Novel Ideas and Therapies: 1997 Retrovirus Conference

CONFERENCE COVERAGE

Editor’s note: The following summaries represent a selection of papers from those presented at the Fourth Conference on Retroviruses and Opportunistic Infections held on January 22-26, 1997, in Washington, DC. It is important to recognize that many of these summaries are extracted only from the published abstract, and it is possible that some of the material presented at the conference may have differed.

The abstracts, as well as other information presented at the conference, are available on the internet at www.retroconference.org.—scd

Opportunistic Infections and Malignancies

Pneumocystis carinii pneumonia. A number of papers dealing with Pneumocystis carinii pneumonia (PCP) prophylaxis were presented, many of them addressing the problem of intolerance to trimethoprim-sulfamethoxazole (TMP-SMX).

In ACTG, 268 patients initiating PCP prophylaxis were randomized to TMP-SMX DS qd immediately or after first undergoing gradual dose escalation with TMP-SMX suspension (8 mg trimethoprim/40 mg sulfamethoxazole per mL). In the latter group, 1 cc was given daily for the first three days followed by 2 cc, 5 cc, 10 cc (each for 3 days), then 20 cc daily until the two-week visit. Gradual initiation of TMP-SMX was associated with a reduced incidence of treatment-limiting adverse effects in the first 12 weeks of administration. (Abstract #2)

It has been hypothesized that the high incidence of adverse reactions to the administration of TMP-SMX is the result of an inability to detoxify reactive metabolites of the sulfa component because of a reduction of levels of glutathione, the major intracellular antioxidant. Glutathione is a tripeptide whose limiting component is cysteine. Two hundred thirty-four patients were randomized to receive TMP-SMX (1 double-strength bid) alone or with nacetylcysteine (NAC), a cysteine prodrug (3 g po bid, 1 hour prior to TMP-SMX). Patients were followed for five months of prophylaxis, but the NAC was discontinued after two months. The incidence of discontinuation of TMP-SMX due to adverse events was 24% in those receiving TMP-SMX alone and 22% in those receiving TMP-SMX plus NAC. The study was terminated at interim analysis because of lack of efficacy. (Abstract #292)

Ironically, the improvement of the immune response in association with institution of potent antiretroviral therapy was suspected by one group to predispose to TMP-SMX intolerance. Five patients with previous hypersensitivity reactions to TMP-SMX who were now receiving this drug after "desensitization" had a recurrence 7-21 days after institution of protease inhibitor therapy. (Abstract #535)

When a patient develops an adverse drug reaction while receiving TMP-SMX, it is generally assumed that the sulfa component is the villain, but this may not always be true. For instance, a study of traveler’s diarrhea prophylaxis performed some years ago in which individuals received either TMP-SMX or trimethoprim alone, showed that many reactions could be attributed to the latter drug (DuPont HL, et al. N Engl J Med 1982; 307:841-844). Thus, it is of interest that a study in the corticosteroid-treated rat model of PCP found that trimethoprim adds no prophylactic benefit to the results achieved with sulfamethoxazole alone. (Abstract #289)

Despite the availability of a number of prophylactic recommendations, PCP has not disappeared. The reasons for the continued occurrence of this infection are several. The occurrence of PCP in 82 (20%) of 414 patients receiving prophylaxis was associated with a CD4 count less than 50, the use of regimens other than TMP-SMX, and non-compliance. In addition, however, seven (9%) of the breakthroughs occurred in patients with CD4 count greater than 50 who were on TMP-SMX and were believed to be compliant. This suggests that the development of resistance to TMP-SMX may have played a role in these cases. (Abstract #294) This is consistent with a study in which molecular analysis of the dihydropteroate synthase gene of Pneumocystis carinii recovered from AIDS patients with PCP found that prophylaxis breakthroughs were associated with mutations in this gene. (Abstract #3)

Adjunctive corticosteroid administration is a standard of care in patients with moderate to severe PCP, although some continue to worry about potential adverse effects resulting from the immunosuppressive properties of such therapy. A retrospective study of 173 PCP patients found that 53 (31%) were given adjunctive adrenal corticosteroid therapy. Life-table analysis found that there were no differences between those who received and did not receive corticosteroids with regard to time to CMV, MAC, cryptococcal meningitis, toxoplasmosis, Kaposi’s sarcoma, or death subsequent to treatment for PCP. (Abstract #293)

Candida infections. Candida resistance to fluconazole is being encountered with increasing frequency in many clinics, presenting often difficult management problems.

