Original Article
Monosomal Karyotype Predicts Inferior Survival Independently of a Complex Karyotype in Patients With Myelodysplastic Syndromes Zoe K. McQuilten, MBBS1; Vijaya Sundararajan, MD2; Nick Andrianopoulos, MBBS, MBiostat1; David J. Curtis, MD, PhD3; Erica M. Wood, MBBS1,3; Lynda J. Campbell, MBBS2,4; and Meaghan Wall, MBBS, PhD4
BACKGROUND: Conflicting data exist about the impact of a monosomal karyotype (MK) on overall survival (OS) for patients with myelodysplastic syndromes (MDSs) and particularly for those with a complex karyotype (CK). This study was aimed at determining whether an MK is associated with OS independently of the number of cytogenetic abnormalities (CAs) in a population-based MDS cohort. METHODS: Cancer registry data on incident MDS cases were linked with cytogenetic data and hospital administrative data from 2000 to 2010 for the Australian state of Victoria. RESULTS: Between 2000 and 2010, 1404 incident MDS cases with cytogenetic results were identified. A CK, defined as 3 or more abnormalities, was present in 126 (9%). A very complex karyotype (vCK), defined as 5 or more abnormalities, was present in 95 (7%). An MK was associated with worse OS in the whole cohort (median 6 vs 39 months, P < 0.001) including those with a coexisting CK (6 vs 17 months, P < 0.001) or vCK (6 vs 9 months, P 5 0.02). After adjustments for the number of CAs, an MK remained independently associated with OS, although its effect size decreased with increasing cytogenetic complexity (hazard ratio for an MK, 4.81; 95% confidence interval, 3.08-7.52; hazard ratio for the number of CAs, 1.22; 95% confidence interval, 1.15-1.30; and hazard ratio for the interaction between an MK and CAs, 0.83; 95% confidence interval, 0.770.89). CONCLUSIONS: These results support the clinical utility of an MK as an independent predictor of adverse outcomes for MDS patients, even among CK and vCK groups, although its prognostic effect decreases with increasing cytogenetic complexity. Cancer C 2015 American Cancer Society. 2015;121:2892-9. V KEYWORDS: cytogenetics, monosomal karyotype, myelodysplastic syndromes, survival.
INTRODUCTION Myelodysplastic syndromes (MDSs) are a heterogeneous group of clonal disorders characterized by peripheral blood cytopenias, marrow dysplasia, and an increased risk of progression to acute myeloid leukemia (AML). They predominantly affect older individuals with a median age at diagnosis of 76 years,1 and the median 3-year survival rate is 45%.2 The prognosis of MDS patients can be estimated with the revised International Prognostic Scoring System (IPSS-R), which is based on bone marrow cytogenetics, the percentage of bone marrow blasts, and peripheral blood cytopenias.3 Cytogenetic abnormalities (CAs) in the IPSS-R are classified into 5 groups, with karyotypic complexity predictive of inferior outcomes. In the IPPS-R, the complex karyotype (CK) group undergoes further subdivision: a CA count of 3 is classified as poor risk, whereas more than 3 CAs confer very poor risk.4 Recently, the presence of a monosomal karyotype (MK), defined as 2 or more autosomal monosomies or 1 autosomal monosomy and 1 structural abnormality,5 has been reported to be associated with worse overall survival (OS) for patients with AML,6-10 including those with a CK and those undergoing allogeneic stem cell transplantation.9,10 However, in MDS patients, there are conflicting data about whether an MK is an independent predictor of worse OS, particularly for those patients with a CK, in part because there is considerable overlap between an MK and a CK in MDS patients. An MK has been reported to be independently associated with OS in MDS patients with 3 or more CAs11,12;
Corresponding author: Zoe K. McQuilten, Department of Epidemiology and Preventive Medicine, Monash University, Level 6, 99 Commercial Road, Melbourne, Victoria, Australia 3004; Fax: (011) 61 3 9903 0576; [email protected] 1 Department of Epidemiology and Preventive Medicine, Monash University, Melbourne, Australia; 2Department of Medicine, St Vincent’s Hospital, University of Melbourne, Melbourne, Australia; 3Australian Centre for Blood Diseases, Monash University, Melbourne, Australia; 4Victorian Cancer Cytogenetics Service, St Vincent’s Hospital, Melbourne, Australia
