GMHC Treatment Issues, Volume 10, Number 11 - November 1996 ----------------------------------------------------------- published by Gay Men's Health Crisis, Treatment Education, 129 West 20th Street, New York, NY 10011 Outrunning HIV to Protect Immune Defenses by Dave Gilden Last month we looked at the reasons people fail to respond to the new aggressive anti-HIV combination therapies. One of the striking features behind lack of response in the face of drugs that should have a tremendous impact on HIV replication is the immune system's inability to control the virus, leaving the drugs to do the entire job alone. But we also looked last month at various anti-HIV strategies based on further suppressing the immune system. These approaches stem from a philosophy that the central cause of AIDS lies in the progression of HIV infection and that the immune system response is on balance counterproductive. Many researchers are beginning to think that HIV establishes an equilibrium in the body based only on the susceptibility of target cells, and this equilibrium gradually breaks down. The chronic activation of the immune system to fight HIV helps the virus more than it hinders it by providing more and more target cells susceptible to HIV infection and by concentrating HIV in the lymph nodes, where those susceptible CD4 cells also converge. A recent theoretical paper went so far as to suggest that antiviral therapy plus immune suppressive chemotherapy that reduces the number of CD4 cells to slightly below the natural HIV-CD4 cell balance will cause the HIV infection to gradually wind down due to lack of sufficient target cells in which to reproduce.1 A possibly related observation is the experience of people who have their spleens removed (splenectomy) to correct HIV- related conditions (principally to improve low platelet counts). The spleen, which is packed with both macrophages and CD4 lymphocytes, is a major site of HIV replication. According to a French group who reported their observations in September,2 patients with HIV who had their spleens removed had only one-eighth the risk of death and one-quarter the risk of an AIDS-defining infection as a comparable group who still had their spleens. Splenectomy is not for the faint of heart: previous reports have indicated that taking out this piece of the immune system increases the risk of certain bacterial infections as well as carrying with it the various risks of infection and trauma inherent in a surgical procedure. Splenectomy also produces a misleading boost in CD4 counts in the blood3 while effecting little or no reduction in HIV levels.4 There might be a useful lesson from the performance of splenectomized patients, but most scientists think that suppressive therapy of this nature is highly questionable. The traditional view is that the immune system does play an active role in initially limiting the spread of HIV infection. Mario Clerici, M.D., of the Universitˆ degli Studi di Milano in Italy, notes, for example, that "the switch from NSI to SI HIV variant [a common evolution within the body that portends accelerated disease progression] occurs preferentially when CD4 counts fall below a critical threshold, suggesting that immune impairment is a prerequisite for cytopathic HIV variants to emerge." The pattern of immune response seems to indicate who is stable and who has progressive HIV disease. Long-term nonprogressors have CD4 T-cells that, when activated, produce IL-2 and other so-called TH-1 cytokines, which stimulate the cytotoxic CD8 T-lymphocytes (CTLs) that kill virus-infected cells. CD4 T-cells from those whose immune systems are declining tend to produce more IL-10 and other TH-2 cytokines that promote antibody production.5 In fact, the concentration and type of HIV in the body (less virulent NSI or more virulent SI) is closely correlated with the stimulated CD4 cells' IL-2 to IL-10 ratio. (Note that in a therapeutic context, IL-10 might actually help block HIV replication -- for the subtleties of this immune suppressive approach, see the October, 1996 Treatment Issues, pages 5-6.) It also seems that the quality of the CD8 cell defense shortly after infection with HIV influences people's prognosis: A reaction by CD8 lymphocytes that vigorously searches out relatively few HIV protein sequences leads to quicker disease progression, whereas progression is slower when the CTL response is more generalized, involving moderate proliferation by many CD8 lymphocyte subfamilies.6 A vicious cycle may be going on here, whereby a deficient immune response allows for more HIV, which in turn leads to further immune defects, such as a waning of anti-HIV CTL activity. If an increasingly improper immune response to HIV plays a central role in allowing disease progression, the trick would be to carefully manipulate that response to make it more effective or to restore it as immune defenses degrade in the later, symptomatic stages of HIV infection. "CD4 Lymphocyte Growth Factor" Of course, the greatest change in the immune system during HIV infection is the