In a prospective study of 57 HIV-infected patients (median CD4 < 50) with oropharyngeal candidiasis, 200 mg fluconazole was given on the first day of therapy, followed by 100 mg daily for the next six days; in non-responders, the dose was doubled weekly until a response was achieved or until a maximum dose of 800 mg daily was reached. Clinical resolution was achieved in 72 (46%) of 155 episodes in one week, 50 (32%) in two weeks, 21 (14%) in three weeks, and 12 (8%) in four weeks or longer. Resistant yeast (MIC > 16 mcg/mL) were detected in 51 (33%) of the episodes. While a response to fluconazole therapy was nonetheless achieved in 151 (97%) episodes, the duration of therapy required was longer for those infected with a resistant yeast; resolution occurred within two weeks in 91 (88%) of 104 episodes with susceptible yeasts but in only 31 (61%) of 51 episodes with resistant yeast. (Abstract #324)

Itraconazole oral solution (100 mg bid), which is not yet available in the United States, was administered to 25 HIV-infected patients with oropharyngeal candidiasis that had failed to resolve after therapy with fluconazole. Clinical cure was achieved within 28 days in 40 (59%) of 60 patients, and clinical improvement was acheived in 50 (71%). A mycological response was seen by day 28 in only 17 (27%) of 62 patients. This may account for the fact that all 22 clinical responders who entered follow-up relapsed an average of 13 days after clinical cure. (Abstract #327)

Cryptococcal infection. The optimal dose of either fluconazole or flucytosine in the treatment of cryptococcal meningitis in AIDS has remained an unanswered question. The efficacy of varying doses of fluconazole with or without flucytosine was evaluated in 89 AIDS patients with first episode cryptococcal meningitis by the California Collaborative Treatment Group. A successful outcome (survival with negative CSF culture) after 10 weeks of therapy was observed in 11%, 37%, 62%, and 62% of those who received 800, 1200, 1600, and 2000 mg of fluconazole daily, respectively (P < 0.01). The addition of flucytosine (150 mg/kg/d) for the first four weeks improved the outcome at each dose of fluconazole; 10-week success rates with the addition of flucytosine were 75%, 87%, 69%, and 83% at each of the above doses (P < 0.01). This suggests that the optimal doses for wholly oral treatment of cryptococcal meningitis may be 1200 mg of fluconazole daily plus flucytosine 150 mg/kg per day in divided doses. (Abstract #5)

Increased intracranial pressure is a common accompaniment of cryptococcal meningitis, the management of which varies from center to center. Rapid and profound improvement in level of consciousness with resolution of headache, nausea, and emesis was reported in 10 HIV-infected patients with elevated intracranial pressure (mean, 44 cm) in whom external lumbar drains were placed for continuous CSF drainage. In eight of these patients, elevated pressure persisted with discontinuation of temporary drainage; all responded to the placement of lumbar-peritoneal shunts for permanent CSF diversion. (Abstract #331)

The value of monitoring serum and CSF cryptococcal antigen titers in evaluating the response to therapy of cryptococcal meningitis was studied in the context of an ACTG trial comparing fluconazole and itraconazole after initial therapy with amphotericin B. There was no correlation between baseline serum antigen titers or changes in serum titers with the outcome of therapy. While a high baseline CSF antigen titer was predictive of a persistently positive culture after two weeks of therapy (P = 0.0002), there was no correlation with clinical or mycological outcome at 10 weeks. There was a significant correlation between the occurrence of a decrease (geometric mean reduction or the group of at least 0.5 log10) and favorable mycological and clinical responses at 10 weeks. Thus, while measurement of the serum cryptococcal antigen titer is not useful in the management of AIDS patients with cryptococcal meningitis, monitoring of CSF antigen may have some modest benefit in some circumstances. (Abstract #7)

Bacterial infections. The importance of the relatively modest levels of neutropenia often observed in AIDS patients with regard to possible increased risk of bacterial infections has been debated, although a recent study from San Francisco General Hospital suggests that the risk is real. G-CSF and, to a lesser extent, GM-CSF have been widely used to reverse this neutropenia, albeit in the absence of data indicating clinical benefit.

