Paroxysmal nocturnal hemoglobinuria (PNH) is a rare acquired disorder caused by the somatic mutations of phosphatidylinositol glycan A (PIGA). The consequent defect of glycosyl phosphatidylinositol (GPI)-anchored proteins on red blood cell (RBC) surface increases the susceptibility of PNH cells to complement-mediated destruction, leading to intravascular hemolytic anemia, which is the main clinical feature of the disease (1, 2). The natural history of PNH was burdened by high morbidity due to chronic anemia and considerably increased mortality, mainly related to fatal thrombotic events (3). With the advent of complement inhibitors, PNH patients significantly ameliorated their quality of life and survival (4). PNH has been described in the context of bone marrow failure (BMF) syndromes, namely, aplastic anemia (AA) and myelodysplastic syndrome (MDS) (4). However, with the development of more sensitive cytofluorimetric techniques (5), PNH clones of various sizes are increasingly being detected in various onco-hematologic and autoimmune disorders (6–9). The coexistence of PNH and myeloproliferative neoplasms (MPNs) has been reported, but its clinical/prognostic significance and therapeutic management are still poorly known. Moreover, these two conditions share an overlapping clinical presentation, represented by thrombotic events at usual and unusual sites (3, 10–13). Here, we provide the description of a patient with MPN who was subsequently diagnosed with PNH and required specific treatment for hemolytic anemia. In addition, we searched for the available evidence in literature about the association of PNH and MPN, collecting data over the last 40 years in MEDLINE via PubMed and the National Library of Medicine. In detail, we reviewed data about the coexistence of clinically overt PNH and MPN, the prevalence of PNH clones in MPN patients, and the prevalence of MPN driver mutations in PNH subjects.

A 62-year-old Caucasian male was diagnosed with Janus kinase (JAK)2-negative essential thrombocythemia (ET) in May 2007 due to isolated asymptomatic thrombocytosis ( Table 1 ). His medical history was unremarkable, except for moderate arterial hypertension on regular treatment; no previous thrombotic events were registered. Bone marrow (BM) evaluation showed normocellularity (40%), increased mature megakaryocytes, slight increase of reticulin fibers (MF-1), and normal karyotype. Once-daily acetylsalicylic acid (ASA) and low-dose hydroxyurea (HU) were started with adequate control of platelet count. From March 2016, a trend to increased lactate dehydrogenase (LDH) levels was noticed, and from February 2019, a mild macrocytic anemia [hemoglobin (Hb) 10.2 g/dL, mean corpuscular volume (MCV) 103 fL; normal iron and vitamin status] developed ( Figure 1 ). Direct antiglobulin test (DAT) was negative, with mild elevation of unconjugated bilirubin (UB, 1.3 mg/dL), consumption of haptoglobin, and increased reticulocytes. HU dose was decreased, but anemia worsened (Hb nadir 7 g/dL), LDH rose to 4× upper limit of normal (ULN), and peripheral CD34-positive cells increased. The patient became strongly symptomatic for anemia, requiring about 1–2 RBC units/month. His physical examination was unremarkable. In October 2019, BM reevaluation showed increased cellularity (>95%) with dystrophic megakaryocytes and increased fibrosis (MF-2) and was therefore interpreted as fibrotic evolution of ET. Molecular tests on peripheral blood revealed the presence of a type-2 calreticulin (CALR) mutation at exon 9. Abdomen ultrasonography displayed normal spleen and liver size. HU and ASA were stopped, and an attempt with steroids (oral prednisone 50 mg/day) was made with a transient response and reappearance of transfusion dependency during tapering (2–3 packed RBC units/month). In February 2020, danazol was administered, again without anemia improvement. In November 2020, due to persistent intravascular hemolytic anemia and referred dark-colored urines, a flow cytometry for PNH was made and turned positive with a clone size of 95%/94% on neutrophils/monocytes. Low-molecular weight heparin (LMWH) prophylaxis (enoxaparin 4,000 units/day) was started, and the patient was referred to our center for treatment indications. No major PNH-related symptoms were registered. In December 2020, the patient started eculizumab after recommended vaccinations. In the subsequent months, Hb stabilized at 7.5–8 g/dL, LDH progressively lowered to 1.1×–1.2× ULN, and transfusion need returned to 1–2 units/month, with subjective amelioration of the quality of life and disappearance of hemoglobinuria. LMWH was stopped. In June 2021, due to a minor response to eculizumab, a re-evaluation was performed: DAT result turned positive for complement without evidence of cryoagglutinins (as frequently observed under eculizumab treatment), and UB levels slightly increased, indicative of extravascular hemolysis (EVH) during terminal complement inhibitor treatment. BM biopsy confirmed increased age-adjusted cellularity, granulocytic hyperplasia, and numerous megakaryocytes in loose and dense clusters, including both mature and atypical, dystrophic cells; fibrosis was stable with diffuse increase in reticulin and focal bundles of thick collagen fibers (MF-2); of note, hyperplasia of the erythroid lineage in the absence of dysplastic features was described. Karyotype analysis was normal. Analyzed by an expert hemopathologist, the BM trephine was consistent with MPN unclassifiable (MPN-U). A targeted next-generation sequencing (NGS) myeloid panel confirmed the presence of type-2 CALR mutation [variant allele frequency (VAF), 45%] and showed an additional somatic mutation in ten-eleven translocation 2 (TET2) gene (VAF, 9.7%). The patient is continuing regular fortnightly eculizumab infusions, with subjective benefit, although with persistent transfusion dependence.

