|Subject:||Genotype Testing for Genetic Polymorphisms to Determine Drug-Metabolizer Status|
|Document #:||GENE.00010||Current Effective Date:||10/14/2014|
|Status:||Revised||Last Review Date:||08/14/2014|
Genotype testing for polymorphisms can identify variants of specific genes associated with abnormal and normal drug metabolism. This document addresses the use of such testing, based on the theory that individuals with certain gene variants may potentially be able to receive higher or lower doses of some drugs, or should avoid some drugs altogether, to improve the likelihood of achieving clinical goals as well as lessening the risk of adverse drug effects.
Note: For additional information regarding pharmacogenomics, please see:
Genotype testing for genetic polymorphisms of Human Leukocyte Antigen B*1502 (HLA-B*1502) to determine the drug-metabolizer status of individuals for whom the use of carbamazepine is being proposed is considered medically necessary when the criteria below have been met:
Genotype testing for identification of the CYP2C19 variant of Cytochrome P450 to determine the drug-metabolizer status of individuals: a) who are currently undergoing treatment with clopidogrel and have not been tested, or b) for whom the use of clopidogrel is being proposed, is considered medically necessary.
Genotype testing for Human Leukocyte Antigen B (HLA-B*5701) is considered medically necessary before beginning treatment with abacavir (Ziagen®) for persons infected with HIV-1.
Genotype testing in hepatitis C virus (HCV) genotype 1a for the presence of the NS3 Q80K polymorphism is considered medically necessary before beginning treatment with Olysio™ (simeprevir) plus peginterferon and ribavirin.
Investigational and Not Medically Necessary:
Genotype testing for genetic polymorphisms to determine drug-metabolizer status is considered investigational and not medically necessary in all other circumstances, including but not limited to:
The use of testing panels for genetic polymorphisms to determine drug-metabolizer status is considered investigational and not medically necessary unless all components of the panel have been determined to be medically necessary based on the criteria above. However, individual components of a panel may be considered medically necessary when criteria above are met. Examples of such panels include but are not limited to the following:
Current evidence regarding the use of genotyping tests for the determination of drug metabolizer status indicates that while available testing methods may accurately identify genetic variations in an individual, there is insufficient data to demonstrate that such testing, and the clinical decisions made based on the testing, results in a significant impact on health outcomes. Specifically, clinical trials have not yet adequately demonstrated that such testing results in either enhanced clinical effectiveness, or in decreased short-term or long-term serious adverse events.
Critical elements of assessing the effectiveness of such genetic tests include: 1) analytic (diagnostic) validity, 2) clinical validity, and 3) clinical utility. Analytic validity measures the technical performance of the test, in terms of accurately identifying the genetic markers to be measured. Clinical validity measures the strength of association between genetic test results and clinical parameters such as dose, therapeutic efficacy, or adverse events. Clinical utility, the ultimate goal of genetic testing, measures the ability of the test to improve clinical outcomes, such as whether prescribing or dosing based on information from genetic testing improves therapeutic efficacy or adverse event rate as compared with treatment without genetic testing.
Therefore, when considering whether or not a test to determine drug metabolizer status is appropriate in the treatment of individuals prescribed certain medications, specific issues need to be evaluated, including:
There has been investigation into the role of HLAB*1502 mutations in the occurrence of toxic epidermal necrolysis (TEN) and Stevens-Johnson syndrome in ethnic Han Chinese individuals receiving treatment with the anticonvulsant drug carbamazepine. A molecular study by Hung et al. (2006) identified this genetic variation as a contributor to this reaction. Based on data reviewed by an expert panel, the FDA decided to place a black-box warning on the label of carbamazepine as follows:
Serious and sometimes fatal dermatologic reactions, including toxic epidermal necrolysis (TEN) and Stevens-Johnson syndrome (SJS), have been reported during treatment with carbamazepine. These reactions are estimated to occur in 1 to 6 per 10,000 new users in countries with mainly Caucasian populations, but the risk in some Asian countries is estimated to be about 10 times higher. Studies in patients of Chinese ancestry have found a strong association between the risk of developing SJS/TEN and the presence of HLA-B*1502, an inherited allelic variant of the HLA-B gene. HLA-B*1502 is found almost exclusively in patients with ancestry across broad areas of Asia. Individuals with ancestry in genetically at risk populations should be screened for the presence of HLA-B*1502 prior to initiating treatment with carbamazepine. Patients testing positive for the allele should not be treated with tegretol unless the benefit clearly outweighs the risk.
Chen and colleagues conducted a study of 4877 carbamazepine-naive subjects who were genotyped for the HLA-B*1502 allele (2001). B*1502 allele-positive subjects were given an alternative medication while negative subjects were treated with carbamazepine. The authors then compared the incidence of SJS and TEN in the study population to historical controls. Results demonstrated that a mild, transient rash developed in 4.3% of B*1502 positive subjects; more widespread rash developed in 0.1% of subjects, who were hospitalized. SJS–TEN did not develop in any of the HLA-B*1502–negative subjects receiving carbamazepine. In contrast, the estimated historical incidence of carbamazepine-induced SJS–TEN (0.23%) would translate into approximately 10 cases among study subjects (p<0.001).
Other genetic mutations have also been investigated as having clinical impact on the outcomes of individuals who may undergo treatment with carbamazepine. McCormack and others described a study investigating the association of the HLA-A*3101 allele and the incidence of carbamazepine-related complications (2011). This study included 65 subjects who had experienced carbamazepine-related complications and 3987 control subjects. An independent genome-wide association study demonstrated a significant association between subjects with the HLA-A*3101 allele and the incidence of carbamazepine-induced hypersensitivity reactions among subjects of Northern European ancestry. Further study is warranted to understand the impact of genetic testing on the rate of occurrence of complications in subjects carrying the HLA-A*3101 allele.