Administration of a single subcutaneous 300 mcg dose of G-CSF to eight HIV-infected patients with CD4 counts less than 100 resulted in an 1800% increase in WBC and a 525% increase in oxidative burst activity of neutrophils while restoring reduced bacterial killing to the level of non-HIV-infected controls. (Abstract #432) This study, however, did not examine clinical benefit.

Some clinical benefit, however, was demonstrated in a randomized trial of 258 patients with actual neutrophil counts (ANC) of 750-1000/mcL. Subjects were randomized (1:2) either to an observational control or to receive G-CSF titrated to achieve an ANC of 2000-10,000/mcL for 24 weeks. In the unadjusted intent-to-treat analysis, G-CSF administration was associated with a 31% reduction in incidence of bacterial infections, which, however, did not reach statistical significance (P < 0.07), and with a 54% reduction in severe bacterial infections (P = 0.02). There was also a 45% reduction in hospital days (P = 0.02). However, the unblinded nature of the trial makes interpretation of differences in hospital days problematic. It is likely that patients would be more likely to be admitted and kept in the hospital if they were known to have neutropenia, independent of their clinical status. (Abstract #365)

A case-control study found that primary prophylaxis with TMP-SMX was associated with a decreased incidence of development of bacterial pneumonia, while secondary prophylaxis was not. (Abstract #295) Another retrospective analysis also found that TMP-SMX administration, as well as clarithromycin administration and a history of having received Hemophilus influenzae type B vaccine, were each associated with a reduction in the incidence of bacterial infections. (Abstract #364)

Mycobacterium avium complex infections. Five patients who developed acute mycobacterial lymphadenitis (3 due to Mycobacterium avium complex [MAC] and 2 unknown) in the absence of mycobacteremia within 1-3 weeks of beginning multidrug antiretroviral therapy, which included a protease inhibitor were described. (Abstract #351) Another group reported similar events in three additional patients, all of whom had associated fever and leukocytosis. (Abstract #352) These observations suggest that the immune enhancement seen with such antiretroviral therapy may cause an inflammatory response that unmasks previously asymptomatic mycobacterial lymphadenitis. (see also CMV.)

Gastrointestinal disease. Paromomycin has previously been reported, in a randomized trial, to provide some clinical and microbiological benefit in patients with cryptosporidiosis. In this study, 35 patients with cryptosporidiosis with a mean CD4 count of 38 were randomized to blindly receive either placebo or paromomycin (500 mg qid) for 21 days followed by open-label administration to all participants for another 21 days. There was no beneficial effect associated with paromomycin administration in this group of patients with advanced HIV infection. (Abstract #4)

Eighteen patients (mean CD4, 88) with chronic diarrhea despite therapy for an identified pathogen (9 patients) or with no identified pathogen were randomized to receive either thalidomide or placebo. The mean number of daily stools decreased in the former group from 4.6 to 2.0 (P = 0.005) with no significant change noted in the placebo group. Significant weight gain, however, was not observed. (Abstract #682)

Perhaps the optimal means of dealing with chronic gastrointestinal infections in AIDS patients is to improve their immunological status by the administration of effective antiretroviral therapy. Nine patients had resolution of chronic (5-18 months) diarrhea and gained a mean of 10 kg within 12 weeks of initiation of combination antiretroviral therapy, which included at least one protease inhibitor. (Abstract #688) Similarly, diarrhea resolved in 12 of 13 patients with chronic cryptosporidiosis or microsporidiosis after initiation of "triple therapy," which included a protease inhibitor. Cryptosporidia completely disappeared from the stool in nine of nine patients, and microsporidia disappeared in four of six. (Abstract #358)

