Seventh International Conference on AIDS: Treatments for Opportunistic Infections

There was little truly new, practical treatment information presented in Florence at the Seventh International Conference on AIDS. The bulk of the research on opportunistic infections focused on the incidence and diagnosis of these diseases; most of the clinical information came from relatively small confirmatory studies of treatments already in use. An occasional poster presentation discussed the use of a new drug, or a new use for an old drug, but these were all very small studies from which confident conclusions cannot easily be drawn.

We present in this article a synopsis of the potentially relevant opportunistic infection treatment and prophylaxis news from Florence. The pneumocystis, toxoplasmosis and cryptococcal meningitis prophylaxis information is important for everyone with advanced HIV. (A comprehensive review of pneumocystis prophylaxis can be found in AIDS TREATMENT NEWS #114.) Unfortunately, providing a complete overview of the status of treatments for each opportunistic infection would require several issues of the newsletter. Instead, we hope this article will be of use to individuals who are already dealing with a given infection, and are familiar with the basic treatment information, but would like to know about potentially important treatment advances presented at the conference.

[Unless otherwise stated, the references in parenthesis (for example, "W. B. 2245") refer to the abstracts of the Seventh International Conference on AIDS; these abstracts, together with the program of the conference and subject and author indices, have been published as a set of three books. Four-digit numbers refer to poster presentations; smaller numbers refer to oral talks. The first letters (M, TU, W, TH, or F) refer to the day of the week; the second letter (A, B, C, or D) refers to one of four "tracks" of the conference: Track A, Basic Science; Track B, Clinical Science and Trials; Track C, Epidemiology and Prevention; or Track D, Social and Behavioral Science.

A note earlier in this issue tells how to purchase a copy of the conference abstracts, as well as tapes of the oral sessions.]

Pneumocystis Prophylaxis

Over fifteen studies compared two or more of the common drugs used for pneumocystis (PCP) prophylaxis, including aerosolized pentamidine, Septra (also called Bactrim, co- trimoxazole, or trimethoprim-sulfamethoxazole), and dapsone with or without pyrimethamine. Figuring out which of these drugs is the most effective for preventing pneumocystis is extremely important. However, none of the small studies presented at the conference will truly answer that question because of their design (chart reviews as opposed to controlled clinical trials) and/or the relatively small numbers of patients studied. Hopefully, the two large, on-going ACTG pneumocystis prophylaxis studies will provide reliable answers within the next year. (The ACTG, or AIDS Clinical Trials Group, is the primary AIDS research arm of the Federal government's National Institutes of Health). In the meantime, some useful information can be gleaned from the conference presentations.

Of the two studies which compared aerosolized pentamidine, Septra and dapsone, one found that Septra was the most effective in preventing pneumocystis, in spite of frequent side effects (M. B. 160); the second concluded that all three were equally effective, but that aerosolized pentamidine was more easily tolerated (W. B. 2245).

Five studies compared aerosolized pentamidine with Septra. Four found no significant difference in the occurrence of pneumocystis (W. B. 2236, W. B. 2247, W. B. 2251, W. B. 2224) while one concluded that Septra was a more effective prophylaxis (W. B. 2246). One study looked at the occurrence of both pneumocystis and cerebral toxoplasmosis and found two cases of each in the aerosolized pentamidine group and none in the Septra group (W. B. 2224).

A larger study of over one thousand people appeared to show that although both aerosolized pentamidine and Septra were effective in preventing pneumocystis, only aerosolized pentamidine was associated with a statistically significant increase in over-all survival time (W. B. 2236). We spoke with Richard Chaisson, M. D., the principal investigator from Johns Hopkins University. He explained that because the study was conducted before Septra was in common use, only a small percentage of the patients were taking it, possibly explaining the failure to achieve statistical difference in survival rates. Although Dr. Chaisson believes that Septra is a better first choice for pneumocystis prophylaxis in patients who can tolerate it, he feels that his study is important in demonstrating that aerosolized pentamidine remains a very useful drug for patients who cannot tolerate Septra or other systemic drugs. Dr. Chaisson emphasized that comparisons of the two drugs cannot be made from a study designed as this one was; such comparisons must await the
large, randomized ACTG trials.

