Isolated del(5q) in myeloid malignancies: Clinicopathologic and molecular features in 143 consecutive patients

Abstract

World Health Organization (WHO) criteria were used to identify 143 consecutive patients (median age 73 years; 90 females) with myeloid neoplasms and isolated del(5q) seen between 1989 and 2009. We have previously reported on 88 (61%) of these patients who met criteria for WHO defined “myelodysplastic syndromes (MDS) with isolated del(5q).” The remaining 55 patients were classified as having “other” MDS variants (n = 29; 20%), acute myeloid leukemia (AML; n = 14; 10%), or myeloproliferative neoplasms (MPN; n = 12; 8%). DNA was available in 138 patients and mutation screening revealed 20 cases with JAK2, 6 with IDH, and 3 with MPL mutations; JAK2 and MPL mutations were seen mostly in MPN or “MDS with isolated del(5q)” whereas IDH mutations were frequent in other MDS variants. Overall median survival for the 143 patient cohort was 35 months and leukemic transformation (LT) was documented in 19 (∼13%) cases. “MDS with isolated del(5q)” had the best prognosis with median survival of 66 months and LT rate of ∼6%. Survival was poor among the other myeloid neoplasm subgroups regardless of specific morphologic diagnosis. Multivariable analysis identified higher leukocyte count and percentage of bone marrow and circulating blasts as independent predictors of shortened survival. The first two parameters and the presence of IDH mutations predicted inferior leukemia-free survival. The current study validates the prognostic relevance of considering “MDS with isolated del(5q)” as a separate WHO subcategory and identifies leukocytosis, higher blast count, and IDH mutations as being prognostically detrimental, in myeloid neoplasms associated with isolated del(5q). Am. J. Hematol. 86:393–398, 2011. © 2011 Wiley-Liss, Inc.

Introduction

Myeloid neoplasms associated with an isolated del(5q) present with a broad clinical and pathological spectrum that includes low, intermediate, and high-risk myelodysplastic syndromes (MDS), myeloproliferative neoplasms (MPN), and acute myeloid leukemia (AML) [1, 2]. Treatment with lenalidomide is US FDA-approved for the management of red cell transfusion-dependent patients with lower risk MDS and del(5q). Some patients with a myeloid neoplasm and isolated del(5q) undergo clonal evolution and progression to AML [3, 4]. We and others have previously shown that patients belonging to the World Health Organization (WHO) MDS subcategory of “MDS with isolated del(5q)” enjoy relatively long survival and a lower rate of leukemic transformation (LT) [1, 5, 6]. This group of patients was also more likely to achieve a sustained erythroid and cytogenetic remission from treatment with lenalidomide [3, 6, 7].

The long arm of chromosome 5 (5q), especially the commonly deleted region (CDR) spanning 5q31-32, is believed to contain important tumor suppressor genes: RPS14 (ribosomal processing S14), SPARC (secreted protein, acidic, cysteine rich), CTNNA1 (alpha catenin), and EGR-1 (early growth response gene-1). Haploinsufficiency of such genes has been implicated in the pathogenesis of del(5q)-associated MDS [8–13]. It is assumed that additional cytogenetic and molecular abnormalities, such as mutations in JAK2, MPL, IDH1, and IDH2, modify disease phenotype and outcome [14, 15]. Current literature on del(5q)-associated myeloid neoplasms is confounded by the inclusion of patients with additional cytogenetic abnormalities [4, 14, 16]. This study is restricted to patients with myeloid malignancies in whom a cytogenetic study at the time of their diagnosis revealed an isolated del(5q). The objectives were: (i) define the spectrum of clinicopathologic phenotype, (ii) describe the prevalence and distribution of JAK2, MPL, and IDH mutations, and (iii) identify prognostic factors for overall and leukemia-free survival.

Materials and Methods

After study approval by the Mayo Clinic Institutional Review Board, we queried an institutional database for the date range 1989 to 2009; to identify all patients karyotyped during that period whose bone marrow (BM) revealed at least two metaphases with isolated deletions of chromosome 5q, including band q31. Medical records, BM, and cytogenetic specimens were reviewed to ensure accuracy of diagnosis. Only those patients that met the 2008 WHO pathological criteria for myeloid neoplasms along with an isolated del(5q), were included in the data analysis [17]. In patients who were alive, every attempt was made to update follow-up information, by means of a questionnaire/telephone call sent to both patients and their primary doctors, and the “date of last contact” reflected this time point, and not the last time they were seen at the Mayo Clinic.

