Late response to rosuvastatin and statin-related myalgia due to SLCO1B1, SLCO1B3, ABCB11, and CYP3A5 variants in a patient with Familial Hypercholesterolemia: a case report.

Statins are the most widely used cholesterol-lowering drugs for cardiovascular diseases prevention. However, some patients are refractory to treatment, whereas others experience statin-related adverse events (SRAE). It has been increasingly important to identify pharmacogenetic biomarkers for predicting statin response and adverse events. This case report describes a female patient with familial hypercholesterolemia (FH) who showed late response to rosuvastatin and experienced myalgia on statin treatment. In the first visit (V1), the patient reported myalgia to rosuvastatin 40 mg, which was interrupted for a 6-week wash-out period. In V2, rosuvastatin 20 mg was reintroduced, but her lipid profile did not show any changes after 6 weeks (V3) (LDL-c: 402 vs. 407 mg/dL). Her lipid profile markedly improved after 12 weeks of treatment (V4) (LDL-c: 208 mg/dL), suggesting a late rosuvastatin response. Her adherence to treatment was similar in V1 and V3 and no drug interactions were detected. Pharmacogenetic analysis revealed that the patient carries low-activity variants in SLCO1B1*1B and*5, SLCO1B3 (rs4149117 and rs7311358), and ABCB11 rs2287622, and the non-functional variant in CYP3A5*3. The combined effect of variants in pharmacokinetics-related genes may have contributed to the late response to rosuvastatin and statin-related myalgia. Therefore, they should be considered when assessing a patient's response to statin treatment. To the best of our knowledge, this is the first report of a pharmacogenetic analysis on a case of late rosuvastatin response. 2021 Annals of Translational Medicine. All rights reserved.

Introduction

Familial hypercholesterolemia (FH) is a genetic metabolic disease that leads to increased high low-density lipoprotein (LDL) cholesterol, which is a risk factor for early atherosclerosis and cardiovascular diseases (1). FH is usually treated with high-dose statins, which are inhibitors of 3-hydroxy-3-methyl-glutaryl coenzyme A reductase (HMGR), a key enzyme in cholesterol biosynthesis pathway. Rosuvastatin is one of the most effective statins, probably due its hydrophilicity, that confers selectivity to hepatic cells, higher affinity to HMGR, and lower rates of statin-related adverse events (SRAE) compared to other statins. It is poorly metabolized by CYP2C9 and CYP2C19, while 72% of the non-metabolized molecules are excreted via biliary system. Therefore, rosuvastatin blood levels rely on the activity of membrane transporters, mainly of solute carrier (SLC) and ATP-binding cassette (ABC) families, highly expressed in intestine, liver, and kidney (2).

Pharmacogenetic studies have shown that loss-of-function variants in genes encoding OATPs, such as SLCO1B1, SLCO2B1, and SLCO1B3, and ABCs have been associated with variability in low-density lipoprotein cholesterol (LDL-c) reduction and higher risk of SRAE (3). The importance of considering the combined effect of variants in key genes for pharmacogenetic analyses has been increasingly evident (4). In this case report, we discuss how variants in genes participating in different stages of statin pharmacokinetics pathway possibly affected the time to response to rosuvastatin and the risk of SRAE in a female FH patient. To the best of our knowledge, this is the first report of a pharmacogenetic analysis on a case of late rosuvastatin response. This case is reported in accordance with the CARE reporting checklist (available at http://dx.doi.org/10.21037/atm-20-5540).

Case presentation

A 26-year-old Caucasian female patient with definite diagnosis of FH according to Dutch Lipid Clinic Network MEDPED criteria (5) was invited to participate in an intervention study in June 2019. She was previously included in a FH sequencing study (May 2018), in which a panel of 84 genes involved in lipid homeostasis and drug metabolism was sequenced using exon-targeted gene sequencing (NGS). All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional research committees and with the Helsinki Declaration (as revised in 2013). Written informed consent was obtained from the patient for both studies.