Recent focus has been placed on the impact of drug metabolizer status testing for individuals prescribed clopidogrel. Several published non-randomized, controlled studies addressed the use of testing for genetic variants in CYP4502C19, ABCB1, CYP2A5, and P2RY12 (Collet, 2009; Mega, 2009; Simon, 2009). These studies found that mutations in these genes, especially CYP2C19 variants, have significant effects on cardiovascular health outcomes. Mega and colleagues conducted a study addressing the impact of CYP-450 gene variants on clinical response to clopidogrel treatment (2009). This study included 162 healthy subjects and 1477 subjects with acute coronary disease being treated with clopidogrel. Carriers of at least one CYP2C19 allele had a 32.4% reduction in the active metabolite of clopidogrel, a 9% decrease in maximal platelet aggregation response, and 300% increase in the risk of stent thrombosis, and relative increase of 53% in the composite primary efficacy outcome of the risk of death from cardiovascular causes, myocardial infarction, or stroke, as compared with non-carriers.
A study by Simon and others enrolled 2208 subjects with acute MI who were receiving clopidogrel therapy (2009). The authors reported a significantly increased risk of adverse cardiovascular events in individuals with CYP2C19 variants when compared to those with no mutations (21.5% vs. 13.3%). Among the 1535 participants who also underwent percutaneous coronary intervention during hospitalization, the rate of cardiovascular events among individuals with two CYP2C19 loss-of-function alleles was 3.58 times the rate among those with none.
In March 2010, the U.S. FDA announced that it was requiring a black-box warning on the label of clopidogrel that addresses the use of pharmacogenetic testing. The warning has four specific points:
Mega and colleagues published the findings of a large meta-analysis conducted in 2010. This report included 9685 subjects who were treated with clopidogrel in 9 studies. The findings included that subjects with 1 or 2 loss-of-function (LOF) CYP2C19 alleles had a significantly increased risk of composite end-point events (HR 1.55, p=0.01; HR 1.76, p=0.002, respectively). Additionally, these subjects had an increased risk of stent thrombosis when compared to non-carriers of loss-of-function alleles.
A study by Simon and colleagues involved 2210 subjects being treated for acute MI who were genotyped for CYP2C19 polymorphisms (2011). They reported that the presence of 2 CYP2C19 LOF alleles was significantly associated with the risk of in-hospital death and major myocardial events at 1 year for individuals with acute MI (adjusted odds ratio 6.67) and those undergoing percutaneous coronary interventions (PCI) (adjusted odds ratio was 6.87). They also investigated the association of PON1 polymorphism with major myocardial events, but reported that no statistically significant association was found.
Mega and others conducted a randomized double-blind trial that enrolled 333 subjects with cardiovascular disease and genotyped for CYP2C19*2 loss of function (LOF) allele status (2011). Non-carriers of the allele received either 75 mg or 150 mg daily dose of clopidogrel in 1 of 2 blinded 28-day long blocks. Carriers of the CYP2C19*2 allele received 75 mg, 150 mg, 225 mg, or 300 mg doses of clopidogrel in a blinded sequence of four 14-day long blocks. For the 75 mg dosage, both CYP2C19*2 hetero- and homozygotes had significantly higher on-treatment platelet reactivity than non-carriers (p<0.001 for both groups). Higher doses of clopidogrel in CYP2C19*2 heterozygotes significantly reduced the proportion of non-responders to 10% in both 225 mg (8 of 75 subjects, p<0.001) and 300 mg (7 of 73 subjects, p<0.001). In CYP2C19*2 homozygotes, higher doses of clopidogrel did not provide similar benefits, with 80% of this group being non-responders at 75 mg and 60% still being non-responders at the 300 mg dosage. The authors reported that in CYP2C19*2 heterozygotes, a dose of 225 mg provided similar platelet reactivity scores to that found with non-carriers receiving a 75 mg dose. In CYP2C19*2 homozygotes, not even the 300 mg dose provided equivalent platelet reactivity to non-carriers. There were no deaths, cerebrovascular events, or Thrombolysis in Myocardial Infarction (TIMI) major or minor events reported in either group at any dose level. This study provides significant evidence to demonstrate that CYP2C19*2 guided dosing of clopidogrel can provide significant benefits in platelet reactivity measures. Further data would be helpful in determining if this also results in significant health outcomes in terms of decreased cardiovascular disease-related deaths and complications.
The results of two large placebo-controlled randomized, controlled trials (RCTs) were published by Pare et al. (2010). The two studies included a total of 5059 subjects randomized to receive either clopidogrel or placebo and followed for the occurrence of primary and secondary composite outcomes. The authors concluded that "no significant difference in the effect of clopidogrel treatment on clinical outcomes was observed when patients were stratified according to metabolizer status." However, some increase in efficacy was seen in subjects with gain-of-function alleles in terms of reduced ischemic events.
In 2011, three meta-analyses were published looking at the health-related outcomes of CYP2C19 genotype testing for individuals receiving clopidogrel. One of these reported that CYP2C19*2 carrier status was significantly associated with increased risk of cardiovascular events. The other two found no such benefit.
The first study, by Jin and colleagues, included a total of 8 prospective cohort studies including 2345 subjects carrying the CYP2C19*2 LOF allele and 5935 wild-type controls. The authors reported that the summary odds ratio demonstrated a statistically significant association in increased cardiac mortality (p=0.007), myocardial infarction (p=0.002), and stent thrombosis (p=0.0001). However, while these findings point to a major role of the CYP2C19 allele in the incidence of major cardiovascular events, the study itself was comparatively small and did not include any RCT data.
In the second study, Bauer et al. looked at the data collected in 15 studies encompassing 28,368 subjects. They found the random effects summary odds ratio for stent thrombosis in carriers of at least one CYP2C19 LOF allele vs. non-carriers was 1.77 (p<0.001). However, the authors note that this finding is subject to significant small study bias and replication diversity. When adjusted for these factors, the significance of this finding was nullified. Furthermore, the odds ratio for major cardiovascular events and stent thrombosis was likewise non-significant. The overall quality of the epidemiological evidence reviewed was graded as low, and the authors' conclusion was that "… at the current state of accumulated information, there is no sufficiently robust and consistent evidence that CYP2C19 represents a strong susceptibility gene modifying the clinical efficacy of clopidogrel."
