The role of hydroxyurea in sickle cell disease

Although the molecular basis for the sickling disorders was identified more than 50 years ago (reviewed in Weatherall 2001), progress towards definitive therapy for sickle cell disease (SCD) has been frustratingly slow. Until the mid-1990s, treatment was almost entirely supportive with no clinically useful drugs available to prevent or reverse the polymerization of HbS. However, a number of key observations about the role of HbF (Stevens et al, 1981; de Simone et al, 1982, Platt et al, 1984, 1991; Veith et al, 1985; Noguchi et al, 1993) led, in 1995, to the publication of a crucially important study in which a new therapeutic agent, hydroxyurea, was heralded as a major breakthrough in the management of SCD (Charache et al, 1995; Schechter & Rodgers, 1995). The initial clinical trials of hydroxyurea in SCD were stopped early as a result of outstanding results in the treatment arm and the drug was rapidly licensed in the USA for the treatment of patients with severe SCD. Since then, hydroxyurea has been used successfully in a variety of situations in SCD, including new studies in very young children (Wang et al, 2001). However, doubts remain both about its efficacy and safety in the long term. This review aims to outline current information about the mechanisms of action of hydroxyurea in SCD, its clinical efficacy in sickling disorders, the approaches to monitoring treatment and to update the evidence about its short-term and long-term toxicity. The principal rationale for treatment with hydroxyurea is its ability to induce HbF (Letvin et al, 1984). However, hydroxyurea appears to exert its beneficial effects in SCD via a number of additional mechanisms, including modification of red cell–endothelial cell interactions and the rheological properties of HbS-containing red cells and via its myelosuppressive effects, particularly on neutrophils. The increase in HbF induced by hydroxyurea appears to interfere with HbS polymerization both by preventing effective contact between adjacent HbS molecules and also by forming mixed hybrids with HbS that have greater solubility than HbS polymers. However, the exact mechanism by which hydroxyurea induces HbF is still unclear. Unlike several of the other HbF-inducing agents, such as 5-azacytidine and butyrates, hydroxyurea does not appear to directly modulate globin gene expression by altering methylation or binding to transcriptionally active elements. That hydroxyurea is a ribonucleotide reductase inhibitor and arrests cells in S-phase as a result of impairment of DNA replication is well known. It may therefore increase HbF indirectly by killing rapidly dividing late erythroid cells, causing recruitment of more primitive erythroid precursors which in turn produce high levels of HbF, or by acting directly on the primitive precursors thereby stimulating HbF production (Kolata, 1984). However, even in the early trials, it was clear that induction of HbF was unlikely to explain all the clinical effects of hydroxyurea; treatment was often associated with clinical improvement well before any detectable laboratory rise in HbF and the level of HbF did not always predict response, other variables such as neutrophil counts and mean cell volume (MCV) often being more strongly correlated with reduction in symptoms (Charache et al, 1995). An important experimental observation is that within 2 weeks of commencing hydroxyurea therapy, prior to any rise in HbF, sickle erythrocytes show reduced adhesion to endothelial cells under low shear flow conditions (Bridges et al, 1996). Other investigators have shown that hydroxyurea reduces adhesion molecule expression on sickle erythrocytes, including very late activation antigen-4 and CD36 (Styles et al, 1997). Erythrocytes from hydroxyurea treated patients also show reduced adhesion in vitro to thrombospondin and laminin (Hillery et al, 2000). As transit times through hypoxic postcapillary venules are critical in promotion of sickling and microvascular occlusion, any reduction in sickle–endothelial adhesion would be predicted to have a beneficial effect. Other rheological properties of sickle erythrocytes may also be important, including erythrocyte hydration status and whole cell deformability, both of which can be increased by hydroxyurea (Ballas et al, 1989). An animal model of sickle cell disease, the transgenic sickle (SAD) mouse, has been used to investigate these properties further. Hydroxyurea causes increased cellular potassium levels and reduced erythrocyte density. Subsequent exposure to chronic hypoxia led to reduced cellular dehydration compared with untreated animals (de Franceschi et al, 1999). Human studies have not always shown consistent results with respect to cell hydration, although there appears to be an alteration in the density profile with a reduction in both low-density reticulocytes and high-density sickle cells (Charache et al, 1987; Orringer et al, 1991). Myelosuppressive effects of hydroxyurea may also be important in reduction of symptoms. The fall in neutrophil count appears especially important, and neutrophilia has long been identified as a marker of severity in SCD (Platt et al, 1994). Neutrophils release powerful pro-inflammatory mediators which are important in endothelial damage and cytokine release, both of which trigger sickling (Noguchi et al, 1993). Neutrophils may also contribute to slow transit time via their adhesive properties and an increase in whole blood viscosity. Evidence that hydroxyurea may modulate neutrophil activation