- Top of page
- Authorship contributions
- Conflict of interest
The characteristics of Gaucher disease (GD) associated with persistent thrombocytopenia despite imiglucerase enzyme therapy in type 1 GD (GD1) were investigated by retrospective analysis of International Collaborative Gaucher Group (ICGG) Registry data. The study involved 1016 GD1 patients with an intact spleen for whom date of diagnosis, therapy initiation, and platelet counts were known, and who received continuous imiglucerase therapy for 4 to 5 years. These patients were stratified by last platelet count: ≥120 × 109/l (n = 772); ≥100 to <120 × 109/l (n = 94); ≥80 to <100 × 109/l (n = 80); and <80 × 109/l (n = 70; 20 with <60 × 109/l) and characterized by initial and cumulative average imiglucerase dose, body mass index, platelet count, anaemia, hepatomegaly, splenomegaly, and skeletal assessments at baseline and after 4–5 years of therapy. Statistically significant associations were found between persistent thrombocytopenia and baseline platelet count (<80 × 109/l), splenomegaly, and anaemia (all P < 0·0001). After 4–5 years, statistically significant associations were found with splenomegaly (P < 0·0001), anaemia (P < 0·0001), white blood cell count (P = 0·049), hepatomegaly (P = 0·004) and bone pain (P = 0·035). Exponential platelet decay in relation to splenomegaly suggests that platelets increase only when spleen volume decreases substantially.
Gaucher disease (GD) is characterized by deficient acid β-glucosidase (E.C. 3·2·1·45; gluocerbrosidase) activity leading to the accumulation of glucocerebroside in the lysosomes of mononuclear phagocytes. The accumulation of glucocerebroside-engorged cells within organs, most commonly the spleen, liver, skeleton and lungs, results in a complex, multisystemic and progressive disease (Cox & Schofield, 1979). Prominent disease manifestations of type 1 (non-neuronopathic) GD (GD1) include splenomegaly (with associated hypersplenism and thrombocytopenia, anaemia and leucopenia), hepatomegaly, and bone disease including bone pain, bone crises, abnormal bone (re)modelling, osteopenia, osteonecrosis and pathological fractures (Wenstrup et al, 2002; Sims et al, 2008).
Enzyme replacement therapy for GD1 has been available since 1991, firstly as human placenta-derived enzyme (alglucerase, Ceredase®, Genzyme, a Sanofi Company, Cambridge, MA, USA) and subsequently, in 1994, as imiglucerase (Cerezyme®, Genzyme) a human recombinant form of the enzyme. Alglucerase and imiglucerase have been shown to be therapeutically equivalent in a randomized, two-arm clinical trial (Grabowski et al, 1995); and from hereon both enzymes are referred to as ‘imiglucerase’. Treatment of GD1 patients with imiglucerase enzyme replacement therapy ameliorates or resolves the predominant disease manifestations including splenomegaly, hepatomegaly and some aspects of skeletal disease (Weinreb et al, 2002; Charrow et al, 2007; Andersson et al, 2008; Elstein & Zimran, 2009). A second recombinant human enzyme replacement therapy, velaglucerase-alfa (Shire Human Genetic Therapies, Dublin, Ireland) (Elstein et al, 2011) was approved in 2010 and another, taliglucerase-alfa (Protalix, Carmiel, Israel), is in late phase development (Aviezer et al, 2009).
