Challenges in Myelodysplastic/Myeloproliferative Neoplasms (MDS/MPN)

Abstract

Myelodysplastic/Myeloproliferative neoplasms (MDS/MPN) are hybrid group of chronic myeloid neoplasms combining features of both MDS and MPN. The world health organization (WHO) classification coined this group designation in 2008 to include chronic myelomonocytic leukemia (CMML), atypical chronic myeloid leukemia (aCML), juvenile myelomoncoytic leukemia (JMML), refractory anemia with ring sideroblasts and thrombocytosis (RARS-T) as provisional entity, and MDS/MPN unclassified (MDS/MPN-U). In this review we highlight the challenges in diagnosing this group of the diseases, summarize the updates in classification and recent evolving understanding of the genetic landscape. We review risk stratification models and overview the current management largely adapted from current MDS or MPN therapies. We define clinical benefit of therapy based on new proposed response criteria developed specifically for those groups of neoplasms. Finally we introduce future opportunities including the planned international ABN MARRO clinical study led by the MDS/MPN international working group.

Introduction

The spectrum of chronic myeloid neoplasm spans from myelodysplastic syndromes (MDS) characterized by morphologic dysplasia and ineffective hematopoiesis to myeloproliferative neoplasms (MPN) with proliferative features.¹ A group of diseases shares clinical, pathological and molecular aspects of both and had been recognized as MDS/MPN. The world health organization (WHO) classification coined this group designation in 2008 to include chronic myelomonocytic leukemia (CMML), atypical chronic myeloid leukemia (aCML), juvenile myelomoncoytic leukemia (JMML), refractory anemia with ring sideroblasts and thrombocytosis (RARS-T) as provisional entity, and MDS/MPN unclassified (MDS/MPN-U).¹

Over the past decade we learned much more about MDS/MPN group of myeloid neoplasms setting diagnostic criteria, risk stratification models, exploring underlying biology, borrowing treatment strategies from MDS and MPN, and attempting to define the clinical benefit of those therapies specifically for this group of patients.

In this review we highlight those efforts to better understand MDS/MPN, explore the challenges encountered diagnosing and managing MDS/MPN and define the unmet needs.

Challenges in diagnosis and update on classification

By definition MDS/MPN has features of both MDS and MPN which raises the first challenge of making the diagnosis and assigning the appropriate classification. The WHO diagnostic criteria are summarized in table-1.¹ The revised WHO 2016 classification retained the original MDS/MPN subtypes with further refinement and nomenclature table-1.² For CMML persistent monocytosis more than 3 month is adequate for making diagnosis even with lack of morphologic dysplasia. The challenge is excluding other causes of monocytosis, cytogenetic abnormalities and somatic mutations may help identifying clonal hematopoiesis ruling out reactive process. Co-existence of TET-2 and SRSF2 somatic mutations are suggested to be highly specific for CMML.³ A clonal marker can be identified in 93% using next generation sequencing on peripheral blood or bone marrow.^3,4

Table-1.

WHO MDS/MPN diagnostic criteria

	CMML	aCML	RARS-T	JMML	MDS/MPN-U
WHO 2016	CMML-0 <2% blasts in the blood and <5% blasts in the marrow CMML-1 2–4% blasts in the blood and 5–9% blasts in the marrow CMML-2 6–19% blasts in the blood and 10–19% blasts in the marrow	aCML BCR/ABL1 negative	Renamed: MDS/MPN with ring sideroblasts and thrombocytosis
Diagnostic Criteria	Persistent peripheral blood monocytosis >1×109/L No BCR-ABL or PDGFR fusion gene <20% blasts in the blood and bone marrow Dysplasia in one or more lineages, if no dysplasia then: ∙An acquired clonal cytogenetic or genetic abnormality ∙The monocytosis has persisted for >3 months ∙All other causes of monocytosis have been excluded	WBC >13×109/L with increased and dysplastic neutrophils No BCR-ABL or PDGR fusion Gene <20% blasts in the blood and bone marrow Minimal absolute basophilia No or minimal monocytosis Hypercellular BM with granulocyte dysplasia Neutrophil precursors >10% of leukocytes	Platelet count >450×109/L ∙15% ring sideroblasts in the bone marrow or >5% with SF3B1 mutation ∙Presence of megakaryocytic atypia resembling EF or MF	Peripheral blood monocytosis >1×109/L No BCR-ABL or PDGFR fusion gene <20% blasts in the blood and bone marrow Two of the following must be present: ∙Hemoglobin F increase ∙Immature granulocytes in peripheral blood ∙WBC >10×109/L ∙Clonal chromosome abnormality ∙GM-CSF hypersensitivity of myeloid progenitors in vitro	Features of MDS category and <20% blasts in blood and bone marrow Prominent myeloproliferative Features No preceding history of MPN or MDS, no recent cytotoxic or growth factor therapy No BCR-ABL or PDGFR or FGFR fusion and no isolated del(5q), chr 3 inversion or Features of mixed MDS MPN and cannot be assigned MDS, MPN or MDS/MPN Category