The third meta-analysis was published by Holmes and others and included 32 studies encompassing 42,016 subjects. Six of the included studies were RCTs. As with the Bauer study previously discussed, this study concluded that "this systematic review and meta-analysis does not demonstrate a clinically important association of genotype with cardiovascular outcomes with the possible exception of stent thrombosis." In many instances, the report stated that when statistically significant analyses were re-run with only studies that included greater than 200 subjects, the original statistically significant findings were nullified. The authors concluded that significant small study bias existed in the body of evidence. This is supported by positive results of the Harbord test for small-study bias (p=0.001). The authors also state that selective outcome reporting and genotype misclassification errors impair the available evidence.
The results of these large well-done meta-analyses call into question earlier assessments regarding the efficacy of CYP2C19 genotyping for individuals receiving clopidogrel. Additional data from large-scale, well done prospective RCTs is needed to further clarify this issue.
The role of genetic polymorphisms in the metabolism and tolerance of various drugs used to treat HIV-1 infection has been of major interest. The most widely studied of these interactions is between the histocompatibility complex allele for HLA-B*5701 and the occurrence of abacavir hypersensitivity reactions (ABC-HSR). Since shortly after the U.S. Food and Drug Administration (FDA) approval of abacavir in 1999, studies began to arise associating the presence HLA-B*5701 with the occurrence of ABC-HSR. The largest study currently available addressing the incidence of ABC-HSR was conducted by Hetherington and colleagues (2001). Using data from approximately 200,000 subjects enrolled in various abacavir clinical trials, the authors conducted a retrospective review of pooled adverse events. Of the 31,096 subjects identified as having hypersensitivity reactions, 1302 (4.3%) were identified as having ABC-HSR. Of these, 176 (9.8%) were considered definitive ABC-HSR cases after failing rechallenge with abacavir. These findings were supported by a later study by the same authors that found the incidence of ABC-HSR to be approximately 4% in a case-control study of 197 subjects from the Glaxo-SmithKlein database (Hetherington, 2002). Mallal and others were the first to publish the results of a trial demonstrating a positive correlation between ABC-HSR and the presence of HLA-B*5701 (2002). This small study of 200 HIV-1 subjects exposed to abacavir identified 18 individuals with definitive ABC-HSR (9%). However, the Mallal study went further and typed all subjects for HLA loci. They reported that HLA-B*5701 occurred in only 4% of abacavir tolerant subjects and 78% in subjects with ABC-HSR (p<0.0001), strongly supporting their hypothesis that the presence of the HLA-B*5701 haplotype was strongly associated with ABC-HSR. They went on to calculate that the presence of HLA-B*5701 had a positive predictive value for ABC-HSR of 100% and a negative predictive value of 97%. These findings were supported by a retrospective study by Rauch and colleagues, who performed genotyping on 131 individuals with suspected ABC-HSR. While these authors did not conduct confirmatory re-challenge to confirm ABC-HSR, they did conduct a blind case review of subjects' medical records, sorting them into likely ABC-HSR (n=27, 21%), unlikely ABC-HSR (n=43, 33%), and uncertain ABC-HSR (n=61, 47%). They found that HLA-B*5701 was present in 31% of likely cases compared to 1% of unlikely cases (p<0.0001). A retrospective case control study investigating the sensitivity and specificity of HLA-B*5701 genotyping in subjects receiving abacavir enrolled 130 white and 69 black subjects for suspected ABC-HSR (Saag, 2008). Positive skin-patch testing identified 42 (33.2%) white and 5 (7.2%) black subjects with confirmed ABC-HSR. All confirmed ABC-HSR subjects were HLA-B*5701 positive (sensitivity = 100%), regardless of race. Among all subjects with clinically suspected ABC-HSR, sensitivity was 44% for white subjects and 14% for black subjects. Specificity for white control subjects was 96% and 99% for black subjects. In the most rigorous study to investigate this issue, Mallal and colleagues conducted a prospective randomized, double-blind study involving 1956 subjects with HIV-1 who were abacavir naïve (2008). Subjects were randomized to undergo prospective HLA-B*5701screening, with positive subjects forgoing abacavir treatment. The control group received routine care with abacavir without HLA-B*5701 screening. Similar to the previous studies, the prevalence of HLA-B*5701 was 5.6%. The authors reported that immunologically confirmed ABC-HSR occurred in 2.7% of subjects in the experimental group vs. none in the control group. The calculated negative predictive value reported to be 100% and positive predictive value was 47.9%. The existing data, discussed above, adequately demonstrate that the HLA-B*5701 genotype is strongly associated with ABC-HSR, and that screening for this genotype significantly decreases the occurrence of ABC-HSR in individuals who have been prescribed abacavir.
In 2011, the DHHS Panel on Antiretroviral Guidelines for Adults and Adolescents published its Guidelines for the Use of Antiretroviral Agents in HIV-1-Infected Adults and Adolescents. The guidelines recommend the following:
Two studies indicate that the presence of hepatitis C virus (HCV) with the NS3 Q80K polymorphism could impart significant resistance to the effects of simeprevir (Lenz, 2013; Palanisamy, 2013). The Q80K polymorphism is a mutation of the NS3/4A protease – a viral enzyme involving in protein processing and replication of HCV. These findings were supported by several RCTs evaluating the efficacy of simeprevir. Forms et al (2014) published the results of a double-blind, placebo controlled trial of 393 subjects infected with confirmed genotype 1 HCV who had failed previous therapy. Enrolled subjects were randomized on a 2:1 basis to receive treatment with either simeprevir in conjunction with peg-interferon α-2a and ribavirin (PR) (n=260) or placebo in conjunction with PR (n=133). In the experimental group, the NS3 Q80K polymorphism was detected in 13.1% (31/257) of subjects. In the control group, 15% (20/133) of subjects were found to have the NS3 Q80K polymorphism. The authors reported that in simeprevir-treated patients with HCV genotype 1a infection, the presence of the Q80K polymorphism at baseline was associated with a lower rate of sustained virologic response at 12 weeks (SVR12) compared with those without this polymorphism at baseline (46.7% [14 of 30] vs. 78.5% [62 of 79], respectively). However; they also noted that the SVR12 rate was high among the 13 simeprevir-treated subjects with baseline Q80K polymorphism who achieved rapid virologic response (RVR, 76.9% vs. 23.5% among subjects without RVR). They concluded that:
Although the Q80K variant itself only has limited effect on simeprevir activity, the resistance barrier for Q80K-carrying variants appears to be lower. This potentially facilitates the emergence of additional mutations, resulting in a higher treatment failure rate in Q80K patients compared with patients without Q80K when treated with simeprevir in combination with PR.
