Anyone who has studied the official information sheets that detail the approved uses of each protease inhibitor will be aware that these drugs can have significant interactions with other prescribed drugs. But while there is quite a lot of information about these interactions, it is much harder to find reliable advice about possible problems if they are taken with recreational drugs.

The issue came to the fore in autumn 1996 when a London club-goer, Philip Kay, died after taking ecstasy. An autopsy found that he had unusually high levels of ecstasy in his blood, which may be at least partially explained by the fact that he was also taking ritonavir (Norvir). But ecstasy is not the only illegal drug that might interact dangerously with prescription drugs, and ritonavir is not the only anti-HIV drug that may interact with illegal drugs.

Shortcomings in research

Packet inserts for drugs such as the protease inhibitors contain extensive information about possible interactions with other prescribed drugs. In some cases this information has been derived from clinical tests in which people took both drugs and were studied through blood tests. But these well-studied interactions are the minority. More often, drug companies' researchers decide which interactions are likely based on the way their drug is metabolised.

Even for some prescribed drugs which are used by substantial numbers of people with HIV, such as the heroin-substitute methadone used in drug treatment programmes, surprisingly little reliable research on interactions with anti-HIV drugs has been conducted. Most of the guidelines on possible interactions are 'informed guesswork' based on the way the drugs in question are metabolised by liver enzymes.

It is hard to find any published guidance on possible interactions with illegal drugs. All the information that exists comes from the same process of 'informed guesswork', and from anecdotal information. It is made more complicated still by several additional problems:

• On rare occasions, deaths have been reported even after people have taken only a single ecstasy tablet and no other drugs that could cause interactions.

• Researchers are often unsure how some illegal drugs are processed in the body.

• Any available information may relate to the pure form of the drug, such as the chemical 3,4-methylenedioxymethamphetamine (MDMA; ecstasy), but 'street' drugs are rarely pure (an ecstasy tablet may well contain methylenedioxyethamphetamine [MDE] or methylenedioxyamphetamine [MDA] rather than MDMA). The same applies to amphetamines and lysergic acid diethylamide (LSD).

• The doses found in 'street' drugs are not controlled. For example, if a drug interaction led to a three-fold increase in blood levels of MDMA, the effect might be barely noticeable to someone who took a tablet containing very little MDMA, but very substantial for someone who took a tablet that consisted of pure MDMA.

• Research on these issues is sometimes hampered because the government and drug companies are anxious not to be seen as 'condoning' illegal drug use.

• Many warnings and recommendations concerning interactions with illegal drugs are based on isolated case reports, rather than prospective or controlled studies. As these describe individual cases, it may not be possible to extrapolate their conclusions to the wider population.

Several of these issues were relevant to the ritonavir / ecstasy case. No formal interaction studies had taken place, although knowledge of the metabolism of the two drugs made it likely that ritonavir would boost ecstasy levels two to three-fold. Philip Kay's partner is sure that Kay took no more than 2.5 ecstasy tablets, yet at post mortem he had blood levels of MDMA equivalent to taking 22 tablets - nearly a tenfold increase. One possibility is that these particular tablets contained unusually high amounts of MDMA.

Alternatively, if Kay was naturally a poor metaboliser of MDMA on account of his genetic make-up, that too could have had a bearing on the case. However, Abbott Laboratories has subsequently suggested that abnormally high peak levels of ritonavir during the early weeks of therapy may cause elevations in ecstasy levels if it is taken at this time. The Chelsea and Westminster Hospital has reported two admissions due to adverse reactions to ecstasy among people taking protease inhibitors.

On the same basis, it is possible that the non-nucleoside reverse transcriptase inhibitor (NNRTI) delavirdine (Rescriptor, which also inhibits the 3A4 enzyme, could also increase levels of some recreational drugs. Delavirdine is no longer licenced for HIV in the United Kingdom. Another NNRTI, nevirapine, has the opposite effect. It induces 3A4, so could lead to reduced levels of any recreational drugs that are metabolised by that enzyme.

Gamma hydroxybutyrate interactions

Levels of the recreational drug gamma hydroxybuytrate (GHB) may be increased to life-threatening levels if it is taken alongside a protease inhibitor, according to a single case report from Seattle. A 29 year-old man receiving treatment with ritonavir and saquinavir became unconscious after taking half a teaspoonful of GHB. He had taken a previous half teaspoonful of GHB and two ecstasy tablets in the preceding 24 hours. The man had not experienced adverse reactions to GHB or ecstasy prior to protease inhibitor treatment. Reporting doctors suggested that ritonavir and saquinavir slowed down metabolism of both drugs and caused the near-fatal reaction.

Usual symptoms of GHB overdose include vomiting, breathing problems, seizures, stupor and coma, and the dose normally associated with severe distress is greater than 50mg/kg of body weight. In this case the dose has been estimated at less than 10mg/kg body weight (Harrington 1999).

Methamphetamine interactions

Australian doctors have reported what they believe to be a fatal interaction between ritonavir and methamphetamine, commonly known as 'crystal' or 'crystal meth'. Methamphetamine is a popular recreational drug amongst gay men, especially in North America and Australia, but this is the first report of a potentially harmful interaction since the introduction of protease inhibitors more than four years ago, leading some to suggest that methamphetamine may not be responsible.

