An audit of UK patients who switch from their first- to second-line anti-HIV therapy has found that toxicity was the main reason for switching therapy. In addition, one in four patients who changed regimens due to virological 'failure' did not obtain a resistance test result before switching therapy. The results were presented at last week's British HIV Association (BHIVA) conference, held in central London, and are now available for download from BHIVA's clinical audit website. The audit also found that some patients with virological 'failure' remained on therapy with detectable viral loads for long periods before switching.
A total of 134 HIV treatment centres responded to the survey, providing data on 504 patients. Of these centres, 100 (75%) said they relied on the BHIVA treatment guidelines, 28 (21%) used local guidelines in addition to BHIVA's, and six (4%) did not answer.
Reasons for switching
A case note review of patients switching therapy excluded 67 of the 504 patients because their switch occurred less than twelve weeks after starting therapy and/or it was their second or subsequent switch. More than one reason could be given for each patient, resulting in a higher total than 100%.
Toxicity or side-effects were the main reasons for switching for more than half (51%, 223/437) of the case notes analysed. Just under one third (30%, 132) switched due to virological 'failure'.
Another fourteen percent (63) switched because of difficulties with adherence; 43 (10%) switches were due to "patient choice"; and 42 (10%) were due to treatment simplification.
A further 21 (5%) switched due to a poor CD4 response; 22 (5%) due to drug interactions or co-morbidity due to other drugs; and 15 (3%) switched because their current therapy wasn't meeting current recommendations.
In addition 14 (3%) women switched because they were planning pregnancy and another five (1%) switched because they were already pregnant.
Metabolic problems
Metabolic problems were the most common toxicities cited as a reason for switching therapy, affecting 71 (16%) of patients including:
- 44 with lipoatrophy
- 26 with high blood pressure
- 17 with high triglycerides
- 12 with central obesity
- and one patient with hyperglycaemia.
Some patients had more than one condition.
Of those with metabolic toxicity, 38 (54%) were on d4T (stavudine, Zerit) prior to their switch. A total of 28% with metabolic toxicity were on a protease inhibitor (PI) before their switch, compared with 18% of all patients; 66% were on a non-nucleoside reverse transcriptase inhibitor (NNRTI), compared with 74% of all patients; 1% were on both, compared with 1% of all patients; and 4% were on neither, compared with 7% of all patients.
However, differences in pre-switch drug use patterns between those with metabolic toxicity and other patients may partially reflect longer duration on therapy. Fat loss due to d4T, for example, can take two years or more to be noticeable. In fact, 70% of those patients who switched had been on their regimen for longer than two years, compared with 40% of all patients.
Other side-effects
A total of 41 patients switched due to central nervous system -related side-effects; 40 (98%) were on efavirenz (Sustiva) prior to switching. Another 25 switched due to gastrointestinal problems, of which 13 (52%) were on a PI prior to switching.
Eighteen patients experienced peripheral neuropathy. Fourteen were on 3TC (lamivudine, Epivir) before the switch, eight were on d4T, and five were on ddl (didanosine, Videx/Videx EC). All of the 18 patients were on at least one of these three drugs.
Other reported toxicities included anaemia (16 patients), hepatitis or liver-related side-effects (nine patients), nail or skin discolouration (six patients) and kidney-related side-effects (three patients).
In addition, five experienced drug hypersensitivity prior to switching: four switched away from nevirapine (Viramune), and one from efavirenz.
There were also three patients with hyperlactataemia/lactic acidosis, all of whom were on ddl before switching.
Virological 'failure'
Virological 'failure' was defined as viral load rebound after reaching less than 50 copies/ml, or never reaching less than 50 copies/ml, and/or an increase in viral load. A total of 132 (30%) patients switched for this reason.
Of the 70 patients who had ever reached less than 50 copies/ml ('undetectable'), a third (24 patients) waited more than six months to switch after the first consistently detectable viral load measurement. The majority, however, (43%, 30 patients) switched less than four months after their first detectable viral load. Almost 60% (41 patients) had been on their prior combination for more than two years before switching.
Of the 62 patients who never reached 'undetectable', remarkably almost 30% (exact numbers not given) had been on their prior therapy for more than 78 weeks. About 20% had been on their therapy for 17-26 weeks before switching, and around 15% each had been on their therapy for between 26-39 weeks or 39-52 weeks before switching. Less than ten percent switched after either 12-17 or 52-65 weeks of therapy.
Although 72% (95 patients) had their therapy guided by resistance testing, one in four did not, despite clear guidance from BHIVA. Twelve patients (9%) switched while resistance resting was being done but before results were available and four patients (3%) had a sample stored for future resistance testing. However, fourteen patients (11%) were neither tested for resistance nor had a sample stored. Five of these fourteen had a viral load over 1000 copies/ml.
Clinic switching policy
The BHIVA guidelines recommend considering switching therapy due to loss of virological control when two viral load tests at least two weeks apart both show that viral load is above 400 copies/ml. Resistance testing is recommended at this point, but due to current technological limitations, viral loads must be over 1000 copies/ml for a reliable result.
Consequently, 77 (57%) of the treatment centres delay switching drugs until viral loads reach 1000 copies/ml. Another two centres (3%) report delaying switching drugs for other, unspecified reasons. Seventeen (13%) switch their patients' drugs regimes after a second viral load test is above 400 copies/ml, and eleven (8%) switch after a second viral load is above 50 copies/ml, the limit of detectability. Twenty-six centres (19%) were not sure, gave no answer, or had no preferred practice.
Therapeutic drug monitoring
BHIVA considers therapeutic drug monitoring (TDM) a niche investigation, and recommends it for particular scenarios, including drug interactions, liver impairment, pregnancy and adherence issues. A total of 59 centres (44%) use TDM only if drug interactions are suspected. Another 17 (13%) use TDM routinely if adherence problems are suspected. A further 25 (18%) never or rarely use TDM, and 33 centres (25%) did not answer, or gave other responses.
Cost of antiretrovirals
The latest BHIVA guidelines mention cost for the first time, and suggest that "the cost of the regimen should also be considered." Surprisingly, no respondent told BHIVA that cost was a main or major consideration in the choice of antiretroviral therapy. However, 72 centres (54%) said that they took cost into account; 59 (44%) said cost was not a consideration and three (2%) did not answer
Key messages
The key messages to clinics from BHIVA was that "clinical centres should reassess their practice so as to minimise delay before changing therapy in patients with virological failure", and that they should "ensure appropriate use of resistance testing."
BHIVA National Clinical Audit Report, BHIVA Autumn Conference, London, 2005.