Thrombocytopenia is a prominent abnormality in GD1 and is frequently the first or sole haematological manifestation of the disorder (Hughes et al, 2007). The platelet count has proven to be a sensitive marker of overall treatment responsiveness and many studies evaluating treatments for GD have used platelet count as an efficacy parameter (Weinreb et al, 2002; Schiffmann et al, 2008; Lukina et al, 2010; Elstein et al, 2011; Zimran et al, 2011). Thrombocytopenia frequently normalizes within 1 to 2 years of imiglucerase treatment in patients with an intact spleen and moderate baseline thrombocytopenia (Pastores et al, 2004), although there is variability in response among individuals (Weinreb et al, 2002, 2008; Pastores et al, 2004; Andersson et al, 2008). Very rarely, the platelet response to therapy is slow or not evident (Pastores et al, 2004; Weinreb et al, 2008).It has been suggested that some patients may have a minimal or highly attenuated response to therapy because of continuing severe splenomegaly (Pastores et al, 2004), focal splenic lesions (Stein et al, 2010), and/or bone marrow involvement (Pastores et al, 2004). Apart from serving as a general disease marker in non-splenectomized patients, severe thrombocytopenia may increase the risk of spontaneous, surgical and obstetrical bleeding in GD patients (Lutsky & Tejwani, 2007; Zimran et al, 2009). Platelet counts of <20 × 109/l have been reported (Zimran et al, 2005), incurring a risk of serious bleeding complications.
The underlying cause of persistent thrombocytopenia (defined in this study as platelet count <120 × 109/l after 4–5 years), despite several years of optimal therapy, is poorly understood and has been highlighted by the International Collaborative Gaucher Group (ICGG) Registry Board of Advisors as an area requiring further investigation (Charrow et al, 2000). The objective of the present study was to identify characteristics of GD that are associated with (and may be potentially predictive of) a lack of platelet response following 4–5 years of therapy by retrospective analysis of ICCG Gaucher Registry data. As the focus was on platelet count as a treatment response parameter, bleeding risk was not specifically addressed in this study. The specific aims were: (i) to identify characteristics at the start of therapy associated with persistent thrombocytopenia after 4–5 years of therapy; (ii) to identify characteristics after 4–5 years of therapy associated with persistent thrombocytopenia after 4–5 years of therapy; and (iii) to elucidate a better understanding of persistent thrombocytopenia by additional analysis in patients with severe thrombocytopenia (<60 × 109/l).
- Top of page
- Authorship contributions
- Conflict of interest
The ICCG Gaucher Registry was created in 1991 to record clinical, genetic, biochemical and therapeutic characteristics of GD patients worldwide, irrespective of disease severity, treatment status or treatment choice (Charrow et al, 2000). The aims are to define the clinical spectrum of GD, its natural history, and the effects of treatment. Governance and scientific direction is provided by an independent, international group of physicians in GD with operational support from Genzyme, a Sanofi Company. To date, data on over 5500 patients have been submitted by physicians from 62 countries resulting in approximately 50,000 patient years of follow-up.
Of the 5571 patients in the ICGG Gaucher Registry as of 5 June 2009, 1016 met the study inclusion criteria: a reported date of GD diagnosis; a reported date of initiation of imiglucerase enzyme replacement therapy; no reported treatment interruptions (defined as no treatment for ≥ 2 months) within 4–5 years after the initiation of therapy; platelet count measurement within 4–5 years of therapy; and no splenectomy. The selected interval of 4–5 years for uninterrupted therapy reflects the clinical experience of imiglucerase therapy, which suggests that 79·5% of patients achieve therapeutic goals for improvement in platelet count within this timeframe (Weinreb et al, 2008).
Based on their most recent platelet counts at 4–5 years following the initiation of therapy, patients were stratified into four groups: platelet count ≥120 × 109/l (n = 772; 76%); ≥100 to <120 × 109/l (n = 94; 9%); ≥ 80 to <100 × 109/l (n = 80; 8%); <80 × 109/l [n = 70 (7%), of which 20 (2%) had a platelet count <60 × 109/l].
For the purposes of this study, thrombocytopenia was defined as a platelet count of <120 × 109/l. This definition is based on proposed treatment goals in GD, which aim to achieve platelet counts of >120 × 109/l for patients with moderate baseline thrombocytopenia (defined as >60–120 × 109/l) and a twofold or greater increase in platelets for patients with more severe baseline thrombocytopenia (<60 × 109/l) within 2 years of treatment initiation (Pastores et al, 2004; Weinreb et al, 2008). The clinical consequence of varying degrees of thrombocytopenia in these patients was not assessed.
