Patient willingness to undergo efavirenz dose reduction based on pharmacogenetic testing

Abstract

Aim:

Among HIV-infected adults receiving efavirenz fixed-dose combination tablets, genotyping could guide efavirenz dose reduction but would require more pills.

Methods:

We assessed willingness to dose reduce among 129 patients at an HIV primary care clinic in the southeastern USA.

Results:

When told that switching from one pill to two or three pills “might make you feel a little better,” 47% expressed definite or possible willingness. This decreased to 9% if there was “a small chance it might not control your HIV as well.” Clinical variables were not associated with willingness.

Conclusion:

Many patients receiving a fixed-dose combination tablet may be willing to take more pills in order to dose reduce, guided by genetic testing, but only if virologic control is not compromised.

In 1998, the US FDA first approved efavirenz as a component of multidrug therapy for HIV-1 infection. Since then efavirenz has been extensively prescribed in the USA and worldwide. Although generally safe and effective, some patients who initiate efavirenz experience CNS symptoms [1] which may include increased risk for suicidality [2]. In April 2015, largely driven by the availability of newer agents with more favorable tolerability profiles, guidelines in the USA were updated to change efavirenz-containing regimens from a recommended option to an alternative option as initial therapy for HIV-1 infection [3].

Efavirenz is metabolized primarily by CYP450 2B6 [4], and several CYP2B6 loss-of-function polymorphisms predict increased plasma efavirenz exposure [5]. Of greatest impact is CYP2B6 516G→T (rs3745274) [5–10], which is more frequent with African ancestry than with European ancestry [11]. A less frequent polymorphism, CYP2B6 983T→C (rs28399499), also predicts increased plasma efavirenz exposure [5,12–14] and appears to be present only with African ancestry [11]. A third polymorphism, CYP2B6 15582C→T (rs4803419), predicts a modest increase in plasma efavirenz exposure and is most frequent with European ancestry [5]. Various combinations for these three polymorphisms predict an approximately tenfold range of interindividual variability in plasma efavirenz exposure [5].

Higher plasma efavirenz concentrations have been reported to correlate with increased CNS symptoms [1,9,14–18], suggesting that reducing the efavirenz dose might lessen such symptoms. Among adults in the USA, efavirenz is typically prescribed at 600 mg once daily in a fixed-dose combination tablet that also contains tenofovir disoproxil fumarate and emtricitabine. In CYP2B6 intermediate metabolizers and CYP2B6 slow metabolizers, once-daily efavirenz doses of 400 and 200 mg, respectively, would maintain plasma efavirenz trough concentrations comparable to extensive metabolizers [5], so should maintain efficacy. Virologic efficacy of reduced-dose efavirenz is supported by pilot data on efavirenz dose reduction based on CYP2B6 genotype [19], and by results of a clinical trial in which patients were randomized to initiate either 600 or 400 mg of efavirenz once daily regardless of genotype [18,20].

Based on minor allele frequencies of these CYP2B6 polymorphisms, if genetic testing were implemented, the majority of patients would receive <600 mg of efavirenz, regardless of race or ethnicity (i.e., in the USA, over 50% of individuals are either CYP2B6 slow or CYP2B6 intermediate metabolizers, regardless of whether they are white, black or Hispanic) [5]. Such dose reduction may also be cost effective in the USA, even if generic efavirenz becomes available [21]. However, efavirenz is typically prescribed as a single once-daily fixed-dose combination tablet that contains 600 mg of efavirenz. Mono-formulated efavirenz is only available as 200 and 600 mg tablets. Therefore, to dose reduce efavirenz would require increasing the total number of pills (including co-formulated tenofovir disoproxil fumarate/emtricitabine) to two (for 200-mg dosing) or three (for 400-mg dosing), and with one additional prescription. Furthermore, patients without major side effects while receiving 600 mg of efavirenz may be unwilling to dose reduce. Individualized dosing of efavirenz therefore raises implementation questions that may also be generally relevant to personalized medicine with other fixed-dose combinations tablets.

The present study, based on the above considerations, assessed willingness of patients to dose reduce one component of a once-daily combination antiretroviral tablet among clients at a large HIV primary care clinic in the southeastern USA. We also assessed whether willingness to undergo dose reduction was associated with various demographic or clinical factors.

Methods

Study participants

Study participants were enrolled during clinic visits at a large HIV primary care clinic in Nashville (TN, USA). In 2014, a total of 3125 different HIV-infected adults had at least one clinic visit, of whom 56.7% were white, 41.8% black, 1.4% were other race, 5.1% were Hispanic, 76.3% were male, 23.0% were female, 0.7% were transgender and median age was 46 years. Each participant was surveyed only once. On days during which surveys were administered, as many potential study participants as was feasible were approached for possible study participation. Participants were enrolled consecutively based on clinic appointment times. Eligible participants were patients whose only current HIV-1 treatment was a once-daily combination tablet that contains at least three antiretrovirals, willing and able to provide written informed consent or a waiver of documentation of consent, and whose primary language was English. Patients were excluded if the patient’s primary care provider felt that participation was not in the patient’s best interest. There were no other eligibility criteria.