The other challenge is distinguishing presence of monocytosis in other myeloid neoplasms. Presence of myeloblasts and or promonocytes 20% or more is diagnostic for acute myeloid leukemia (AML). Rearrangements of PDGFR or FGFR1 are diagnostic of myeloid neoplasms with PDFR/FGFR-1 rearrangement.⁵ Repartition of monocytes to classical monocyte subset by flowcytometry CD14+/CD16 – is suggested to be specific for patients with CMML distinguishing it from reactive or other myeloid neoplasms.^6,7

For aCML the key is ruling out CML by lack of BCR/abl rearrangement which had been emphasized with the new nomenclature of WHO 2016 MDS/MPN classification.² Presence of granulocytic dysplasia is key to distinguish from MPN as well as presence of more than 10% neutrophilic precursors.²

RARS-T can be diagnosed as essential thrombocythemia if anemia is subtle and iron stains were not performed on bone marrow to detect ring sideroblasts or as RARS-T if the focus was on presence of ring sideroblasts. SF3B1, JAK-2 mutations are commonly seen in RARS-T and CALR mutation is been associated particularly with this MDS/MPN subtype.⁸

The MDS/MPN-U remains most challenging to diagnose relaying heavily on shared clinic-pathological features of MDS and MPN which cannot be classified otherwise.²

The evolving genetic landscape

Cytogenetic abnormalities are detected in almost 70% of MDS/MPN patients using conventional cytogenetic and high resolution single nucleotide polymorphism array karyotyping.⁹ Somatic mutations observed in MDS/MPN fall in 4 major categories including signaling, splicing, epigenetic and transcription gene mutations.^4,10 Table-2

Table-2.

somatic gene mutations in MDS/MPN

Pathway	Gene	Frequency %
		CMML	JMML	aCML	MDS/MPN-U	RARS-T
Epigenetic
	DNMT3A	2	0		4	15
	TET-2	58	0	30	30	25
	IDH 1/2	6	0		5–10
	ASXL-1	40	0	69		15
	EZH-2	5	0	13	10
Signaling
	RAS	19	39	35	10
	JAK-2	8	0	4–8		57
	CBL	10	15	8	>10	4
	SETBP1	6–15	8	48	10	1
	CSF3R			<10
	PTPN11		44
	NF1		13
RNA Splicing
	SF3B1	6	0			93
	SRSF2	46	0			7
	U2AF1	5	0
	ZRSR2	8	0
Transcription
	RUNX1	15	0	6	14
	CEBPA	4–20		4	4

In CMML almost 90% of patients will exhibit one or more somatic mutation. The most frequently observed mutations inlcudeTET2 (50%−60%), SRSF2 (40%−50%), ASXL1 (35%−40%) and RUNX1(15%). Signaling mutations correlate with proliferative CMML. TP53 mutations are rare. The concurrent mutations in TET2 and SRSF2 appear to be highly specific for CMML.^11–14

In aCML, SETBP1 mutations are enriched while CSFR3 mutations are more often observed in chronic neutrophilic leukemia (CNL) where morphological and clinical features can overlap with aCML.^15,16

In patients with RARS-T, 60–80% of patients harbor SF3B1 splicing mutation. JAK-2 and MPL mutation are reported in almost half of the patients. CALR mutation had been also described in RARS-T particularly among the MDS/MPN categories.^8,17

Somatic mutations in ASXL1, TET2, JAK2 and SRSF2 are common (>20%) in patients with MDS/MPN-U.¹⁸ In JMML almost all children will harbor a somatic or germline mutation in the RAS pathway.¹⁹

Risk stratification and prognostic models

Understanding risk of progression and death from disease is critical in the care of patients with MDS/MPN. The International Prognostic Scoring System (IPSS), the revised International Prognostic Scoring System (IPSS-R), the WHO-based Prognostic Scoring System, and the Dynamic IPSS have been used for patients with MDS/MPN, and the global MDACC prognostic model was developed to account for patients with proliferative CMML, and patients with MDS who received prior therapy. These systems have improved risk stratification and have informed management approaches.^20–24 The most commonly used is the IPSS-R, however, it has become clear that this approach is limited as it does not incorporate molecular genetic information, and is only validated for adult patients with MDS and dysplastic CMML at the time of diagnosis.