At baseline (i.e. initiation of therapy), patients in each group were characterized by: age and year of diagnosis; age and year of initiation of therapy; Gaucher diagnosis before/after 1991 (availability of enzyme replacement therapy); genotype; gender; ethnicity; geographic region; body mass index (BMI); platelet count; anaemia (corrected for age and gender as outlined previously (Pastores et al, 2004); hepatomegaly classified as severely enlarged (>2·5× normal); moderately enlarged (>1·25× to <2·5× normal); mildly enlarged (<1·25× normal); splenomegaly classified as severely enlarged (>15× normal); moderately enlarged (>5 to <15× normal); mildly enlarged (<5× normal); and skeletal disease (bone pain, bone crisis, osteopenia, avascular necrosis, lytic lesions, Erlenmeyer flask deformity, fractures, and bone mineral density (BMD) by dual-energy X-ray absorptiometry scan of the spine and femora). Patients were also characterized by the dose of imiglucerase therapy at initiation of treatment, and whether or not they had an interval between diagnosis and treatment of ≥2 years or <2 years (Mistry et al, 2009). Of the patients included in this analysis, 594 (58%) initiated therapy with alglucerase. Most of these patients (n = 555) switched to imiglucerase treatment between 1995 and 2000.
At 4–5 years after the initiation of therapy, patients in each cohort were characterized by: anaemia; hepatomegaly; splenomegaly; skeletal assessments; and average cumulative dose of enzyme replacement therapy. Associated biomarkers (chitotriosidase, angiotensin converting enzyme, tartrate-resistant acid phosphatase) were not analysed because the serial data were too limited.
Data are shown according to platelet count category with percentages for categorical variables. Data for continuous measures are presented as mean and/or median and 75th and 25th percentiles. Possible associations with persistent thrombocytopenia were tested using a multivariate proportional odds regression, which adjusted for age at diagnosis and therapy initiation, genotype, gender, year of diagnosis, and year of therapy initiation. P-values indicate statistical difference in frequencies across the platelet count categories. A P-value of ≤0·05 was considered statistically significant.
An analysis of the relationship between spleen volume and platelet count was performed across all patients in the ICGG Gaucher Registry with reported spleen volumes in the study (n = 660; 2299 observations). Statistical analysis of the relationship between spleen volume and platelet count was conducted by scatter plots, with a smoothing curve produced by the Loess method.
As the imaging modality used to determine spleen volume could vary between centres [ultrasound, computed tomography (CT), or magnetic resonance imaging (MRI)], the relationship between spleen volume and platelet count was also investigated in patients from the Amsterdam Medical Centre (AMC) in the Netherlands (n = 50; 538 observations), for whom spleen volume had been prospectively determined only by CT or MRI. Data from this group of patients, irrespective of treatment status, were analysed in a separate scatter plot. A curve estimation function (ibm spss statistical software, NY, USA) was used to test several mathematical models of the relationship between spleen volume and platelet count.
- Top of page
- Authorship contributions
- Conflict of interest
Of the 5571 patients entered in the ICGG Gaucher Registry by mid-2009, 1016 met the criteria of being non-splenectomized, having platelet counts at baseline and at latest follow-up and continuous enzyme replacement therapy throughout a defined follow-up period of 4–5 years. Of these, 24% had varying degrees of persistent thrombocytopenia (<120 × 109/l).
Table 1 shows the number of patients in each platelet category after 4–5 years of treatment by baseline characteristic and mean number of years between diagnosis and first imiglucerase infusion within each group. Of the 1016 patients, 660 had documented spleen volumes before and during the 4- to 5-year treatment period: 500 (1653 observations) in the ≥120 × 109/l platelet category; 66 (276 observations) in the ≥100 to <120 × 109/l platelet category; 52 (195 observations) in the ≥80 to <100 × 109/l platelet category; and 30 (131 observations) in the <80 × 109/l platelet category, which included 12 (44 observations) in patients with platelets of <60 × 109/l.
