Survivorship and prognosis in myelofibrosis with myeloid metaplasia

Myelofibrosis with myeloid metaplasia (MMM) is currently defined by a set of bone marrow histological and cytogenetic criteria that are fulfilled either de novo (agnogenic myeloid metaplasia; AMM) or in a setting of antecedent polycythemia vera (post-polycythemic myeloid metaplasia; PPMM) or essential thrombocythemia (post-thrombocythemic myeloid metaplasia; PTMM) [1]. Pathogenetic studies have thus far established MMM as a stem cell-derived clonal myeloproliferation that is associated with cytokine-mediated bone marrow stromal reaction that includes fibrosis and osteosclerosis [2]. However, the primary molecular event(s) that leads to MMM remain obscure, despite the recent description of an activating JAK2 mutation (JAK2^V617F) that occurs in approximately one half of the patients with either AMM or PTMM and in more than 90% of those with PPMM [3],[4]. Currently, there is no effective drug therapy in MMM, and most patients are not suitable candidates for allogeneic hematopoietic stem cell transplantation (AHSCT) because of either their advanced age or associated comorbidities [5].

Reported median survival in AMM ranges from 3.5-10 years [6-8] and prognostic factor-adjusted survival in both PPMM and PTMM might be similar to that of AMM [9]. Among the many factors that are associated with shortened survival, hemoglobin level 10 g/dl, constitutional symptoms, circulating blasts ≥1–3%, leukocyte count >30 × 10⁹/l or <4 × 10⁹/l, platelet count <100 × 10⁹/l, circulating immature granulocytes ≥10%, age above 60 years, male sex, and cytogenetic abnormalities have been utilized in different prognostic scoring systems for AMM [6-8],[10],[11]. In this issue of Leukemia & Lymphoma, Strasser-Weippl et al. [12] confirm the independent prognostic value for survival of both age and hemoglobin level in MMM and propose a yet another prognostic scoring system that is based on age (> or ≤60 years) and hemoglobin level (< or ≥10 g/dl). They also show the applicability of their prognostic model in their sub-cohort of patients with AMM. The authors assert that their prognostic model is better than other previous models in separating high- from intermediate-risk patient groups. Whereas the observations from the current study are not surprising taking into account the age distribution of the study population (median age = 68 years), the question is whether or not an age-based prognostic scoring system has any therapeutically relevant practical utility in MMM.

As mentioned above, AHSCT has the potential to alter the natural history of MMM but is associated with a relatively high complication rate that includes death and chronic graft-vs.-host disease [5]. Therefore, at present, an effective prognostic scoring system in MMM should enable accurate identification of patients in whom the benefit of AHSCT outweighs its risk. To that effect, such a system should be applicable to transplant-eligible patients, who are often younger than 60 years. This has been accomplished by two previous studies where prognostic factors were evaluated in patients below the age of either 60 years [13] or 56 years [10] (Table I) [14]. In the most recent of these two studies [13], a complete blood count-based scoring system accurately identified high-risk as well as intermediate-risk disease categories, thus allowing the respective consideration of AHSCT or experimental drug therapy. Furthermore, two more recent studies have suggested an additional prognostic information from cytogenetic studies in transplant-eligible patients with either AMM or secondary myelofibrosis [9],[15].

Table I.  Prognostic scoring systems in transplant-eligible patients (age <60 years from Dingli et al. [12] and ≤55 years from Cervantes et al. [10]) with agnogenic myeloid metaplasia.

	Risk category	Score sum	Median survival (months)	Score for Hgb <10 g/dl	Score for WBC 4 < or >30 × 10^9/l	Score for Plt <100 × 10^9/l	Score for constitutional symptoms	Score for circulating blasts ≥1%
Dingli et al. [13]	Low	0	155	1	1	1	NA	NA
	Intermediate	1	69
	High	≥2	24
Cervantes et al. [10]	Low	0 or 1	176	1	NA	NA	1	1
	High	≥2	33

Hgb, hemoglobin; WBC, white blood cell count; Plt, platelet count. Adapted with permission [14].

What determines the worse prognosis of elderly patients with AMM? In his classic monograph published in 1975, the late Murray N. Silverstein (1928–98) highlighted the contribution of age-associated vascular and other comorbid events as causes of death in patients with AMM [16]. Consistent with this notion, advanced age is associated with inferior survival also in the clinically more indolent forms of myeloproliferative disorders, including essential thrombocythaemia and polycythemia vera [17],[18]. On the other hand, because drug therapy in MMM is currently suboptimal, it is unlikely that differences in either treatment approaches or patterns of care could explain the age-related difference in survival. Finally, the recent demonstration of a significantly higher incidence of both JAK2^V617F mutation and abnormal cytogenetic profile in older patients with AMM supports the potential role of differences in underlying disease biology in accounting for the difference in survival between young and older patients.