A 49 year-old Melbourne man taking ritonavir, soft gel saquinavir and d4T (stavudine, Zerit) was found dead the morning after injecting methamphetamine and sniffing amyl nitrite (poppers). A toxicology analysis showed that the dead man had methamphetamine levels of 0.5mg/l in his blood, a level seen in many people who have died of methamphetamine overdose.

Methamphetamine is metabolised by the liver enzyme 2D6, which is inhibited by ritonavir. The protease inhibitor could have slowed the metabolism of methamphetamine, thus causing the overdose.

The authors speculate that amyl nitrite use could also have contributed to the overdose, because amyl nitrite is metabolised to nitric oxide, another cytochrome P450 inhibitor (Hales 2000).

Methadone interactions

Interactions between AZT and methadone have been studied. Methadone increases levels of AZT by about twofold, so people taking both drugs need only take half the standard dose of AZT to get the same anti-HIV effect. The same applies to other opiate drugs.

There has been very little research into interactions between methadone and other nucleoside analogues such as ddI, ddC and 3TC (lamivudine, Epivir), although no problems have been reported with combining these drugs.

Merck is currently conducting a formal study of indinavir and methadone. The company suggests that on the basis of their liver metabolism, levels of both drugs might increase and so dose reductions may be necessary. The same may be true of saquinavir.

Nelfinavir reduces methadone levels by up to 40% but a US study of 34 individuals who started twice daily nelfinavir treatment whilst receiving methadone did not show any evidence of withdrawal effect (Hsyu 2000).

Of the NNRTIs, delavirdine is likely to substantially increase levels of methadone, while nevirapine is likely to decrease methadone levels. A study in Dublin reported that nevirapine reduced methadone levels by 46% within two to three weeks of commencing the NNRTI. Patients began to report opiate withdrawal symptoms eight to ten days after starting nevirapine, but it is not recommended that methadone dosage be increased at the same time as starting nevirapine. Instead, it may better to monitor withdrawal symptoms and increase the methadone dose by 10mg per day if withdrawal does begin to occur (Clarke 2000).

Efavirenz reduces exposure to methadone by about 60% within 24 hours of commencing efavirenz. This occurs because efavirenz speeds up, or induces, the metabolism of methadone. However, dosage adjustment should not take place immediately because the neurological side-effects of efavirenz may be mistaken for opiate withdrawal. As with nevirapine, close monitoring for withdrawal is recommended. Gradual 10 mg dose increases of methadone are advised if symptoms of withdrawal occur.

Ketoconazole (Nizoral) and itraconazole (Sporanox) may increase the level of methadone, and rifabutin (Mycobutin) may reduce it. Rifampicin (Rifadin / Rimactane) dramatically reduces levels of methadone, so patients have to have their methadone increased about three-fold if they are to avoid symptoms of withdrawal. It is important to educate these patients about the interaction. If they were to stop taking rifampicin while still taking three times the normal methadone dose, they could die from methadone overdose.

Practical advice

Needless to say, the safest solution is not to mix protease inhibitors and recreational drugs at all.

But if you do use illegal drugs, or plan to do so, it really is sensible to tell your doctor as this could affect your treatment decisions. Most HIV doctors will be quite happy to discuss drug use, not least because most have been treating HIV-positive injecting drug users for years.

Even so, you should not expect any easy answers to questions about possible interactions. No-one can predict with certainty how a prescribed drug and a recreational drug will interact for a given individual. For example, anecdotal accounts indicate that a significant number of people taking ritonavir are also using ecstasy or other drugs without experiencing any problems. But the case of Philip Kay highlights the possibility that some unlucky individuals may suffer potentially fatal interactions.

Some doctors recommend leaving as many hours as possible between a dose of a protease inhibitor and a drug such as ecstasy. There is no hard evidence that this will reduce the risk of an interaction, but it is unlikely to increase the risk either. It is not a good idea to skip doses of your protease inhibitor altogether, as this may make it easier for resistance to emerge.

Finally, do not overlook the ways in which illegal drugs may affect your lifestyle. It may be harder to remember to take treatments at the right times, or your eating habits may be disrupted, affecting drugs that should be taken with or without food.

Reference

Clarke S et al. Managing methadone and non-nucleoside reverse transcriptase inhibitors: guidelines for clinical practice. Seventh Conference on Retroviruses and Opportunistic Infections, San Francisco, abstract 88, 2000.

Hales G et al. Possible fatal interaction between protease inhibitors and methamphetamine. Antivir Ther 5: 19, 2000.

Harrington RD et al. Life-threatening interactions between HIV-1 protease inhibitors and the illicit drugs MDMA and gamma hydroxybuytrate. Arch Intern Med 159: 2221-2224, 1999.

Hsyu PH et al. Pharmacokinetic and pharmacodynamic interactions between nelfinavir and methadone. Seventh Conference on Retroviruses and Opportunistic Infections, San Francisco, abstract 87, 2000.