Two studies compared aerosolized pentamidine with dapsone (W. B. 2228 and W. B. 2248). Neither showed a statistically significant difference in the occurrence of pneumocystis. However, one of the studies also looked at the occurrence of toxoplasmosis and found five cases in the aerosolized pentamidine group and none in the dapsone group (W. B. 2248). The authors of this study concluded that as a dual prophylactic regimen, dapsone may be superior. A small study compared patients receiving sulfadiazine and pyrimethamine as standard maintenance treatment for toxoplasmosis with a placebo control group from a recently completed study; they found that the toxoplasmosis maintenance treatment also prevented pneumocystis in these patients, again suggesting a possible dual prophylactic approach (W. B. 2218).

Different dosage schedules of Septra were compared in a study which looked at twice daily vs. three times weekly dosing. Both regimens were equally effective, but the thrice weekly dosing appeared to be more easily tolerated (W. B. 2199). Another group administered pentamidine by monthly intramuscular injections in a group of 89 patients; there were three cases of pneumocystis (W. B. 2227). Although there was no control group, this method of administering the drug appears to be approximately as effective as the other regimens currently in use. Important side effects included abscess formation at the injection site, glucose intolerance similar to diabetes and low blood pressure when standing up.

We spoke with Sam Bozzette, M. D., one of the leading pneumocystis researchers working with the ACTG, about this approach. In his experience, the occurrence of the sterile abscesses, which often require surgical intervention, makes intramuscular injections of pentamidine an unattractive alternative. In addition, the diabetes-like side effects of systemic pentamidine therapy can be quite serious. For rare patients who are unable or unwilling to regularly take pills or use aerosolized pentamidine, intramuscular injections may be a less than ideal but reasonable alternative.

What is the bottom line about pneumocystis prophylaxis? All the treatments currently in widespread use seem to have approximately the same effectiveness in preventing the disease. The systemic treatments like Septra and dapsone may well have the added advantage of preventing toxoplasmosis, as well as pneumocystis infections in organs other than the lungs. Even if the large ACTG studies do confirm that the systemic treatments are a superior first choice, aerosolized pentamidine will remain an important second-line prophylaxis with equal or slightly less effectiveness in preventing pneumocystis, but fewer side effects.

Pneumocystis Treatment

The usual treatments for acute pneumocystis include intravenous Septra or pentamidine. Alternatives to these drugs are important because serious side effects often limit their use. One regimen being tested is oral clindamycin plus primaquine. Two small studies using this combination confirm its usefulness and tolerability in mild to moderate pneumocystis, either as the initial therapy (TH. B. 42) or as salvage in patients who have failed or are intolerant of the other drugs (W. B. 2229).

566C80 is an experimental drug being studied as a treatment for pneumocystis and toxoplasmosis. A small pilot study reported on its effectiveness in the initial treatment of mild to moderate pneumocystis (W. B. 2239). 79% of the patients were successfully treated with 566C80. Side effects included rash, drug related fever, and an increased liver enzyme (alanine aminotransferase) in some patients. Six of the 34 patients experienced a recurrence of their pneumocystis within 6 months of completing treatment. The abstract does not state what kind of prophylaxis they were using.

ACT UP/Boston has been working with Burroughs Wellcome, the developers of 566C80, to make this unapproved compound available to patients who have failed or are intolerant of standard pneumocystis treatments. Burroughs Wellcome confirmed that the drug is available on a case-by-case basis for patients who are intolerant of all other treatment options and do not qualify for an existing study of 566C80. Although Burroughs Wellcome has not yet formalized an expanded access program, physicians may call the medical department at 800/722-9292 for information on compassionate access to 566C80 for the treatment of pneumocystis. For information on clinical trials of 566C80, call 1-800/TRIALS- A.