Survey administration & data collection

Study participants were approached by research personnel who were not involved in patient care. Of 132 participants, 118 (89%) participated in 2014, and 14 (11%) in 2013. Research personnel explained the study, obtained informed consent, and administered oral surveys verbatim. Each survey required <10 min to complete. Because survey responses were completely anonymous, we did not attempt to validate responses against the medical record. Full survey text is provided in the Supplementary Materials. If queried by participants during the survey, study personnel could provide clarification to make sure participants understood the particular question, but did not otherwise stray from the script. Data were managed using a secure REDCap database [22]. There were no financial incentives to participants. This study was approved by the Vanderbilt University Institutional Review Board.

Statistical analyses

Spearman rank test was used to assess correlations between continuous or ordinal variables, and proportional odds test was used for categorical variables. The study targeted a sample size to generate a 95% CI that spans no more that 20 percentiles for proportion of HIV-infected individuals willing to dose reduce. Assuming that 50% of individuals would be willing to undergo dose reduction, a sample size of 100 would give a 95% CI that spans 19.2% (40.4–59.6%). To allow for some patients with unevaluable data, we targeted enrollment of 120 subjects.

Results

Study participants

A total of 132 patients were surveyed, of whom 129 provided evaluable data. Data from three patients, judged by the interviewer to be invalid, were excluded (e.g., a patient hurrying to finish, not paying attention to questions). Characteristics of evaluable participant are shown in Table 1. There were slightly more white participants than black participants, and most participants were male. Two-thirds of participants were receiving an efavirenz-containing regimen. Of the 129 participants, 117 (91%) reported feeling ‘pretty good’ or ‘great’ on a typical day, 111 (87%) reported having an undetectable viral load, and 57 (44%) reported insurance that required co-payments (a co-payment is a fixed fee required by a health insurer to be paid by the patient for each office visit, outpatient service or filling of a prescription).

Table 1.

Characteristics of evaluable study participants.

Characteristics	All evaluable subjects (n = 129)	Efavirenz recipients (n = 84)
Race, n (%):
– White	66 (51.2)	45 (53.6)
– Black	50 (38.8)	33 (39.3)
– Other	13 (10.0)	6 (7.1)
Sex, n (%):
– Male	113 (87.6)	73 (86.9)
– Female	16 (12.4)	11 (13.1)
Age, median (IQR); years	45 (34–52)	49 (39–54)
Years on ART, median (IQR)	2.4 (0.9–5.0)	4.6 (2.2–7.1)
Once-daily ART, n (%):
– Atripla®†	84 (65.1)	84 (100)
– Stribild®‡	29 (22.5)	–
– Complera®§	14 (10.9)	–

^†Atripla^® is a three-drug co-formulation of efavirenz, tenofovir disoproxil fumarate and emtricitabine.

^‡Stribild^® is a four-drug co-formulation of elvitegravir, cobicistat, tenofovir disoproxil fumarate and emtricitabine.

^§Complera^® is a three-drug co-formulation of rilpivirine, tenofovir disoproxil fumarate and emtricitabine.

ART: Antiretroviral therapy; IQR: Interquartile range.

Willingness to undergo dose reduction

The process of dose reduction from a once-daily co-formulated tablet was explained to participants (Table 2, question 1), and only patients who endorsed understanding proceeded to subsequent questions. To feel as good as possible and to have an undetectable viral load were considered to be equally important to 118 (91%) of the participants. When told that changing from one pill once daily to two or three pills once daily “might make you feel a little better” (Table 2, question 2), 60 (47%) of the participants said they were ‘yes’ (27%) or ‘maybe’ (19%) willing to change their regimen. When told that the same change “might make you feel a little better, but with a small chance it might not control your HIV as well” (Table 2, question 3), only 11 (9%) said they were ‘yes’ (2%) or ‘maybe’ (7%) willing to change their regimen. Responses are shown in Figure 1. To undergo genetic testing to select the best dose of a drug was acceptable to 117 (91%).

Table 2.

Selected survey questions.

