Hyperferritinemia and diagnosis of type 1 Gaucher disease

EuroBloodNet Referral Center for Iron Disorders and Gruppo Interdisciplinare Malattie del Ferro, Internal Medicine Unit, Azienda Ospedaliera Universitaria Integrata, Verona, Italy Regional Referral Center for Lysosomal Storage Diseases, Division of Internal Medicine and Metabolism, Azienda Ospedaliero-Universitaria di Modena Ospedale Civile, University of Modena and Reggio Emilia, Modena, Italy Department of Medicine and Medical Specialities, Fondazione IRCSS Cà Granda, Milan, Italy Department of Clinical Sciences and Community Health, University of Milan, Milan, Italy Division of Internal Medicine and Center for Hemochromatosis and Heredometabolic Liver Diseases, EuroBloodNet Referral Center for Iron Disorders, Azienda Ospedaliero-Universitaria di Modena Policlinico, Modena, Italy Department of Medical and Surgical Sciences, University of Modena and Reggio Emilia, Modena, Italy EuroBloodNet and MetabERN Referral Center, Department of Medicine and Surgery, University of Milano-Bicocca, Medical Genetics, ASST Monza S. Gerardo Hospital, Monza, Italy


| CASE PRESENTATION
Initially, Gaucher disease (GD) was not suspected. The patient had ≥2 alterations of the metabolic syndrome, liver steatosis and normal TSat, fitting the criteria for the diagnosis of DIOS, 1 which represents a far more frequent cause of hyperferritinemia than GD. A possible Ferroportin Disease, suggested by iron accumulation in Kupffer cells, 2 was ruled out by SLC40A1 sequencing.
The patient reported worsening back and knee pain. Lumbar spine MRI revealed diffuse bone marrow infiltration, multiple bulging disks, and osteophytosis.
The initial diagnosis of DIOS was questioned since a more pronounced decrease of serum ferritin would have been expected due to the substantial metabolic improvement. Considering the whole clinical phenotype characterized by worsening thrombocytopenia, polyclonal gammopathy, and bone pain, GD was eventually suspected and confirmed by the reduced β-glucocerebrosidase activity, on both dried blood spot (DBS) and leukocytes (0.6 nmol/mg/h; normal range 10-30). GBA1 sequencing detected homozygosity for the pathogenic N370S mutation. A critical revision of the liver biopsy allowed re-defining the "hypertrophic" Kupffer cells as Gaucher cells.
Abdominal MRI showed hepatomegaly (2043 mL; normal value <2000), normal liver iron concentration (35 ± 20 μmol/g, normalvalue <36), and mild splenomegaly (510 mL; normalvalue 110-340) with several nodular lesions (the so-called "Gaucheromas") Gaucher Disease is frequently underdiagnosed because of lack of awareness, even when all typical clinical signs and symptoms are present. 3 This case was challenging since the classical hematological signs of GD were initially mild and incomplete (normal Hb, barely detectable thrombocytopenia and splenomegaly). Over the 15 years follow-up, the platelet count dropped and the patient developed severe and symptomatic bone marrow infiltration, justifying the starting of ERT. 4,5 Regarding hyperferritinemia, it was initially attributed to metabolic syndrome, and improvement of lifestyle was recommended. Of note, obesity is uncommon in untreated GD patients. 6 Although liver biopsy and normal TSat substantially excluded a true iron overload (IO), the patient was initially treated by unnecessary phlebotomies.
We subsequently collected 8 GD type 1 (GD1) patients referred to four Italian specialized centers for iron disorders, because of "unexplained hyperferritinemia" ( Table 1). All subjects were Caucasians, with a median age at diagnosis of 45 years. Both sexes were represented (five males and three females). At least one N370S mutation was present in seven out of eight patients. Median ferritin was 1285 μg/L, median TSat was 27%, and only one of the patients had evidence of significant IO, in line with previous case series. 7,8 Noteworthy, the classical GD hematological alterations were mild or even absent at first evaluation, underscoring the need for a high degree of suspicion.
Thrombocytopenia and splenomegaly were present in all patients, but were both mild in the majority. In particular, splenomegaly was barely noticeable (spleen longitudinal diameter ranging from 12 to 13 cm by