The patient carries the variant LDLR rs28941776 (c.1646G>A, p.Gly549Asp), which has been associated with FH and is classified as pathogenic according to the American College of Medical Genetics (ACMG) guidelines (6).

Her clinical history included high levels of total cholesterol and LDL-c since childhood. In 2008, at the age of 15 years, she had an abnormal lipid profile even under a daily treatment with simvastatin 10 mg and ezetimibe 10 mg. Laboratory analyses showed a total cholesterol of 324 mg/dL, LDL-c 264 mg/dL, high-density lipoprotein cholesterol (HDL-c) 46 mg/dL, and triglycerides 71 mg/dL. In 2014, she was diagnosed with hypothyroidism and treated with levothyroxine 25 µg/day, which was gradually increased to 100 µg/day in 2019. She also had a pregnancy history in January 2017.

Her therapy history included simvastatin, which led to severe myopathy in 2008, with marked increase in serum creatine kinase (CK) to 1,080 U/L (4.7-fold the upper reference value). The cholesterol-lowering therapy was changed to pravastatin 20 mg and ezetimibe 10 mg daily until May 2011, when she reported another episode of myalgia. Pravastatin was withdrawn and atorvastatin 20 mg was introduced, also associated with ezetimibe 10 mg. Three months later, in August 2011, she reported interrupting atorvastatin treatment due to myalgia. Rosuvastatin 10 mg was then introduced, also associated with ezetimibe 10 mg, after which she showed an LDL-c level of 125 mg/dL and never reported myalgia again. However, her lipid profile worsened throughout the years even under rosuvastatin treatment, with her LDL-c reaching 194 mg/dL with rosuvastatin 20 mg.

The patient had no history of liver or kidney impartment, HIV, coronary artery disease (CAD), diabetes, obesity, cardiovascular events, and did not smoke or drink. Her mother and grandmother had a history of FH, but not CAD or cardiovascular events, while her father had hypertension and type 2 diabetes.

In the intervention study, the patient was seen four times (V1 to V4) in 5 months, and clinical history and therapy data were obtained. The protocol consisted of a 6-week rosuvastatin wash-out period, after which rosuvastatin was reintroduced for additional 6 weeks, when treatment response was evaluated. Adherence to treatment was assessed in each timepoint using the translated and validated version of the Brief Medication Questionnaire (BMQ) (7) and blood samples were taken in each visit for laboratory testing.

The lipid profile during the follow-up is shown in . In April 2019 (V1), the patient was taking rosuvastatin 40 mg, ezetimibe 10 mg, and levothyroxine 88 µg daily. She reported experiencing muscle pain after recently increasing rosuvastatin dose from 20 to 40 mg/day. Her lipid profile was altered (total cholesterol 376 mg/dL, LDL-c 263 mg/dL, HDL-c 67 mg/dL, triglycerides 234 mg/dL) without increase in CK levels. She reported being active, running 2 km 2–3 times a week, and had a healthy diet, eating more than five portions of vegetables daily. Her TSH and T4 levels were normal. Rosuvastatin 40 mg was then discontinued for wash-out, ezetimibe was maintained, and levothyroxine dose was increased to 100 µg/day.

Figure 1

Plasma lipid profile and pharmacotherapy of the FH patient throughout the study period. EZT, ezetimibe; LVT, levothyroxine; RSV, rosuvastatin; SRAE, statin-related adverse events.

In June 2019 (V2), after undergoing a 6-week rosuvastatin wash-out period between V1 and V2, her lipid profile worsened (total cholesterol 512 mg/dL, LDL-c 405 mg/dL, HDL-c 65 mg/dL, triglycerides 213 mg/dL). Because the patient reported myalgia in V1 (rosuvastatin 40 mg), the physician prescribed rosuvastatin 20 mg/day for six weeks. Surprisingly, in August 2019 (V3), the lipid profile (total cholesterol 531 mg/dL, LDL-c 407 mg/dL, HDL-c 67 mg/dL, triglycerides 286 mg/dL) did not change compared to V2. The patient reported experiencing no myalgia to rosuvastatin 20 mg. In September 2019 (V4), her lipid profile improved (total cholesterol 299 mg/dL, LDL-c 208 mg/dL, HDL-c 59 mg/dL, triglycerides 158 mg/dL) and she continued not experiencing myalgia to rosuvastatin.