Two multicenter, randomized, double-blind, parallel-group, placebo-controlled, phase 3 clinical trials designed to assess the efficacy, safety, and tolerability of simeprevir in combination with peginterferon alfa plus ribavirin were published in 2014. The first, the QUEST-1 study (Jacobson, 2014), involved the randomization of subjects in a 2:1 ratio to receive treatment with either simeprevir in combination peginterferon alfa 2a plus ribavirin (n=264) or placebo plus peginterferon alfa 2a plus ribavirin (n=130). In the experimental group, 21 subjects discontinued participation in the study (7.9%). For the control group, 10 subjects discontinued participation (7.6%). In the experimental group, the presence of the Q80K polymorphism at baseline was associated with a lower SVR12 vs. subjects without the Q80K polymorphism (74% [28/38] of experimental subjects vs. 92% [59/64] for controls). With regard to RVR, 63% (38/60) of experimental subjects with the Q80K polymorphism achieved RVR vs. 74% (64/86) of control subjects without the Q80K polymorphism. No p-values are provided for this data. The authors stated that in subjects with HCV genotype 1a who had the Q80K polymorphism at baseline, SVR12 was not significantly higher in the simeprevir-treated group than the placebo group. The QUEST-2 (Manns, 2014) study also involved the randomization of subjects in a 2:1 ratio to receive treatment with either simeprevir in combination with peginterferon alfa 2 plus ribavirin (n=257) or placebo plus peginterferon alfa 2 plus ribavirin (n=134). In an interesting methodological twist, where randomization to peginterferon alfa 2b was allowed, subjects in both groups were randomly assigned in a 1:1 ratio to peginterferon alfa 2a plus ribavirin or peginterferon alfa 2b plus ribavirin. The aim was to achieve a maximum of 30% of the overall study population assigned to the regimen containing peginterferon alfa 2b. When such methods were not possible, subjects were treated with peginterferon alfa 2a plus ribavirin. The authors reported that in the experimental group, 75% (18/24) of subjects with the Q80K polymorphism achieved SVR12 vs. 82% (65/79) without the Q80K polymorphism. In the control group, 50% (7/14) of subjects with the Q80K polymorphism achieved SVR12 vs. 43% (17/40) of subjects without the Q80K polymorphism. No p-values are provided for this data. In the experimental group, 23% (24/103) of subjects had sequencing data available and had Q80K polymorphisms at baseline. Of these 24 subjects, 63% (15/24) achieved RVR and 93% (14/15) achieved SVR12. The authors noted that, with the exception of the Q80K subgroup, significantly more subjects had SVR12 in the experimental group vs. the control group. In additional analyses, significant difference between SVR12 in subjects with the Q80K polymorphism in the experimental group and SVR12 in all placebo group subjects with HCV genotype 1a (i.e., with and without Q80K) (p=0.005). It is this finding that the FDA based their recommendation that all patients with HCV genotype 1a be screened for the presence of the Q80K polymorphism before beginning triple therapy with simeprevir plus peginterferon plus ribavirin and to consider an alternative treatment if this polymorphic variant is detected.
Research is currently underway investigating the impact of the Q80K mutation on the efficacy of other drug regimens, including the combination of simeprevir and sofosbuvir (with or without the addition of ribavirin). However, to date no peer-reviewed published studies have been published addressing this issue. As such, the utility of Q80K mutation testing in individuals receiving drug regimens other than simeprevir plus peginterferon and ribavirin is unknown.
Perhaps the most studied area regarding the use of genotype polymorphism testing involves genetic variations in enzymes key to the metabolism and operation of the drug warfarin. A significant amount of evidence has shown that the enzymes cytochrome P450 2C9 (CYP2C9) and vitamin K epoxide reductase enzyme subunit C1 (VKORC1) have the most significant role in warfarin metabolic variability (Higashi, 2002; Kirchheiner, 2005; Osman, 2006; Sconce, 2005.) Some reports attribute approximately 55% of warfarin dose variability to these two variants (Sconce, 2005; Wadelius, 2005).
A report published by McClain and colleagues, conducted for the American College of Medical Genetics (ACMG, 2007), evaluated the use of CYP4502C9 (CYP2C9) and VKORC1 testing of individuals receiving warfarin due to increased risk of thrombotic events. The authors concluded the following:
Caraco and colleagues (2007) describe a randomized, controlled trial evaluating CYP2C9-guided warfarin therapy. This study included 191 participants (n=96 controls, n=95 in the experimental groups) prescribed warfarin therapy. While this study did find significant benefits in some secondary outcomes, such as time to stable dosing, more time spent in therapeutic range, and lower rates of minor bleeding, the small study population did not permit assessment of significant differences in serious bleeding, thrombotic events, major morbidity or mortality. The authors state that further research is warranted.