Morphology

Morphological evaluation of the BM and peripheral blood (PB) was performed according to the standard evaluation protocol at the Mayo Clinic. BM and PB slides were available for review in 143 patients that had a myeloid neoplasm with an isolated del(5q). All pathological diagnoses were made in accordance to the 2008 WHO criteria [17]. For the estimation of overall survival (OS) and leukemia free survival (LFS), patients were subdivided into the following six morphological categories: (i) WHO defined “MDS with isolated del(5q),” (ii) low-risk MPN including polycythemia vera (PV) and essential thrombocythemia (ET), (iii) high-risk MPN including primary myelofibrosis (PMF) and post ET/PV MF, (iv) MDS or MDS/MPN without excess blasts but not meeting diagnostic criteria for “MDS with isolated del(5q),” (v) MDS or MDS/MPN with excess blasts, and (vi) AML. Dysgranulopoiesis was defined by the presence of nuclear hypolobation (pseudo-Pelger-Hüet anomaly), hypogranulation, and atypical nuclear segmentation.

Cytogenetic studies

Cytogenetic studies were performed on all 143 patients utilizing short term BM cultures along with conventional G or Q banding fluorescence. When possible a goal number of 20 metaphases were evaluated [18]. The findings were described according to the International Society of Cytogenetic Nomenclature (ISCN) [18].

Molecular studies

Archived BM and PB cell pellets on 168 patients with an isolated del(5q) anomaly were stored at −70°C in methanol: glacial acetic acid (2:1) fixative. After appropriate washing techniques the DNA on these patients was extracted by using the Qiagen DNA Mini extraction kit (Qiagen, Santa Clarita, CA). The DNA was then analyzed for JAK2 MPL, IDH1, and IDH2 mutations, using previously described methods [1, 19–21].

Statistical analysis

Descriptive and statistically analyzed data were based on parameters collected at the time of initial diagnosis. Statistical procedures utilized were conventional and all data were analyzed by using StatView (SAS Institute, Cary, NC). All P values were two-tailed and statistical significance was set at the level of P < 0.05. Continuous variables were summarized as medians and ranges. Categorical variables were described as count and relative frequency. Comparison between categorical variables was performed by χ² statistics. Comparison between categorical and continuous variables was performed by either the Mann-Whitney U test or Kruskal-Wallis test. Survival analysis was performed by the Kaplan-Meier method taking the interval from the date of diagnosis to death, time of allogeneic hematopoietic cell transplantation or last contact. LFS was calculated by considering leukemic transformation as the uncensored variable. The log-rank test was used to compare survival data. Cox regression model was used for multivariable analysis.

Results

Of 24,262 unique patient cytogenetic studies performed at our institution between 1989 and 2009, isolated del(5q) associated with a myeloid neoplasm was identified in 143 patients (∼0.5%). Table I details the spectrum of myeloid neoplasms associated with isolated del(5q) and provides information on their JAK2, IDH1, IDH2, and MPL mutational status. Median age for the entire study population was 73 years (range, 28–91) and 90 (63%) patients were females. Eighty-eight (62%) patients (60 females) met the 2008 WHO criteria for “MDS with isolated del(5q)”; the median age for this cohort was 74 years (range, 28–89). The remaining 55 patients were classified as having “other” MDS variants (n = 29; 20%), acute myeloid leukemia (AML; n = 14; 10%), or myeloproliferative neoplasms (MPN; n = 12; 8%). Table II includes the main clinical and laboratory features among the different myeloid neoplasm subgroups.