Question	Text	Options
1	“Once-a-day pills for HIV contain three different HIV medicines, each at a specific dose. It is possible that some people on a once-a-day pill for HIV might be getting more of one of the medicines than they need. But to lower the dose of only one of the medicines in the pill would mean taking two or three pills once a day, not one pill once a day. Do you understand this?”	Yes, no
2	“Do you think you would be willing to change your one pill once a day to two or three pills once a day, and with two prescriptions instead of one prescription, if this might make you feel a little better? For example, you might have more energy, sleep better or be able to concentrate better. It would be the same medicines that are in the once-a-day pill, but one medicine would be at a lower dose”	Yes, no, maybe, do not know
3	“Do you think you would be willing to change your one pill once a day to two or three pills once a day, and with two prescriptions instead of one, if this might make you feel a little better, but with a small chance it might not control your HIV as well?”	Yes, no, maybe, do not know

None of the questions in the full survey mentioned a specific medicine by name.

Figure 1. Participant willingness to undergo dose reduction.

The graph represents responses from 129 evaluable participants. Hashed bars represent responses to question 2 in Table 2 (“do you think you would be willing to change your one pill once a day to two or three pills once a day, and with two prescriptions instead of one prescription, if this might make you feel a little better? For example, you might have more energy, sleep better or be able to concentrate better. It would be the same medicines that are in the once-a-day pill, but one medicine would be at a lower dose”). Solid bars represent responses to question 3 in Table 2 (“do you think you would be willing to change your one pill once a day to two or three pills once a day, and with two prescriptions instead of one, if this might make you feel a little better, but with a small chance it might not control your HIV as well?”).

Associations with willingness to undergo dose reduction

We explored whether willingness to undergo dose reduction was associated with selected demographic or clinical factors. Considering that dose reduction “might make you feel a little better,” willingness to undergo dose reduction was not significantly associated with race, sex, age, years on antiretroviral therapy (ART), specific once-daily ART, insurance co-payment status, feeling or wellness on a typical day, number of non-ART, and number of non-ART prescriptions (p > 0.05 for each variable, Table 3). In analyses limited to the subset of 84 efavirenz recipients, associations were also not significant (data not shown).

Table 3.

Willingness to undergo dose reduction in relation to patient characteristics.

Characteristics	Survey response†
	Yes	Maybe	No
Race, n (%):
– White	17 (26)	17 (26)	30 (45)
– Black	12 (24)	8 (16)	30 (60)
– Other	6 (46)	0 (0)	6 (46)
Sex, n (%):
– Male	31 (27)	23 (20)	56 (50)
– Female	4 (25)	2 (13)	10 (63)
Age, median (IQR); years	46 (33–52)	37 (33–49)	46 (36–52)
Years on ART, median (IQR)	2.4 (1.0–4.9)	1.5 (0.7–3.8)	4.0 (1.0–7.1)
Once-daily ART, n (%):
– Atripla®‡	22 (26)	13 (15)	49 (58)
– Stribild®§	7 (24)	10 (34)	11 (38)
– Complera®¶	6 (43)	2 (14)	5 (36)
Insurance co-pays, n (%):
– Yes	14 (25)	13 (23)	29 (51)
– No	20 (29)	12 (18)	35 (51)
Feeling on typical day, n (%):
– Good or great	31 (27)	22 (19)	61 (52)
– Bad or terrible	5 (50)	2 (25)	2 (25)
Other pills, median (IQR)	3 (1–6)	2 (0–4)	2 (0–5)
Other prescriptions, median (IQR)	2 (1–4)	2 (0–3)	1 (0–3)

^†p > 0.05 for each variable (three participants whose response was ‘do not know’ are not represented in the table).

^‡Atripla^® is a three-drug co-formulation of efavirenz, tenofovir disoproxil fumarate and emtricitabine.

^§Stribild^® is a four-drug co-formulation of elvitegravir, cobicistat, tenofovir disoproxil fumarate and emtricitabine.

^¶Complera^® is a three-drug co-formulation of rilpivirine, tenofovir disoproxil fumarate and emtricitabine.

ART: Antiretroviral therapy; IQR: Interquartile range.

Regarding the question stating that dose reduction “might not control your HIV as well,” too few participants expressed willingness to consider dose reduction to meaningfully assess associations with demographic or clinical factors (data not shown).

Discussion

The present study demonstrated that a substantial proportion of HIV-infected patients would be willing to undergo dose reduction of one component of their once-daily co-formulated ART if this might make them feel better, even though this would require switching from one pill to multiple pills, and even though they reported feeling pretty good or great on a typical day. However, very few patients were willing to consider such a change if there was a small chance than dose reduction would compromise virologic control. In addition, we did not find associations between willingness to undergo dose reduction and selected demographic and clinical factors. From these data we conclude that many patients would be willing to undergo such dose reduction of efavirenz, and regardless of many clinical factors, but only if there is virtually no risk of virologic failure. It is remarkable that a substantial proportion of patients expressed willingness to switch from one pill to multiple pills, considering the considerable marketing by pharmaceutical companies focused on one pill, once-daily regimens. In the coming years, as generic efavirenz and Atripla^® become available, this approach may become more attractive.