Finally, a Canadian study raised questions about the usefulness of adding steroids to standard treatment of moderate to severe pneumocystis. A consensus statement published in the New England Journal of Medicine in November 1990 recommends routine use of corticosteroids in certain cases of acute pneumocystis, based on the consensus panel's analysis of five studies (NEJM, November 22, 1990; volume 323, number 21, pages 1500-1504). The randomized, placebo controlled study of 78 patients presented at the Florence conference found no statistically significant difference in survival between the steroid treated and untreated groups.

Dr. Bozzette raised several issues to consider when interpreting the data from this study. First, although there was no statistically significant difference between the treated and untreated groups, there was a trend toward better survival in the treated group. Second, individuals who were not treated with steroids on the study, but became sicker, were eligible for steroid treatment off the study. Ten of the thirty-eight placebo patients did in fact receive steroids. Finally, the study was too small to detect a significant difference since the death rate even without steroids was low. When the data from this study are added to those from the other studies on which the consensus statement were based, the conclusion remains valid that steroid therapy improves survival in very sick patients.

In an article published in the daily conference newspaper, Sharon Walmsley, the principal investigator of the steroid study from the University of Toronto, warns that steroid therapy could worsen the course of lung disease which is misdiagnosed as pneumocystis but is actually tuberculosis or another viral or bacterial infection. Dr. Bozette agrees that patients should only be treated with steroids after a complete and definitive diagnostic work-up because of the dangers of steroid use with diseases other than pneumocystis. Although this particular study found no statistically significant benefit with the use of steroids, when the study is interpreted in the context of all the available data about steroid use in severe pneumocystis, use of the steroids appears to remain an important recommendation.

CMV Infections

Intravenous ganciclovir (DHPG) is the only FDA-approved treatment for CMV infections. Foscarnet, manufactured by Astra Pharmaceuticals, is an experimental treatment in the United States which is available on expanded access to patients who have failed or are intolerant of ganciclovir; Foscarnet is also an intravenous drug. Information was presented at the conference about both of these drugs, used separately and in combination, and on various methods of administering them (for example, orally, directly into the eye using an implant, or three times weekly rather than daily). In addition, minimal information on other experimental compounds such as oral FIAU and the monoclonal antibody TI-23 was also presented. Unfortunately, there were no clinical studies of compounds which we have not already discussed in past issues of AIDS TREATMENT NEWS.

There were, however, two studies on the treatment of CMV colitis and esophagitis, problems which have been inadequately studied in the past. Treatment of CMV esophagitis was effective in approximately 80 percent of the patients using foscarnet in two separate studies (W. B. 2262 and W. B. 2293), with resolution of symptoms and clearing of CMV organisms in biopsy tissue. In one study, patients were only treated for three weeks; three of the fifteen participants experienced a recurrence of their CMV within seven months. Two of these three were successfully retreated and remained asymptomatic after nine months (W. B. 2262). The second study utilized on-going maintenance treatment, with a 79% success rate, although three of the seven patients experienced complications of their esophagitis (strictures) and survival did not seem to be greatly improved with treatment.

The rate of improvement of CMV colitis symptoms and biopsies was similar to the esophageal improvements in the maintenance study, but less successful in the three week only treatment study. Only eleven of 22 patients experienced biopsy clearance of CMV after the three week treatment, and most continued to have diarrhea. The authors of the presentation attributed the diarrhea to the presence of other organisms in addition to the CMV.

An encouraging report on the treatment of a five year old girl who had failed sequential ganciclovir and foscarnet therapies described success using a combination of the two drugs (W. B. 2286). There was no significant toxicity with this combination, and it was effective in preserving her vision. The ACTG is currently implementing a study to compare the use of ganciclovir and foscarnet concurrently, as in this case, or on an alternating basis. There are theoretical advantages and disadvantages to each approach, including the hope for decreased toxicity and increased effectiveness. One of the disadvantages of concurrent therapy is the need to use two intravenous infusions each day. However, until effective oral drugs are readily available, we consider the development of creative uses of ganciclovir and foscarnet together and/or in alternation to be important.