Table I. Spectrum of Myeloid Neoplasms Associated with Isolated del(5q), JAK2, IDH1, IDH2, MPL Mutational Status, and Information on Subsequent Disease Evolution

Diagnosis	Total no. cases (n = 143)	JAK2 V617F mutation (%) (n = 19)	JAK2 exon 12 mutation (%) (n = 1)	IDH1 mutationsa (%) (n = 2)	IDH2 mutationsb (%) (n = 4)	MPL mutationsc (%) (n = 3)	Subsequent disease evolution (n)
WHO, World Health Organization; MDS, myelodysplastic syndrome; RAEB, refractory anemia with excess blasts; CMML, chronic myelomonocytic leukemia; MDS-U, myelodysplastic syndrome unclassifiable; t-MDS, therapy related myelodysplastic syndrome; PV, polycythemia vera; ET, essential thrombocytosis; PMF, primary myelofibrosis; MPN-U, myeloproliferative neoplasm-unclassifiable; AML, acute myeloid leukemia; SM, systemic mastocytosis. JAK2, MPL, and IDH mutational analysis was performed in 83 of the 88 patients with the WHO defined “MDS with isolated del(5q).” a The identified IDH1 mutations included IDH1R132C and IDH1R132G. b The identified IDH2 mutations included IDH2 R140Q. c The identified MPL mutations included MPL W515L.
WHO defined “MDS with isolated del(5q)”	88	5 (6)	0	0	0	3 (3)	AML (5); RAEB-1 (1); CMML-1 (1)
RAEB-1	13	2 (2)	0	1 (1)	1 (1)	0	AML (5)
RAEB-2	9	1 (1)	0	0	2 (2)	0	AML (6)
MDS-U	4	2 (50)	0	0	0	0	AML (1)
CMML-1	2	0	0	0	0	0	AML (1)
t-MDS	1	0	0	0	1 (100)	0
PV	4	3 (75)	1 (25)	0	0	0	Myelofibrosis (1)
ET	3	3 (100)	0	1 (33)	0	0	Myelofibrosis (1); AML (1)
PMF	3	2 (67)	0	0	0	0
MPN-U	1	1 (100)	0	0	0	0
SM	1	0	0	0	0	0
AML	14	0	0	0	0	0

Table II. Clinical and Laboratory Features of 143 Patients with Myeloid Neoplasms and Isolated del(5q) Stratified by Disease Pathology

Variable	Entire cohort of patients with myeloid neoplasms and isolated del(5q) (n = 143)	WHO defined “MDS with isolated del(5q)” (n = 88)	MDS with isolated del(5q), other than “WHO defined MDS with del(5q)” (n = 29)	MPN with isolated del(5q) (n = 12)	AML with isolated del(5q) (n = 14)	P
MCV, mean corpuscular volume; AML, acute myeloid leukemia; CMML, chronic myelomonocytic leukemia; RAEB-2, refractory anemia with excess blasts-2.
Age (yr) (median and range)	73 (28–91)	74 (28–89)	68 (39–91)	66 (54–88)	70 (52–88)	0.05
Sex (M/F)	53/90	28/60	10/19	8/4	7/7	0.17
Hemoglobin (gm/dl, median and range)	9.5 (4.7–17.4)	9.2 (4.7–13.3)	9.5 (7–16.9)	11.7 (8.2–17.4)	9.7 (5.8–11.9)	0.004
MCV (fl, median and range)	100 (79–130)	103 (79–130)	103 (79–130)	98 (81–100)	98 (81–102)	0.45
Leukocyte count (×10^9/l, median and range)	4.7 (0.9–218)	4.7 (1.5–27.3)	4.1 (1.9–56)	14.8 (6.7–218)	4.1 (0.9–112)	0.0002
Platelet count (×10^9/l, median and range)	190 (12–1,800)	235 (47–1,800)	109 (12–443)	374 (76–923)	93 (27–203)	<0.0001
Circulating blasts (%, median and range)	0 (0–70)	0 (0)	0 (0–4)	0 (0–9)	10 (0–70)	<0.0001
Bone marrow blasts (%, median and range)	1 (0–90)	0 (0)	8 (3–15)	2 (0–19)	40 (3–90)	<0.0001
Mutation analysis (%)
JAK2 V617F	19 (14)	5 (6)	5 (17)	9 (75)	0	0.001
JAK2 Exon 12	1 (1)	0	1 (3)	0	0
IDH1	2 (1)	0	1 (3)	1 (8)	0
IDH2	4 (3)	0	4 (14)	0	0
MPL	3 (2)	3 (3)	0 (0)	0	0
Disease transformation (n, %)	AML (19, 13); CMML-1 (1, 1); RAEB-2 (1, 1)	AML (5, 6); CMML-1 (1, 1); RAEB-2 (1, 1)	AML (12, 36)	AML (2, 17)	NA
No. deaths (%)	95 (66)	50 (64)	23 (78)	9 (75)	13 (93)	0.01
Duration of follow-up in months (median and range)	35 (0–246)	31 (0–246)	16.8 (0.5–97)	24 (0.26–150)	16 (1.6–53)	<0.0001