Table 1. Patient baseline characteristics versus platelet count category after 4–5 years of enzyme replacement therapy
|Platelet count (109/l)||Platelet count after 4–5 years of enzyme replacement therapy|
|≥120||≥100 to <120||≥80 to <100||<80||Total|
|Male||380 (49·2)||56 (59·6)||43 (53·8)||33 (47·1)||512 (50·4)|
|Female||392 (50·8)||38 (40·4)||37 (46·3)||37 (52·9)||504 (49·6)|
|N370SN370S||165 (21·4)||23 (24·5)||23 (28·8)||20 (28·6)||231 (22·7)|
|N370S/other||401 (51·9)||46 (48·9)||37 (46·3)||34 (48·6)||518 (51·0)|
|Other/other||85 (11·0)||8 (8·5)||4 (5·0)||3 (4·3)||100 (9·8)|
|Not reported||121 (15·7)||17 (18·1)||16 (20·0)||13 (18·6)||167 (16·4)|
|Age of diagnosis (years) (%)|
|<20||486 (63·0)||46 (48·9)||36 (45·0)||3042·9)||598 (58·9)|
|20–<40||170 (22·0)||24 (25·5)||29 (36·3)||21 (30·0)||244 (24·0)|
|40–<60||98 (12·7)||20 (21·3)||11 (13·8)||18 (25·7)||147 (14·5)|
|60+||18 (2·3)||4 (4·3)||4 (5·0)||1 (1·4)||27 (2·7)|
|Diagnosis year (%)|
|Before 1991||215 (27·8)||50 (53·2)||44 (55·0)||45 (64·3)||354 (34·8)|
|1991–1999||367 (47·5)||26 (27·7)||26 (32·5)||13 (18·5)||432 (32·5)|
|2000 or later||190 (24·6)||18 (19·1)||10 (12·55)||12 (17·1)||230 (22·6)|
|Age at start of treatment (years) (%)|
|<20||419 (54·3)||29 (30·9)||21 (26·3)||23 (32·9)||492 (48·4)|
|20–<40||167 (21·6)||28 (29·8)||28 (35·0)||17 (24·3)||240 (23·6)|
|40–<60||144 (18·7)||24 (25·5)||21 (26·3)||24 (34·3)||213 (21·0)|
|60+||42 (5·4)||13 (13·8)||10 (12·5)||6 (8·6)||71 (7·0)|
|Year of 1st infusion (%)|
|Before 1995||145 (18·85)||35 (37·2)||34 (42·5)||33 (47·1)||247 (24·3)|
|1995–1999||311 (40·3)||28 (29·8)||30 (37·5)||19 (27·1)||388 (38·2)|
|2000 or later||316 (40·9)||31 (33·0)||16 (20·0)||18 (25·7)||381 (37·5)|
|Interval between diagnosis and 1st infusion|
|<2 years (%)||409 (53·0)||29 (30·9)||25 (31·3)||25 (35·7)||488 (48·0)|
|>2 years (%)||363 (47·0)||65 (69·1)||55 (68·8)||45 (64·3)||528 (52·0)|
|Years between diagnosis and first infusion (mean ± SD)a||5·36 (8·11)||8·97 (9·32)||10·38 (11·21)||8·90 (9·42)||6·33 (8·77)|
|Cumulative mean dose (4–5 years after initiation)|
|Mean (± SD)||37·2 (15·86)||33·4 (14·67)||332·9 (15·94)||32·2 (14·26)||36·2 (15·79)|
At 4–5 years of therapy, persistent thrombocytopenia was highly associated (P < 0·0001) with several baseline characteristics, including platelet count <80 × 109/l, degree of splenomegaly and anaemia (Fig 1a–c). There was a statistically significant association with the presence of the Erlenmeyer flask deformity at baseline [P = 0·045; n = 319 (72%) of 445 patients]. No associations were found for genotype, BMI, other parameters of bone disease, demographic characteristics, and enzyme dose at treatment initiation [initial enzyme doses were broadly similar across all groups (median 30·0 u/kg per 2 weeks for each platelet category; mean 38·4 ± 15·79 u/kg per 2 weeks across all groups)]. No association was found between an interval of greater than 2 years versus an interval of less than 2 years between diagnosis and initiation of therapy (see Table 1).