An alternative approach to oral or intravenous therapy is the deposition of the drug directly into the eye. Ganciclovir can be administered by intravitreal injection to patients who cannot tolerate systemic therapy. Initial animal studies of intravitreal foscarnet have suggested a dose which is expected to be effective and minimally toxic (W. A. 66). This approach is obviously not an attractive one, but may be an important option for those who cannot tolerate systemic therapy.

An intraocular device which contains ganciclovir was placed in the eyes of eight patients with CMV retinitis (W. B. 2255). The device releases ganciclovir over 120 days and is placed surgically, using local anesthesia. This approach was effective, with stabilization or regression of the retinitis in all of the treated eyes. There were some complications, however, including two retinal detachments and a change in the ability to focus (astigmatism) in one patient. It is unclear from this small number of people whether the rate of retinal detachments was similar to, or higher than, the usual detachments seen with CMV retinitis (approximately 15-20%). This approach may offer an alternative to intravitreal injections for people who cannot tolerate systemic therapy. One serious concern in the case of CMV, however, is that it often causes disease in multiple organs, making systemic therapy preferable in those who can tolerate it.

We reported from last year's Sixth International Conference on AIDS in San Francisco on a small series of patients from Australia who used ganciclovir three times a week (see AIDS TREATMENT NEWS #108). This year, there was another report using the same regimen, i.e. 30 mg/kg/week divided into three doses, in contrast to the standard 5 to 6 mg/kg/day, five or seven days per week (W. B. 2354). This approach would be more convenient than a daily intravenous infusion.

We spoke with Stephen Follansbee, M. D., an Infectious Disease consultant in San Francisco, who said that he would be uncomfortable using this regimen. He explained that early pharmacokinetic studies with the 6 mg/kg dose demonstrated that there was no drug in the bloodstream at the end of 24 hours. Although he was unsure of similar data with the larger dose, from his experience he feels that it is necessary to use the drug seven days per week in order to adequately control CMV progression. In addition, the U. K. study showed an unexpectedly high rate of thrombocytopenia (decreased number of platelets, a type of blood cell), which Dr. Follansbee believes might indicate a toxicity from the higher dose.

William Buhles, D. V. M., Ph.D., from Syntex, the manufacturers of ganciclovir, explained that Syntex has never directly compared daily vs. thrice weekly ganciclovir therapy. He did express concern about the potential danger of infusing this higher dose, since some central nervous system toxicity has been seen with high plasma levels and rapid infusion of the drug. (Dr. Follansbee does not feel that central nervous system toxicity is a significant concern.) However, the retinitis progression rates seem similar to those seen with daily therapy. Dr. Buhles believes a controlled trial would be warranted if it was believed that enough benefit would be gained by patients with this dosing schedule. Syntex does not plan to run such a trial, however, because they are focusing on developing oral ganciclovir, about which they are very optimistic.

CMV cultures were shown to clear in the majority of patients treated with oral ganciclovir (W. B. 2309), suggesting that enough of the drug is being absorbed to be effective, and studies of this compound are ongoing for CMV retinitis (see AIDS TREATMENT NEWS #124). Syntex is also planning an important CMV prophylaxis study with oral ganciclovir. Unfortunately, a small study of oral foscarnet suggested that it is not adequately absorbed to treat CMV infections; adjusting the acidity of the stomach did not improve absorption (W. B. 2115).

FIAU is an oral compound which is being studied at increasing doses for safety and tolerance. Nausea and vomiting seem to be common side effects. Only a single study was presented on this compound, and no efficacy information was included (W. B. 2290).

The drugs discussed above are all antivirals with activity against CMV. Another approach to treating CMV involves the use of a concentrated solution of antibodies, similar to the antibodies produced by the healthy immune system to fight disease. Several studies have been conducted in other immunosuppressed patients, including bone marrow and renal transplant recipients, with a mixture that is rich in CMV antibodies. These studies suggest a positive prophylactic and treatment effect in these HIV negative patients. (For a good review article, refer to Snydman, D. R., Cytomegalovirus immunoglobulins in the prevention and treatment of cytomegalovirus disease. Reviews of Infectious Diseases, September-October 1990; volume 12, supplement 7.)