comes from a recent study by Benkerrou et al (2002), who showed that hydroxyurea treatment in patients with SCD corrected the dysregulated neutrophil l-selectin expression and normalized the increased hydrogen peroxide production characteristic of neutrophil activation in SCD. Finally, it has been suggested that hydroxyurea may mediate some of its effects via nitric oxide. Studies in rats showed that metabolism of hydroxyurea leads to nitric oxide release (Jiang et al, 1997). In humans, nitrosylhaemoglobin can be detected in blood within 30 min of ingestion of hydroxyurea (Glover et al, 1999) and steady state levels of nitric oxide metabolites can be detected in hydroxyurea-treated subjects (Nahavandi et al, 2000, Gladwin et al, 2002). This leads to the possibility that some of the clinical effects are mediated by nitric oxide-induced vasodilatation or reduced platelet activation. Initial reports of the use of hydroxyurea in SCD emerged in the 1980s (Platt et al, 1984; Veith et al, 1985). These early studies focused on the ability of hydroxyurea to increase HbF and improve laboratory parameters rather than clinical efficacy. In these studies, using small numbers of patients with homozygous SCD, hydroxyurea consistently led to increases in the number of erythrocytes and reticulocytes containing HbF (F cells and F reticulocytes respectively), in the overall percentage of HbF, and in the MCV. Some studies also showed an increase in steady-state haemoglobin with reductions in neutrophils and reticulocytes, improved red cell survival and improved red cell filterability (Rodgers et al, 1993). These encouraging laboratory changes were accompanied by a lack of evidence of serious toxicity (Charache et al, 1992) and led to the Multicenter Study of Hydroxyurea in Sickle Cell Anemia (MSH) study (Charache et al, 1995). This study ran from January 1992 until June 1994 (Charache et al, 1995). The trial was stopped 10 months early because of a significant reduction in events in the hydroxyurea group. The study was a double-blind, randomized controlled trial of 299 patients with severe SCD (defined as at least three painful crises per year) recruited from 21 centres in the USA and Canada. All patients had HbSS. Treatment was started at a dose of 15 mg/kg/d and escalated according to 2 weekly blood counts aiming for the highest dose possible without myelosuppression (maximum tolerated dose; MTD). The results were impressive and included: reduced annual rate of crises (2·5 in the treatment arm vs 4·5 in the placebo arm) P < 0·001 increased median time interval to first crisis (3·0 vs 1·5 months) P = 0·001 increased median time to second crisis (8·8 vs 4·6 months) P < 0·001 reduced incidence of acute chest syndrome (25 events vs 51 events) P < 0·001 reduced transfusion requirements (48 events vs 73 events) P = 0·001 The MSH trial showed that hydroxyurea could ameliorate the clinical course of SCD in some adults without serious short-term adverse effects and that the MTD was not always needed to achieve a clinical response. Although higher HbF levels were associated with lower crisis rates, the association was not statistically significant and neutrophil counts appeared to have the most impact, being associated with lower crisis rates throughout the whole study (Charache, 1997). In addition, the final HbF levels in the hydroxyurea group did not differ markedly from their pretreatment levels (5%vs 9%) (Steinberg et al, 1997a) and overall neither the initial HbF nor the phenotype of the X-linked F-cell production locus (FCP locus) predicted outcome, although the Central African Republic globin haplotype was associated with a reduced response. Use of hydroxyurea was also shown to be cost effective, despite the rigorous monitoring used with 2 weekly clinic visits and blood tests. Average annual costs were $16 810 in the hydroxyurea group vs$22 020 in the placebo group with the vast majority of cost being accounted for by inpatient admissions. This difference failed to reach statistical significance. No account was taken of possible reductions in long-term morbidity with treatment (Moore et al, 2000). Follow-up data after 6–7 years showed that there had been 51 deaths, 21 in the hydroxyurea group and 30 in the placebo group (P = 0·13); although because of the unblinding and subsequent switch of many patients to the treatment arm, these data are difficult to interpret. Analysis according to hydroxyurea usage rather than original treatment allocation suggests that the odds ratio for mortality during hydroxyurea use compared with no hydroxyurea use ranges from 0·59 (P = 0·08) to 0·39 (P = 0·001), depending upon assumptions made about patients with incomplete data. There was no difference in the incidence of stroke in the hydroxyurea-treated patients. None of the patients have developed malignancies and, although the incidence of acute chest syndrome was reduced by hydroxyurea, pulmonary disease was the commonest cause of death (Steinberg et al, 1999). A Cochrane review looking at the use of hydroxyurea in SCD was published in February 2001 (Davies & Olujohungbe 2001). Twenty randomized or quasi-randomized trials comparing hydroxyurea with placebo, standard therapy or other interventions were found. Only two of these, the MSH study and the Belgian paediatric trial (see below), were assessable. Problems with study design or data excluded the other 18. Despite these methodological concerns, none of the published data conflict in any substantive way with the MSH findings. The vast majority report increases in HbF, MCV