|Home | About | Journals | Submit | Contact Us | Français|
Sickle cell disease (SCD) is characterized by progressive vascular injury and its pathophysiology is strikingly similar to that of atherosclerosis. Statins decrease inflammation and improve endothelial function in cardiovascular disease, but their effect in SCD is not known. In this pilot study, we examined the safety and effect of short-term simvastatin on biomarkers of vascular dysfunction in SCD. We treated 26 SCD patients with simvastatin, 20 or 40 mg/d, for 21 d. Plasma nitric oxide metabolites (NOx), C-reactive protein (CRP), interleukin-6 (IL-6), vascular cell adhesion molecule-1 (VCAM-1), tissue factor (TF) and vascular endothelial growth factor (VEGF) were analyzed and responses to simvastatin were compared between the two treatment groups. Simvastatin increased NOx levels by 23% in the low-dose (P = 0.01) and 106% in the moderate-dose (P = 0.01) groups, and by 52% overall (P = 0.0008). CRP decreased similarly in both dose groups and by 68% overall (P = 0.02). Levels of IL-6 decreased by 50% (P = 0.04) and 71% (P < 0.05) in the low- and moderate-dose groups, respectively. Simvastatin had no effect on VEGF, VCAM1 or TF. Simvastatin was well-tolerated and safe. Our preliminary findings showing a dose-related effect of simvastatin on levels of NOx, CRP and IL-6 suggest a potential therapeutic role for simvastatin in SCD.
Sickle cell disease (SCD) is now recognized as a chronic, progressive vasculopathy that can lead to irreversible organ damage. This chronic organ damage is the most frequent cause of death beyond childhood (Powars et al, 2005). Although red blood cell sickling and rheologic abnormalities are central to the pathophysiology of SCD, the endothelium and its response to injury cells plays a major role in the vaso-occlusive manifestations of this disease. Reperfusion injury resulting from transient occlusive events, oxidative stress and chronic inflammation lead to a state of sustained endothelial activation (Hebbel et al, 2004). Systemic inflammation has been demonstrated experimentally and clinically by increased leukocyte production and elevations in pro-inflammatory cytokines (Belcher et al, 2003, 2000; Blann et al, 2003; Bourantas et al, 1998; Schnog et al, 2004a). Nearly every manifestation of SCD correlates with leukocytosis and, even at baseline, patients exhibit an acute-phase response marked by elevated levels of tumour necrosis factor (TNF), interleukins and high sensitivity C-reactive protein (hsCRP) (Bourantas et al, 1998; Platt, 2000; Singhal et al, 1993). Increased expression of leukocyte adhesion molecules has also been associated with more severe disease and experimental studies support a direct role for leukocytes in microvascular occlusion (Okpala et al, 2002; Turhan et al, 2002).
In addition to inflammation, pathophysiologic responses to endothelial injury in SCD also involve endothelial adhesion and activation of coagulation. Cellular interactions with the endothelium trigger increased expression of tissue factor (TF) and vascular cell adhesion molecule (VCAM-1), causing further adherence of sickle red cells and leukocytes, as well as platelet activation (Kaul et al, 1995; Solovey et al, 1998, 2001). At the same time, production of cytoprotective mediators, such as nitric oxide (NO), are consumed by the oxidative by-products of ischemia/reperfusion injury (Chiang & Frenette, 2005; Kato & Gladwin, 2008; Makis et al, 2000).
Decreased NO bioavailability has been observed at ‘steady state’, and in association with vaso-occlusive pain and acute chest syndrome (ACS) (Gladwin et al, 1999; Morris et al, 2000; Sullivan et al, 2008). Related studies have shown that plasma NO consumption and the uninhibited production of reactive oxygen species exacerbate vascular injury through activation of inflammatory responses (Klings & Farber, 2001; Osarogiagbon et al, 2000). Restoring normal endothelial function with therapeutic agents that target inflammation and NO dysregulation may thus have a favourable impact on the vasculopathy of SCD.
