;51: 61–65.
ORIGINAL ARTICLE: CLINICAL
Sensitivity and specificity of cerebrospinal fluid flow cytometry for the diagnosis of leukemic meningitis in acute lymphoblastic leukemia/lymphoma
Leuk Lymphoma Downloaded from informahealthcare.com by University of Maastricht on 06/18/14 For personal use only.
Zahi Mitri1, Momin T. Siddiqui2, Fuad El Rassi3, Jeannine T. Holden2, Leonard T. Heffner3, Amelia Langston3, Edmund K. Waller3, Elliott Winton3, Morgan McLemore3, Leon Bernal-Mizrachi3, David Jaye2, Martha Arellano3 & Hanna Jean Khoury3 1Department of Medicine, 2Department of Pathology and Laboratory Medicine and 3Department of Hematology and
Medical Oncology, the Winship Cancer Institute of Emory University, Emory University School of Medicine, Atlanta, GA, USA
which dictates the number of prophylactic intrathecal (IT) chemotherapy injections [5]. The diagnosis of leukemic meningitis is based on the presence of leukemic cells in the CSF by light microscopy [6–8]. Given the limitations associated with cytology, flow cytometry immunophenotyping (FCI) is often applied to CSF samples [9]. The sensitivity and specificity of CSF FCI for the diagnosis of leukemic meningitis in ALL are unknown. We therefore conducted this retrospective analysis.
Abstract The presence of leukemic blasts detected by light microscopy in cerebrospinal fluid (CSF) establishes the diagnosis of leukemic meningitis in acute lymphoblastic leukemia/lymphoma (ALL). Flow cytometry immunophenotyping (FCI) is a very sensitive method that detects a minute number of aberrant cells, and is increasingly performed on CSF samples. We sought to determine the sensitivity and specificity of CSF FCI for the diagnosis of leukemic meningitis in ALL. Between November 2007 and August 2011, 800 CSF samples from 80 patients with ALL were available from diagnostic lumbar punctures (LPs; n ⫽ 80), followup LPs (n ⫽ 687) and at the time of relapse (n ⫽ 33). FCI was performed on 267 samples, and only identified aberrant cells in cytologically confirmed cases of leukemic meningitis. A blinded review of all cases with detectable CSF nucleated cells confirmed these findings. We conclude that CSF FCI has a 100% sensitivity and specificity for the detection of lymphoblasts. However, additional studies are needed to define the role this procedure plays in the diagnosis of leukemic meningitis.
Materials and methods Basic demographics, disease characteristics, CSF and bone marrow findings, systemic and IT chemotherapy treatments, complications, remission, relapse and survival of patients diagnosed with ALL between November 2007 and August 2011 at the Winship Cancer Institute of Emory University were captured onto study-specific case report forms. CSF status at diagnosis was classified according to well-accepted criteria: number of red blood cells, number of white blood cells (WBCs) and presence of blasts. Briefly, central nervous system (CNS) 1 was defined by the presence of fewer than 10 red blood cells per microliter and no identifiable leukemic blast cells after cytocentrifugation, and CNS 2 by the presence of ⱕ 5 WBCs per microliter with leukemic blast cells after cytocentrifugation. A traumatic lumbar puncture (TLP) was defined as TLP (⫺) if it had 10 red blood cells per microliter and no leukemic blast cells after cytocentrifugation [8]. Leukemic meningitis was defined by the presence of ⬎ 5 blasts per microliter [10]. Patients were classified into high or intermediate risk for CSF leukemic involvement at diagnosis based on LDH levels (⬎ or ⬍ 600 U/L) only [5], given that mitotic indices to calculate S ⫹ G2M were not available. Patients
Keywords: Lymphoblastic lymphoma, cerebrospinal fluid, flow cytometry, sensitivity, specificity
Introduction Cerebrospinal fluid (CSF) involvement by leukemic blasts occurs in less than 10% of adult patients with newly diagnosed acute lymphoblastic leukemia/lymphoblastic lymphoma (ALL) [1,2], and in 5–15% at the time of relapse [3,4]. Risk factors for CSF involvement include a high serum lactate dehydrogenase (LDH) level, and a high percentage of proliferative cells as measured by the S ⫹ G2M phases. Based on these findings, patients are classified into low, intermediate and high risk for CSF involvement,
Correspondence: Hanna Jean Khoury, MD, FACP, Emory University School of Medicine, 1165 C Clifton Road, Room C1152, Atlanta, GA 30322, USA. E-mail: [email protected] There is an accompanying commentary that discusses this paper. Please refer to the issue Table of Contents. Received 26 March 2013; revised 26 September 2013; accepted 29 September 2013