Given growing outcome data on CMML, several CMML-specific models recently have been developed for risk stratification, yet none of these models are used consistently in practice.^24,25 In 2002, Gonzalez-Medina et al. reported that although CMML could be classified into two groups, myelodysplastic and myeloproliferative, prognostic variables including bone marrow blast percentage, and elevated WBC count were more predictive of outcome.²⁶ The MD Anderson Prognostic Scoring System (MDAPS) has been used since 2002, and incorporates anemia, peripheral blood absolute lymphocyte count greater than 2.5 × 10⁹/L, bone marrow blasts < 10%, and presence of circulating immature myeloid cells (IMC).²⁵ In a group of CMML patients from the Mayo clinic, elevated peripheral monocyte count, anemia and thrombocytopenia were found to be significant prognostic factors. In the Mayo Clinic study, they reported that their model was more predictive of prognosis than MDAPS, G-MDAPS and Spanish cytogenetic risk models.²⁷ Prognostic scores that incorporate cytogenetics and specific gene mutations to determine risk stratification have more recently been developed. In 2013, Itzykson et al developed a model that included both clinical and genetic factors, and identified ASXL1 mutations, age >65 years, leukocytosis, thrombocytopenia, and anemia as independent prognostic variables.¹² The CMML-specific prognostic scoring system (CPSS) was also developed and published in 2013, and demonstrated prognostic impact for CMML FAB classification, WHO sub-classifications, CMML-specific cytogenetic risk classification and transfusion dependence.²³ The revised Molecular Mayo Model was developed in 2014, and differed from the original Mayo Score in that it accounted for the prognostic significance of ASXL1 mutations (both nonsense and frameshift).²⁸ Most recently, a unique clinical/molecular CPSS model (CPSS-mol) was developed and incorporates both a “genetic score” as well as clinical characteristics previously included in the CPSS to determine an overall prognostic score using both clinical and molecular variables.²⁹ Table-3

Table 3:

Prognostic Classification Systems for MDS/MPN

	Blasts	Cyto	Hgb	Plts	Circulating IMC	ANC	ALC	LDH	Age	Transfusion Dependency	PS	ASXL-1 mut	Cytogenetics
IPSS(1997)	+	+	+	+		+			+
WPSS	+	+	+							+
MDA-LR	+	+	+	+					+
MDAPS (2002)	+		+		+		+
MDAPS-M1 (2007)			+		+		+	+
MDAS (2008)	+	+	+	+		+			+	+	+
FPSS	+	+								+	+
IPSS-R (2012)	+	+	+	+		+			+		+		+
CPSS (2013)	+	+	+			+				+
Molecular Mayo Model (2014)			+	+	+							+
CPSS-mol (2016)	+									+			+

IPSS: International Prognostic Scoring System

WPSS: WHO Classification-Based Prognostic Scoring System

MDA-LR: M.D. Anderson Lower-Risk MDS Prognostic Scoring System

MDAPS: M.D. Anderson Prognostic Scoring System

MDAPS-M1: Modified M.D. Anderson Prognostic Scoring System

MDAS: Global MD Anderson Risk Model Score for MDS

FPSS: French Prognostic Scoring System

IPSS-R: Revised International Prognostic Scoring System

CPSS: Chronic Myelomonocytic Leukemia (CMML) Prognostic Scoring System

CPSS-mol: CMML-specific Prognostic Scoring System Molecular

While the presence of several mutations seem to correlate with decreased survival, more comprehensive genotyping of large data sets will need to be conducted to validate these mutations in scoring systems of risk for CMML.