Figure 1. Association between baseline platelet count (×109/l) (A), baseline splenomegaly (B) and anaemia (C) with persistent thrombocytopenia after 4–5 years of imiglucerase therapy.
Download figure to PowerPoint
At 4–5 years of therapy, there were statistically significant associations between persistent thrombocytopenia and persistent anaemia (P < 0·0001), degree of splenomegaly (P < 0·0001) and hepatomegaly (P = 0·004). Statistically significant associations were also observed between persistent thrombocytopenia and bone pain (P = 0·035) (but not with other skeletal assessments), low white blood cell count (P = 0·049), and cumulative enzyme dose (dose range 18·6–52·2 u/kg per 2 weeks; doses <25th percentile to 75th percentile; P = 0·043) (Fig 2 a–f).
Figure 2. Association between anaemia (A), splenomegaly (B), hepatomegaly (C), bone pain (D), white blood cell count (WBC) (E) and cumulative imiglucerase dose with persistent thrombocytopenia after 4–5 years of imiglucerase therapy. 25th P, 25th percentile; 75th P, 75th percentile.
Download figure to PowerPoint
Clinical case studies
Participating haematologists received all GD specific data on the 20 ICGG Registry patients with platelet counts <60 × 109/l at 4–5 years following treatment initiation for further analysis in order to gain a better understanding of the basis for persistent thrombocytopenia. These records covered up to 10 years of treatment history. The average interval between diagnosis and treatment in these patients was 11·29 years (standard deviation: 12·16 years) (Table 1). Eleven of 20 patients had baseline spleen data, which showed spleen volume ranging from 1350 to 3345 ml (mean 2514 ml) (approximately 10–25 times normal). Follow-up of spleen and platelet responses in these 20 patients for up to 10 years is shown in Table 2. Of 13 patients with adequate data for analysis, three patients showed no sustained improvement in either spleen volume or platelet counts; three showed both a slow spleen response and slow platelet response; seven showed a slow spleen response without improvement in platelet count, and no patient showed an improved platelet count in the absence of a spleen response. Three patients received increased enzyme doses. The last recorded platelet counts for these patients were: 69 × 109/l (2–3 years after dose increase); 103 × 109/l (approximately 1 year after dose increase); and 116 109/l (unspecified time period after dose increase).
Table 2. Up to 10-year treatment follow-up of spleen and platelet responses in patients from the ICGG Registry with platelets <60 × 109/l following the first 4/5 years of treatment
|No appreciable change in spleen size or platelet counts over 5–10 years of treatment||3||One patient received a dose increase after 5 years of 30–60 u/kg per month which was followed by a slow spleen response and maximum platelet count of 69 × 109/l (2–3 years after dose increase)|
|Slow spleen response: Slow platelet response||3||Dose increase in 1 patient after 7 years (unspecified) with further slow improvement in spleen and platelet count (116 × 109/l)In one patient, the maximum platelet countwas 60 × 109/l, despite spleen reduction to 793 ml (other causes involved?) One patient: platelet count 50 × 109/l after 5 years (baseline 21 × 109/l), spleen 1100 ml after 10 years|
|Slow spleen response: No platelet response for 5–10 years||7||In 1 patient, enzyme dose was increased from 30 to 60 u/kg after 9 years with gradual increase in platelet count to 103 × 109/lIn 1 patient, platelet count began to increase after 6 years after spleen had reduced by 30%In 1 patient, platelet count started to increase after 5 years after spleen had reduced to 800 ml|
|Platelet response; No spleen response||0|| |
|Inadequate data for analysis||7|| |
Relationship between spleen volume and platelet count
Analysis of the relationship between spleen volume and platelet count across all ICGG Gaucher Registry patients with evaluable data (n = 660; 2,299 observations) suggested an exponential relationship between the two parameters (Fig 3). Analysis of spleen volume versus platelet counts in patients from the AMC in the Netherlands whose spleen volumes had been accurately determined (50 patients; 538 associations) demonstrated a similar curve, which was best described by a logarithmic relationship (Fig 4). In this cohort, the effect of a change in spleen volume on platelet count appeared to be minimal above a threshold spleen volume of approximately 1000–1500 ml.
