By Harvey S. Bartnof, MD, Staff Physician at AVERI, AIDS Virus Education and Research Institute in San Francisco, California

PART 1:
     ATHENA Study
     PharmAdapt Study

Introduction
A leading edge of the HIV/AIDS epidemic is the clinical pharmacology of anti-HIV therapy, which correlates the measurement of anti-HIV drug levels in blood with viral load outcomes and toxicity. This incorporates the concept of "TDM," or Therapeutic Drug Monitoring, when drug dosing is increased (or decreased) if measured blood levels are too low (or too high). TDM is not new in clinical medicine. It is the "standard-of-care" in drug therapy for certain bacterial infections that use "aminoglycoside" antibiotics, seizures (convulsions or "fits"), asthma treated with theophylline and for certain heart conditions that require digoxin. Pharmacology is the study of drugs.

Two countries in Europe are very involved in clinical research of TDM for HIV patients: the Netherlands and France. In a personal conversation this reviewer had with David M. Burger, PharmD, who is on the Organizing Committee of the Workshop and from University Hospital Nijmegen in the Netherlands, he said that approximately 70-80% of all HIV patients taking anti-HIV therapy in his country currently have regular TDM. Results are forwarded to the treating physician. In a personal conversation this reviewer had with Dominique Breilh, PharmD, PhD of Bordeaux University Hospital in France, a presenting author at the Workshop, she said that approximately 90% of HIV patients taking anti-HIV therapy in her country have regular TDM. The results are then reviewed at regular interdisciplinary meetings with a patient's physician, a pharmacologist and virologist. Based upon other research presented at the Workshop, other European countries also are actively involved in TDM research for HIV/AIDS patients, including the United Kingdom, Belgium, Germany, Switzerland, Spain and Italy. The US overall appears to have had less research in the area, although two of six of the Organizing Committee members of the Workshop were exclusively from the US. Less US research in this area may be related to less interest and/or bias up to this point in time by members of the ACTG (AIDS Clinical Trials Group) and other key HIV research professionals, including physicians, pharmacologists and virologists. However, this is changing, since presenters from the NIH announced at the Workshop a new "NIH HIV Drug Interaction and Information Resource," and expanded support for new studies.

Most of the pharmaceutical manufacturers of anti-HIV drugs have a major interest in the area of TDM, particularly since basic pharmacology is required during the pre-approval and development process of drugs. In addition, they have a legitimate financial interest that will likely also benefit patients. Specifically, in theory, TDM will allow for a greater number of patients to be taking anti-HIV drugs with individualized dosing associated with less toxicity, less discontinuation of therapy due to toxicity and, hopefully, higher rates of viral undetectability. However, with the exception of one study below, such benefits have not been demonstrated in any prospective studies to date.

The Workshop was organized by Virology Education in the Netherlands and was sponsored by 11 pharmaceutical or diagnostic companies. There were approximately 162 registrants and 77 abstracts, representing an increase from last year's 1st Workshop where there were approximately 122 registrants and 40 abstracts. This year, there were 30 oral presentations, 93% of which, along with the remaining abstracts, were presented as posters. The abstracts and many of the posters, in addition to other relevant sections are available on the Internet at www.HIVpharmacology.com (although no site came up when I tried to access it).

Currently, in the US, all tests to measure drug levels of anti-HIV drugs generally are "for research only." However, even though many reference clinical laboratories do perform the tests, many insurance companies and other third party payers do not pay for anti-HIV TDM. (Some physicians have found a way around this by using TDM "billing codes" for digoxin or other drugs that are commonly a part of TDM. Of interest, Virco, which performs HIV genotype and phenotype resistance testing, also does anti-HIV drug level testing using its VircoPK test. Remember that blood plasma (no cells) generally is used to measure drug levels of the PI (protease inhibitor) and NNRTI (non-nucleoside) drugs. However, plasma levels of the NRTI drugs do not provide much information, since it is the concentration of NRTI "tri-phosphate" levels (active components) in blood immune white cells that would be the appropriate measurement in TDM. (Editorial note: researchers are trying to develop testing which would correlate NRTI blood levels with intracellular NRTI levels, and although promising it's preliminary). Also, be aware that there still are many unknowns in the area of anti-HIV TDM, including which measurement is best: the minimum, maximum or total ("area-under-the-curve") drug concentration?