A study by Anderson and colleagues (2007) indicates some promise for the use of genetic polymorphism testing for individuals receiving warfarin therapy. In their randomized, blinded study of 206 participants, the investigators compared pharmacogenetic-guided therapy vs. standard dosing methodology. While the results indicated that the pharmacogenetic-guided therapy more closely approximated stable doses resulting in significantly smaller and fewer dosing changes, the primary endpoint of reducing out-of-range INRs was not significantly different. However, in a post-hoc subset analysis, the authors reported that in wild-type individuals and those with multiple variant carriers the differences were significant between groups. These authors also indicate that additional research is warranted based upon their findings.
In February 2009, the International Warfarin Pharmacogenetics Consortium published a study that describes the development and modeling of two warfarin therapy algorithms that aid in the prediction of the ideal therapeutic dose. The first algorithm uses both clinical and pharmacogenetic information from a retrospective cohort of 4043 individuals (2009). The second algorithm uses the same population and methodology, excluding the pharmacogenetic data. Using data from a separate retrospective cohort of 1009 individuals, the consortium created a model testing the use of these two algorithms against a standard fixed treatment approach of 5 mg warfarin/day. While this study is of interest, it is only a model and does not provide real-world clinical results. As has been discussed earlier, clinical validity data is needed for the proper evaluation of the clinical role of pharmacogenetic testing methods. This report does not provide data on adverse events such as thromboembolic events or bleeding. The next step is to see how this algorithm functions in a clinical setting with outcomes data reported.
In early 2008, based upon the information provided above, the ACMG published a position statement regarding the use of CYP2C9 and VKORC1 testing, which concluded:
The group determined that the analytical validity of these tests has been met, and there is strong evidence to support association between these genetic variants and therapeutic dose of warfarin. However, there is insufficient evidence, at this time, to recommend for or against routine CYP2C9 and VKORC1 testing in warfarin-naive patients. Prospective clinical trials are needed that provide direct evidence of the benefits, disadvantages, and costs associated with this testing in the setting of initial warfarin dosing… Although the routine use of warfarin genotyping is not endorsed by this work group at this time, in certain situations, CYP2C9 and VKORC1 testing may be useful, and warranted, in determining the cause of unusual therapeutic responses to warfarin therapy.
However, selection criteria or specific algorithms were not described based upon clinical study evidence.
The Agency for Healthcare Research and Quality (AHRQ) published a technology assessment addressing the use of pharmacogenetic testing for warfarin and statin therapy (2008). In this assessment it evaluated the available evidence regarding the clinical impact and outcomes related to the use of pharmacogenetic testing for variants of CYP2C9, VKORC1, and MTHFR. The report concludes:
Overall, studies evaluating associations between the pharmacogenetic test results and the patient's response to therapy for non-cancer and cancer conditions showed considerable variation in study designs, study populations, medication dosages, and the type of medications. This variation warrants caution when interpreting our results. Data on the relationships among pharmacogenetic test results and patient- and disease-related factors and on the patient's response to therapy are limited. We found no data on the benefits, harms, or adverse effects from subsequent therapeutic management after pharmacogenetic testing. Detailed patient-level analyses are needed to adjust estimates for the effects of modifiers, such as age or tumor stage.
In 2011, Burmester and others conducted a double-blind RCT with 203 subjects randomized to receive warfarin therapy guided by either standard algorithm (n=112) or by genotype guided algorithm (n=113). Only 184 subjects (80%) completed the 60 day trial period. The results indicated that the genotype-based algorithm was almost more than twice as accurate at predicting final effective dose compared to the standard model (p<0.0001). However, no difference was noted between groups for time spent in the therapeutic range, time to stable therapeutic dose, time to INR >4, or adverse events. The authors concluded that their data was not able to demonstrate that genotype-based initial warfarin dosing is superior to clinical-based dosing with respect to time in therapeutic range through the first 14 days of therapy. However, the impact of this benefit on the incidence of adverse events remains to be evaluated in a large well-designed study.
A large double-blind RCT conducted by Kimmel (2013) involved 1015 subjects initiating warfarin treatment who were randomized to receive treatment guided by a protocol which included Genotype data for CYP2C9 and VKORC1 variants plus clinical variables (n=514) or a protocol that included clinical variables only (n=501). Subjects had their initial dose and dose adjustments for the first 4-5 days of therapy guided by the assigned protocols. Subsequent adjustments were per standard protocol for the next four weeks. All subjects were followed for 6 months. The results show no significant differences between groups with regard to mean percentage of time within therapeutic range during the first 4 weeks (p=0.91). Overall, there were no significant between-group differences in the mean percentage of time above or below the therapeutic range (INR, <2 or >3). The time to determination of the maintenance dose did not differ significantly between the two groups overall or according to race or total number of genetic variants. The authors did note a significant difference between groups when a pre-specified sub-analysis was conducted for race. For black subjects, the mean time in the therapeutic range in the first 4 weeks was less in the genotype-guided group (p=0.01), and overall, black subjects in the genotype-guided group took longer on average to reach the first therapeutic INR than did those in the clinically-guided group. Black subjects in the genotype-guided group also took longer on average to reach the first therapeutic INR than did those in the control group. No differences between groups were reported with regard to time of INR ≤4, major bleeding, or thromboembolism. The authors concluded that the genotype-guided algorithms performed better at predicting maintenance dose among non-black subjects. However, there was no overall benefit of genotype-guided dosing with respect to percentage of time in the therapeutic INR range. The authors end their report by stating, "Our results emphasize the importance of performing randomized trials for pharmacogenetics, particularly for complex regimens such as warfarin."