BM derived DNA was available for mutation analysis in 138 patients. Nineteen (13%) patients displayed JAK2V617F mutation, one (1%) JAK2 exon 12, 2 (1%) IDH1 (IDHR132C and IDH1R132G), 4 (3%) IDH2 mutations (IDH2R140Q), and 3 (2%) MPL (MPL W515L) mutations. Among 78 patients with WHO-defined “MDS with isolated del(5q),” four (∼5%) tested positive for the JAK2V617F (JAK2V617F allele burden <10% in all instances), three (∼4%) tested positive for MPLW515L, and one tested positive for both. None of the patients with the MPLW515L displayed thrombocytosis, either at diagnosis or during follow up. Five (83%) of the six IDH mutations occurred in patients with higher risk MDS (two in RAEB-1, two RAEB-2, and one therapy related-MDS). The remaining one IDH-mutated patient had ET that transformed to AML after 35.7 months from diagnosis. Supporting Information, Table I outlines the clinical and pathological details of the six IDH-mutated patients.

Median follow-up for the entire study population was 35 months (range, 0–246). During this period, 95 (66%) patients had died and 21 (15%) transformations into AML (n = 19; 13%), CMML-1 (n = 1), or RAEB-1 (n = 1) were documented. Transformations into AML occurred mainly in patients with MDS variants other than the “WHO defined MDS with isolated del(5q)” (n = 12; 36%). Among these 12 patients with other MDS variants, five (42%) had mutations involving IDH1 or IDH2 and the median time to LT was 12 months (range, 0.5–39). Fourteen of the 19 patients with LT had additional cytogenetic abnormalities at the time of AML transformation. None of the patients with LT are currently alive.

Interestingly, there was one patient who was diagnosed with AML with isolated del(5q) in 1993 and received standard induction chemotherapy with daunorubicin and cytarabine, followed by three cycles of consolidation chemotherapy with high-dose cytarabine. The patient had achieved complete morphological and cytogenetic remission. This patient subsequently developed refractory anemia in 2005 and was eventually diagnosed with therapy-related MDS (t-MDS) and an isolated del(5q). DNA analysis on paired samples revealed IDH2R140Q mutation at the time of initial diagnosis of AML as well as at the time of diagnosis of his t-MDS. The patient is currently being treated with lenalidomide and has achieved red cell transfusion independence. The one patient with JAK2 exon 12 mutation (H538Q) had a long standing history of isolated erythrocytosis requiring intermittent phlebotomies. He eventually developed refractory macrocytic anemia without any prior chemotherapy or radiation treatment and BM examination revealed dyserythropoiesis and isolated del(5q) in 16 of 20 metaphases.

There were 14 (10%) patients who presented with AML and isolated del(5q). Four (29%) of these patients carried a history of antecedent MDS with normal cytogenetics (one RCMD-1, one RAEB-1, two RAEB-2). One patient had received lenalidomide therapy for RCMD for approximately 9 months before transformation. None of these patients had mutations involving JAK2, IDH1, IDH2, or MPL. Supporting Information, Table II outlines the clinical, pathological, cytogenetic, and molecular profile of the 14 patients with del(5q)-associated AML.

Forty-seven (53%) patients with the WHO-defined “MDS with isolated del(5q)” were prescribed pharmacological therapy in an attempt to ameliorate red cell transfusion dependancy. Treatment included erythropoiesis stimulating agents (ESAs) in 28 (32%) cases, lenalidomide in 18 (20%), and one patient each received thalidomide, 5-azacytidine, and danazol, respectively. Among the 18 lenalidomide-treated patients, 5 (28%) became transfusion independent.