A more specific approach is to make a monoclonal antibody, i.e. many copies of a single antibody which is believed to be effective in neutralizing the virus. For the past two years, small studies of a monoclonal antibody called TI-23 have been reported on at the International Conference. Last year, an initial safety study suggested that TI-23 was safe. This year, there was a report of a study in 13 patients, eight with CMV retinitis and four with gastrointestinal CMV disease (W. B. 2291). Patients were treated with the intravenous formulation weekly for eleven weeks, and then continued treatment on a compassionate access basis if the treatment was effective, for up to 350 days. Unfortunately, retinitis progressed in four of the eight patients; gastrointestinal symptoms were reduced in two of the three who were treated on the extended compassionate basis.

While TI-23 does not yet seem to be more effective than ganciclovir or foscarnet, it may prove to be useful in combination with these other drugs. It may be possible to identify more effective antibodies than TI-23. Monoclonal antibody technology is important for AIDS treatment activists with an interest in CMV to follow closely.

Two case reports described patients with symptoms of chronic sinusitis who were found to have CMV infections of their sinuses (M. B. 2182). These patients responded to ganciclovir therapy and surgical drainage. Sinus cultures, unfortunately, were negative, and diagnosis required microscopic examination of biopsy tissue. It is important to consider that CMV may be a treatable cause of chronic sinus symptoms in some patients.

Toxoplasmosis

Common treatments for toxoplasmosis include pyrimethamine plus sulfadiazine or pyrimethamine plus clindamycin. An oral presentation of a study in 216 people with toxoplasmosis found no statistically significant difference in the effectiveness of these two combinations, although there was a trend towards fewer treatment failures with the sulfadiazine-pyrimethamine combination (W. B. 30). The types of side effects were different with the two regimens, occurring slightly more often with sulfadiazine.

A trial of 566C80 in eight patients who had failed or were intolerant of standard therapy showed improvement in 6 patients, stable disease in one, and progressive symptoms but a stable MRI in one (W. B. 31). See AIDS TREATMENT NEWS #123 for information on ongoing clinical trials with 566C80 for people with toxoplasmosis or call 1-800/TRIALSA.

Two very small studies of FDA-approved drugs which are not normally used for the treatment of toxoplasmosis, including doxycycline (M. B. 2027) and a combination of clindamycin and 5- fluorouricil (5-FU; M. B. 2003) were also presented. 5-FU is a cancer chemotherapy drug with significant side effects. In combination with clindamycin, it was claimed to be effective in nine out of ten cases of toxoplasmosis in eight patients who had failed or were intolerant of standard therapy. Doxycycline is an antibiotic which was associated with clinical and radiologic improvement in two patients.

Several studies of pyrimethamine for the prevention of toxoplasmosis are currently ongoing. Clindamycin was also being studied but has been dropped from the protocol because of an unexpectedly high rate of side effects. A study using twice weekly dapsone (100 mg) plus pyrimethamine (25 mg) for the prevention of pneumocystis and toxoplasmosis was presented (W. B. 2185). Two of 109 patients developed pneumocystis. Twenty-six individuals were antibody positive for toxoplasmosis, indicating that they have been exposed to the organism and are at risk of developing disease; none developed toxoplasmosis with a mean follow up of 16 months. The incidence of toxoplasmosis will also be assessed in the on-going ACTG pneumocystis prophylaxis trials. The development of agents which are effective in preventing both infections must be a high priority in current and future prophylaxis trials.

MAC (or MAI)

Rifabutin is currently being tested in clinical trials for the prevention of MAC and for treatment in combination with a multi-drug regimen. Results from a study of 90 patients who received rifabutin under a compassionate access program showed an over-all symptom improvement rate of 72% (W. B. 2300). Most of these people used rifabutin in combination with one or more other anti-mycobacterial drugs. Treatment and prophylaxis studies using rifabutin are on-going (800/TRIALSA). According to Beverly Wynne, Ph.D., rifabutin project leader and associate director of anti-infectives from Adria Labs, compassionate access to rifabutin will only be granted by Adria on a case by case basis. Physicians can call Dr. Wynne at 614/764-8307.

