Characteristics and clinical correlates of MPL 515W>L/K mutation in essential thrombocythemia

Ninety-five percent of polycythemia vera^1-5  and approximately 60% of essential thrombocythemia (ET)^6-9  or primary myelofibrosis (PMF) patients^10-13  harbor a JAK2V617F mutation. Another recurrent molecular abnormality, involving a W to L or K transversion at MPL codon 515, has been reported in 5% of PMF and 1% of ET patients, sometimes coexisting with JAK2V617F.^14-17  The W>L allele conferred factor-independent growth to Ba/F3 cells and induced a disorder recapitulating human myelofibrosis in a murine bone marrow transplant model.¹⁷ 

Patients with myelofibrosis harboring MPLW515L/K mutations presented more severe anemia compared with MPL wild-type (wt), but no other meaningful characteristic was ascertained¹⁶ ; on the other hand, only a few ET patients with MPLW515L/K mutation have been reported to date.^14,17  The aim of this study was to describe the prevalence, characteristics, and clinical and laboratory features associated with MPLW515L/K mutation in a large population of ET patients.

This multicenter study was conducted within the Gruppo Italiano per le Malattie Ematologiche dell'Adulto (GIMEMA) myeloproliferative disease (MPD) Working Party; 6 centers reported an unselected population of 994 patients, for whom hematologic and clinical data at diagnosis and during follow-up (median follow-up, 59 months) were available. The study received formal institutional review board approval at the Coordinating Center in Florence (#2007/0047/08). Informed consent was obtained in accordance with the Declaration of Helsinki. Original diagnosis of ET was according to World Health Organization (WHO)¹⁸  or polycythemia vera study group (PVSG)¹⁹  criteria in 680 (68%) and 314 (32%) patients, respectively; in all of the latter, appropriateness of ET diagnosis according to WHO criteria was confirmed by ad-hoc reevaluation of bone marrow (BM) biopsies.

Splenomegaly was defined as a palpable organ below the left costal margin. Major thromboses at diagnosis or in the 2 preceding years²⁰  were recorded if objectively documented²¹ ; diagnostic criteria for thrombosis and major bleeding have been defined elsewhere.⁵  Microvessel disturbances (headache, acral paresthesia, erythromelalgia, transient neurologic and visual disturbances) were considered only if they were described from the patient as nonoccasional, of recent onset, and often ameliorated by aspirin; superficial venous thrombosis/thrombophlebitis was ruled out in case of erythromelalgia. Pruritus was recorded when it was described as diffuse, nonoccasional, itching exacerbated by water contact. Diagnosis of evolution to post-ET myelofibrosis fulfilled recent criteria.²² 

The JAK2V617F allele burden was measured by real time quantitative–polymerase chain reaction (RTQ-PCR) in granulocytic DNA.^23,24  Genotyping for MPLW515L/K was performed using a novel RTQ-PCR assay, based on locked nucleic acid probes, which reliably detected less than or equal to 0.1% mutant allele in a wild-type background²⁵ ; the procedure is described in Document S1 (available on the Blood website; see the Supplemental Materials link at the top of the online article).

BM biopsies from MPLW515L/K and MPLwt patients were analyzed blind by 3 hemopathologists. In addition, digital pictures were taken for to measure cellularity, fibrosis, number of megakaryocytes, small megakaryocytes, and megakaryocytic clusters (as described in Document S1).^26,27 

Statistical analysis was performed with GraphPad InStat (San Diego, CA) and SPSS (Chicago, IL) software. We used the chi-square or Fisher exact test (2 × 2 table) to compare categorical variables, the Mann-Whitney U test for continuous variables, and unconditional logistic regression models with backward stepwise where appropriate. Significance level was P less than .05 in 2-sided tests.

Characteristics of the 994 ET patients, categorized according to their genotype, are shown in Table 1. We found 30 patients (3.0%) harboring MPLW515L/K mutation, of whom 18 (60%) had the W515L and 12 (40%) the W515K allele. MPL mutant allele burden greater than 50% was found in 50% of W515K patients compared with 17% of W515L patients (P = .04). JAK2V617F mutation was harbored by 55.7% of the patients; 25 of them (4.5%) had greater than 50% V617F allele. MPLW515 and JAK2V617F mutation coexisted in 3 patients with W515L and 5 with W5151K allele, while no patient contemporarily harbored both MPL mutant alleles. The median V617F allele burden in double MPLW515L/K/JAK2V617F mutant was significantly lower than in MPLwt/JAK2V617F patients (8% vs 26%, P = .002), and for the purposes of this study MPLW515L/K patients were considered irrespective of their JAK2 genotype. Prevalence of MPLW515L/K mutation was 3.3%, 3.8%, and 2.5% in samples genotyped within 1 year from diagnosis (30%), within 3 years (21%) or later (49%), respectively, suggesting that the mutation is present at diagnosis rather than being acquired during the course of the disease.