Figure 3. Relationship between spleen volume and platelet count in 660 GD1 patients from the ICGG Gaucher Registry (2299) observations. The scatter plot shows all data points from all patients within the respective observation period (baseline before treatment to last observation).
Download figure to PowerPoint
Figure 4. Relationship between spleen volume (determined by computed tomography or magnetic resonance imaging) and platelet counts in the entire cohort of 50 GD1 patients (538 observations) from the Amsterdam Medical Centre, Amsterdam. The relationship between spleen volume and platelet count was best described by a logarithmic curve-fitting model. Results of curve estimate models were as follows: linear: R2 = 0·440; logarithmic: R2 = 0·561; inverse: R2 = 0·503; quadratic: R2 = 0·520; cubic: R2 = 0·540; compound: R2 = 0·540; power: R2 = 0·566; S: R2 = 0·389; growth: R2 = 0·540; exponential: R2 = 0·540; logistic: R2 = 0·540 and are not shown. The analysis was also repeated without outliers (top-left quadrant) to confirm that the results were not biased by these data points (results not shown).
Download figure to PowerPoint
- Top of page
- Authorship contributions
- Conflict of interest
This study confirmed that, despite therapy, severe persistent thrombocytopenia occurs in a small percentage of GD1 patients (2% of patients with a platelet count of <60 × 109/l after 4–5 years of therapy). Highly statistically significant associations between persistent thrombocytopenia after 4–5 years of therapy and baseline low platelet count, splenomegaly, and anaemia (all P < 0·0001) support a major pathophysiological role of hypersplenism before the initiation of therapy (Pastores et al, 2004). All patients with persistent platelet counts <60 × 109/l had severe splenomegaly before therapy was initiated.
The exponential relationship between platelet response and spleen volume demonstrated here (Figs 3 and 4) suggests that in patients with severe splenomegaly, a substantial reduction in spleen size may be necessary before platelet numbers increase significantly. This is supported by clinical data from patients with <60 × 109 platelets/l after 4–5 years, where platelet counts did not appear to improve in any patient independently of a decrease in spleen size. This finding may help physicians manage patient expectations in the achievement of therapeutic goals for thrombocytopenia (Pastores et al, 2004).
In most cases, imiglucerase therapy is effective in reducing splenomegaly in patients with GD (Weinreb et al, 2002). A decrease in spleen size is generally associated with a correction of low haemoglobin levels and platelet counts (Weinreb et al, 2002). While a dose-response relationship for imiglucerase therapy has been demonstrated for haematological and visceral parameters (Grabowski et al, 2009), there is little evidence from this study to indicate that enzyme dose was the major differentiating factor in severe persistent thrombocytopenia (platelet count <60 × 109/l). While there was a statistically significant difference between cumulative enzyme dose in patients with platelet counts of <80 × 109/l (P = 0·04), enzyme dose was similar for all groups at therapy initiation. Patients with platelet counts <60 × 109/l showed little or no response to treatment for over 9 years, even with increasing doses. Increasing enzyme doses thus may not always be warranted when spleen volumes remain above approximately 1000–1500 ml. Exceptions would include patients at serious risk of bleeding, for example, when platelet counts are below 30 × 109/l, or when patients have recurrent bleeding problems, and in whom an increased enzyme dose may be considered to shorten the duration of severe thrombocytopenia. An alternative approach may be to vary the frequency of infusions, for example, from biweekly to weekly infusions (Hollak et al, 2009).