(Editorial note from Jules Levin: The minimum is often referred to as the Cmin or trough, and is the lowest level of drug in the blood which usually occurs at the end of the dosing period. For example, if a person is taking a twice a day drug like ritonavir the minimum is at 12 hours just before taking the next dose. The maximum is often referred to as the Cmax and is the peak or highest drug level in the blood, and usually occurs several hours after taking a drug. The AUC is the total amount of drug in the blood during the entire dosing period whether it be 8, 12, or 24 hours (as in the case of efavirenz or any once daily dosed drug).

ATHENA Study Shows Benefits for Therapeutic Drug Monitoring (TDM) in Treatment-Naïve Patients

• First, prospective study to show viral load benefits.
• Protease inhibitor drugs were indinavir- or nelfinavir-based HAART.

David Burger, PharmD presented the results of the first prospective study to show virologic benefits with TDM, therapeutic drug monitoring. He and his colleagues are from University Medical Center in Nijmegen, the Netherlands. Dr. Burger presented only the results for TDM of the PI drugs nelfinavir (Viracept) and indinavir (Crixivan). Future analyses will encompass the other PI drugs and the NNRTI drugs nevirapine and efavirenz. He first reviewed that previous studies have shown that low blood levels of either nelfinavir or efavirenz have been associated with virologic failure. Also, previous studies have shown that high blood plasma levels of indinavir have associated with toxicity, in this case kidney stones. "Nephrolithiasis" due to indinavir crystallization in urine can cause severe back/groin pain and blood in urine.

(Editorial note: a recent publication showed data that efavirenz-related CNS side effects can be associated with higher blood levels. The study authors suggested that blood levels can be monitored and adjusted while viral suppression is maintained. This data is preliminary and needs further study).
AIDS Jan. 2001, 15: 71-75
http://www.natap.org/2001/jan/efavirenz_plasma011501.htm

While ATHENA enrolled 600 patients, the TDM part was a sub-study. Overall ATHENA results have been presented previously. Among 92 treatment-naïve (no previous therapy) patients randomized to nelfinavir (1,250 mg twice daily) combination therapy within ATHENA, patients in the sub-study were further randomized to TDM report group or to a TDM "blinded group" whereby results were not given to the treating physician. TDM patients had their TDM results given to their doctors, along with "yes/no" advice about changing therapy. If the first nelfinavir level was low, physicians were recommended to discuss the importance of taking the drug with food. If the second test still showed low blood levels, the physicians were recommended to increase the nelfinavir dose to 1,500 mg twice daily. If the 3rd test sill showed levels that were low, the recommendation was to increase nelfinavir to 1,500 mg twice daily or to add low dose ritonavir (Norvir, PI drug) to boost nelfinavir blood levels.

Nelfinavir blood levels were tested along with HIV RNA levels at regular intervals; blood samples were drawn "randomly." Dr. Burger defined "low" nelfinavir levels by using the "concentration ratio" (CR). The CR compares each patient's blood concentration of nelfinavir to a "time-adjusted" value from a comparable population. For nelfinavir, a CR less than 0.9 was considered low. (Dr. Burger described the CR for indinavir in detail in a separate poster presentation.)

In the TDM and no-TDM results arms, randomization was equal, with both arms balanced for age, gender (sex), baseline CD4 counts, viral load and use of other anti-HIV drugs. All 92 patients were taking 2-3 other anti-HIV drugs as a part of HAART (highly active antiretroviral therapy), in addition to nelfinavir.