and steady-state haemoglobin along with clinical improvement in the number of painful crises, hospital attendance rates and acute chest syndrome (el-Hazmi et al, 1992; Kutlar et al, 2000). Studies from the developing world suggested hydroxyurea was no less effective in these populations (Lima et al, 1997; el-Alfy 2000). While these studies have been useful in establishing the clinical effectiveness of hydroxyurea, they leave a number of unanswered questions including: the optimum dose and schedule, the potential for prospective identification of responders and non-responders to avoid exposure to a potentially toxic drug, the impact of treatment on long-term organ damage, morbidity and mortality, and the safety of hydroxyurea in children. A registry of European patients on hydroxyurea has recently been established with the aim of addressing these important questions (Davies & Roberts-Harwood, 1998). Hydroxyurea has many theoretical benefits in the treatment and prevention of stroke. It lowers white cell count, improves steady-state haemoglobin and reduces the risk of acute chest syndrome: all risk factors for stroke (Ohene-Frempong et al, 1998). It also leads to improved rheological properties of red cells and reduction in endothelial adhesion. The only theoretical disadvantage is the increased whole blood viscosity secondary to the rise in steady-state haemoglobin. Despite these promising properties, the MSH study did not show a statistical reduction in stroke although the numbers involved (two in treatment group and three in placebo) were very small (Charache et al, 1995). It is possible that hydroxyurea may have a role to play in patients coming off transfusion regimes, although preliminary data in children suggest that this may be associated with a significant risk of stroke recurrence (Ware et al, 1999). However, as most patients are unwilling or unable to continue transfusions lifelong, many clinicians recommend stopping transfusions once patients reach adulthood (Powars, 2000). Prospective randomized trials of hydroxyurea versus continuing transfusions will be required to define the role of hydroxyurea in this setting. The use of hydroxyurea in children potentially brings the greatest rewards in terms of prevention of end organ damage but also carries greater potential risks (Vichinsky, 1997). The possibility of adverse effects on growth and development, and the risks of secondary malignancy in groups exposed to the drug for long periods led to caution. However, a number of promising phase I/II studies, showing profound and sustained increases in HbF without serious adverse events, led to larger scale trials (Scott et al, 1996; Kinney et al, 1999). The published studies of hydroxyurea therapy in children with SCD are summarized in Table II. Of the published studies, only four have focused on measures of clinical improvement in the severity of SCD (Ferster et al, 1996, 2001; Jayabose et al, 1996; Koren et al, 1999) while the other studies have mainly assessed laboratory parameters of response and The Belgian paediatric trial of hydroxyurea in severe SCD first in (Ferster et al, with the in 2001 (Ferster et al, 2001). The first study children in a design with months hydroxyurea and months The study patients including a small number of who had a median of There was evidence in both studies of a reduction in hospital and in and evidence of a reduced of acute chest increases in HbF and MCV with reduction in white cell and counts were also MTD was not needed for efficacy and no important adverse events were beneficial results with hydroxyurea have been in the other clinical studies in children et al, 1996; Koren et al, 1999). In all the studies to growth and appear to be 1997; et al, even in children under et al, and in (Wang et al, 2001). However, there is no evidence that hydroxyurea stroke or the recurrence of stroke in children. The study three of a stroke within weeks of from transfusions to hydroxyurea, although the dose and of hydroxyurea therapy may have to these results (Ware et al, 1999). The of end organ damage is associated with morbidity and mortality in SCD (Platt et al, 1994). This damage often early with evidence of by months of This the possibility that hydroxyurea be to exert a significant to very young children with SCD. An early report suggested that hydroxyurea use in two young adults with sickle cell disease had led to of & 1996). This was not by a trial in children because may well have by this & 1998). However, in a trial of hydroxyurea in children months 15 months) with and et al a lower of compared with the from data of as assessed by While this may a clinical of hydroxyurea in very young children with SCD, this study also showed that hydroxyurea is not on its to prevent the major of the disease at this of the children in this study had events, three had acute chest syndrome and two had despite hydroxyurea These results the still as to its role in young a randomized controlled trial would The dose of hydroxyurea needed for clinical appears not to be the as the Although myelosuppression is associated with adverse clinical events in the of or it to to the risk of long-term including secondary Hydroxyurea is available as of and can be made for paediatric use but is not available The MSH study started patients at once for blood count, and along with HbF level were taken 2 The dose was increased by mg/kg/d weeks in the of count < 2 haemoglobin < 4·5 reticulocytes or < had the dose was at mg/kg/d less than the toxic dose and this was as the patients a dose of