For MDS/MPN-U no accurate model exists for prognosis and unfortunately majority of patients have poor outcome even when lower risk by IPSS or R-IPSS.³⁰

Defining benefit of treatment:

In 2006, the MDS International Working Group (MDS-IWG) published consensus guidelines for the measurement of response to therapy in MDS.^31,32 Accordingly, the International Working Group-Myeloproliferative Neoplasms Research and Treatment (IWG-MRT) and European LeukemiaNet (ELN) followed with a set of defined criteria for treatment response in myelofibrosis.³³ These guidelines have been applied to clinical studies that include patients with MDS/MPN^34,35 and in a small number of trials conducted specifically for MDS/MPN.^36–38 However, as patients with MDS/MPN exhibit both features of dysplastic and proliferative disease, it became clear that development of disease-specific guidelines for MDS/MPN would be necessary to best develop new therapy for MDS/MPN. The MDS/MPN International Working Group (MDS/MPN IWG) published MDS/MPN proposed response criteria for MDS/MPN in 2015. Complete remission (CR) was defined as all of the following: bone marrow with <5% myeloblasts, return of normal cellularity in the marrow, absent or < grade 1 osteomyelofibrosis, peripheral blood findings including WBC ≤ 10× 10⁹ cells/L, hemoglobin ≥ 11 g/dL, platelets ≥ 100 × 10⁹/L and ≤ 450 × 10⁹/L, neutrophils ≥ 1.0 × 10⁹/L, blasts 0%, neutrophil precursors reduced to ≤ 2%, and monocytes ≤ 1 × 10⁹/L, and resolution of extramedullary disease prior to therapy. Complete cytogenetic remission was defined as resolution of a previously present chromosomal abnormality that is known to be associated with MDS/MPN. Partial remission (PR) was outlined as normalization of peripheral blood counts and hepatosplenomegaly with bone marrow blasts reduced by at least 50%, but still greater than 5% of cellularity as required for CR. The group also proposed including a category of clinical benefit that includes blood cell counts, spleen size reduction, and improved functional status which are all meaningful metrics to be considered when evaluating patient outcomes.³⁹

Current therapies for MDS/MPN

Treatment strategies for MDS/MPN are largely borrowed from MDS or MPN available therapy. A crucial step in management is assuring the right diagnosis and risk stratification.

CMML management

Allogeneic hematopoietic stem cell transplant (AHSCT) is the only curative option and all higher risk CMML patients with good performance status and no comorbidities should be evaluated for transplant.^40,41 The long term disease free survival ranges from 18–47%.⁴¹ In retrospective study, complete response to hypomethylating agents (HMA) prior to AHSCT was associated with better outcome.⁴² For patients who are lower risk or those not candidate for AHSCT based on age, comorbidities or functional status then treatment is geared symptomatically with no clear evidence that early treatment alters the natural history of the disease. Observation is acceptable strategy for those patients with no cytopenia or proliferative symptoms. Cytoreductive therapy is offered to symptomatic patients with proliferative symptoms including constitutional symptoms, splenomegaly and MPN like symptoms. Hydroxyurea was superior to etoposide in a randomized clinical trial.⁴³ There is ongoing phase 3 trial comparing decitabine +/− hydroxurea to hydroxyurea alone. A recent phase 1/2 trial of the JAK inhibitor ruxolitinib demonstrated efficacy, particularly in patients with splenomegaly and constitutional symptoms.³⁶ For patients with cytopenia hypomethylating agents are commonly used, the impact of HMA on natural history of disease is controversial with no clear superior choice between azacitidine and decitabine although some retrospective studies suggested higher response rate with decitabine, however, similar overall survival. The overall response to HMA varies between 16 to 75% in the series reported. Leukocytosis and splenomegaly are associated with worse outcome. ASXL-1 mutation is negative predictor to HMA response while TET-2 MT/ASXL-1 WT are associated with higher overall response rate.^44–48 (Figure-1 How do I treat CMML)

Figure 1.

How do I Treat CMML?

Abbreviations: AHSCT = autologous hematopoietic stem cell transplant; CMML = chronic myelomonocytic leukemia; HMA = hypomethylating agent; NGS = next generation sequencing; WHO = World Health Organization.