1
2
Z. Mitri et al.
diagnosed with leukemic meningitis received twice-weekly IT chemotherapy until CSF blasts cleared, followed by consolidation with 4-weekly IT injections [10]. Patients’ risk for relapse was determined according to cytogenetic results [11,12].
Leuk Lymphoma Downloaded from informahealthcare.com by University of Maastricht on 06/18/14 For personal use only.
Flow cytometry immunophenotyping CSF samples were not obtained at night or weekends. Upon completion of the procedure at the bedside, samples were hand delivered to the laboratory by the person performing the procedure. No stabilizers were added, and shortly after receipt, samples underwent standard brief hypotonic red cell lysis and were resuspended in RPMI medium with 10% fetal calf serum before staining with antibodies. Antibodies were used according to the manufacturer’s instructions and purchased from Becton Dickinson (San Jose, CA). The target number of events per tube was 10 000 using an open gate with threshold set on forward scatter just below lymphocytes. Data acquisition on the cytometer ceased once 10 000 events were acquired or when the tube contents had been depleted. For final analyses, gates were placed on lymphocytes and blasts on the forward by side scatter plot. The volume of CSF for the flow tube was 2–4 mL. Median WBCs in CSF were 9/μL, with range 1–29. When available, blood or marrow blast immunophenotype was used to customize an assay that incorporated antibodies of interest; otherwise, the antibody combination included for B-cell ALL: CD10– fluorescein isothiocyanate (FITC)/CD19–phycoerythrin (PE)/CD34–allophycocyanin (APC)/CD45–peridinin chlorophyll protein (PerCP), and for T-ALL: CD2, CD5, CD7, CD45 and/or CD3, CD4, CD8 and CD45. After centrifugation, the CSF supernatant was discarded, and residual cells were mixed with lineage-specific color combinations of four antibodies conjugated to the following fluorochromes: FITC, PE, PerCP and APC. Samples were acquired on a FACSCanto™ cytometer and analyzed for phenotypic aberrancies using Cellquest™ or Diva™ software (BD Biosciences, San Jose, CA) [13]. Statistics were descriptive, and performed using SPSS 16.0. The results were reported as a proportion. This retrospective study was approved by Emory University Institutional Review Board.
Results Patients Table I summarizes the characteristics of the 80 newly diagnosed patients with ALL. The median number of IT injections was 8 (range, 1–28) for intermediate risk and 12
Table I. Patient characteristics (n ⫽ 80). Age (years) Gender: M/F (%) Cytogenetics, n (%) Normal Poor prognosis Philadelphia chromosome Failed or not available WBC count, median (range) Lineage: B-cell/T-cell (%) LDH ⬎ 600, n (%) LDH ⬎ ULN, n (%) Treatment, n (%) HCVAD [14] COG AALL 0232 [15]
46 (16–77) 64/36 37 (46) 6 (8) 27 (34) 10 (12) 8.3 ⫻ 109/L (0.4–138) 81/19 29 (36%) 65 (81%) 63 (79%) 17 (21%)
M, male; F, female; LDH, lactate dehydrogenase; ULN, upper limit of normal; HCVAD, hyper-fractionated cyclophosphamide, vincristine, adriamycin and dexamethasone regimen; COG AALL 0232, Children’s Oncology Group Protocol AALL 0232.
(range, 0–22) for high risk patients. Median follow-up for all patients was 16 months (range, 2–45).