The underlying cause of thrombocytopenia related to spleen size is unclear from this study, but may not necessarily involve only increased platelet destruction in the spleen. Other contributory factors may include a pooling effect where the enlarged spleen acts as a reservoir to take platelets out of the general circulation (Aster, 1966); and/or prolonged delay before the initiation of therapy leading to irreversible spleen changes, such as fibrosis (Lee, 1982; Hill et al, 1986; Iwanami et al, 1992) or changes in vascularization, which could prevent the therapeutic enzyme from reaching Gaucher cells (Pastores et al, 2004; Stein et al, 2010; Mistry et al, 2007; de Fost et al, 2008). A study has shown that focal lesions in the spleen are associated with suboptimal splenic and platelet responses to imiglucerase therapy (Stein et al, 2010). An enlarged spleen is also thought to act as a ‘sink’ for enzyme, preventing it from reaching other compartments (Pastores et al, 2004). This is consistent with further associations between persistent thrombocytopenia and hepatomegaly (P = 0·004) and bone pain (P = 0·035) after 4–5 years, which suggest more severe manifestations of GD at other locations. The association with low white blood cell count (P = 0·049) after 4–5 years is also consistent with continuing hypersplenism and patterns of cytopenia in more advanced GD1 (Zimran et al, 2005). It should be noted, however, that identified variables with statistically weaker associations may have been influenced by an element of chance, due to the multiple variables examined in the analysis and the paucity of serial data in the respective categories. It is likely that for patients with persistent thrombocytopenia, several of the possible causative factors described above may be involved to differing extents.
While bone marrow infiltration may contribute to thrombocytopenia by compromising the production of platelet progenitors, it was not possible to investigate a potential relationship with persistent thrombocytopenia because bone marrow infiltration is not routinely captured by the ICGG Gaucher Registry. The presence of bone disease might be expected in established or severe GD1 although there is no firm correlation between the severity of haematological and skeletal manifestations (Taddei et al, 2009). At baseline, only a statistically weak association was found with the Erlenmeyer flask deformity, and after 4–5 years, with bone pain. The Erlenmeyer flask deformity is a common bone modelling abnormality found in all but the most mildly affected Gaucher patients and is unlikely to be of significance for persistent thrombocytopenia.
A number of other factors, not measured in this study, may contribute to persistent thrombocytopenia in some GD patients, such as advanced liver involvement, folate or vitamin B12 deficiency, and altered immune and/or inflammatory status. Hepatomegaly is a common manifestation in GD, and in some patients hepatic fibrosis and portal hypertension may occur (Cox et al, 2008) with exacerbation of thrombocytopenia as a result of splenic sequestration of platelets. Vitamin B12 deficiency is a frequent finding in GD (Gielchinsky et al, 2001), which may compromise platelet production in the bone marrow. Comorbidities related to immune dysfunction, such as idiopathic autoimmune thrombocytopenic purpura (ITP), have been noted in GD and may contribute to low platelet numbers (Lester et al, 1984). Pro- and anti-inflammatory factors are released in GD and may play a role in pathophysiology (Cox, 2001). A direct effect on platelet production is unclear, although a correlation between protein activated receptor 1 and thrombocytopenia has been described (Altarescu et al, 2010). Apparent persistent thrombocytopenia, or ‘pseudo’ thrombocytopenia may also arise because of measurement error in platelet counting. A number of different manual, semi-automated and automated platelet counting methods are available with varying levels of measurement precision. Apparently reduced platelet counts may occur through platelet aggregation by EDTA-dependent agglutinins, and/or by the presence of large platelets by some automated methods (Briggs et al, 2007). Practical considerations in this respect involve further investigations for such contributory factors when a patient has a relatively small spleen but fails to improve with respect to platelet counts. If there is an overall failure of response, the physician should check patient compliance, presence or absence of neutralizing antibodies (Zhao et al, 2003; Starzyk et al, 2007), or enzyme dose should be re-evaluated.
Platelet count alone may not be a reliable indicator of bleeding risk in GD patients (Givol et al, 2012). Other factors, such as platelet quality and coagulation deficiencies, may also be involved (Hollak et al, 1997; Deghady et al, 2006). Management, therefore, must be based on clinical assessment of bleeding, and an understanding of each patient's haematological profile.