Cytomegalovirus infections. A variety of new drugs or prodrugs active against cytomegalovirus (CMV) are entering clinical trials. For oral prophylaxis, however, the clinician only has ganciclovir, a drug that is not uniformly successful. The less than ideal bioavailability of ganciclovir when given orally suggests that monitoring of the plasma concentration of this drug may be useful. However, a study of 14 patients in France failed to find a correlation between the failure of oral ganciclovir prophylaxis and trough plasma concentrations of the drug. (Abstract #310)

Most patients with CMV retinitis have very low CD4 counts. Thus, a retrospective analysis of cases of CMV retinitis in patients who had been enrolled in various trials found that only one of 27 had a CD4 count at onset of infection higher than 50, and none had CD count higher than 100. This may, however, be changing. Five cases of CMV retinitis that developed within 4-8 weeks of initiating antiretroviral therapy with multiple drugs, including a protease inhibitor, were reported. Furthermore, their CD4 counts, which had been less than 85 in all prior to the initiation of that therapy, were higher than 200 at the time of diagnosis of CMV retinitis. (Abstract #353) In addition, eight of 210 patients developed CMV disease after a mean interval of 37 days following the initiation of similar antiretroviral therapy. (Abstract #354)

A group from Rennes reported four patients (CD4 count 3-100) who developed CMV retinitis with onset within the first four months of protease inhibitor therapy. This represented relapse in two. The mean increase in CD8 count in the four patients had been 953, while the CD4 increase was 97. The occurrence of retinitis in these subjects despite the massive CD8 increase could mean that CMV immunity had been permanently lost and that the CD8 increase did not contain cells with activity against CMV. (Abstract #315) Alternatively, it could mean that the CMV became apparent shortly after the initiation of effective antiretroviral therapy as the result of the return of effective anti-CMV cytoxicity with its resultant inflammatory response becoming apparent in the retina.

A study of 26 patients with CMV retinitis found that the total annual undiscounted billings per patient for inpatient and home care averaged $126,996 with pharmacy charges accounting for 52% of this total. (Abstract #307)

Placement of a ganciclovir implant in the vitreous has been demonstrated to be effective in the primary treatment of CMV retinitis; new evidence indicates that this approach is effective in the management of relapsed CMV retinitis. In a randomized treatment trial in patients with primary CMV retinitis, the median time to progression in 23 eyes in which relapse was observed was 181 days in implanted eyes and 55 days in 23 eyes of patients treated with IV ganciclovir (P = 0.001). These 46 eyes were then treated with the ganciclovir implant. The median time to progression in eyes receiving a second implant was 259 days, while it was 115 days in the eyes of patients previously treated with IV ganciclovir (P = 0.068). The overall incidence of endophthalmitis in 107 implanted eyes was 1.9%, and that of retinal detachment was 15.9%. Extraocular CMV disease occurred in 12 (16.7%) of 72 patients. Thus, a second ganciclovir implanted after relapse is as effective as the first and that an implant is also effective in individuals who relapse after treatment with IV ganciclovir. The high incidence of extraocular CMV disease in patients treated with the implant alone is, however, reason for concern. (Abstract #303) Nonetheless, a study from Munich found no apparent increase in extraocular CMV in patients treated with the implant alone, but they did report an increased incidence of CMV involvement of the other eye. (Abstract #312)

Foscarnet is an effective agent in the treatment of CMV disease, but maintenance of hydration is necessary to prevent renal complications associated with the use of this drug. It is generally recommended that this extra fluid be administered intravenously. However, it was demonstrated that oral hydration is as effective as intravenous hydration in preventing the nephrotoxicity associated with indication therapy with foscarnet. (Abstract #299)

Cidofovir was recently approved by the FDA for the treatment of CMV retinitis. Cidofovir is a nucleotide analog and, consequently, its activity is not affected by mutations in the thymidine kinase gene, a major means of resistance development to ganciclovir. Unfortunately, cidofovir resistance may result from mutations at other sites, some shared with ganciclovir. CMV isolates obtained from five patients after a mean of 16.4 weeks of primary treatment with cidofovir for CMV retinitis were completely susceptible in vitro to cidofovir, as well as to ganciclovir, foscarnet, and adefovir. Isolates were also obtained from the bloodstream of patients receiving cidofovir as second-line therapy; eight isolates obtained after a mean time of nine weeks on cidofovir were tested in vitro. Four were completely susceptible to all drugs tested, while the other four demonstrated moderately reduced susceptibility to cidofovir and high level resistance to ganciclovir; in two of the four, this phenotype was present prior to the start of cidofovir therapy. Studies with the other two isolates suggest that the reduced in vitro susceptibility to cidofovir (which was not reflected clinically) was the result of polymerase mutations that had been selected by prior ganciclovir therapy. (Abstract #304)