Our patient, along with the others described, allows several clinical and pathogenic considerations about the rare coexistence of PNH and MPN. Besides AA and MDS, PNH clones have been detected also in the context of lymphoid disorders, such as acute lymphoblastic leukemia and lymphomas (9, 36), and in autoimmune/idiopathic cytopenias (6–8, 37). The review of the literature highlighted that about 10% of MPN patients harbor a PNH clone (26, 29, 30), and this frequency rises up to 17% if clones smaller than 1% are considered (27). More importantly, the disregarded association of these two conditions may cause a significant delay in PNH diagnosis, as observed in our case. Additionally, in the MPN setting, the differential diagnosis of hemolytic anemia may be hampered by several confounders: haptoglobin can be decreased in more than 30% of MF (38), LDH is often elevated as a consequence of disease burden (31), and reticulocytosis can be observed in case of myeloid metaplasia (39). Furthermore, the appearance of anemia in MPN should prompt the exclusion of fibrotic evolution, which was in fact observed in our patient, who met the diagnostic criteria for post-ET MF (40). However, at variance with post-ET MF, BM trephine revealed erythroid hyperplasia, which may be attributed to the concomitant peripheral hemolytic process and indicative of bone marrow compensation. Consultation with an expert hemopathologist may be thus advised when morphological findings are not fully consistent with a clear diagnosis and confounding factors coexist.

An important clinical issue is the thrombotic risk in MPN-PNH patients and its management. In our literature review, 35% of MPN-PNH subjects had a severe thrombotic presentation, often refractory to combined anticoagulant, cytoreductive, and anti-complement treatments. This frequency appears higher than that reported for isolated untreated PNH (18.8%) (41) and for MPN (complessively 20%, higher in PV vs. ET/MF and JAK2-mutated patients) (12, 13, 42, 43), possibly due to the association of two thrombophilic conditions. With regard to therapy, it is well established that cytoreductive and anti-coagulant/platelet therapies are the cornerstones of thrombosis treatment in MPN, although recurrencies interest about 20% of treated patients (44, 45). In PNH, complement inhibition (Ci) has proven to significantly reduce the thrombotic risk, while anticoagulation alone is poorly effective (4, 46). Whether primary thromboprophylaxis is indicated in untreated PNH is still debated. Given the higher risk observed in patients with a larger clone size (i.e., >50%), they are generally candidates to primary prophylaxis if there are no contraindications. Prophylaxis may be then discontinued once complement inhibitors are started, as in the case described (47). With regard to secondary prophylaxis in patients on complement inhibitors, some experts discontinue anticoagulants when Intravascular hemolysis (IVH) is well controlled by anti-complement therapy (4, 48, 49). Finally, anti-platelet prophylaxis had been stopped in our patient when CALR-positive post-ET MF diagnosis was made. In fact, JAK2-negative ET is apparently associated with lower rates of thrombosis (50, 51); additionally, the indication to anti-thrombotic primary prophylaxis in MF is not clear-cut (52).

Another relevant clinical issue in MPN-PNH patients may be the infectious risk due to the treatment with anti-complement therapy and the known infectious diathesis observed in MPN (53). Despite the known risk of capsulated bacterial infections under Ci (54), no infectious complications occurred in our patient and in only one out of 23 patients (4%) in the literature (18).

With regard to therapy, in our patient, eculizumab showed effectiveness in controlling IVH but was not able to resolve transfusion-dependent anemia. Accordingly, the review of the literature showed that only a fraction of MPN-PNH patients responded to eculizumab ( Table 2 ) (18, 25). In classic PNH, persistent anemia under Ci treatment can be caused by residual IVH, concomitant BMF, and EVH. The management of the latter still remains an unmet need, but promising results are coming from clinical trials with proximal complement inhibitors that avoid deposition of C3 on RBC surface (55).

Many speculations can be made regarding the pathogenic significance of the association of MPN and PNH clones. The evidence of MPN driver mutations, particularly JAK2, selectively in GPI-deficient cells (18, 21) has raised the hypothesis that they may confer an intrinsic growth advantage to PNH cells. This cooperative effect has also been proposed for other somatic mutations, such as TET2 (33, 35, 56, 57), also present in our patient. This view would provide a different explanation to the more common notion that PNH cells have an extrinsic growth advantage secondary to an autoimmune, GPI-selective process against bone marrow precursors (58, 59). Interestingly, MPN diagnosis preceded that of PNH in more than half of cases, and Shen et al. (33) demonstrated that PIGA mutation was subclonal to JAK2-mutated clone in their report. It is tempting to speculate that the MPN-associated inflammatory microenvironment (60–63) can impair normal hematopoiesis, exerting an “immune pressure” that favors PNH clone expansion, similarly to what happens in AA.

In conclusion, the association of MPN and PNH deserves attention because of the unpredictable clinical course often affected by dramatic thrombotic complications. Since PNH diagnosis is based on highly sensitive and cost-effective techniques, testing for PNH clones should be prompted in any MPN patient showing unexplained anemia with/without frank hemolysis, recurrent thrombosis, and/or atypical BM morphologic findings. Anti-complement treatment in the setting of MPN relies on a careful case-by-case evaluation, weighing the contribution of intravascular hemolysis to anemia and the thrombotic risk.

The original contributions presented in the study are included in the article/supplementary files. Further inquiries can be directed to the corresponding author.

The studies involving human participants were reviewed and approved by Comitato Etico Milano Area 2. The patients/participants provided their written informed consent to participate in this study.

JG, BF, and WB followed the patient, described his clinical history, and revised the literature and the paper for intellectual content. All authors contributed to the article and approved the submitted version.

WB received consultancy honoraria from Agios, Alexion, Novartis, Annexon, Sanofi, and Sobi. BF received consultancy honoraria from Amgen, Alexion, Novartis, Momenta, Annexon, and Apellis.

The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.