An unblinded RCT published in 2013 by the EU-PACT study group (Pirmohamed, 2013) reported contradictory findings to those by Kimmel. The study used point-of-care genotype-guided dosing in 455 subjects with either atrial fibrillation (72.1%) or venous thromboembolism (27.9%) receiving initial treatment with warfarin. Subjects were randomized to receive management with either a genotype-guided algorithm which included data for CYP2C9 and VKORC1 variants plus clinical variables (n=227) or management with an algorithm which included clinical variables only (n=228). Similar to the Kimmel study, subjects had their initial dose and dose adjustments for the first 4-5 days of therapy guided by the assigned protocols. Subsequent adjustments were per local clinical practice standards. All subjects were followed for 3 months. The presented analysis included only those subjects with at least 13 days of INR data (genotyped-guided group, n=211 vs. control group, n=216). The percentage of time with an INR of 2.0 to 3.0 was 67.4% in the genotype-guided group vs. 60.3% in the control group when adjusted for center and indication (p<0.001). In the per-protocol analysis, values in the genotype-guided group (n=166) and control group (n=184) were 68.9% and 62.3% (p=0.001). The difference between the two groups with regard to mean percentage of time in the therapeutic range was significantly different at weeks 1-4 (p<0.001) and 5-8 (p<0.001), but not for weeks 9-12 (p<0.6). Subjects in the genotype-guided group were less likely to have an INR of 4.0 or higher vs. the control group (p<0.03). A total of 173 subjects (82.0%) in the genotype-guided group reached a stable dose by 3 months vs. 52 subjects (70.4%) in the control group (p<0.003). Fewer dose adjustments were required in the genotype-guided group (p=0.02). No significant differences in bleeding or other adverse events were reported. The authors concluded that genotype-based dosing at the initiation of warfarin therapy increased the time in the therapeutic range by 7% and reduced the incidence of excessive anticoagulation, the time required to reach a therapeutic INR, the time required to reach a stable dose, and the number of adjustments in the dose of warfarin.
Acenocoumarol and Phenprocoumon
A second EU-PACT study (Verhoef, 2013) described the results of a single-blind RCT of genotype-guided dosing of acenocoumarol (n=381) and phenprocoumon (n=127). Subjects were randomized to receive management with either a genotype-guided algorithm which included data for CYP2C9 and VKORC1 variants plus clinical variables (n=273) or management with an algorithm which included clinical variables alone (n=275). The percentage of time in the therapeutic range during the first 4 weeks after the initiation of treatment was 52.8% in the genotype-guided group vs. 47.5% in the control group (p=0.02). However, this difference did not persist through the 3 month follow-up period with the percentage of time in the therapeutic INR range being 61.6% for genotype-guided group vs. 60.2% in the control group (p=0.52). No significant differences between the two groups were reported for several secondary outcomes, including number of subjects with INR ≥4, percentage of time with INR ≥4, percentage of time with INR <2, time to reach therapeutic INR and number of subjects with stable dose within 12 weeks. Additionally, no significant differences were reported with regard to the incidence of bleeding or thromboembolic events. The authors concluded that genotype-guided dosing of acenocoumarol or phenprocoumon did not improve the percentage of time in the therapeutic INR range during the 12 weeks after the initiation of therapy.
The Evaluation of Genomic Applications in Practice and Prevention (EGAPP) Working Group published its recommendations regarding the use of UGT1A1 testing for individuals undergoing treatment with irinotecan (2009). This paper concluded:
The evidence is currently insufficient to recommend for or against the routine use of UGT1A1 genotyping in patients with metastatic colorectal cancer who are to be treated with irinotecan, with the intent of modifying the dose as a way to avoid adverse drug reactions (severe neutropenia).
Dihydropyrimidine Dehydrogenase (DPYD)
Deenen and others conducted a retrospective nested case-control study of 45 subjects with colorectal cancer (CRC) who had capecitabine-related toxicity and 100 randomly selected controls (2011). All subjects were selected from a sample of 568 individuals with previously untreated CRC enrolled in the CAIRO2 trial and were tested for DPYD genetic variants. From this data genotype frequencies of polymorphisms were calculated. The authors reported that 4 variant alleles (IVS14+1G>A, 1236G>A, 2846G>A, and 2194G>A) were significantly associated with severe diarrhea when carriers were treated with capecitabine-based chemotherapy. Furthermore, heterozygous carriers of IVS14+1G>A were significantly at risk for developing grade 3to 4 toxicity. No association with overall survival was noted for any specific allele. While this study did identify a role of several alleles in capecitabine-related toxicity, no data regarding outcomes benefit of screening for DYPD genotypes was provided. Further investigation is warranted.
The use of CYP2D6 genotyping testing to determine drug metabolizer status and predict breast cancer-related outcomes in individuals with breast cancer treated with tamoxifen has been a topic of significant debate for the past few years. Several large-scale RCTs have been published addressing this issue. Abraham and colleagues reported the results of their study, which used data from the Studies of Epidemiology and Risk Factors in Cancer Heredity (SEARCH) study (2010). This study included 6640 subjects with invasive breast cancer, with 3155 subjects receiving tamoxifen therapy and genotyped for CYP2D6. Along with genetic data, survival data was used to calculate breast cancer specific survival (BCSS) in this population. The authors concluded that there was weak, if any, effect of CYP2D6 on BCSS in tamoxifen-treated subjects. These findings were corroborated in two separate large scale double-blind RCTs published in 2012. The first was a study using subjects enrolled in the Arimidex, Tamoxifen, Alone or in Combination (ATAC) study, which involved 1203 women genotyped for CPY2D6 and 1,209 genotyped for UGT2B7 (Rae, 2012). No statistically significant associations were observed between CYP2D6 and disease recurrence. Additionally, a near-null association was noted between UGT2B7and recurrence in tamoxifen treated subjects. The second study, by Regan et al. involves subjects enrolled in the Breast International Group (BIG) 1-98 study, which involved 1243 subjects with breast cancer treated with tamoxifen (2012). As with the previously mentioned studies, the BIG study authors found no association between CYP2D6 metabolism phenotype and breast cancer-free interval.