Cause of death was documented in 60 (63%) of the 95 patients that had died during the study period and included 23 deaths from complications related to AML or its treatment (neutropenic sepsis, lung injury/ARDS, diffuse alveolar hemorrhage, intracranial hemorrhage and multiorgan failure syndrome), 10 deaths from angioinvasive fungal infections, 10 deaths from pneumonia, 7 from congestive cardiac failure, 4 from cardiac arrest, 2 from metastatic lung cancer, and 1 each from metastatic breast cancer, myocardial infarction, stroke, and lower GI bleed.

Prognostic variables for overall survival and leukemic transformation

Univariate and multivariable analyses of parameters at diagnosis identified higher leukocyte count and percentage of PB or BM blasts as independent predictors of inferior survival (Supp. Info., Table III, Fig. 1). In univariate analysis, higher leukocyte count, increased percentage of BM blasts, and presence of IDH mutations were significantly associated with inferior LFS; leukocyte count and IDH mutational status sustained their significance during multivariable analysis (Supp. Info., Table IV, Figs. 2 and 3).

Figure 1. Survival data on 143 patients with myeloid neoplasms and isolated del(5q) stratified by morphological diagnosis. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Figure 2. Leukemia-free survival data on 129 patients with myeloid neoplasms and isolated del(5q) stratified by morphological diagnosis. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Figure 3. Leukemia-free survival data on 129 patients with myeloid neoplasms and isolated del(5q) stratified by the presence of or absence of IDH mutations. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Among patients with “MDS with isolated del(5q),” median survival was 66 months and LT rate only ∼6%. We have previously shown that in this group of patients, age ≥70 years, red cell transfusion dependancy, and the presence of dysgranulopoiesis were independent predictors of inferior survival [1]. In this study, we show that other myeloid malignancies with isolated del(5q) do not fare as well those with “MDS with isolated del(5q),” in terms of both OS and LFS (Figs. 1 and 2); median survival was 13 months for AML, 16 months for MDS or MDS/MPN without excess blasts, 21 months for MDS or MDS/MPN with excess blasts, 35 months for low-risk MPN and 10.5 months for high-risk MPN.

Discussion

The current study confirms the validity of special categorization of “MDS with isolated del(5q)” by the WHO classification system and underscores the importance of clearly separating this group of patients from other MDS variants with isolated del(5q). It is also important to recall that advanced age, red cell transfusion need, and dysgranulopoiesis identify poor-risk patients among this group [1, 21]. The current study suggests that the presence of isolated del(5q) in myeloid malignancies other than “MDS with isolated del(5q)” is a poor prognostic factor regardless of the clinicopathologic phenotype. However, the presence of higher leukocyte count or blast percentage in the PB or BM appeared to increase the risk even further. Obviously, additional studies with larger number of patients are required to validate these preliminary observations and clarify the prognostic relevance of IDH mutations in this group of patients.

In a recent multicenter study of 541 patients with MDS and del(5q), Mallo et al. identified additional chromosomal abnormalities, lower platelet count, and increased BM blast percentage as significant predictors of inferior OS and LFS [14]. In this cohort of patients, 144 were classified as having “5q-syndrome” and displayed an OS of 68 months and a LT rate of 17% [14]. It is difficult to interpret these findings because reported cases were not systematically reviewed to conform to the 2008 WHO criteria for “MDS and isolated del(5q)” and the presence or absence of dysgranulopoiesis was not addressed [1, 21].

Mutations involving IDH1 and IDH2 have been described in a variety of myeloid neoplasms, including chronic phase MPN (≈4%), blast phase MPN (≈21%) [22, 23], high risk MDS (≈20%) [15], and AML [15, 24]. Overall the data from these studies have indicated that although IDH mutations may arise early in the disease course, they are generally associated with LT in patients with chronic phase MDS or MPN. Approximately 16% of patients with cytogenetically normal AML carry IDH mutations [24, 25]. IDH mutations predict poor outcome in the subset of patients with mutated NPM1 (nucleophosphomin 1) without the FLT3-ITD (fms like tyrosine kinase-internal tandem duplication), and are infrequent in AML patients with activating FLT3 mutations [FLT3 ITD and TKD (tyrosine kinase domain mutations)] [24]. The IDH1 R132C, IDH1 RI32G, and the IDH2 R140Q mutations identified in this study have been shown to cause loss of physiologic enzyme function for isocitrate dehydrogenase [26, 27]. However, the exact role that this aberration plays in myeloid neoplasms remains to be elucidated.