CSF status at diagnosis Among the 80 CSF samples obtained at the time of diagnosis, one was diagnostic for leukemic meningitis, 40 (51%) were classified as CNS 1, one (1%) as CNS 2 and 38 (48%) were TLP (⫺) [8]. FCI was performed on 66/80 (83%) diagnostic CSF samples, and confirmed the presence of aberrant cells in the one positive sample. No aberrant cells were identified in the remaining 65 samples.
Follow-up CSF analyses A total of 720 CSF samples were collected beyond the initial diagnostic LP. The median number of CSF samples available per patient was 10 (range 1–29). Of these 720 samples, 687 were collected during treatment that included: induction, consolidation, intensification, maintenance and in the post-transplant setting. Additionally, 33 CSF samples were collected from 15 patients at the time of relapse. FCI was performed on 186/687 follow-up samples (27%) and 15/33 (45%) samples at the time of relapse. During ALL treatment, analysis of two successive CSF samples from the patient who had leukemic meningitis at diagnosis showed a decreasing number of leukemic blasts by both cytology and FCI, which was followed by clearance of these leukemic cells after the third CSF analysis. One sample was suspicious for leukemic meningitis by light microscopy, and led to a repeat LP that failed to show leukemic cells by both cytology and FCI. The other 683 samples were negative by cytology, including the 183 samples with available FCI that also did not detect aberrant cells.
Table II. Results of CSF analyses in 80 adult patients with ALL. CSF samples from diagnostic LP (n ⫽ 80) Diagnostic test result Negative Positive
CSF samples during treatment and post-transplant (n ⫽ 687)
CSF samples at relapse (n ⫽ 33)
Cytology (n ⫽ 80)
FCI (n ⫽ 66)
Cytology (n ⫽ 687)
FCI (n ⫽ 186)
Cytology (n ⫽ 33)
FCI (n ⫽ 15)
79/80 1/80
65/66 1*/66
685/687 2/687
184/186 2*/186
23/33 10/33
8/15 7†/15
CSF, cerebrospinal fluid; ALL, acute lymphoblastic leukemia/lymphoma; LP, lumbar puncture; FCI, flow cytometry immunophenotyping. *CSF samples with leukemic cells detected by FCI are the same samples positive by cytology. †CSF FCI was not done in 3/10 leukemic meningitis CSF samples diagnosed by cytology.
CSF flow cytometry in adult ALL
Leuk Lymphoma Downloaded from informahealthcare.com by University of Maastricht on 06/18/14 For personal use only.
Fifteen patients experienced ALL relapse at a median of 12 months from diagnosis (range, 7–41): one had isolated CSF relapse, and 14 had systemic relapse that included CSF leukemic involvement in 5/14. Thirty-three CSF samples were collected from these 15 patients, with FCI available in 15/33 samples. FCI was performed on 7/10 samples collected from the five patients who relapsed with leukemic meningitis, and confirmed the presence of leukemic cells in all seven samples. Of the 23 CSF samples collected from the nine patients with systemic relapse only, FCI was performed on eight samples and failed to detect aberrant cells. Table II summarizes results of CSF analyses in these 80 adult patients with ALL.
Discussion In the absence of published reports, this retrospective study was designed to determine the sensitivity and specificity of CSF FCI for the detection of leukemic meningitis in adult patients with ALL. A large number of CSF samples collected at various stages in the course of ALL treatment were analyzed (n ⫽ 800), with FCI performed in 83% of diagnostic LPs, and 45% of relapsed samples. The sensitivity and specificity of FCI was 100%, as FCI as performed at our institution confirmed the presence of leukemic blasts in all CSF samples diagnosed with leukemic meningitis, and failed to detect aberrant cells in any of the samples that were negative by cytology. Of interest, in two patients who relapsed with CSF involvement, the absence of aberrant cells by FCI in the diagnostic LP was not predictive of leukemic meningitis-free survival. FCI was not performed on diagnostic samples in the other three patients with CSF involvement at relapse. In this study, we considered cytology as the gold standard against which FCI was evaluated. Given that it is not possible to ascertain whether the cytopathologist’s microscopic evaluation was influenced by the readily available FCI results, all cases with detectable CSF nucleated cells at diagnosis (n ⫽ 28) or follow-up/relapse (n ⫽ 14) were reviewed by a blinded cytologist (M.T.S.). None of these cases were considered positive for leukemic meningitis by this microscopic review. In an era when cost-effective medicine will likely reflect quality of care and influence reimbursement, this study provides an insight into a technology that is very sensitive, widely available and commonly used. Our analysis suggests that FCI has a very high sensitivity and specificity for the diagnosis of leukemic meningitis in adults with ALL.