decimation of CD4 cells. An obvious immune-boosting therapy would be one that elevates CD4 T- lymphocyte proliferation so as to stabilize the CD4 count. And one of the most obvious substances to look at is IL-2, which causes replication of a broad number of white blood cells, including CD4 and CD8 lymphocytes, monocytes and natural killer cells. "IL-2 is just CD4 lymphocyte growth factor," claims Clifford Lane, M.D., of the National Institute of Allergy and Infectious Diseases (NIAID). When you give it to people and their CD4 counts go up, they don't get sick." Dr. Lane's group has just published the results of a 12-month controlled trial of IL-2 plus nucleoside analogs in 60 volunteers.7 The starting IL-2 dose was 18 million international units (IU) per day, administered as a continuous five-day infusion every two months. After the 12 month trial period, members of the control group, which had been receiving only nucleoside analogs, were permitted to receive IL-2 also. Follow-up data were reported out to 30 months. Starting with an average CD4 count of 427 cells/mm3, the IL-2 group gained about 37 cells per month, ending with an average count of over 900 after one year. During the follow-up period, this group's count stabilized (the IL-2 was only administered as needed to maintain the previous gains) while the CD4 counts of the initial control group caught up. The price of these gains was high: 90% of the volunteers on IL-2 at some point suffered severe fatigue, headache, diarrhea or other flu-like symptoms during and after an IL-2 infusion. By the end of the trial's initial phase, the IL-2 dose had been reduced to an average of about 8 million IU per day of infusion. According to Joseph Kovacs, M.D., the study's lead author, "We went for the maximum tolerated dose and then worked our way down." That optimum dose has yet to be determined and may differ for different people, in particular those with higher CD4 counts may require smaller amounts. One proposal to reduce the onerous nature of IL-2 therapy is to replace the five-day infusions with twice daily subcutaneous (under the skin) injections. Rather than be hospitalized, patients could administer IL-2 themselves at home. One study presented at the International AIDS Conference in Vancouver this summer found that after the first three months, CD4 counts had risen by 50% or 80% among the volunteers getting 15 million IU per day for five days every four or eight weeks, respectively.8 People on a lower dose (3 million IU/day) received less benefit, but since the starting CD4 counts were over 500 (i.e., near normal), it is questionable whether an 80% increase in CD4 cells is necessary. Just a slight gain or merely stabilization of CD4 count may be all that is necessary. The cost in terms of side effects remained high for subcutaneous IL-2, though less than with IV administration in other studies, and was dose- related. It is noteworthy that the one study that has directly compared intravenous and subcutaneous IL-2 found that side effects were distinctly fewer from subcutaneous administration, but the CD4 increases were a lot less, too.9 This study used a less active and now disfavored formulation, though, that joined the IL-2 molecules with polyethylene glycol for improved stability. While Dr. Lane's group works from the highest tolerable dose of IL-2 back down to a more benign therapeutic intervention, others are starting with lower doses in an effort to create a nontoxic regimen. Kendall Smith, M.D., and colleagues at Cornell Medical Center reported in September on what they call a "rational" IL-2 therapy.10 These investigators use comparatively small doses designed to limit the IL-2 to the high affinity IL-2 receptors present on activated CD4 T cells, B cells and a small subset of NK cells (a type of antiviral lymphocyte that works independently of CD4 cells). In this way, the lower affinity receptors present on most NK cells would be missed, leaving these cells in a quiescent state. The NK cells activated only by high-dose IL-2 produce many of the inflammatory compounds (such as tumor necrosis factor) that are implicated in IL-2's toxic effects. In the Cornell group's recent report, ten HIV-positive volunteers who received about 270,000 to 375,000 IU of IL-2 daily for six months as subcutaneous injections responded with an average CD4 count boost of 27 cells per month (from a baseline mean of 383 to a final count at six months of 543). This is a figure remarkably similar to the gains reported with high dose IL-2 at the NIH. Volunteers who received higher amounts of IL-2 in the Cornell study suffered from side effects and had their dose reduced. Those receiving lower doses had no benefit -- in fact their average CD4 count declined by 28 cells per month. The difference in CD4 cell response between the higher and lower doses in this trial was statistically significant, but only because of the surprising decline in the low-dose arm of the trial. Other trials using "low dose" IL-2 have not observed CD4 cell gains. Just last spring, the AIDS Research