Emerging data indicate that many of the clinical benefits of statins are independent of their cholesterol-lowering effects and are largely conferred through modulation of NO, via inhibition of Ras/Rho proteins and direct upregulation of endothelial nitric oxide synthase (eNOS) (Hernandez-Perera et al, 1998; Kano et al, 1999; Laufs et al, 1998). The hypothesis that statins afford protection from endothelial injury through their direct action on NO production has been confirmed by studies in eNOS deficient mice subjected to stroke and myocardial ischemia (AminHanjani et al, 2001; Asahi et al, 2005; Endres & Laufs, 2004; Endres et al, 1998; Sironi et al, 2003). In subsequent studies, statins have been shown to down-regulate inflammation, adhesion and thrombosis through NO-mediated pathways (Diomede et al, 2001; Egashira et al, 2000; Kimura et al, 1997; Lefer et al, 1999; Niwa et al, 1996; Pruefer et al, 1999; Sparrow et al, 2001). The clinical benefit derived from statins has been demonstrated by improved endothelial function in several disease populations, including diabetes, stroke and metabolic syndrome as well as cardiovascular disease (Cucchiara & Kasner, 2001; Danesh & Kanwar, 2004; Endres et al, 1998; Feske et al, 2009; Hernandez-Perera et al, 1998; Kureishi et al, 2000; Laufs et al, 1998; Rikitake et al, 2005; Tan et al, 2002; van de Ree et al, 2003; Yamada et al, 2000).
The mechanistic potential of statins to improve vascular function may be extended to the pathophysiology of SCD. Data from transgenic mouse models indicate that statins attenuate endothelial activation in SCD. Solovey et al, (2004) found that in sickle mice subjected to hypoxia-reperfusion injury, lovastatin inhibited an expected increase in TF expression. In a more recent study, pretreatment with statins improved survival in pneumococcus-infected sickle mice with direct demonstration of an anti-inflammatory effect. Preliminary data from a limited study in SCD patients with pulmonary hypertension showed improvement in measures of NO-dependent vascular dysfunction after treatment with atorvastatin (Kato et al, 2004).
Although simvastatin has been extensively studied and marketed for use in other patient populations, its clinical safety and potential efficacy has not been studied in SCD. The aim of this pilot study was to examine the effect of treatment with short-term simvastatin (20 and 40 mg/d) on plasma NO and selected biomarkers of endothelial dysfunction (IL-6, hs-CRP, VCAM-1, TF, and VEGF) in subjects with SCD.
Patients at least 13 years of age, with all genotypes of SCD were eligible. Patients were recruited at their baseline status without acute pain or SCD-related complications requiring hospitalization. Patients treated with hydroxyurea (HU) at a stable dose (>3 months) were eligible.
Exclusion criteria consisted of a pretreatment serum total cholesterol <100 mg/dl, triglycerides (TG) below lower normal limits (LNL) (<40 mg/dl), creatine kinase (CK) greater than upper normal limits (UNL) (>232 U/l), serum creatinine >1.5-fold UNL (1.8 mg/dl), alanine transaminase (ALT) >2-fold UNL (>130 U/l). Other exclusion criteria were pregnancy/lactation, red cell transfusion or hospitalization within the 30 d prior to enrollment, current treatment with statins, amiodarone or other drugs with known metabolic interactions with statins (e.g. cytochrome P450 3A4 metabolism), an underlying musculoskeletal disorder, a positive urine toxicology screen or known history of cocaine or amphetamine use. This study was approved by the Institutional Review Board at Children's Hospital & Research Center Oakland (CHRCO). All subjects or their parents provided written informed consent.
The study was a phase I/II safety and dose escalation trial in which participants were treated with one of two doses (20, 40 mg/d) of oral simvastatin daily for 21 d, followed by a taper over 4 d. A dose-escalating approach was used, whereby demonstration of clinical safety in all subjects enrolled in the low dose (20 mg/d) group was required prior to enrollment of subjects in the moderate dose (40 mg/d) group. Serial non-fasting blood samples were collected at baseline, days 7, 14 and 21 during treatment, and days 25 and 39 after treatment, for laboratory evaluation of clinical safety and measurement of plasma biomarker levels.
Lipid profiles, complete blood counts and routine chemistries were conducted using standard laboratory techniques in the clinical laboratory at CHRCO. Plasma samples were assayed for the concentration of NO metabolites, nitrite/nitrate (NOx), using an NO chemiluminescence analyzer (Sievers Model 280 NOA; Sievers Instruments, Boulder, Colorado, USA), as previously described (Braman & Hendrix, 1989). Hs-CRP was measured by a high-sensitivity (0.01 mg/dl) Plasma hs-CRP was determined by a latex particle enhanced immunochemistry assay (Vitros Model 5,1 FS Chemistry Systems, Johnson & Johnson Clinical Diagnostics Inc., Rochester, NY, USA). Plasma IL-6 and VCAM-1 were measured by ultrasensitive ELISA according to the manufacturer's instructions (Quantikine HS Human IL-6 Immunoassay, Parameter Human sVCAM-1 Immunoassay; R&D Systems, Minneapolis, MN, USA). Plasma TF and VEGF were similarly measured with a commercial ELISA kit (R&D Systems).