During the follow-up period, serum TSH and T4 levels remained unchanged, suggesting that her hypothyroidism was controlled and did not influence the lipid profile. Moreover, serum CK did not show any abnormality, which indicates no muscle damage due to statin treatment.

The patient also reported being adherent to treatment. In the BMQ adherence questionnaire, she reported forgetting the lipid-lowering medications 2 days in the week before V1 (71.4% adherence) and 1 day in the week before V3 (85.7% adherence).

The genetic profile of the patient is shown in . She carries five missense variants in SLCO1B1, SLCO1B3, and ABCB11. She is also homozygote for the CYP3A5*3 (rs776746) splicing variant. No other missense variants described as impacting rosuvastatin response were found in CYP3A4, CYP2C9, CYP2C19, or other drug transporters, such as ABCG2 (data not shown).

Table 1

Variants in pharmacokinetic-related genes of the FH patient with late response to rosuvastatin

Gene	Variant code	Variant type	Nucleotide change (Amino acid change)	Patient genotype	Allele frequency (1,000 genomes, %)	Functional impact	Effects on rosuvastatin pharmacokinetics	References
SLCO1B1	rs2306283 (SLCO1B1*1B)	Missense	c.388A>G (p.Asn130Asp)	AG	*1B: 54.4	Comparable to *1A	No effect on plasma rosuvastatin levels	Ho et al., 2006; Lee et al., 2013
SLCO1B1	rs4149056 (SLCO1B1*5)	Missense	c.521T>C p.(Val174Ala)	TC	*5: 8.8	Reduced activity	Increased rosuvastatin plasma levels; Reduced hepatic uptake	Kameyama et al., 2005; Lee et al., 2013
SLCO1B1	rs2306283, rs4149056 (SLCO1B1*15)	Missense	c.388A>G, c.521T>C (p.Asn130Asp, p.Val174Ala)	AG, TC	*15: 7.8	Reduced activity	Increased rosuvastatin plasma levels; reduced hepatic uptake	Kameyama et al., 2005; Birmingham et al., 2015
SLCO1B3	rs4149117	Missense	c.334T>G (p.Ser112Ala)	GG	G: 70.2	Reduced activity	Reduced hepatic uptake	Schwarz et al. 2011
SLCO1B3	rs7311358	Missense	c.699G>A (p.Met233Ile)	AA	A: 70.2	Reduced activity	Reduced hepatic uptake	Schwarz et al. 2011
ABCB11	rs2287622	Missense	c.1331T>C (p.Val444Ala)	TC	C: 58.9	Reduced activity	Increased rosuvastatin plasma levels	Soko et al. 2019
CYP3A5	rs776746 (CYP3A5*3)	Splicing	c. 6986A>G	GG	*3: 62.1	No activity	No rosuvastatin metabolism; Reduced LDL-c response	Bailey et al. 2010

FH, familial hypercholesterolemia; LDL-c, low-density lipoprotein cholesterol.

Discussion

In heterozygous FH patients, LDL-c level reductions of 47.1% have been observed after a 6-week treatment with rosuvastatin 20 mg (8). The patient, however, did not experience any changes in LDL-c levels at week 6 (V3) of rosuvastatin 20 mg treatment, with a 48.9% LDL-c reduction only at week 12 (V4) of therapy.

The delayed rosuvastatin response could be explained by modifications in the therapy scheme during the follow-up period. However, the only change was in levothyroxine dose, that was increased from 88 to 100 µg in V1. It is unlikely that the late response is due to an adaptation to the new levothyroxine dose. The patient was already on treatment with levothyroxine 88 µg before V1; moreover, changes in cholesterol due to an adaptation period should be reflected in her lipid profile in V3, not only in V4. Another possible explanation is a lack of adherence from V2 to V3; however, the patient showed a similar treatment adherence in V3 and V1, which should lead to a similar lipid profile between visits. Furthermore, drug interactions between rosuvastatin, levothyroxine, and ezetimibe that could affect treatment response were not detected, excluding this possibility.