The results of these trials have been somewhat controversial, with several editorials pointing out significant methodological flaws. Kelly and Pritchard commented that the power of these studies was insufficient to show a positive association between CYP2D6 and outcomes in subjects taking tamoxifen (2012). Pharoah and others pointed out that both the Rae and Regan studies were not properly randomized to control for the exposure of interest (2012). Both these studies were randomized for treatment regimen, not CYP2D6 genotype. They continue, criticizing both studies for the use of tumor samples to determine genotype, and the Regan study in particular for failure to report consistency of genotype quality and Hardy-Weinberg equilibrium (HWE) data. The Rae article did not provide data on power calculation, and Pharoah indicates that the study was probably underpowered. Nakamura and colleagues also point out the inadequacies in the Regan study in relation to HWE issues, and insufficient data provided with regard to genotype data quality (2012). Finally, as Pharoah pointed out, the use of tumor samples to determine genotype is flawed. They comment that CYP2D6 is frequently deleted in some common cancers, leading to misclassification of the subject's actual phenotype, as the unaffected cells in their body may contain a different genotype than their cancer cells.
The results of these studies indicate that the use of CYP2D6 genotyping does not provide data that significantly affects breast cancer-related health outcomes. However, as the editorials accompanying these studies indicate, there are many flaws in these trials that leave important questions unanswered. Further investigation is warranted in fully assessing the use of genotyping in this population of individuals.
Testing for genetic polymorphisms has also been proposed for a wide array of other drugs, involving many different conditions and enzymes. At this time the available literature addressing such testing is limited and insufficient to allow any assessment of clinical utility in the treatment of individuals. The outcomes that require further research attention include major adverse events, utilization of health resources, and time to clinically significant changes in condition using appropriate and validated measures.
While the potential of pharmacogenomics is intriguing for many clinical applications, it is not yet clear which are most likely to yield clinical benefit in the near future. As this field evolves and matures, and if pre-prescription testing can be shown to be of clinical utility for specific drugs and individuals, it will be imperative to establish evidence-based guidelines for health care professionals delineating the most effective courses of action based on such genotype testing results.
Several commercial laboratories market multi-test panels to test for genetic polymorphisms related to drug metabolizer status. While the use of some individual tests included in these test panels may be reasonable under specific circumstances, the use of all the tests within a panel is rarely justified unless there is clinical evidence that an individual has used or been exposed to multiple substances, and knowledge of such exposure provides information that leads to meaningful impact on treatment. At this time there are no peer-reviewed published studies addressing the use of such test panels.
Drug efficacy and toxicity vary substantially between individuals. Because drugs and doses are typically adjusted to meet individual requirements as needed by using trial and error, clinical consequences may include a prolonged time to optimal therapy and serious adverse events. It has been found that inherited DNA sequence variation (polymorphisms) in genes for drug-metabolizing enzymes may have a significant effect on the efficacy or toxicity of a drug. This field of research is referred to as pharmacogenomics.
It has been proposed that genotype testing for certain genes to detect polymorphisms will allow physicians to predict side effects to drugs and to tailor a drug regimen based on an individual's genetic make-up. It may be that genotype testing will improve the choice of drug, or the dose of the drug, when the drug in question has a narrow therapeutic dose range, when the consequences of treatment failure are severe, and/or when serious adverse reactions are more likely in individuals with certain polymorphisms.
One of the drugs where this approach has been most extensively investigated is for the anticoagulant drug warfarin. This is because the most appropriate dose of warfarin for an individual varies widely, and individuals must be periodically monitored to ensure that a proper level of anticoagulation is maintained. Warfarin is primarily metabolized by the enzymes in the CYP450 family, and is heavily impacted by the activity of vitamin K epoxide reductase. Determination of polymorphisms of these genes has been proposed as an aid to help physicians tailor anticoagulant therapy.
The impact of polymorphisms has been the focus of study with a wide variety of drugs and for many diseases and conditions. The use of this type of science is just starting to be investigated, and its impact on actual medical practice is not yet fully understood.
Cytochrome P450: Refers to a family of 60 different enzymes involved in drug and toxin metabolism.
Genotype testing: Determining the DNA sequence in genes.
Metabolize: Refers to breaking down a drug so that it is no longer clinically active.
Polymorphisms: Refers to genetic variation between individuals resulting in differences in gene expression, in this case differing activity of various enzymes.
Uridine diphosphate glucuronosyltransferase 1A1 (UGT1A1): An enzyme that is involved in drug metabolism.
Vitamin K epoxide reductase subunit C1 (VKORC1): An enzyme involved with the metabolism of vitamin K; its C1 subunit (VKORC1) is the target of the anticoagulant warfarin.
Warfarin: A commonly prescribed anticoagulant, i.e., blood thinner.
The following codes for treatments and procedures applicable to this document are included below for informational purposes. Inclusion or exclusion of a procedure, diagnosis or device code(s) does not constitute or imply member coverage or provider reimbursement policy. Please refer to the member's contract benefits in effect at the time of service to determine coverage or non-coverage of these services as it applies to an individual member.
When Services may be Medically Necessary when criteria are met:
|81225||CYP2C19 (cytochrome P450, family 2, subfamily C, polypeptide 19) (eg, drug metabolism), gene analysis, common variants (eg, *2, *3, *4, *8, *17)|
|81479||Unlisted molecular pathology procedure [when specified as genotype testing for polymorphisms of Human Leukocyte Antigen B*1502 (HLAB*1502) for carbamazepine metabolism]|
|87902||Infectious agent genotype analysis by nucleic acid (DNA or RNA); Hepatitis C virus|
|ICD-9 Diagnosis||[For dates of service prior to 10/01/2015]|
|ICD-10 Diagnosis||[For dates of service on or after 10/01/2015]|
When Services may also be Medically Necessary when criteria are met:
|81381||HLA Class I typing, high resolution (ie, alleles or allele groups); one allele or allele group (eg, B*57:01P), each [when specified as Human Leukocyte Antigen B*57:01P (HLA-B*5701) or Human Leukocyte Antigen B*1502 (HLA-B*1502)]|
|ICD-9 Diagnosis||[For dates of service prior to 10/01/2015]|
|042||Human immunodeficiency virus [HIV] disease|
|345.00-345.91||Epilepsy and recurrent seizures|
|ICD-10 Diagnosis||[For dates of service on or after 10/01/2015]|
|B20||Human immunodeficiency virus [HIV] disease|
|G40.001-G40.919||Epilepsy and recurrent seizures|
When Services are Investigational and Not Medically Necessary:
For the procedure codes listed above when criteria are not met, or when the code describes a procedure indicated in the Position Statement section as investigational and not medically necessary.