RARS-T management

Patients with RARS-T tend to have a good overall prognosis. Clinical manifestations include thrombocytosis where management recommendations are similar to essential thrombocythemia with cytoreductive therapy and aspirin as indicated bases on risk of thrombosis while for patients with anemia treatment is borrowed from MDS. Anemia treatment includes erythroid stimulating agents, lenalidomide and hypomethylating agents.⁸ Case series suggest higher responses and long duration of response to lenalidomide compared to non del 5 q MDS.^49–51 Recently phase III study with luspatercept (an activin 2 B receptor fusion trap protein which improves terminal erythroid differentiation) finished accrual where patients with RARS-T as well as MDS patients with ring sideroblasts where included and may provide future option for anemia treatment.^52,53

aCML management

There is no standard of therapy for patients with aCML, only AHSCT offers cure and long term outcome.⁵⁴ Hypomethylating agents, interferon and hydrea have been used for symptomatic management.^55,56 Recent report for patients with CSFR3 mutation indicates potential activity for dasatinib for those with truncating mutations and ruxolitinib for those with proximal mutations.⁵⁷

MDS/MPN-U management

The outcome for MDS/MPN-U is very poor and challenges in treatment stems from challenges in diagnosis and exclusion of those patients from most of clinical studies. Treatment is often symptomatic, retrospective studies suggest potential survival advantage for hypomethylating agent use and AHSCT should be considered for those patients.⁵⁸

Future opportunities:

While there have been many trials for MDS or MPN that do include patients with CMML, the number of CMML patients on these trials are admittedly limited. Despite low representation on the AZA001 study (18 patients),⁵⁹ CMML was included in the approval label by the FDA for 5’azacitidine, leading to DNA methyltransferase inhibition (DNMTi) as the standard of care in CMML. This standard will not be accepted for future drug approval in CMML, however, and future dedicated trials for CMML and other MDS/MPN are necessary. Likewise, trials for relapsed MDS/MPN have been considerably underrepresented and there are no specific, approved, recommended treatment options. The MDS/MPN IWG has worked to advance therapies for MDS/MPN and is soon to begin a basket trial ABNL MARRO for patients with de novo and relapsed/refractory disease. This study incorporates novel therapy onto DNMTi backbone therapy in several different open label combination arms for patients with MDS/MPN previously treated or treatment naïve, and is due to enroll patients in 2019. Ruxolitinib, a JAK 1/2 inhibitor, has recently been approved by the FDA for use in myelofibrosis and has been tested both as monotherapy and in combination with azacitidine in the treatment of MDS/MPN.^36,38 Padron and colleagues conducted a multi-institution phase 1 trial of ruxolitinib in patients with CMML, which established the recommended phase II dose of ruxolitinib and reported promising activity in this cohort of patients.³⁶ To follow, Assi et al published results of a phase II trial using ruxolitinib in combination with azacitidine in MDS/MPN patients, and found that the combination was not only well tolerated, but demonstrated a response rate of 57% in patients with MDS/MPN.³⁸ JAK1 specific inhibition which purportedly avoids JAK2 specific dose dependent thrombocytopenia and anemia has been tested as a single agent and in combination with 5’azacitidine in MDS/MPN with early preliminary efficacy,^60,61 and this combination will be tested in early cohorts of ABNL MARRO by the MDS/MPN IWG.

In addition, compounds providing responses in other myeloid diseases face upcoming developments in MDS/MPN. Luspatercept is an activin FcVIII receptor agonist which has led to considerable clinical benefit in patients with refractory anemia, most impressively in patients with SF3B1 mutations and MDS-RS, with minor adverse events.⁵² Patients with MDS/MPN RS-T were specifically excluded thus far in the study of luspatercept, but given the transfusion-dependence of many MDS/MPN patients and high prevalence of SF3B1 mutations, specifically in MDS/MPN RS-T, this is imminent. It is unclear the role immune checkpoint inhibitors or the BCL-2 inhibitor venetoclax, will play in chronic myeloid diseases despite striking responses in acute myeloid leukemia (AML). Similarly, therapy targeted at genetic aberrations seen with considerable frequency in MDS/MPN (specifically, IDH inhibition, splicing inhibition for SRSF2, ZRSF2, SF3B1, or U2AF1 mutant disease) holds appeal in MDS/MPN but has yet to be tested.^62–64

A novel agent directed to the interleukin-3 receptor-alpha (CD123), SL-401, is currently being tested in clinical trial for patients with relapsed/refractory CMML. In a small subset of patients, this drug has demonstrated significant reduction in spleen size in majority of patients and in some, complete remission.⁶⁵ Pevonedistat (PEV), a NEDD8-activating enzyme inhibitor that has shown activity in recent clinical trials for AML and MDS, may also have promise in MDS/MPN overlap syndromes.⁶⁶