Deletion mapping of the CDR involved in patients with the “5q-syndrome” identified a 1.5 megabase segment containing 44 genes, the haploinsufficiency of which, has been implicated in disease pathogenesis [13]. This CDR contains genes such as, SPARC, RPS14, CTNNAI, and EGR-1 (9–12) [28]. Decreased expression of the RPS14 gene in cultured normal myeloid progenitor cells causes a phenotype that mimics the 5q-syndrome, whereas the forced expression of this gene in vitro reverses the defect in erythropoiesis in del(5q) cells [8]. RPS14 is part of the 40S subunit of ribosomes. Haploinsufficiency of the RPS14 gene in the 5q-syndrome (5q33.1) is associated with deregulation of ribosomal and translation-related genes Moreover, loss-of-function mutations involving other ribosomal components (e.g. RPS19, RPS24) occur in certain congenital syndromes (e.g. Diamond-Blackfan anemia) which share histological and clinical features with MDS [29]. Presumably, haploinsufficiency of ribosomal genes causes inadequate formation of ribosomal subunits, which in turn alters translation of genes and activation of proteins involved in differentiation and apoptosis (e.g. p53) [30]. SPARC is a tumor suppressor gene which is affected by haploinsufficiency in 5q33.1 [12]. In vitro, the gene is specifically up regulated in 5q-erythroblasts by lenalidomide [31]. Haploinsufficiency of EGR1 renders mice treated with a DNA alkylating agent more susceptible to myeloid malignancies than their wild-type counterparts [9]. Another gene, CTNNA1, is down-regulated in 5q-leukemia-initiating stem cells [11]. Furthermore, in HL-60 cells, which are 5q-, CTNNA1 expression in the retained allele is epigenetically suppressed and its re-expression reduces the proliferation rate of the cells and increases apoptosis [11].

In December 2005, based on encouraging results seen with lenalidomide in the treatment of transfusion dependent, lower risk patients with MDS and del(5q), the United States FDA approved lenalidomide therapy for this indication [6, 7]. Patients who received lenalidomide achieved a red cell transfusion independence rate of 66%, and a complete cytogenetic response rate of 44% [6]. Lenalidomide however was not approved for this indication by the EMEA, the European counterpart of the US FDA, partly due to concerns of enhanced leukemic transformation is some patients [32]. These concerns mainly emanate from a smaller prospective study that has indicated an enhanced risk of LT in patients with MDS and del(5q) treated with lenalidomide, especially those that failed to achieve sustained erythroid or cytogenetic remission [3]. At a median follow-up of 40 months, 15 (36%) of 40 patients with MDS and del(5q) had transformed to AML. This study, similar to other retrospective studies, however, did not strictly adhere to the WHO criteria for “MDS with isolated del(5q),” and included patients with advanced MDS and associated clonal cytogenetic abnormalities, thus confounding the results [3, 4, 16, 33, 34]. Tehranchi et al. have recently shown that MDS del(5q) stem cells (CD34+, CD90+, CD38−/low) remain distinctly resistant to lenalidomide therapy, even in patients who have otherwise achieved a complete clinical and cytogenetic remission [35]. Over time lenalidomide resistance develops in most patients, with recurrence or expansion of the del(5q) clone, resulting in clinical and cytogenetic progression [35].

This study is unique in its strict adherence to WHO criteria for selecting study patients with myeloid neoplasms and isolated del(5q) and providing original data on their clinical and molecular phenotype, survival, prognostic factors and baseline risk of LT. Patients with the WHO defined “MDS with isolated del(5q)” indeed have a unique disease biology with a relatively indolent course. It is this group of patients that potentially would benefit the most from lenalidomide therapy, given their lower inherent risk for LT. However, due to the presence of resistant del(5q) stem cell clones, most patients over time will develop disease progression, and newer treatment strategies/clinical trials are the need of the hour. Analysis for IDH mutations helps identify a subgroup of chronic phase MDS/MPN patients with a higher risk for LT, and should be incorporated in the planning of future clinical studies.