| DISCUSSION
Gaucher Disease is a pleiotropic lysosomal storage disorder due to biallelic mutations in the GBA1 gene (1q21), or, in exceedingly rare cases, in the PSAP gene, which encodes for its activator protein (saposin C). Reduced or absent activity of β-glucocerebrosidase results in the accumulation of glucocerebroside in macrophages of several organs, mainly the spleen, liver, and bone marrow. 9 The current phenotypic classification of GD into different subtypes is likely an oversimplification of a continuum of the same enzymatic defect that can be variably severe and clinically expressed. Such classification is based on the presence of central nervous system involvement, being type 1 the "non-neuronopathic", type 2 the "acute neuronopathic", and type 3 the "chronic neuronopathic" form of the disease. 10 A certain degree of overlap is possible, particularly regarding the occurrence of Parkinsonism in GD1 patients. 11 Gaucher Disease type 1 accounts for about 90% of all GD cases, and has a heterogeneous clinical expression, ranging from overt visceral and bone involvement in childhood/ early-adulthood to more faded manifestations in late-adulthood.
Unrecognized GD often leads to significant disability due to bone pain, fractures, and bleeding. N370S homozygosity in GBA1, the most common genotype in Ashkenazi Jewish, and the second most common one in European non-Jewish patients, is traditionally related to a mild phenotype but a clear genotype-phenotype correlation is lacking, and acquired factors can also influence the phenotype. 12 East Italy. 14 Indeed, such new findings have corroborated the belief that the disease is often under-recognized. The increasing availability of "dried blood spot" tools for widespread first-level analysis of enzymatic activity represents a useful help in the diagnostic process. 15 The confirmation of low enzymatic activity is virtually diagnostic, and has almost completely superseded bone marrow examination, 16 which could still be needed only when the clinical differential diagnosis includes a hematological malignancy. 17 Noteworthy, effective treatments are available for GD1 and for non-neurological manifestations of GD3, that is, enzyme replacement therapy (ERT) and substratereduction therapy (SRT). 10 In such a context, it is essential to increase the awareness of the disease in order to make the right diagnosis as early as possible, hence reducing the disability burden. Hematologists are the specialists more likely involved in the diagnosis of GD, but, even when all classical clinical features are present, only a minority (~20%) of them consider GD in the differential diagnosis. 3,18 Consultation of several specialists, misdiagnosis, and diagnostic delay are frequent features of the patient's journey towards the correct diagnosis of GD, eventually resulting in a poor quality of life, complications, and sometimes irreversible disabilities. 18 Common misdiagnoses include hematological malignancies, immune thrombocytopenic purpura, autoimmune disease, liver cirrhosis, idiopathic avascular necrosis, and "idiopathic" splenomegaly. 18 Since hematologic findings like splenomegaly, thrombocytopenia and anemia are among the most prevalent signs of GD, diagnostic algorithms starting from these findings have been proposed, with dedicated algorithms for subjects of Jewish ancestry due to the higher prevalence of GD. 19 For example, the application of one of these algorithms to patients referred to Italian hematology outpatient clinics because of splenomegaly and/or thrombocytopenia detected GD in 7 out of 196 cases (prevalence 3.6%). 20 Nonetheless, because of phenotypic heterogeneity and lack of awareness, other "high-risk patterns" need to be considered. Biochemical abnormalities commonly found in GD include hyperferritinemia, hypolipidemia with low HDL-cholesterol levels, low vitamin B12, polyclonal hypergammaglobulinemia, and MGUS. 6,7 Hyperferritinemia is a frequent finding in clinical practice, potentially associated with many different etiologies. 21 iron homeostasis), generally associated with increased TSat, or acquired (eg, due to repeated blood transfusions). More frequently, hyperferritinemia does not reflect a true IO, but it is due to inflammation, chronic liver disease, excessive alcohol intake, components of metabolic syndrome or malignancies. In these cases, TSat is usually normal. To make the correct diagnosis is essential because the management depends on the underlying disorder. Investigations of raised serum ferritin (ie, >200 μg/L in females, >300 μg/L in males) should include an accurate collection of patient's medical history and some simple laboratory test, as summarized in Figure 3. In some cases, liver MRI or biopsy can be useful to assess liver iron stores. If hereditary hemochromatosis is suspected, a genetic analysis is recommended, including rare mutations in HFE and non-HFE genes (eg, HAMP, HJV, TFR2, BMP6, etc.) when a first-level genetic test is negative. Among the different causes of hyperferritinemia, GD is often overlooked.
Hyperferritinemia (mean elevation 3-4 x upper limit of normal) with normal TSat is a common finding in naïve patients with type 1 GD. 23 Recent case series reported a prevalence of 63-81%. 7,8 While ferritin levels in GD patients can be markedly elevated, they are typically disproportioned with respect to liver iron accumulation, which is usually absent or unremarkable. 23 For this reason, hyperferritinemia per se is not an indication for phlebotomies in GD patients, who may even experience worsening anemia and fatigue. Only sporadic cases reports in literature described significant iron deposits in GD patients, generally associated with cofactors like HFE mutations or chronic hepatitis C virus infection. 23 An elevated TSat or liver iron accumulation at MRI can be of value in these rare cases.
Pathophysiology of hyperferritinemia in GD is still debated, and several explanations have been proposed, including long-lasting lowgrade chronic inflammation, impaired function of macrophages involved in iron recycling, and local dysregulation of the hepcidin-ferroportin axis. 7 Ferritin levels typically decrease during treatment, and have been proposed as a biomarker of disease activity, 17 although not as meaningful as glucosylsphingosine, which directly reflects substrate accumulation. 24 The GD specific treatment with ERT or SRT must be initiated according to international recommendations and consensus, not solely on the basis of mutation analysis and regardless of the ferritin levels.
Our series of 8 GD cases diagnosed at different Italian referral centers for iron disorders highlights the possibility of diagnosing GD starting from hyperferritinemia. In this context, even mild or incomplete classical manifestations of GD (ie, thrombocytopenia, splenomegaly, anemia, bone involvement) should lead to include GD in the differential diagnosis. Of note, our series shows that clinically mild phenotypes are not restricted to N370S homozygotes. In Figure 3 we propose a diagnostic flow-chart that could help physicians to detect GD when facing patients presenting with apparently "unexplained hyperferritinemia".
The combination of a careful clinical history with simple first level investigations may be sufficient for guiding a differential diagnosis, also including rare conditions, like GD. When GD is suspected, subsequent investigations should be tailored according to the strength of clinical suspicion. A DBS testing could be used as a first-line screening strategy, followed by confirmatory diagnostic approaches in the case of pathologic DBS results (Figure 4). Bone marrow examination may still play a role in the diagnostic process, particularly considering that GD per se is a risk factor for the development of hematological malignancies.

| CONCLUSIONS
Given the difficulties in diagnosis of type 1 GD in adults because of disease heterogeneity and lack of awareness, appropriate diagnostic algorithms or flow-charts starting from non-specific findings may help.
Further studies may establish the usefulness of our proposed flowchart in patients presenting with "unexplained hyperferritinemia".