When Services are Investigational and Not Medically Necessary:
Molecular pathology procedure, Level 1 (eg, identification of single germline variant [eg, SNP] by techniques such as restriction enzyme digestion or melt curve analysis):
Molecular pathology procedure, Level 2 (e.g., 2-10 SNPs, 1 methylated variant, or 1 somatic variant [typically using nonsequencing target variant analysis], or detection of a dynamic mutation disorder/triplet repeat) [when specified as the following]:
|ICD-9 Diagnosis||[For dates of service prior to 10/01/2015]|
|230.0-234.9||Carcinoma in situ|
|290.0-319||Mental, behavioral and neurodevelopmental disorders|
|ICD-10 Diagnosis||[For dates of service on or after 10/01/2015]|
|D00.00-D09.9||Carcinoma in situ|
|F01.50-F99||Mental, behavioral and neurodevelopmental disorders|
When Services are also Investigational and Not Medically Necessary:
When the code describes a procedure indicated in the Position Statement section as investigational and not medically necessary.
|81226||CYP2D6 (cytochrome P450, family 2, subfamily D, polypeptide 6) (eg, drug metabolism), gene analysis, common variants (eg, *2, *3, *4, *5, *6, *9, *10, *17, *19, *29, *35, *41, *1XN, *2XN, *4XN)|
|81227||CYP2C9 (cytochrome P450, family 2, subfamily C, polypeptide 9) (eg, drug metabolism), gene analysis, common variants (eg, *2, *3, *5, *6)|
|81350||UGT1A1 (UDP glucuronosyltransferase 1 family, polypeptide A1) (eg, irinotecan metabolism), gene analysis, common variants (eg, *28, *36, *37)|
|81355||VKORC1 (vitamin K epoxide reductase complex, subunit 1) (eg, warfarin metabolism), gene analysis, common variants (eg, -1639/3673)|
|81479||Unlisted molecular pathology procedure [when specified as drug metabolism testing for all other drugs listed, as a panel of tests for drug metabolism or individual gene tests for drug metabolism for all other drugs listed]|
|G9143||Warfarin responsiveness testing by genetic technique using any method, any number of specimen(s)|
|ICD-9 Diagnosis||[For dates of service prior to 10/01/2015]|
|ICD-10 Diagnosis||[For dates of service on or after 10/01/2015]|
Peer Reviewed Publications:
Government Agency, Medical Society, and Other Authoritative Publications:
AmpliChip™ Cytochrome P450 (CYP450) Genotype Test
Cytochrome P450 (CYP450)
Cytochrome P450 2C9 (CYP2C9)
Polymorphisms, Drug Testing
Proove® Drug Metabolism test
Verigene® Warfarin Metabolism Nucleic Acid Test
Vitamin K Epoxide Reductase
Vitamin K Epoxide Reductase Subunit C1 (VKORC1)
The use of specific product names is illustrative only. It is not intended to be a recommendation of one product over another, and is not intended to represent a complete listing of all products available.
|Revised||08/14/2014||Medical Policy & Technology Assessment Committee (MPTAC) review. Added to medically necessary section: "Genotype testing for the presence of hepatitis C virus (HCV) genotype 1a with the NS3 Q80K polymorphism is considered medically necessary before beginning treatment with Olysio™ (simeprevir) plus peginterferon and ribavirin". Added investigational and not medically necessary statement regarding testing panels. Updated Coding, Rationale, Reference, and Index sections.|
|Reviewed||02/13/2014||MPTAC review. No change to position statement. Updated Rationale and Reference sections.|
|Reviewed||02/14/2013||MPTAC review. No change to position statement. Updated Rationale and Reference sections.|
|01/01/2013||Updated Coding section with 01/01/2013 CPT changes; removed 88384-88386 deleted 12/31/2012.|
|Revised||02/16/2012||MPTAC review. Added medically necessary statement for genotype testing for Human Leukocyte Antigen B (HLA-B*5701) for persons infected with HIV-1 before starting treatment with abacavir. Updated Rationale, Coding and Reference sections.|
|01/01/2012||Updated Coding section with 01/01/2012 CPT changes.|
|Reviewed||05/19/2011||MPTAC review. No change to position statement. Updated Reference section.|
|Revised||05/13/2010||MPTAC review. Added testing for CYP2C19 variant of Cytochrome P450 as medically necessary for individuals receiving clopidogrel therapy and who have not been previously tested or those for whom clopidogrel therapy has been proposed. Updated Rationale, Coding, Reference and Index sections.|
|01/01/2010||Updated Coding section with 01/01/2010 HCPCS changes.|
|Revised||05/21/2009||Hematology/Oncology Subcommittee review. Added use of Human Leukocyte Antigen B*1502 (HLAB*1502) as medically necessary with criteria. Added Clopidogrel and HLAB*1502 to investigational and not medically necessary section. Updated Rationale, Coding, Reference and Index sections.|
|Reviewed||02/26/2009||MPTAC review. Updated Rationale and Reference sections.|
|Reviewed||02/21/2008||MPTAC review. Updated Rationale and Reference sections.|
|Revised||11/29/2007||MPTAC review. Altered title to replace "Cytochrome P450" with "Genetic." Revised the investigational/not medically necessary position statements to include all genetic polymorphism testing for drug metabolizer status. The phrase "investigational/not medically necessary" was clarified to read "investigational and not medically necessary." Updated Rationale, Background, Reference, and Index sections.|
|Revised||03/08/2007||MPTAC review. Added tamoxifen to investigational/not medically necessary section. References and Coding updated. Document number changed from LAB.00013 to GENE.00010.|
|Reviewed||03/23/2006||MPTAC review. References updated.|
|New||04/28/2005||MPTAC initial document development.|