We thank Helen Farrugia and the Victorian Cancer Registry for data access. We also thank the Victorian Department of Health for data access and the Victorian Data Linkage Unit and Ying Chen for supervising the linkage. DOI: 10.1002/cncr.29396, Received: November 16, 2014; Revised: February 11, 2015; Accepted: February 19, 2015, Published online June 4, 2015 in Wiley Online Library (wileyonlinelibrary.com)
2892
Cancer
September 1, 2015
Monosomal Karyotype in Myelodysplastic Syndromes/McQuilten et al
however, 2 recently published studies found that an MK was not independently associated with worse OS after the number of CAs was also taken into account.13,14 Given the conflicting evidence, we aimed to investigate whether an MK was independently associated with OS in a large population-based cohort of MDS patients identified through cancer registry data. MATERIALS AND METHODS Patients and Data Sources Victorian Cancer Registry (VCR)
Patients between 2000 and 2010 were identified through the VCR, which collects data on all cancer diagnoses in Victorian residents. Victoria is Australia’s second most populous state with a population of more than 5.3 million people. MDSs have been notifiable to the VCR since 2002. Data held within the VCR include demographic and cancer details and the date of death, which is sourced from the Victorian Death Registry. Victorian Cancer Cytogenetic Service (VCCS)
The VCCS is the single provider of cytogenetic services for adult hematological malignancies in Victoria. The VCCS has been routinely providing cytogenetic results for all newly diagnosed MDS patients to the VCR since 2005. Before this, results for a limited number of patients were provided to the VCR by the VCCS. Metaphase cytogenetic analysis was performed according to standard techniques. Karyotypes were described with the International System for Cytogenetic Nomenclature.15 Two pathologists with expertise in cytogenetic analysis (L.J.C. and M.W.) independently reviewed the karyotypes to classify individual abnormalities and the number of CAs. Cases with discordant classification were reviewed, and a consensus was reached between the 2 pathologists. Hospital Admission Data
Data on hospitalizations were sourced from the Victorian Admitted Episode Dataset (VAED), which is compiled from data submitted by all private and public hospitals within Victoria. Clinical information is coded by a trained staff of professional coders according to the Australian modification of International Statistical Classification of Diseases and Related Health Problems, Tenth Revision (ICD-10-AM),16 and the Australian Classification of Health Interventions.17 Data standards are maintained with regular, independent coding audits against corresponding medical records.18 All hospitalization episodes for VCR-notified patients from 1998 to 2011 were extracted from the VAED and linked with the VCR-VCCS database with a Cancer
September 1, 2015
strategy that included a unique hospital identifier, a Medicare number, the first 3 letters of the given name, and the date of birth. Definitions
The histological MDS subtypes are reported to and classified by the VCR according to the third edition of International Classification of Diseases for Oncology19 and include refractory anemia, refractory anemia with ringed sideroblasts, refractory anemia with an excess of blasts, refractory anemia with an excess of blasts in transformation, refractory cytopenia with multilineage dysplasia, MDS with 5q deletion, therapy-related MDS, and MDS not otherwise specified. An MK was defined as the presence of 2 or more autosomal monosomies or 1 monosomy with at least 1 additional structural abnormality.5 A CK was defined as 3 or more CAs, and a very complex karyotype (vCK) was defined as 5 or more CAs in line with previous publications.11,13,14 Patients were classified as red cell transfusion–dependent at diagnosis if they had at least 4 prior hospitalizations of any length with an Australian Classification of Health Interventions red cell transfusion code (1370602) and at least 2 occurred within an 8-week period.20 Comorbidities at the time of the MDS diagnosis were classified according to the Charlson comorbidity index with a coding algorithm validated for ICD-10-AM.21 Because there were no data available on peripheral blood counts in the linked data set, variables for thrombocytopenia and neutropenia at diagnosis were created with ICD10-AM diagnostic codes (D695, D696, and D70) if they were recorded at or before the MDS diagnosis date. Statistical Analysis