These findings were consistent with those of another study from the University of Minnesota. Twenty-one CMV isolates from 10 patients with CMV end organ disease who had been treated with ganciclovir (8) or ganciclovir followed by foscarnet (2) were screened for ganciclovir UL97 resistance mutations by restriction digest analysis and tested for phenotypic resistance. All isolates were susceptible to foscarnet. Thirteen (62%) of the isolates were resistant to ganciclovir, and four of these had decreased susceptibility to cidofovir, with three of the four being fully resistant. These four isolates had DNA polymerase mutations. Ten (77%) of the ganciclovir-resistant isolates had one or more ganciclovir-resistance UL97 mutations, but it was demonstrated that ganciclovir-resistant CMV strains may contain UL97 mutations, polymerase mutations, or mutations in both genes. Consistent with the results described above, ganciclovir therapy may select for DNA polymerase mutant strains of CMV that are cross-resistant to cidofovir. (Abstract #319)

Twelve patients with CMV encephalitis with focal findings were described. Eight had a brainstem encephalitis presenting as isolated cranial neuropathies; gadolinium enhancement of the brainstem was noted on MRI in all eight, and one of the eight also had ventriculoencephalitis. Four cases mimicked a CNS tumor. Eleven of the 12 patients improved with treatment. (Abstract #313)

Kaposi’s sarcoma-associated herpes virus. In vitro studies indicate that Kaposi’s sarcoma-associated herpes virus (KSHV) is susceptible to foscarnet (IC50 = 90 mcM) and to cidofovir (IC50 < 1 mcM) but not to acyclovir (IC50 = 70 mcM), while being intermediately susceptible to ganciclovir (IC50 < 1 mvcM). No clinical benefit from administration of antiviral therapy has been demonstrated, however. (Abstract #LB20)

JC virus infection/progressive multifocal leukoencephalopathy. Progressive multifocal leukoencephalopathy (PML), caused by JC virus (JCV), remains a relatively uncommon but devastating infection in patients with AIDS. JCV was detected by PCR in peripheral blood (plasma, total leukocytes, or B lymphocytes) in 36% of 50 AIDS patients and in only 4% of 50 controls. The virus was most frequently detected in B lymphocytes. Only one AIDS patient and no control, however, had detectable mRNA for the late VP1 gene, which is indicative of viral replication. Urine of AIDS patients, however, frequently contained detectable viral DNA and mRNA. The CSF of three AIDS patients with PML contained both viral DNA and mRNA. Thus, evidence of JCV replication is commonly detected in urine of AIDS patients and in CSF of AIDS patients with PML; while JCV can be detected frequently in the blood of AIDS patients, it is usually present only in a latent, non-replicative state. (Abstract #340)

Because of anecdotal reports, some patients with PML continue to be treated with cytosine arabinoside (ARA-C). It was planned to randomize 90 patients with PML in ACTG 243 to either antiretroviral therapy (ART) alone, ART plus intravenous ARA-C, or ART plus intrathecal ARA-C. The study was halted after randomization of 62 patients when interim analysis indicated that there was little chance that there was a significant difference between treatment arms. This should put the use of ARA-C in AIDS-associated PML to rest. (Abstract #8)

A retrospective comparison of 21 HIV-infected patients with PML and who received interferon-alpha (IFN-a) for at least three weeks to 32 similar historical controls found that the median survivals were, respectively, 325 days and 121 days (P = 0.0002). IFN-a use was associated with a significant delay in onset of memory impairment as well as speed and gait abnormalities. This observation should lead to a prospective randomized trial. (Abstract #341)

PML improved markedly in one patient after initiation of multiple drug therapy that included a protease inhibitor. (Abstract #357)

Next month’s conference coverage concludes with HIV infection.