The results after one year were as follows. Using a strict "intent-to-treat" (ITT) analysis ("non-completer equals failure," including all randomized patients), the TDM results arm achieved a significantly higher level of viral undetectability (81%) than the TDM-blinded arm (59%, limit 500 copies per milliliter, p=0.03). Also after one year, significantly less patients in the TDM results arm discontinued treatment (12%) than in the TDM-blinded arm (35%, p=0.01). This was mainly due to a lower rate of discontinuation in the TDM results arm due to virologic "failure" (2%) than in the TDM-blinded arm (18%, p=0.02). Discontinuation due to toxicity was nearly equal in both arms, approximately 7% each. Discontinuation due to "patient request" was in approximately 8% of the TDM-blinded arm and zero in the TDM results arm. Dr. Burger did not review toxicity differences in the two arms. Dr. Burger and colleagues concluded, "TDM of nelfinavir 1,250 mg BID [twice daily] in treatment-naïve patients improves treatment outcome and should become standard of care.

There are some limitations to the findings. The results would have greater weight if the drug concentrations at one year had been reported and were found to be associated with the viral load differences in the two groups. Undoubtedly, Dr. Burger will be doing this analysis in the future. Also, toxicity results in the two arms would have been helpful, including whether or not there was greater toxicity among those who nelfinavir doses were increased. Also, it would have been helpful to see the mean and range of nelfinavir levels for all time points for each arm and the percentage of patients whose nelfinavir dose actually was increased. In addition, it will be useful to know the number of patients in each arm taking NNRTI therapy and the percentage taking each type of NNRTI drug. This might have affected potency of the HAART regimens. During the question-and-answer session, it became clear that Dr. Burger did not know whether physicians in the TDM results arm followed the recommendations of increasing (changing) doses, when applicable, or whether patients followed the physicians' recommendations. Certainly, all of this information that was not reported is available in the patients' charts and/or in the computer database that has the laboratory results. Nevertheless, the virologic undetectability rate in the TDM results arm is rather impressive.

In the other part of his presentation, Dr. Burger presented the results of 55 treatment-naïve patients who were randomized to indinavir HAART. Within that group, 27 were randomized to the TDM-blinded arm, while the remaining 28 to the TDM results arm. Among the 55 patients, 16 took standard indinavir dosing of 800 mg 3-times daily (subgroup 1), 20 took indinavir 800 mg plus ritonavir (Norvir) 100 mg, both twice-daily (subgroup 2), while the remaining 19 patients took 400 mg each of indinavir and ritonavir, twice daily (subgroup 3). Both the TDM results and TDM-blinded arms were "well balanced" by age, gender, CD4 counts, viral load and concurrent HAART regimen components.

The results of regular, random (within the dosing interval) blood levels and viral loads were reported to the patients' physician. The indinavir target levels for the three subgroups were CR (concentration ratio, see above) 0.75-2.0 (subgroup 1), 0.25-2.0 (subgroup 2), and greater than 0.5 (subgroup 3). Physicians were provided with advice about whether to modify dosing based upon the CR from TDM.

Using a strict "ITT" analysis, the results after one year were as follows. A significantly higher proportion of patients in the TDM results arm achieved viral undetectability (75%) than in the TDM-blinded arm (48%, p=0.04). There was a non-significant trend towards a higher discontinuation rate in the TDM-blinded arm (48%) than in the TDM results arm (25%, p=0.07). This was mainly due to a significantly higher rate of discontinuation due to toxicity (40%) in the TDM-blinded subgroups that took 800 mg 3-times daily or indinavir 800 mg/ritonavir 100 mg twice-daily than in the TDM results subgroups (10%) with the same dosing (p=0.03). In the 400/400 mg subgroup, a non-significantly higher rate of those in the TDM results arm discontinued due to toxicity (approximately 28%) than in the TDM blinded arm (approximately 16%). Discontinuation due to virologic failure was low in both the TDM results (approximately 4%) and TDM-blinded arms (zero).

Dr. Burger concluded, "TDM of indinavir 800 mg TID [3-times daily] or indinavir 800 mg plus ritonavir 100 mg BID [twice daily] in treatment-naïve patients improves treatment outcome, predominantly by allowing a better management of indinavir toxicity." Limitations to the indinavir results are similar to those mentioned above for nelfinavir. Nevertheless, the viral undetectability differences are still impressive.

We look forward to additional analyses that included information discussed in the limitations above, as well as the results of patients randomized to the other PI drugs and by NNRTI drug. Additional larger, prospective studies will be necessary to confirm the results of the ATHENA TDM subgroups.