OS was the primary study endpoint. OS was defined as the time from diagnosis until death, and patients were censored at the end of the study period. Patients who received an allogeneic transplant were censored at the day of admission for the transplant. Means and standard deviations and medians and interquartile ranges (IQRs) were reported for normally and nonnormally distributed continuous data, respectively. Comparison testing was performed with a t test for normally distributed data, with a Wilcoxon rank sum test for nonnormally distributed continuous data, and with a chi-square test for categorical data. Survival was estimated with the Kaplan-Meier method, and groups were compared with a log-rank test. Multivariate analysis was performed with a stepwise Cox proportional hazards regression method, and it included those variables with a P value < .2 on univariate testing. Single CAs were considered for inclusion in the 2893
Original Article TABLE 1. Characteristics of Patients With Cytogenetic Results Characteristic Age, median (IQR), y Male sex, No. (%) MDS type, No. (%) RA RARS RAEB RAEB-T RC-MLD 5q deletion Therapy-related MDS NOS Comorbidity, No. (%)a Red cell TD at diagnosis, No. (%)a Prior nonmyeloid malignancy, No. (%)a Cytogenetics, No. (%) Normal karyotype Abnormal karyotype Number of CAs, No. (%) 1 2 3 4 5 Number of CAs, median (IQR) Cytogenetic risk score, No. (%)b Very good Good Intermediate Poor Very poor Monosomal karyotype, No. (%)
All Patients (n 5 1404)
Monosomal Karyotype (n 5 90)
Complex Karyotype: 3 CAs (n 5 126)
77 (69-83) 866 (62)
77 (66-82) 48 (53)
77 (66-84) 68 (54)
126 (9) 127 (9) 243 (17) 5 (
Monosomal Karyotype Predicts Inferior Survival Independently of a Complex Karyotype in Patients With Myelodysplastic Syndromes Zoe K. McQuilten, MBBS1; Vijaya Sundararajan, MD2; Nick Andrianopoulos, MBBS, MBiostat1; David J. Curtis, MD, PhD3; Erica M. Wood, MBBS1,3; Lynda J. Campbell, MBBS2,4; and Meaghan Wall, MBBS, PhD4
BACKGROUND: Conflicting data exist about the impact of a monosomal karyotype (MK) on overall survival (OS) for patients with myelodysplastic syndromes (MDSs) and particularly for those with a complex karyotype (CK). This study was aimed at determining whether an MK is associated with OS independently of the number of cytogenetic abnormalities (CAs) in a population-based MDS cohort. METHODS: Cancer registry data on incident MDS cases were linked with cytogenetic data and hospital administrative data from 2000 to 2010 for the Australian state of Victoria. RESULTS: Between 2000 and 2010, 1404 incident MDS cases with cytogenetic results were identified. A CK, defined as 3 or more abnormalities, was present in 126 (9%). A very complex karyotype (vCK), defined as 5 or more abnormalities, was present in 95 (7%). An MK was associated with worse OS in the whole cohort (median 6 vs 39 months, P < 0.001) including those with a coexisting CK (6 vs 17 months, P < 0.001) or vCK (6 vs 9 months, P 5 0.02). After adjustments for the number of CAs, an MK remained independently associated with OS, although its effect size decreased with increasing cytogenetic complexity (hazard ratio for an MK, 4.81; 95% confidence interval, 3.08-7.52; hazard ratio for the number of CAs, 1.22; 95% confidence interval, 1.15-1.30; and hazard ratio for the interaction between an MK and CAs, 0.83; 95% confidence interval, 0.770.89). CONCLUSIONS: These results support the clinical utility of an MK as an independent predictor of adverse outcomes for MDS patients, even among CK and vCK groups, although its prognostic effect decreases with increasing cytogenetic complexity. Cancer C 2015 American Cancer Society. 2015;121:2892-9. V KEYWORDS: cytogenetics, monosomal karyotype, myelodysplastic syndromes, survival.