Pharmacokinetics-related genes may have contributed to the late response to rosuvastatin (). The patient carries two variants in SLCO1B1, c.388A>G (SLCO1B1*1B) and c.521T>C (SLCO1B1*5), that are important determinants of rosuvastatin response. SLCO1B1*5 is a loss-of-function variant that decreases the hepatic uptake and increases blood levels of statins (9) (). SLCO1B1*1B has shown comparable activity to the functional *1A variant in in vitro functional studies (10). SLCO1B1*1B and *5 variants are in linkage disequilibrium (LD) and form the SLCO1B1*15 haplotype, that also reduced rosuvastatin uptake in functional studies with HEK293 and HeLa cells (11). The decreased liver uptake caused by these SLCO1B1 variants has been associated with increased plasma levels of rosuvastatin in pharmacokinetics studies (9) ().

Figure 2

Proposed mechanism for patient’s late rosuvastatin response and myalgia. 1. The hepatic uptake of rosuvastatin occurs through SLCO1B1 and SLCO1B3 influx transporters, while atorvastatin and simvastatin are internalized through SLCO1B1. The presence of deleterious variants in these transporters (SLCO1B1*15 and SLCO1B3 c.334T>G and c.699G>A) decreases statin uptake, therefore decreasing their concentration inside the hepatocyte and increasing statin plasma levels. 2. The lack of expression of CYP3A5 due to CYP3A5*3 also decreases atorvastatin and simvastatin metabolization, which contributes to increasing their plasma levels. This enzyme does not participate markedly in rosuvastatin metabolism. 3. The resulting higher blood statin levels increased the patient’s muscular exposure to statins, that are internalized through SLCO2B1 transporter into the skeletal muscle cell. The high concentrations in the skeletal muscle cell possibly caused patient’s myalgia. 4. Rosuvastatin’s bile excretion occurs through ABCB11 efflux protein. ABCB11 c.1331T>C variant results in a reduced activity ABCB11, which decreases rosuvastatin efflux; this increases rosuvastatin intrahepatic levels and blood levels. Although the patient had reduced function influx transporters, we suggest that the small portion of rosuvastatin absorbed in the beginning of the treatment accumulated due to the loss of function of the ABCB11 variant. This, together with rosuvastatin active metabolites generated by the normal function CYP2C9, allowed HMGR inhibition and therefore cholesterol lowering in the last visit.

SLCO1B3 is also an important gene that encodes an influx transporter for rosuvastatin. The patient was homozygous for both SLCO1B3 c.334T>G and c.699G>A, which are in strong LD (12). In an in vitro study, HeLa cells transfected with SLCO1B3 c.334G and c.699A haplotype showed a 13% decrease in rosuvastatin uptake, while for other substrates, such as cholecystokinin-8, an even more marked decrease of 57% was observed (13) ().

Although the effect of SLCO1B3 c.334G and c.699A haplotype in rosuvastatin uptake is not sufficient to explain the delayed response, it might be significant when combined with the effect of the decreased function haplotype SLCO1B1*15. While SLCO1B1*5 and SLCO1B1*15 are associated with higher plasma levels of rosuvastatin, previous studies failed to find an association between these variants and LDL-c reduction in response to short- and long-term rosuvastatin treatments (9). Therefore, the simultaneous presence of decreased function SLCO1B1 and SLCO1B3 haplotypes possibly caused a marked reduction of rosuvastatin intrahepatic concentration, resulting in the lack of response observed in V3.

ABCB11 encodes the efflux protein ABCB11, which plays an important role in rosuvastatin bile excretion. In a recent study, ABCB11 c.1331C allele has been associated to increased plasma rosuvastatin levels in healthy subjects (14) (). This variant possibly causes lower rosuvastatin excretion via bile, which in turn would increase intrahepatic rosuvastatin concentrations. Therefore, this mechanism could explain why even in the presence of low function SLC variants, the patient showed a late but evident LDL-c reduction after 12 weeks of rosuvastatin treatment.