(Editorial note: there will be follow-up with additional information gathering and relevant info will be reported to you).

PharmAdapt Study: No Benefits for TDM after 12 Weeks

• Study has limitations.

The 12-week results of the PharmAdapt prospective study of TDM were re-presented by Dr. Peter Clevenbergh of Nice Hospital in France. (He presented essentially the same results at the recent 8th Annual Retrovirus Conference in February.) PharmAdapt used PI drug genotype resistance test results at baseline among patients failing a PI-drug regimen. They were randomized to TDM for PI drugs (trough or lowest drug level) after four weeks (84 patients) or no TDM (96 patients). At week 8, 17% of physicians of patients in the TDM arm were given PI dose modification recommendation, a significantly higher rate than 6% in the no TDM (control) arm who had a physician-generated PI drug modification (p=0.006).

The results after 12 weeks showed a "comparable" viral load decrease in both arms: -2.6 log in the control arm and -2.3 log copies per milliliter in the TDM arm. The percentage with an undetectable viral load (limit 200 copies per milliliter) was also quite similar in the two arms: 52% in the control arm and 45% in the TDM arm.

The PharmAdapt prospective study did not show a benefit for TDM, at least after 12 weeks of follow-up. There were four major criticisms of the study. First, Dr. Andrew Hill of Roche said that the number of patients randomized in each arm was too small to demonstrate a statistically significant difference. Dr. Hill had an entire poster devoted to the concept of "statistically underpowered" TDM studies. Second, four weeks between the time of the blood test and recommendations for changing the dose possibly was too long and that the recommendations should have been made sooner. (Additional mutations might have accrued necessitating even higher dosing.) Third, the "cut-offs" for optimal PI drug concentrations used the "protein-adjusted IC50" (inhibitory concentration for 50% of HIV growth in the laboratory, adjusted for binding proteins) for "wild-type" (without mutations) HIV. It probably would have been better, if possible, to use the IC50 for mutated (drug-resistant) HIV of each patient, since it is unlikely that most (or all) had wild-type at baseline. And fourth, longer follow-up time might be needed, as in the ATHENA substudy above, where there were only borderline significant differences in viral load reductions between TDM and TDM-blind arms at six months, yet significant differences after 12 months.

References
Burger DM and others. Therapeutic drug monitoring (TDM) of indinavir in treatment-naïve patients improves therapeutic outcome after 1 year: results from ATHENA. Abstract and oral presentation 6.2a at the 2nd International Workshop of Clinical Pharmacology of HIV Therapy; April 2-4, 2001; Noordwijk, the Netherlands.

Burger DM and others. Therapeutic drug monitoring (TDM) of nelfinavir 1,250 mg BID in treatment-naïve patients improves therapeutic outcome after 1 year: results from ATHENA. Abstract and oral presentation 6.2b at the 2nd International Workshop of Clinical Pharmacology of HIV Therapy; April 2-4, 2001; Noordwijk, the Netherlands.

Burger DM and others. Validation of the use of the concentration ratio (CR) to estimate the area-under-the-curve (AUC) of indinavir in HIV-1 infected patients. Abstract and poster presentation 6.6 at the 2nd International Workshop of Clinical Pharmacology of HIV Therapy; April 2-4, 2001; Noordwijk, the Netherlands.

Clevenbergh P and others. PharmAdapt: a prospective multicenter randomized controlled trial to evaluate the usefulness of protease inhibitor drug monitoring: 12 week results. Abstract and oral presentation 6.1 at the 2nd International Workshop of Clinical Pharmacology of HIV Therapy; April 2-4, 2001; Noordwijk, the Netherlands.

Hill AM and Stirnadel HA. Are clinical trials of TDM underpowered to detect effects on HIV RNA?: three-stage modeling analysis of sample sizes. Abstract and poster presentation 6.9 at the 2nd International Workshop of Clinical Pharmacology of HIV Therapy; April 2-4, 2001; Noordwijk, the Netherlands.

Personal communication with David M. Burger, PharmD on April 4, 2001.

Personal communication with Dominique D. Breilh, PharmD, PhD on April 4, 2001.