INTRODUCTION Myelodysplastic syndromes (MDSs) are a heterogeneous group of clonal disorders characterized by peripheral blood cytopenias, marrow dysplasia, and an increased risk of progression to acute myeloid leukemia (AML). They predominantly affect older individuals with a median age at diagnosis of 76 years,1 and the median 3-year survival rate is 45%.2 The prognosis of MDS patients can be estimated with the revised International Prognostic Scoring System (IPSS-R), which is based on bone marrow cytogenetics, the percentage of bone marrow blasts, and peripheral blood cytopenias.3 Cytogenetic abnormalities (CAs) in the IPSS-R are classified into 5 groups, with karyotypic complexity predictive of inferior outcomes. In the IPPS-R, the complex karyotype (CK) group undergoes further subdivision: a CA count of 3 is classified as poor risk, whereas more than 3 CAs confer very poor risk.4 Recently, the presence of a monosomal karyotype (MK), defined as 2 or more autosomal monosomies or 1 autosomal monosomy and 1 structural abnormality,5 has been reported to be associated with worse overall survival (OS) for patients with AML,6-10 including those with a CK and those undergoing allogeneic stem cell transplantation.9,10 However, in MDS patients, there are conflicting data about whether an MK is an independent predictor of worse OS, particularly for those patients with a CK, in part because there is considerable overlap between an MK and a CK in MDS patients. An MK has been reported to be independently associated with OS in MDS patients with 3 or more CAs11,12;
Corresponding author: Zoe K. McQuilten, Department of Epidemiology and Preventive Medicine, Monash University, Level 6, 99 Commercial Road, Melbourne, Victoria, Australia 3004; Fax: (011) 61 3 9903 0576; [email protected] 1 Department of Epidemiology and Preventive Medicine, Monash University, Melbourne, Australia; 2Department of Medicine, St Vincent’s Hospital, University of Melbourne, Melbourne, Australia; 3Australian Centre for Blood Diseases, Monash University, Melbourne, Australia; 4Victorian Cancer Cytogenetics Service, St Vincent’s Hospital, Melbourne, Australia
We thank Helen Farrugia and the Victorian Cancer Registry for data access. We also thank the Victorian Department of Health for data access and the Victorian Data Linkage Unit and Ying Chen for supervising the linkage. DOI: 10.1002/cncr.29396, Received: November 16, 2014; Revised: February 11, 2015; Accepted: February 19, 2015, Published online June 4, 2015 in Wiley Online Library (wileyonlinelibrary.com)
2892
Cancer
September 1, 2015
Monosomal Karyotype in Myelodysplastic Syndromes/McQuilten et al
however, 2 recently published studies found that an MK was not independently associated with worse OS after the number of CAs was also taken into account.13,14 Given the conflicting evidence, we aimed to investigate whether an MK was independently associated with OS in a large population-based cohort of MDS patients identified through cancer registry data. MATERIALS AND METHODS Patients and Data Sources Victorian Cancer Registry (VCR)
Patients between 2000 and 2010 were identified through the VCR, which collects data on all cancer diagnoses in Victorian residents. Victoria is Australia’s second most populous state with a population of more than 5.3 million people. MDSs have been notifiable to the VCR since 2002. Data held within the VCR include demographic and cancer details and the date of death, which is sourced from the Victorian Death Registry. Victorian Cancer Cytogenetic Service (VCCS)
The VCCS is the single provider of cytogenetic services for adult hematological malignancies in Victoria. The VCCS has been routinely providing cytogenetic results for all newly diagnosed MDS patients to the VCR since 2005. Before this, results for a limited number of patients were provided to the VCR by the VCCS. Metaphase cytogenetic analysis was performed according to standard techniques. Karyotypes were described with the International System for Cytogenetic Nomenclature.15 Two pathologists with expertise in cytogenetic analysis (L.J.C. and M.W.) independently reviewed the karyotypes to classify individual abnormalities and the number of CAs. Cases with discordant classification were reviewed, and a consensus was reached between the 2 pathologists. Hospital Admission Data