Alliance in Los Angeles reported that its 16-week trial of IL-2 (at 360,000 or 720,000 IU per day) plus gamma interferon detected no consistent changes in CD4 count.11 Granted, the Los Angeles trial involved advanced AIDS patients (average CD4 count of 47). Increases in NK cells, which are depressed by HIV infection, were dramatic in both the New York and Los Angeles studies. For the Research Alliance cohort on the combined IL-2/IFN-g regimens, the jump in NK cells seemed associated with decreases in HIV viral load. If this could be confirmed, it would be the first time a viral load decrease could be effected by a therapy involving IL-2. The other IL-2 trials have found no decrease in HIV levels, or even a transitory increase in people not on anti-HIV medications. For this reason, IL-2 trials have required participants to be on anti- HIV therapy with approved nucleoside analogs. The lack of viral load reduction with IL-2 can be taken as a sign that the IL-2's expansion of immune system cells is not creating any functional improvement. But it should be noted that a substantial increase in CD4 cells without any enhancement of the anti-HIV defenses should lead to permanently heightened levels of HIV, since the virus' "food supply" -- the proliferating CD4 T-cells -- would be increased. (This could be what is happening when IL-2 is administered to people with actual AIDS: In trial participants with CD4 counts under 200, permanent increases in viral load with little if any CD4 cell count improvement has occurred.) Besides a reduction in HIV levels, another definitive sign that the increased cell counts due to IL-2 are actually doing some good would be a reduction in the number of opportunistic infections contracted by people receiving the cytokine. So far, the only data available is contained in a small "meta- analysis" combining three placebo-controlled high dose IL-2 studies with a total of 157 participants, all of whom received open-label IL-2 after the first year. Since the start of each study, there were six OIs in the IL-2 cohorts and 14 in the placebo groups. This difference is not statistically significant, and the analysis is very vague in any case since the individuals under observation were all on IL-2 for varying lengths of time. Dr. Lane presented this very loose data set on November 3 at the Third International Congress on Drug Therapy in HIV Infection (in Birmingham, UK). He also claimed that all the opportunistic infections occurred at the expected CD4 counts, although he did not present a specific analysis of the correlation between OIs and CD4 counts. According to Dr. Lane, a definitive study of IL-2 documenting disease progression would require as many as 3,000 people with initially asymptomatic HIV infection who would be followed for five years. The total cost could be upwards of $50 million. No one has come forth as yet with this kind of commitment. Although the Chiron Corporation spent at least that much to obtain FDA approval for its Proleukin brand of IL-2 as treatment for a type of kidney cancer, corporate sources say that it would be "very difficult to pull off" this kind of additional investment for HIV. Considering that large clinical endpoint trials have already been successfully completed, such as the one comparing AZT, ddI, AZT/ddI and AZT/ddC (ACTG 175) or saquinavir, ddC and ddC/saquinavir (Roche NV14256), it seems difficult to comprehend why the same cannot be done for IL-2. Then again, if the optimum dose, dosing schedule and route of administration for IL-2 has not been settled, it seems that all the preparatory work for a large clinical endpoint trial has not yet been done. But there are nevertheless late reports of progress, with discussions continuing at both the level of AIDS Clinical Trials Group and internationally to conduct the clinical endpoint trial using subcutaneous IL-2 administration in HIV- positive volunteers with CD4 counts over 400. Adding a Little HAART Laboratory tests of immune cell proliferation do indicate that the additional cells created under IL-2's influence are responsive to foreign antigens, but specific lost immune responses, such as those to vaccines, are only occasionally recovered. As indicated by the lack of decrease in plasma HIV levels, people's immune control of HIV is not restored to the way it was just after primary infection even though their nominal, IL-2-influenced CD4 counts may be normal or even higher than normal. IL-2 may help to ensure stability and even reinforcement of available immunity by increasing the sheer numbers and concentration of existing cells. The new naive CD4 T-cells needed to expand the immune system's "repertoire" of responses to invading pathogens do not seem to appear under the influence of IL-2, perhaps because HIV damages the thymus, where many naive cells mature. Still, over the long- run IL-2 could expand any rare CD4 cells that already exist to produce detectable responses to antigens where previously none existed. The chances of this occurring would be increased if HIV levels were suppressed