Clarithromycin, a new macrolide antibiotic, is also being studied as a treatment for MAC. A very small study tested combination therapy with INH, ethambutol and clofazimine plus placebo or clarithromycin for six weeks, followed by twenty-four weeks of clarithromycin plus rifabutin, followed by clarithromycin alone (W. B. 2358). Although there appeared to be a greater response in the clarithromycin group as opposed to the placebo group, a total of only nine patients were studied, so definitive conclusions cannot be drawn.

Finally, a test-tube study suggested that an experimental antibiotic called sparfloxacin may be as active against MAC as clarithromycin and azithromycin (W. A. 1051).

Cryptococcal Meningitis

Standard treatments for cryptococcal meningitis include amphotericin B and fluconazole. Small studies presented at the conference confirmed the utility of high dose fluconazole (800 mgs; W. B. 2287) and suggested that the addition of flucytosine to 400 mg fluconazole might increase it's effectiveness, in spite of some gastrointestinal and bone marrow toxicity (W. B. 2337).

A new form of amphotericin B, liposomal amphotericin, was discussed in two small studies (W. B. 2305 and W. B. 2177). As expected, this formulation was effective and more tolerable than standard amphotericin.

A prophylaxis study of cryptococcal meningitis using 100 mg fluconazole daily, compared to a historical control group, suggested that fluconazole is effective in preventing crypotococcal meningitis. One case of meningitis developed in 329 treated patients with fewer than 68 T-cells, whereas sixteen cases developed in 337 untreated historical controls (W. B. 2279). Unfortunately, there was no significant difference in the number of patients who developed histoplasmosis, indicating that fluconazole is not effective in preventing this fungal infection.

Cryptosporidiosis

Several agents have been proposed as treatments for cryptosporidiosis, and some patients have been reported in small studies and case reports to respond to each, although none has been found effective in all patients. Paromomycin (Humatin) was reported to improve the symptoms of cryptosporidiosis in 30 of 31 episodes in 22 patients (M. B. 2270). Five individuals experienced a relapse of symptoms; all five improved when retreated with paromomycin. An unconventional approach, using fluconazole, was reported to be effective in four patients. All four improved clinically, with some reduction in the amount of cryptosporidium present in the stool (M. B. 2199). It is questionable, however, whether or not this approach will be useful, as several physicians with whom we have spoken know of patients who are using fluconazole for other reasons, yet still have cryptosporidiosis.

Microsporidiosis

A report about five patients from the U. K. treated with albendazole, a broad spectrum antiparasitic drug which is used for various worm infections, demonstrated clinical improvement in all of them (W. B. 2265). This drug is approved in the U. K. for cestode infections (a type of worm) and is available in some other European countries. It is still investigational in the U. S. and is manufactured by Smith Kline &French. Metronidazole (Flagyl) demonstrated symptomatic benefit in fifteen of twenty- two patients with microsporidium (W. B. 2267). Twelve of these patients relapsed; most responded to re-treatment. Neither of these studies was large or controlled, so it is difficult to assess the true usefulness of these compounds. However, these encouraging preliminary results should certainly be followed up.

PML

A report of three patients with PML who were treated with multiple courses of the cancer chemotherapy drug ARA-C (cytosine arabinoside) demonstrated improved symptoms and decreased lesion size on MRI exam, but only after a minimum of three courses of treatment (M. B. 2037). A controlled study is being planned by these investigators who believe that negative results with ARA-C reported in the past may have resulted from an inadequate course of treatment.

Mycoplasma

Finally, Dr. Luc Montagnier attracted a significant amount of attention at the Sixth International Conference on AIDS in San Francisco last year with his announcement that a mycoplasma was an important co-factor in the development of AIDS. His research team presented results of an antibiotic treatment study which attempted to eradicate mycoplasma with the experimental antibiotic sparfloxacin (W. B. 2176). Unfortunately, markers of mycoplasma presence were not reduced with this treatment, which was potent in test-tube studies, nor were there any improvements in clinical or laboratory markers of the progression of HIV disease.