Descriptive and experimental measures were expressed as mean ± SD. Baseline characteristics were compared between the two treatment groups using unpaired t-tests for continuous variables and Fisher's exact test for discrete variables. Changes in biomarker levels from baseline in response to simvastatin were assessed using matched paired t-tests. Spearman correlations were performed where appropriate. A two-tailed P < 0.05 was considered significant. Statistical analysis was performed with STATA (version 9.0; STATA Corporation, College Station, TX, USA) and GraphPad Prism statistical software (version 5.01; GraphPad, San Diego, CA, USA).
A total of 26 subjects participated in the study, with 14 subjects in the low dose (20 mg/d) group and 12 subjects in the moderate dose (40 mg/d) group.
A greater number of subjects with a HbSS genotype were enrolled in the low dose group (n = 12; 86%) compared to the moderate dose group (n = 5; 42%) (P = 0.04). Of the 17 subjects with HbSS, nine were on stable hydroxyurea (HU) therapy. With the exception of SCD genotype, there were no clinically significant differences between the two dose groups at baseline.
Transient, asymptomatic elevations in CK (2–3X UNL) were observed in two subjects on day 21 of treatment, with a rapid return to baseline within 1 week after discontinuation of simvastatin. Five subjects experienced a mild sickle cell pain episode during the study, but did not require hospitalization. All five subjects were managed in the outpatient setting with oral pain medications. Consistent with previous studies, plasma NOx levels rapidly declined in these five subjects during the vaso-occlusive episode (VOE) (Fig 1) (Morris et al, 2000). Levels of hs-CRP and IL-6 simultaneously increased, returning to baseline after resolution of the VOE. As this study was intended to determine the effects of simvastatin treatment at ‘steady-state’, data points corresponding to these vaso-occlusive pain events were not included in the analysis.
Simvastatin reduced total cholesterol and low-density lipoprotein (LDL) cholesterol levels, indicating general compliance with the study drug. This effect did not appear to be dose-related, as the total cholesterol and LDL cholesterol levels decreased by the same extent in both treatment groups. Serum high-density lipoprotein (HDL) cholesterol and TG levels were unaffected by simvastatin (Table I). Renal, hepatic and hematologic studies, as measured by serum complete metabolic panel and complete blood count (CBC), were also unaffected by simvastatin treatment. Simvastatin had no effect in either treatment group on laboratory markers of hemolysis, including haemoglobin (Hb), reticulocyte count, lactate dehydrogenase (LDH), and total bilirubin.
Changes in biomarker levels for the two treatment groups are shown in Table II. Plasma NOx levels increased by 23% in the low dose group, and by 106% in the moderate dose group (Fig 2). Plasma IL-6 levels decreased by 50% (P = 0.04) and 71% (P < 0.05) in the low and moderate dose groups, respectively. The change in hs-CRP levels did not show the same dose-effect, decreasing by 73% and 60% in the low and moderate dose groups, respectively. After discontinuation of simvastatin on day 21, levels of NOx, IL-6 and hs-CRP returned to baseline within 2 weeks. Analysis of the two treatment groups combined showed a significant impact of simvastatin on the levels of NOx (P = 0.0008), hs-CRP (P = 0.02) and IL-6 (P = 0.01) (Table II). Although VEGF levels decreased overall (P = 0.04), this effect did not reach significance in the analysis of individual treatment groups. Similarly, simvastatin had no effect on plasma levels of sVCAM-1 or TF levels.
Plasma NOx levels varied with respect to HU use in subjects with HbSS. Baseline NOx levels were higher in subjects on HU therapy (40 ± 26 μmol/l; n = 9) compared to those who were not taking HU (15.4 ± 12.4 μmol/l; n = 8) (P = 0.02). However, the observed increase in NOx levels after simvastatin treatment was much greater in subjects who were not taking HU, rising by 98% from baseline in this group (P = 0.002) compared to only 16% in HU-treated subjects (P = 0.13) (Fig 3). Excluding the subjects who were on HU therapy from the overall analysis further increased the magnitude of the simvastatin treatment effect on NOx levels, with NOx increasing by 50% (P = 0.02) in the low dose group and by 120% (P = 0.01) in the moderate dose group. Hydroxyurea therapy had no significant effect on hs-CRP and IL-6 levels at baseline, or after simvastatin treatment.
Levels of hs-CRP correlated with IL-6 (Spearman r = 0.5; P = 0.03) and with NOx (Spearman r = 0.5; P < 0.05). No correlations were found between the increase in NOx level and levels of the other biomarkers measured (hs-CRP, IL-6, VEGF, VCAM, TF). There were no significant correlations between the increase in NOx with simvastatin and hematologic or lipid parameters.