Data on hospitalizations were sourced from the Victorian Admitted Episode Dataset (VAED), which is compiled from data submitted by all private and public hospitals within Victoria. Clinical information is coded by a trained staff of professional coders according to the Australian modification of International Statistical Classification of Diseases and Related Health Problems, Tenth Revision (ICD-10-AM),16 and the Australian Classification of Health Interventions.17 Data standards are maintained with regular, independent coding audits against corresponding medical records.18 All hospitalization episodes for VCR-notified patients from 1998 to 2011 were extracted from the VAED and linked with the VCR-VCCS database with a Cancer
September 1, 2015
strategy that included a unique hospital identifier, a Medicare number, the first 3 letters of the given name, and the date of birth. Definitions
The histological MDS subtypes are reported to and classified by the VCR according to the third edition of International Classification of Diseases for Oncology19 and include refractory anemia, refractory anemia with ringed sideroblasts, refractory anemia with an excess of blasts, refractory anemia with an excess of blasts in transformation, refractory cytopenia with multilineage dysplasia, MDS with 5q deletion, therapy-related MDS, and MDS not otherwise specified. An MK was defined as the presence of 2 or more autosomal monosomies or 1 monosomy with at least 1 additional structural abnormality.5 A CK was defined as 3 or more CAs, and a very complex karyotype (vCK) was defined as 5 or more CAs in line with previous publications.11,13,14 Patients were classified as red cell transfusion–dependent at diagnosis if they had at least 4 prior hospitalizations of any length with an Australian Classification of Health Interventions red cell transfusion code (1370602) and at least 2 occurred within an 8-week period.20 Comorbidities at the time of the MDS diagnosis were classified according to the Charlson comorbidity index with a coding algorithm validated for ICD-10-AM.21 Because there were no data available on peripheral blood counts in the linked data set, variables for thrombocytopenia and neutropenia at diagnosis were created with ICD10-AM diagnostic codes (D695, D696, and D70) if they were recorded at or before the MDS diagnosis date. Statistical Analysis
OS was the primary study endpoint. OS was defined as the time from diagnosis until death, and patients were censored at the end of the study period. Patients who received an allogeneic transplant were censored at the day of admission for the transplant. Means and standard deviations and medians and interquartile ranges (IQRs) were reported for normally and nonnormally distributed continuous data, respectively. Comparison testing was performed with a t test for normally distributed data, with a Wilcoxon rank sum test for nonnormally distributed continuous data, and with a chi-square test for categorical data. Survival was estimated with the Kaplan-Meier method, and groups were compared with a log-rank test. Multivariate analysis was performed with a stepwise Cox proportional hazards regression method, and it included those variables with a P value < .2 on univariate testing. Single CAs were considered for inclusion in the 2893
Original Article TABLE 1. Characteristics of Patients With Cytogenetic Results Characteristic Age, median (IQR), y Male sex, No. (%) MDS type, No. (%) RA RARS RAEB RAEB-T RC-MLD 5q deletion Therapy-related MDS NOS Comorbidity, No. (%)a Red cell TD at diagnosis, No. (%)a Prior nonmyeloid malignancy, No. (%)a Cytogenetics, No. (%) Normal karyotype Abnormal karyotype Number of CAs, No. (%) 1 2 3 4 5 Number of CAs, median (IQR) Cytogenetic risk score, No. (%)b Very good Good Intermediate Poor Very poor Monosomal karyotype, No. (%)
All Patients (n 5 1404)
Monosomal Karyotype (n 5 90)
Complex Karyotype: 3 CAs (n 5 126)
77 (69-83) 866 (62)
77 (66-82) 48 (53)
77 (66-84) 68 (54)
126 (9) 127 (9) 243 (17) 5 (