Compared with MPLwt/JAK2wt, MPLW515L/K patients were older, but did not present any difference in hematologic parameters (Table 1); however, hemoglobin level was significantly lower and platelet count significantly higher compared with MPLwt/JAK2V617F patients. Arterial thrombosis during follow-up occurred at significantly higher rate in MPLW515L/K patients than in MPLwt/JAK2wt, while conversely more arterial events at diagnosis were found in MPLwt/JAK2V617F than in MPLwt/JAK2wt; the low number of events precluded multivariate analysis, and no definite statement in this regard could be made. There was no difference in venous thromboses or hemorrhages between MPL mutant and MPLwt patients. However, microvessel disturbances were significantly more frequent in MPLW515L/K than in MPLwt patients, and both JAK2wt and JAK2V617F; in multivariate analysis, the relative risk was 2.0 (95% confidence interval, 1.5-2.7; P < .001) using MPLwt patients as reference. MPLW515L/K did not impact on transformation to myelofibrosis that occurred in 6.4% of MPLW515L/K compared with 3.2% and 5.0% of MPLwt/JAK2wt and MPLwt/JAK2V617F patients, respectively. Finally, analysis of BM biopsies revealed significantly reduced total and erythroid cellularity in MPLW515L/K patients, associated with increased number of megakaryocytes, megakaryocytic clusters and small megakaryocytes, but no increase in reticulin fibrosis (Table S1).

MPLW515L and W515K mutant patients were then considered separately to ascertain characteristics associated with each mutant allele (Table 2). W515L mutant patients were older than W515K and MPLwt/JAK2wt patients, and displayed significantly lower hemoglobin level compared with W515K and MPLwt patients, particularly if they harbored JAK2V617F mutation. Conversely, platelet count was significantly higher in W515K patients than in W515L and MPLwt/JAK2V617F patients. There was no difference between W515L and W515K patients in splenomegaly, pruritus, hemorrhage, arterial or venous thrombosis or microvessel disturbances that occurred in 55% and 66% of patients, respectively.

In this series of ET patients we found an association of MPLW515L/K mutation with lower hemoglobin level, as also reported in myelofibrosis,¹⁶  and with higher platelet count, especially in comparison to MPLwt/JAK2V617F counterpart. These abnormalities mimic the phenotype of mice transplanted with MPLW515L-transduced cells that presented thrombocytosis and a shift of erythroid cells toward more immature phenotype¹⁷ ; furthermore, MPLW515L/K progenitor cells from PMF patients were found to generate thrombopoietin-independent megakaryocytic colonies but no erythropoietin-independent erythroid colonies, and overall erythroid progenitor cloning efficiency was reduced.^28,29  In keeping with this was the reduced erythroid cellularity with increased megakaryocytes we documented in BM biopsies from MPLW515L/K patients; this would support a preferential expansion of megakaryocytic lineage at the expense of erythroid differentiation due to activating MPL mutation.³⁰ 

A novel finding was the association of MPLW515L/K with microvessel disturbances; because activation of Mpl by thrombopoietin enhances normal platelet function,³¹  and abnormal activation of ET platelets by thrombopoietin preincubation has been described,³²  we hypothesize that platelets from MPLW515L/K patients present constitutively enhanced reactivity.

In summary, this large survey indicated that prevalence of MPLW515L/K mutation in ET is higher than originally described and closer to myelofibrosis,^14,16,17  as also observed in the PT-1 cohort³³ ; however, because MPL mutant alleles other than W515L/K have also been described,³⁴  overall prevalence of MPL mutation in ET might be slightly underestimated. Although our findings did not result in the definition of a discrete clinical phenotype, they underscore the relevance of MPL mutational screening in the diagnostic work-up of suspected ET as a tool to establish the occurrence of a clonal myeloproliferation³⁵  in cases that lack the JAK2V617F mutation.

This study was supported by Associazione Italiana per la Ricerca sul Cancro, Milano, Italy; Ente Cassa di Risparmio di Firenze; MIUR projects (COFIN 2006067001_003) to AMV; and funds from Istituto Toscano Tumori (Firenze, Italy). A.P. was the recipient of a fellowship from Associazione Italiana per le Leucemie, Firenze, Italy. V.G. was supported by fellowship “Schenker-Erico Ghezzi” from the Italian Association for Cancer Research (AIRC).

Contribution: A.M.V. designed research, analyzed data, and wrote the manuscript; E. Antonioli collected and analyzed clinical data, and helped write the manuscript; P.G. collected and analyzed clinical data, assayed JAK2 and MPL mutations, reviewed biopsies, and helped write the manuscript; A.P. performed MPL mutational assay and helped biopsy data collection; V.G., G.B., M.R., G.S., F.L.C., F.D., L.V., S.F., E. Ammatuna, A.R., L.V., and A.B. collected clinical data and biologic samples; V.C., R.A., and S.D.L. reviewed biopsies; G.C. processed bone marrow biopsies; G.B., F.L.C., and T.B. discussed data and helped write the manuscript. All authors reviewed and gave final approval to the manuscript.

Conflict-of-interest disclosure: The authors declare no competing financial interests.

Correspondence: Alessandro M. Vannucchi, MD, UF di Ematologia, Dipartimento di Area Critica Medico-Chirurgica, Università degli Studi, Viale Morgagni 85, 50134 Florence, Italy; e-mail: amvannucchi@unifi.it.