Sickle cell disease is characterized by chronic inflammation and sustained endothelial activation (Hebbel, 2004). With the exception of lipid deposition, the vasculopathy of SCD is remarkably similar to that of atherosclerosis. Markers of endothelial dysfunction, including CRP and IL-6, sVCAM-1, selectins, TF and VEGF are similarly elevated in both diseases (Cleator & Vaughan, 2008; Devaraj et al, 2006b; Fontaine et al, 2003; Grundy, 2002; Hebbel, 2004; Jialal et al, 2006; Li & Fang, 2004).
Numerous experimental studies have shown that statins modulate inflammation, both in vitro and in vivo. Decreased levels of inflammatory mediators have been observed after treatment with statins in clinical trials of atherosclerotic heart disease, diabetes, stroke and metabolic syndrome (Albert et al, 2001; Blake & Ridker, 2000; Kinlay et al, 2008; Koh, 2000; Sacks et al, 1996; Shah & Newby, 2003). However, there have been no studies to date examining the effect of statins on biomarkers of inflammation and endothelial damage in patients with SCD. In this pilot efficacy study, short-term treatment of SCD subjects with simvastatin (20 and 40 mg/d) had a favourable impact on biomarkers of endothelial injury and inflammation, as indicated by a dose-related increase in NOx, and concomitant reductions in hs-CRP and IL-6 levels.
Patients with SCD exhibit elevated levels of circulating CRP and other inflammatory mediators, even at ‘steady state’ (Akinola et al, 1992; Krishnan et al, 2010, Makis et al, 2006; Pathare et al, 2004). We observed similarly elevated levels of hs-CRP and IL-6 in subjects at baseline, prior to treatment with simvastatin. These results indicate a state of sustained low-grade inflammation in SCD during asymptomatic periods. The correlation between baseline hs-CRP and IL-6 found in our study population is not surprising, as IL-6 produced in response to subclinical vaso-occlusion stimulates production of acute phase proteins such as CRP (Yudkin et al, 2000). Moreover, in the five subjects who developed a mild VOE during the study, we observed a sharp decline in NOx and simultaneous increase in hs-CRP and IL-6 levels at the onset of reported symptoms. Decreased NOx levels have previously been reported in association with vaso-occlusive pain events and acute chest syndrome (ACS) (Morris et al, 2000; Reiter et al, 2002; Schnog et al, 2005). These findings also agree with previous studies showing positive correlations between hs-CRP levels and the frequency of hospitalizations for VOE, as well as the severity of SCD (Krishnan et al, 2010, Makis et al, 2006). We cannot rule out the possibility that the increase in NO (via upregulation of neuronal NOS) with simvastatin treatment may have contributed to the VOE observed in this study. With larger numbers and longer duration, a clinical trial is necessary to definitively answer this question. Although it was not possible to isolate the influence of SCD genotype on biomarker levels in this limited study, subjects with an HbSS genotype tended to have lower baseline NOx levels and higher baseline hs-CRP and IL-6 levels. Further studies are needed to address the influence of genotype on these biomarkers of endothelial dysfunction in SCD.
Plasma NOx levels increased with simvastatin in both treatment groups and more than doubled in the moderate dose group, consistent with numerous studies in other vascular disease populations showing statin-associated increases in NO production (Cucchiara & Kasner, 2001; Danesh & Kanwar, 2004; Laufs, 2003; Liao & Laufs, 2004; Romano et al, 2000; Seljeflot et al, 2002). We cannot exclude the possible influence of dietary NO sources on the measured changes in NOx levels, as subjects were not required to fast or restrict their diet prior to collection of blood samples. However, the consistent pattern of NOx changes in response to simvastatin suggests that dietary NO had a nominal effect on our results.
Hydroxyurea has been shown to increase intravascular generation of NO in SCD (Gladwin et al, 2002; Glover et al, 1999) and patients treated with HU demonstrate elevated plasma NOx levels (Cokic et al, 2003; Gladwin et al, 2002; Huang et al, 2002; Sato et al, 1997). In our sub-analysis examining the effect of HU on biomarker responses to simvastatin, the observed increase in NOx levels was restricted almost exclusively to subjects who were not on HU therapy. Despite the small number of subjects included in this analysis, it is conceivable that subjects who were regularly taking HU had achieved a ceiling effect in NOx levels, beyond which simvastatin had little further effect.
