Paratrabecular myelofibrosis and occult mastocytosis are strong morphological clues to suspect FIP1L1-PDGFRA translocation in hypereosinophilia

Abstract

To study the clinico-haematological and histopathological characteristics of FIP1L1-PDGFRA rearranged hypereosinophilia/hypereosinophilic syndrome (F/P+ve HE/HES), a retrospective analysis of patients with F/P+ve HE diagnosed over a period of 43 months was performed. Peripheral blood smears, bone marrow aspirate (BMA) and biopsies (BMB) were reviewed in each case and; reticulin stain and immunohistochemistry for mast cell tryptase (MCT) and CD117 was performed. F/P+ve HE was diagnosed in a total of ten patients during study period. All patients were males with a median age of 36 years (23–44 years). The median duration of presenting complaints was 7 months (2 months–3 years) which included specific symptoms related to various organs (80% of cases). Anaemia, thrombocytopenia and splenomegaly were seen in 60%, 50% and 90% of the cases respectively. Mastocytosis was not obvious in BMA but identified by MCT on BMB in all cases. Myelofibrosis (grade ≥ 1) was seen in 80% of the cases and includes multifocal paratrabecular fibrosis in 50% of the biopsies. Our study shows that bone marrow mastocytosis and myelofibrosis are very useful morphological indicators to suspect F/P+ve HE and suggests the routine use of reticulin staining and MCT immunohistochemistry in all BMBs performed for the evaluation of HE/HES.

Introduction

The WHO recognized “Myeloid/lymphoid neoplasms with eosinophilia and PDGFRA rearrangement” usually manifests as hypereosinophilia (HE; absolute eosinophil count/AEC of ≥ 1.5 × 10⁹/l) or hypereosinophilic syndrome (HES; HE with organ damage) [1] and less frequently as acute myeloid leukemia or T-lymphoblastic leukemia/lymphoma [2]. It involves the constitutive activation of tyrosine kinases due to FIP1L1-PDGFRA (F/P) translocation that are highly sensitive to tyrosine kinase inhibitor (TKI) Imatinib [2, 3]. There are only a few reports of F/P+ve HES from tropical countries [1, 4, 5] which otherwise have a high burden of parasitic infection associated HE. In the absence of well-characterized morphological features, F/P+ve HE continues to be underdiagnosed in resource limited settings. In this background, we aim to explore morphological clues which help the pathologist to suspect F/P translocation in patients with HE/HES who undergo bone marrow examination.

Material and methods

A retrospective analysis of the clinico-haematological spectrum of patients with F/P + ve HE diagnosed and treated over a period of 43 months (January 2016–July 2019) was performed. Individual patient files were screened for clinical and follow-up data. Peripheral blood film (PBF), bone marrow aspirate (BMA) and biopsies (BMB) were reviewed. Testing for F/P translocation by fluorescence in-situ hybridization (FISH) (Vysis PDGFRA, CHIC2, FIP1L1 tricolor re-arrangement probe (Abbott Molecular, Illinois, USA) was routinely performed in-house from March 2018 to July 2019 (17 months), however it was intermittently outsourced for RT-PCR from various other laboratories before March 2018. Report of the FISH results was reviewed from the laboratory records for the percentage of positivity. Additionally, reticulin stain and immunohistochemical analysis of mast cell tryptase (MCT) (clone AA1, DAKO) and CD117 (clone A450229, DAKO) was performed to assess the degree of fibrosis and mast cell (MC) count respectively. All procedures followed were in accordance with the ethical standards of the responsible committee on human experimentation (institutional and national) and with the Helsinki Declaration of 1975, as revised in 2008.

Results

A total of ten cases of F/P + ve HE were diagnosed during the study period. Nine cases were diagnosed between January 2016 and July 2019. The remaining one patient was on empirical Imatinib since 2012 for HES, and F/P translocation was confirmed in March 2019, at the time of relapse. From March 2018 to July 2019 (17 months), regular and routine testing for F/P translocation was performed by in-house FISH testing. During this period, a total of 45 cases of eosinophilia underwent testing; and F/P translocation was diagnosed in six cases (13.3%).