mg/kg/d although some tolerated Some showed toxicity at levels of only studies in children have used of there have been studies looking at In is to at a dose of blood count including reticulocytes, HbF and are prior to commencing and reticulocytes be at 2 weeks and at during the initial of and be at least the is well established on a blood counts The interval between counts be taken in without a recent blood count and not weeks at any are to report any or other symptoms and have a blood count in these monitoring the response to hydroxyurea, the on clinical is more important than any increase in HbF that to clinically to mg/kg/d can have their dose escalated as long as there is no evidence of Although most of response are to (Steinberg et al, there appear to be some who are to hydroxyurea et al, 1997). of has in a small number of (el-Hazmi et al, 1995). of about long-term it is to treatment in any who has had no clinical or response after a trial of months of It is that laboratory for of HbF have been developed to low It is important to that the used is at higher HbF et al, 1996; 1999). The commonest is Although this is there have been reports of several weeks or months & 1994). As as are there has been a of associated with which is and increased especially on the and et al, 2001). be of this potential as it can be Hydroxyurea is associated with of in and some studies have rates of to in SCD patients on hydroxyurea, but the role of hydroxyurea is as of these patients had a prior of et al, 2001). In the MSH of patients had at and new developed with in treatment and placebo groups (Charache et al, 1995). and other have been with hydroxyurea (de et al, 1999) in randomized trials, are no than with placebo (Charache et al, 1995). Hydroxyurea is and small increases in are on be dose reductions are needed in patients with Studies in rats showed an increase in and reduced and et al, 2000). Despite to for both and patients on there have been a number of in published than have been three were was in a with and the other 10 in with no et al, 1999). on hydroxyurea be as to the theoretical risks of and to treatment months before to The Study showed a towards increased second malignancies after years in a small study of 51 patients treated with hydroxyurea in P = et al, 1997). A larger study of patients with before the of showed an risk of of about by the of treatment and a rate of other although this was only greater than the rate of in patients of this group & 1997). A more group be young patients with secondary to a study of patients treated with hydroxyurea for years showed no increase in malignancy et al, 1994). Initial studies of hydroxyurea showed that some adults had in blood but there was no through treatment and no difference from and (Charache et al, are experimental studies of DNA in children treated for between and 30 months that showed an increased rate as by the Although this was greater than with children (P = it was still within the for et al, 2000). There have been four of malignancy in sickle cell patients on hydroxyurea and are to be to the time course and of 2 and are more It is difficult to risk as the in terms of number of sickle cell patients on hydroxyurea is The recruitment to trials in 1992 have 10 years of The risk appears small but is needed & 1995; 1995; de & 2001). Initial studies were to patients with homozygous sickle cell scale studies have of treatment in other sickling conditions especially sickle et al, 1995; et al, et al, 2001). hydroxyurea may be even more effective in this group of disorders than in as HbF production has the of for some of the in disease are more difficult to interpret. An initial study of patients showed an increase in haemoglobin levels and and a in reticulocytes and on treatment but no statistically significant in HbF levels (Steinberg et al, A more recent trial in children with showed a significant rise in HbF and a clinical response with a reduction in hospital rates et al, 2001). It is unlikely that these patients are to hydroxyurea and therefore there no to a trial of reports beneficial effects of hydroxyurea in and disease can also be 1997; et al, 1999). on the evidence hydroxyurea an important role in the management of a number of of SCD. This is summarized in Table the majority of adults with SCD, hydroxyurea can be as the treatment of for preventing crises in patients who have crises especially these to hospital per or to time from or higher monitoring for toxicity is and or about possible long-term may many patients to this there are as no other available with efficacy and the risks associated with chronic transfusion this an for children with painful crises, be made to and the about the to prevent and with crises, as this will often be to the of hospital admissions. with in SCD are often very in more severe Hydroxyurea be for children who have even after these supportive measures have been the mortality and morbidity associated with acute chest syndrome in SCD and the clear evidence of a reduced incidence of acute chest syndrome in patients on hydroxyurea, this be the treatment of to prevent recurrence for all adults and who have more than of acute chest syndrome in the There is no evidence of of transfusion in this and cell be for children who have failed to or are unable to There is no evidence that hydroxyurea is to prevent first or or It is also hydroxyurea early in is to the and possible toxicity of hydroxyurea, its use in these of SCD be within the of controlled clinical

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.