Our results showing a dose-related increase in NOx and a simultaneous decrease in IL-6 levels imply that simvastatin may be more effective at higher doses. The lack of a similar dose-effect of simvastatin on hs-CRP levels agrees with findings from a recent meta-analysis of clinical cardiovascular disease trials showing the absence of correlation between statin dose and hs-CRP level (Balk et al, 2003; Genser et al, 2008; Liem et al, 2008; van de Ree et al, 2003).
Nitric oxide deficiency, resulting from oxidant-induced consumption or loss of biological function, drives the inflammatory, pro-adhesive and thrombotic processes that cause endothelial injury. The vasculoprotective effects of statins beyond cholesterol lowering have been shown in several experimental studies to be modulated by NO through direct upregulation of eNOS (Endres et al, 1998; Hernandez-Perera et al, 1998; Kureishi et al, 2000; Laufs et al, 1998; Yamada et al, 2000). In animal models, pretreatment with simvastatin has been shown to upregulate eNOS, decrease adhesion molecule expression and attenuate leukocyte rolling (Gertz et al, 2003; Hernandez-Perera et al, 1998; Laufs et al, 2000, 1998; Pruefer et al, 2002; Santizo et al, 2002). Similar anti-inflammatory effects were recently demonstrated in statin-treated sickle cell mice showing prolonged survival following pneumococcal infection (Rosch et al, 2010).
In a recent study of patients with metabolic syndrome, simvastatin was found to decrease hs-CRP and IL-6 levels through regulation of nuclear transcription factors, NFkB and Akt (Devaraj et al, 2006a), In vitro studies of human monocytes incubated with lovastatin have documented similar changes in IL-6 levels and NFkB activity (Bustos et al, 1998; Grip et al, 2002; Ortego et al, 1999).
Solovey et al, previously demonstrated that hypoxia/reoxygenation-induced endothelial TF expression could be inhibited by lovastatin in sickle mice and more recently documented that endothelial downregulation of TF is modulated by eNOS (Solovey et al, 2010, 2004). Similar results have been reported in other vascular disease populations showing that statins, through NO-mediated upregulation, decrease TF expression and thereby oppose coagulation activation (Undas et al, 2005). In our study, simvastastin only modestly decreased plasma TF levels without reaching statistical significance. The lack of an effect on TF levels in our study may be explained by the involvement of other transcription factors or mechanistic pathways that are independent of statins. It is also possible that the optimal dose required to have an effect on TF had not been reached in this pilot study. Alternatively, as TF is differentially expressed in various tissues, circulating plasma levels may not reflect tissue levels. Solovey et al, (2004) found that TF expression was localized to the pulmonary vasculature in sickle mice. Moreover, treatment with lovastatin had no effect on TF expression in monocytes, and only modestly decreased pulmonary endothelial VCAM expression after hypoxic challenge. Plasma VCAM levels were similarly unaffected by simvastatin in our study. Additional studies, particularly in humans, are needed to examine tissue-specific biomarker levels and the influence of statins on TF, VCAM and other mediators of endothelial dysfunction.
Despite initial concerns over the cholesterol-lowering effects of simvastatin on red cell membrane biology and the potential for increased hemolysis, we found no relationship between the reduction in cholesterol and laboratory measures of hemolysis, such as Hb, reticulocyte count, LDH and total bilirubin levels in our study population. Moreover, the observed increase in NOx levels after simvastatin did not correlate with Hb, reticulocyte count, LDH or total bilirubin, suggesting that the effect of simvastatin on NOx may be independent of hemolytic activity.
Except for transient, asymptomatic increases in CK levels, simvastatin was well-tolerated with minimal toxicity in our study population. Five subjects reported mild SCD-related pain that did not differ from their clinical history in terms of frequency, intensity and duration. Four subjects were being treated with HU for a history of recurrent vaso-occlusive episodes and two subjects reported weekly episodes that were controlled with oral opioids.
Although confirmation in larger studies is needed, the results of this pilot study show overall improvements in the levels of NOx and downsteam inflammatory markers associated with endothelial dysfunction in SCD. In addition, the positive safety profile of statins observed in the general population is substantiated by the tolerability and lack of serious adverse events in our study population of SCD patients. The present pilot study should lead to larger, randomized, placebo-controlled trials to assess both clinical and mechanistic effects of simvastatin treatment on endothelial function and vascular inflammation in patients with SCD.
This work was supported by the FDA Office of Orphan Products Development Grant # 1R01FD-03080-01-A1 (Clin-icalTrials.gov ID NCT00508027) and Grant #UL1 RR024131-01 from the National Center for Research Resources (NCRR).