The lack of apparent association between willingness to undergo dose reduction and selected clinical variables was somewhat unexpected. We had speculated that patients might be reluctant to consider undergoing dose reduction if their insurance required co-payments, if they had more recently initiated ART, if they were being prescribed a fewer total number of pills or had fewer total prescriptions. Our data suggest that dose reduction may be equally acceptable independent of these factors.

The translation of pharmacogenetic testing into clinical care is a complex undertaking. When evidence is overwhelming for benefit of a genetic test, such as with HLA-B*5701 screening to avoid abacavir hypersensitivity reactions [23], there is strong motivation to implement and acceptance will be high [24]. However, when the benefit of a pharmacogenetic test is less pronounced, any small obstacle to implementation may be magnified. Benefits of CYP2B6 genotyping to guide efavirenz dose reduction may be modest [21], so the need to change from one pill once daily to multiple pills represents a substantial obstacle. Against this backdrop, if direct-to-consumer genetic testing becomes more widespread [25], questions about whether to order a particular genetic test may become less relevant, the focus may become how to best use readily available genetic data, and with patients becoming more involved in decision-making. In addition, a number of commercial laboratories offer CYP2B6 genotyping.

There were limitations to the present study. A larger sample size may have revealed associations that were not apparent in this study. However, the remarkably similar proportions, medians and overlapping interquartile ranges in Table 3 suggest that associations with demographic or clinical factors would likely be modest. The proportion of women in our study was less than in the primary care clinic, reflecting the fact that prescribing of efavirenz to women of childbearing potential was generally avoided. We do not know whether responses to survey questions reliably predict behavior if the patient were to actually be offered dose reduction. In addition, we did not rigorously assess the extent to which participants understood the survey questions. While we included participants who were receiving any once-daily combination tablet that contains at least three antiretrovirals, associations were also not significant in analyses limited to efavirenz recipients.

In summary, among patients at a large HIV primary care clinic in the southeastern USA, a substantial proportion expressed willingness to dose reduce one component of their one pill once-daily ART if this might make them feel a little better, even though this would require switching to multiple pills, but only if there is virtually no risk of virologic failure.

Conclusion

Translation of human genetic testing into HIV clinical practice has the potential to both decrease drug toxicity and optimize efficacy. However, for the widely prescribed antiretroviral drug efavirenz, individualized dose reduction based on pharmacogenetic testing would require switching from a single fixed-dose combination tablet once daily to two or three pills once daily. The present study demonstrated that, among patients receiving primary care at a large HIV primary care clinic in the southeastern USA, a substantial proportion of patients expressed willingness to switch from one pill to multiple pills. This is remarkable given the considerable current emphasis on one pill, once-daily regimens, and supports the potential feasibility of such an approach in clinical practice.

Executive summary.

Background

• Although efavirenz is generally safe and effective, some patients who initiate efavirenz experience CNS symptoms, which may include increased risk for suicidality.

• Higher plasma efavirenz concentrations have been reported to correlate with increased CNS symptoms.

• Several CYP2B6 loss-of-function polymorphisms in combination predict an approximately tenfold range of interindividual variability in plasma efavirenz exposure.

• If genetic testing to dose reduce efavirenz were implemented in clinical practice, the majority of patients would receive less than the standard 600 mg dose of efavirenz, regardless of race or ethnicity.

• To dose reduce efavirenz would require switching from a single fixed-dose combination tablet once daily to two or three pills once daily.

Methods

• We surveyed patients at a large HIV primary care clinic in the southeastern USA to assess willingness to dose reduce one component of a once-daily combination antiretroviral tablet.

Results

• Among 129 patients, when told that changing from one pill once daily to two or three pills once daily “might make you feel a little better,” 47% of the participants said they were ‘yes’ (27%) or ‘maybe’ (19%) willing to change their regimen.

• When told that the same change “might make you feel a little better, but with a small chance it might not control your HIV as well,” only 9% said they were ‘yes’ (2%) or ‘maybe’ (7%) willing to change their regimen.

• Willingness to undergo dose reduction was not significantly associated with race, sex, age, years on antiretroviral therapy (ART), specific once-daily ART, insurance co-payment status, feeling of wellness on a typical day, number of non-ART pills, and number of non-ART prescriptions.

Discussion

• The present study demonstrated that a substantial proportion of HIV-infected patients would be willing to undergo dose reduction of one component of their once-daily co-formulated ART if this might make them feel better, even though this would require switching from one pill to multiple pills, and even though they reported feeling pretty good or great on a typical day. However, very few patients were willing to consider such a change if there was a small chance than dose reduction would compromise virologic control.

• It is remarkable that a substantial proportion of patients expressed willingness to switch from one pill to multiple pills, considering the current emphasis on one pill, once-daily regimens.