The demographic, clinical and laboratory profile of all ten cases is summarized in Table 1. HE was associated with organ-related symptoms (HES) in eight cases (80%). Splenomegaly was present in 90% and hepatomegaly in 10% of the patients. 90% (n = 9) had AEC > 5 × 10⁹/L. Anemia was seen in 60% while thrombocytopenia was seen in 50% of patients. A few immature myeloid precursors (myelocytes and metamyelocytes) or nucleated red blood cells were noted in all cases in PBF with a leukoerythroblastic picture in five cases (50%). Dacrocytes were seen in seven cases. Morphologically, the eosinophils showed hyperlobation with tri- and quadri-lobated forms along with cytoplasmic vacuoles in all the cases (Fig. 1a, b). In all the nine cases (where BMA/BMB was available for review), the bone marrow was hypercellular (80–100% cellularity) and revealed intense granulocytic hyperplasia, particularly of eosinophils and its precursors. None of the cases had evidence of any other hematological malignancy. The MC distribution was evaluated by immunohistochemistry for MCT and CD117. All cases revealed a diffuse interstitial increase of round to spindle-shaped MCs (Fig. 2a). Spindle shaped cells constituted > 25% of the MC in 90% of the cases. Five cases (50%) also showed loose aggregates of MCs (Fig. 2b–d). The aggregates were seen in a perivascular location or paratrabecular location (Fig. 2e, f). The MCs were better highlighted by MCT compared to CD117. Overall, the distribution of MCs was quite varied throughout the biopsy. In focal areas, MC constituted 5–30% of the nucleated cells, whereas in majority of other areas they were singly dispersed and accounted for 1% to less than 5% of all cells. None of the cases showed obvious increase of MCs in BMA or imprint smears or dense clusters which limits the ability for exact quantification. Myelofibrosis was present in eight cases (80%) [grade 1, grade 2 and grade 3 in three, two and three cases respectively, as per WHO grading system 0–3] [2]. The fibrosis was diffuse only in one patient while all other cases had patchy multifocal fibrosis localized to areas showing MC aggregates giving a streaming appearance to the cells (Fig. 1d). Five of these cases (50%) also showed multifocal paratrabecular fibrosis (Fig. 1c and e). 8/10 (80%) cases were positive for F/P rearrangement, detected by FISH and the percentage of positivity ranged from 48 to 95% of interphase cells (Fig. 1F). F/P re-arrangement was detected by RT-PCR in the other two (20%) cases.

Table 1.

Demographic, clinical profile of F/P + ve HE

Median Age (range) (years)	36 (23–44)
Sex ratio (Male: female)	All males
Duration of symptoms—median (range)	7 months (2 months–3 years)
Signs and symptoms	n (%)
Fatigue	6 (60)
Fever	5 (50)
Skin rash and pruritus	5 (50)
Left hypochondriac pain related to massive splenomegaly	2 (20)
Oral ulcers	3 (30)
Cardiac symptoms (dyspnoea, palpitations)	2 (20)
Paraparesis	1 (10)
Lower limb swelling due to deep vein thrombosis	1 (10)
Splenomegaly [n (%)] (Total; mild; moderate; massive)	9 (90%); 4 (40%); 3 (30%); 2 (20%)
Organ involvement [n (%)]—total; skin; heart; blood vessel; vertebrae	8 (80); 5 (50); 1 (10); 1 (10); 1 (10)
Laboratory profile
Haemoglobin (g/l) [Median (range)]	102(71–137)
Total leukocyte count × 109/l [Median (range)]	21.5(8.1–60.6)
Platelet count × 109/l [Median (range)]	138.5(44–271)
Peripheral blood eosinophil % [Median (range)]	60.5 (26–85)
Absolute eosinophil count × 109/l [Median (range)]	13.7(2.6–24.5)
No. of cases with blasts in peripheral blood [n (%)]	Nil
Bone marrow eosinophil (%) [Median (range)]	35 (18–58)
% of blasts in BM [Median (range)]	1(1–3)

Fig. 1.

a, b Peripheral blood film showing tri- and quadri-lobated forms of eosinophils with cytoplasmic vacuoles as well as a myelocyte (100×, May Grunwald Giemsa stain); c Bone marrow shows paratrabecular fibrosis (10×, Hematoxylin and Eosin stain); d Multifocal patchy fibrosis; e Paratrabecular fibrosis (4× reticulin staining); fCHIC2 deletion (red) detected by FISH (Vysis PDGFRA, CHIC2, FIP1L1 tricolour re-arrangement probe, Abott molecular, Illinois, USA)

Fig. 2.

a Mast cell tryptase immunostain showing predominantly interstitial increase of round and spindle shaped mast cells (10×); b–d Multifocal loose aggregates of mast cells (4 × or 20×) in a fibrotic background giving a streaming appearance; e Paratrabecular aggregates of mast cells (20×); f Perivascular aggregates of mast cells (20×)

Treatment and follow-up data were available in 9 cases all of whom received Imatinib (100 mg daily). One patient with deep vein thrombosis also received Dabigatran and another patient with paraspinal extramedullary myeloid sarcoma underwent laminectomy to relieve spinal cord compression. The median duration of follow-up was 8 months (1 month–7 years). This includes the extended follow-up of three cases reported in our previous series [1]. All patients except one (who was not compliant) showed normalization of AEC within one month of therapy and remained asymptomatic on low dose Imatinib (100 mg daily) at the time of last follow-up. One patient experienced relapse on stopping the drug.