A limitation of the current study is that insufficient data on bleeding is available from the Registry. Observational studies suggest that the pattern of bleeding in GD is comparable to that observed in immune thrombocytopenia, where patients show individual variation but generally exhibit a higher risk of bleeding during (orthopaedic) surgery and an increased risk at platelet counts below 20 × 109/l (Zimran et al, 2005, 2009; Lutsky & Tejwani, 2007). International guidelines for ITP are thus frequently used in the management of GD patients (Provan et al, 2010). In GD there also remains a limited role for splenectomy. Before therapy was widely available, splenectomy was carried out frequently to reduce severe cytopenia as a result of functional hypersplenism and/or to reduce the effects of an enlarged spleen on abdominal viscera (Cox et al, 2008). Persistent severe and life-threatening thrombocytopenia may be one of the few circumstances where splenectomy may still be justified in GD (Cox et al, 2008). Imiglucerase treatment following splenectomy usually normalizes the platelet count within 1 year (Weinreb et al, 2008), although splenectomy is itself a risk factor for further complications in GD, such as exacerbation of bone disease as well as increased risk of infections (Cox et al, 2008; DeMayo et al, 2008).
In conclusion, the strong association between splenomegaly, anaemia and thrombocytopenia at the initiation of therapy and persistent thrombocytopenia after 4–5 years of treatment suggests that extensive spleen involvement before therapy initiation may be predictive of thrombocytopenia that is refractory to treatment in a minority of cases. The exponential relationship between platelet count and spleen volume suggests that platelet counts increase only when spleen volume has decreased substantially. A critical threshold of 1000–1500 ml in spleen volume for a platelet count >100 × 109/l arises from this study. Further research should focus on spleen responsiveness to therapy by assessing its degree of fibrosis by non-invasive means.
Conflict of interest
- Top of page
- Authorship contributions
- Conflict of interest
Logistical support for this manuscript was provided by Genzyme Corporation. The database for the International Collaborative Gaucher Group (ICGG) Gaucher Registry is supported by Genzyme.
CH has received reimbursement of expenses from Genzyme, Shire, Actelion, and Protalix/Pfizer. All consulting and speaker fees are either donated to the Gaucher Stichting, which supports research in lysosomal storage diseases or transferred to the AMR (AMC Research BV).
NB has been reimbursed travel and accommodation expenses, and donates all consulting and, speaker fees from Genzyme, Shire, Actelion, Protalix/Pfizer to the APRIMI (Association for the Promotion of Research in Infectiology and Internal Medicine), which supports research in lysosomal storage diseases. The APRIMI has received payment for clinical trials from Genzyme and Actelion. The Referral Centre for Lysosomal Diseases has received grants and funds from Genzyme, Shire, Actelion, Pfizer.
AC is an employee of Genzyme.
SvD receives honoraria from Genzyme, Actelion and Shire, serves on advisory boards for Genzyme and Shire, and receives educational and research grants from Genzyme.
PD has received travel grants to attend conferences, speaker honoraria, and research grants from Genzyme and Shire.
JG has been reimbursed travel and accommodation expenses to attend meetings, from Genzyme, Shire, Actelion, and Protalix/Pfizer.
BR receives an educational grant from Genzyme.
LvD has received travel grants from Genzyme and Protalix on two occasions.
ATS has received speaker fees from Genzyme and is an Advisory Board member for Fabry, Gaucher and MPS I Registries.
NW receives research grant support related to Gaucher disease from Genzyme and from Shire, honoraria from Genzyme for participation in ICGG Scientific Advisory Board meetings and from Shire and Pfizer/Protalix for participating in medical advisory boards. NW is a member of the speaker's bureau for Genzyme, Shire and Actelion. He receives no royalties, has no patents and does not own stock in any corporate entity related to Gaucher disease or any other lysosomal storage disease.
AZ receives expenses related to the ICGG meetings and his clinic receives funding for participation in the ICGG program. AZ serves as a consultant for Shire HGT and Protalix, receives honoraria from Actelion, Genzyme and Shire, serves on an advisory board for Protalix and receives stock options from Protalix.
MDC has received grants for travel and research and speaker fees from Genzyme and from Novartis.