3
However, with the availability of a cytologist with expertise in hematological malignancies, FCI did not appear to provide additional clinically relevant information as compared to cytology. Potential conflict of interest: Disclosure forms provided by the authors are available with the full text of this article at www.informahealthcare.com/lal.
References [1] Thomas X, Le QH. Central nervous system involvement in adult acute lymphoblastic leukemia. Hematology 2008;13:293–302. [2] Lazarus HM, Richards SM, Chopra R,et al. Central nervous system involvement in adult acute lymphoblastic leukemia at diagnosis: results from the international ALL trial MRC UKALL XII/ECOG E2993. Blood 2006;108:465–472. [3] Surapaneni UR, Cortes JE, Thomas D, et al. Central nervous system relapse in adults with acute lymphoblastic leukemia. Cancer 2002;94:773–779. [4] Reman O, Pigneux A , Huguet F, et al. Central nervous system involvement in adult acute lymphoblastic leukemia at diagnosis and/or at fi rst relapse: results from the GET-LALA group. Leuk Res 2008;32:1741–50. [5] Kantarjian HM, Walters RS, Smith TL, et al. Identification of risk groups for development of central nervous system leukemia in adults with acute lymphocytic leukemia. Blood 1988;72:1784–1789. [6] Mastrangelo R, Poplack D, Bleyer A , et al. Report and recommendations of the Rome workshop concerning poor-prognosis acute lymphoblastic leukemia in children: biologic bases for staging stratifi cation and treatment.Med Pediatr Oncol 1986;14:191–194. [7] Mahmoud HH, Rivera GK, Hancock ML, et al. L ow leukocyte counts with blast cells in cerebrospinal fluid of children with newly diagnosed acute lymphoblastic leukemia. N Engl J Med 1993;329:314– 319. [8] Gajjar A , Harrison PL, Sandlund JT, et al. Traumatic lumbar puncture at diagnosis adversely aff ects outcome in childhood acute lymphoblastic leukemia. Blood 2000;96:3381–3384. [9] Subirá D, Castañón S, Román A , et al. Flow cytometry and the study of central nervous disease in patients with acute leukaemia. Br J Haematol 2001;112:381–384. [10] Cortes J. Central nervous system involvement in adult acute lymphocytic leukemia. Hematol Oncol Clin North Am 2001;15:145–162. [11] Rowe JM. Prognostic factors in adult acute lymphoblastic leukaemia. Br J Haematol 2010;150:389–405. [12] Hoelzer D, Gökbuget N, Ottmann O, et al. Acute lymphoblastic leukemia. Hematology Am Soc Hematol Educ Program 2002: 162–192. [13] Li S, Juco J, Mann KP, Holden JT. Flow cytometry in the differential diagnosis of lymphocyte-rich thymoma from precursor T-cell acute lymphoblastic leukemia/lymphoblastic lymphoma. Am J Clin Pathol 2004;121:268–274. [14] Garcia-Manero G, Kantarjian HM. The hyper-CVAD regimen in adult acute lymphocytic leukemia. Hematol Oncol Clin North Am 2000;14:1381–1396, x–xi. [15] Al-Khabori M, Minden MD, Yee KW, et al. Improved survival using an intensive pediatric-based chemotherapy regimen in adults with T-cell acute lymphoblastic leukemia. Leuk Lymphoma 2010;51: 61–65.