Discussion

The exact prevalence of HE associated with FIP1L1-PDGFRA rearrangement (F/P+ve HE) in tropical countries is not known. A previous case series from South India reported a frequency of 19.2% among 26 adults with HE [5]. The frequency of F/P + HE/HES ranges from 3–56% in the literature [1, 5–10]. The current study shows a prevalence of 13.3% in consecutive 45 patients who underwent FISH testing for F/P translocation over a period of 17 months. The huge variation in the frequency is likely related to the large selection biases in these studies [3].

There are no clinical or laboratory features that can completely predict F/P translocation in patients with HE/HES [1]. However, F/P+ve HE has been reported to be associated with elevated MCT and vitamin B12 levels. Increased MCs, blasts, and myelofibrosis are found more frequently in the BMBs of these patients [2, 11]. In our study, all the patients had myelocytes or metamyelocytes or leucoerythroblastic picture in peripheral blood; however, none had blasts in PBF or increased blast count in BMA. Hyperlobated eosinophils, as well as cytoplasmic vacuoles, was observed in all cases. As reported previously, PBF findings in F/P+ve HE is not specific and can be seen in reactive conditions also [1, 2]. Increased MCs (singly scattered in 100% and multifocal loose clusters in 50% cases and composed of both spindle-shaped and round ones) and myelofibrosis (grade ≥ 1) were noted in 100% and 80% of the patients respectively. The increase in MCs were not noticeable on May Grunwald Giemsa stained BMA or hematoxylin and eosin-stained BMB; and required either CD117 or preferably MCT immunostaining (occult mastocytosis). The study by Schwaab et al. also showed mastocytosis and myelofibrosis in 100% and 71% of the F/P+ve HE cases respectively [11]. An earlier report has shown peritrabecular clusters and diffuse infiltration of MC to be frequent with systemic mastocytosis (SM) without F/P translocation [6]. However, multifocal paratrabecular infiltrate was seen in five of our patients (50%).

Being a retrospective analysis, the study has certain limitations. The serum tryptase levels or vitamin B12 levels were not available in any of our patients. The aberrant MC expression of CD2 and CD25 though reported to be frequent in F/P+ve HE [11] were not evaluated by flowcytometry at the time of diagnosis. The immunohistochemistry attempted retrospectively using anti CD25 monoclonal antibodies (4C9 clone) also did not give satisfactory results. Despite these limitations, our study supports the routine use of reticulin staining and MCT immunohistochemistry in all BMBs performed for the evaluation of HE/HES to identify myelofibrosis and occult mastocytosis. In patients with mastocytosis or fibrosis, F/P translocation must be excluded. The presence of scattered or clusters of either normal or spindle-shaped MCs has been reported in previous reports of F/P+ve HE [12, 13] and cannot distinguish SM from F/F/P+ve HES [14, 15]. In the absence of F/P translocation, the closest differential diagnosis is SM with eosinophilia (SM-Eo) which have overlapping morphological features with F/P+ve HE, especially in cases with clusters of MCs. In such cases, KIT mutation studies (especially D816V) must be performed. Though, F/P+ve HES and KIT+ve SM-Eo appear to be distinct entities, an occasional study has shown that F/P translocation favour secondary mutations including KIT mutation in MCs which remains at a low burden detectable only after sorting of MCs by microdissection or other techniques [16].

Like early reports [8, 12, 17], 90% of our patients who received Imatinib regularly achieved and continue to be in morphological remission; and one patient relapsed on stopping the drug. The current challenge is that, despite good response and in-depth molecular response, these patients relapse on discontinuation of Imatinib [3].

Conclusion

Our study despite having a few limitations highlight the importance of identifying myelofibrosis and occult mastocytosis in patients undergoing bone marrow examination in the evaluation of HE/HES. Bone marrow mastocytosis and myelofibrosis were noted in 100% and 80% of the patients respectively with multifocal paratrabecular fibrosis in 50% of cases suggesting that they are very useful morphological indicators to suspect F/P+ve HE.

Funding

There are no source(s) of funding in this study.

Compliance with Ethical Standards

Conflict of interest

All authors declare that they do not have any conflict of interest to declare.

Ethical approval

All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.

Informed consent

Informed consent was obtained from all individual participants included in the study.