The patient also carries the homozygous form of CYP3A5*3, an intronic variant that results in undetectable expression of CYP3A5 (15). The GEOSTAT-1 study reported that dyslipidemic patients carrying CYP3A5*3/*3 had lower LDL-c reduction after three-month rosuvastatin 10 mg treatment compared to carriers of *1/*1 or *1/*3 (). It was suggested that the metabolite produced by CYP3A5 also plays a role in HMGR inhibition, potentiating the response to rosuvastatin, which is why CYP3A5 non-expressors have reduced LDL-c response to rosuvastatin (16). CYP3A5*3 possibly impaired the patient’s response time to rosuvastatin, but in lower extent, as CYP3A5 does not participate markedly in rosuvastatin metabolism.

In addition to the delayed response to rosuvastatin, the patient experienced myalgia associated with rosuvastatin 40 mg/day and other statins, as previously commented. This SRAE may be due to SLCO1B1 variants. SLCO1B1*5 and SLCO1B1*15 have been extensively associated with myopathy to simvastatin. A systematic review and meta-analysis reported that carriers of the C allele of SLCO1B1*5 (c.521T>C) showed a higher risk of myotoxicity (17). Additionally, SLCO1B1*5 has been associated to rosuvastatin myotoxicity in previous studies (18,19). It has been suggested that it causes higher efflux of statins, increasing statin exposure and, therefore, the risk of myalgia (20). Also, a recent case report showed that variants in SLCO1B3 (c.334T>G and c.699G>A) and ABCB11 (c.1331T>C) and the interaction between rosuvastatin and ticagrelor led to rhabdomyolysis in a patient with chronic kidney disease and other chronic conditions (21), but no other reports were found.

CYP3A5*3 may also have contributed to statin myotoxicity, since it has been associated with increased risk to atorvastatin and rosuvastatin-related myalgia in South-Indian dyslipidemic patients (22). However, this variant was not associated to statin intolerance in another study (23). Most studies have evaluated the effect of individual variants in SRAE, and not the interaction between a group of variants in key genes in statin pharmacokinetics pathway. Therefore, we suggest that the combined effect of the low-activity variants in SLCO1B1 and SLCO1B3, the high-activity variant in ABCB11, and the lack of activity of CYP3A5*3 predisposed the patient to low hepatic uptake, metabolization and efflux, respectively. The resulting higher rosuvastatin plasma concentration increased its systemic exposure, which may have caused myalgia ().

Importantly, the patient carries LDLR rs28941776 (c.1646G>A, p.Gly549Asp), a disruptive-missense variant that showed reduced LDL uptake in an in vitro study (24). LDLR variants have been associated with variability in statin response in FH patients (25), but we did not find studies that investigated the association between LDLR variants and time to statin response or myalgia. Nevertheless, this variant could have played a role in patient’s rosuvastatin time to response and it should be considered for further studies.

A limitation of this study is that plasma concentrations of rosuvastatin and its metabolites were not measured. However, the adherence of the patient to the prescribed treatment was ensured using a validated adherence questionnaire and regular follow-up calls.

In summary, the combination of four low-activity variants in SLC genes, a high-activity variant in ABCB11, and a non-functional variant in CYP3A5 may explain the observed late response to rosuvastatin and the statin-related myalgia. With this case report, we have shown the importance of considering a combination of variants in a pharmacogenetic analysis to predict individual responses to statin treatment and prevent adverse drug events. We believe this study contributes to precision medicine in future clinical settings.

Patient perspective

“I have had high cholesterol since I was a child and it has been an issue because of the delayed response to treatments and of many adverse reactions to medications, especially simvastatin. The authors have been very attentive towards me throughout the whole study and discovered possible variants that may delay my response to rosuvastatin and influence the pain that I have felt when using statins. I am very happy for knowing the cause of my problem and I would like to thank the authors for this possible diagnosis. This has improved my perspectives of cholesterol treatment.”

The article’s supplementary files as