to the maximum possible using the new highly active antiretroviral therapy (HAART) drug combinations. The less HIV, the more the existing cells will be able to replicate. Conversely, IL-2 could reinforce HAART by maximizing the increase in CD4 counts. Ronald Mitsuyasu, M.D., at University of California Los Angeles is now planning ACTG 328, a 150-person trial with IL- 2 plus the protease inhibitor indinavir and any two nucleoside analogs (one has to be new to the individual trial participant). This trial's population will have advanced disease, with CD4 counts of 50 to 300, but no prior protease inhibitor treatment. If 12 weeks of the triple drug combination reduces their plasma viral loads to below 5,000, participants will be assigned to receive continued HAART plus either intermittent cycles of intravenous IL-2, intermittent subcutaneous IL-2 or nothing. Therapy will switch to subcutaneous administration after three cycles for those in the IV IL-2 group who are responding to therapy. The trial's goal is to monitor the resulting immunological and viral load changes, not prove a benefit in terms of disease progression. The latter would require a significantly larger trial enrollment. If IL-2 can be made to work in an advanced population, the required size for a trial generating disease progression data would be much smaller than the one discussed above for earlier patients -- only about 300 participants would be required, Dr. Mitsuyasu says. There already is some indication that HAART and IL-2 reinforce each other. A study chaired by Judy Falloon, M.D., of the National Institute of Allergy and Infectious Diseases has reported one-year observational data on 24 people who had previously failed IL-2 therapy and had CD4 counts less than 300. For 14 weeks, participants received either intermittent IL-2 plus daily indinavir (600 mg four times a day), intermittent IL-2 plus the indinavir only on days when actually receiving the IL-2, or daily indinavir alone. After 12 weeks everyone received IL-2, indinavir and whatever other antiviral drugs they requested. Baseline CD4 count averaged 165 and the average initial viral load was about 100,000. By 14 weeks, the IL-2 plus daily indinavir group witnessed their average CD4 count increase from 186 to 382 but their viral load was little changed. In contrast, the daily indinavir monotherapy group saw their CD4 counts increase from 144 to only 229, while mean viral load fell about five-fold. (Little improvement was observed in the intermittent IL- 2/intermittent indinavir group.) At one year, the CD4 count for the total group (all on open-label IL-2/indinavir therapy since week 14) averaged 433, while the overall average viral load had been kept down to 37,000. It is difficult to say in this trial what therapy provided what benefit, but clearly, adding antiviral drugs created a response to IL-2 where it previously did not exist. The IL-2, meanwhile, may have helped boost CD4 counts more than the average 100 to 150 cell gain usually seen with HAART combinations alone. There is a hint here that relieving the pressure from HIV, now possible for the first time, will allow immune therapies to work, aiding the immune system to provide some backup for antiviral drug therapy. Such support could help clear the virus more quickly and thoroughly. A functional immune defense against HIV, even if only partial, would help avoid the drug resistance and breakthrough HIV infection problems that frequently arise in patients now relying only on HAART. Before IL-2 can be combined freely with HAART, however, many basic questions once again have yet to be answered. The toxic nature of IL-2's inflammatory effects will make high doses of the agent difficult for most people to tolerate. HAART should allow IL-2 doses to be dropped to a greater extent than would less effective anti-HIV therapies, but since the optimum CD4 boost has yet to be ascertained, the amount of IL-2 to be used with HAART cannot be settled. Nor can the scheduling of IL-2 administration: One of the hopes expressed by proponents of high dose IL-2 is that only a few bimonthly cycles will prove necessary to achieve acceptable CD4 counts with continued highly suppressive antiviral cover. After that, administration will be needed at most only a few times a year to keep the CD4 count stable. Next month Treatment Issues will look at strategies designed to improve the immune system's specifically anti-HIV activity. 1 De Boer RJ and Boucher CAB. Proceeding of the Royal Society of London. B. July 22, 1996; 263(1372):899-905. 2 Morlat P et al. AIDS. September, 1996; 10(10):1170-2. 3 Zurlo JJ et al. Clinical Infectious Diseases. April, 1995; 20(4):768-71. 4 Henry K et al. Third Conference on Retroviruses and Opportunistic Infections. January, 1996; (abstract 245). 5 Clerici M et al. AIDS Research and Human Retroviruses. 1996; 12(11):1053-61. 6 Fauci AS. Third Conference on Retroviruses and Opportunistic Infections. January, 1996; (abstract S2). 7 Kovacs JA et al. New England Journal of Medicine. Oct. 31, 1996; 335(18):1350-6. 8 Davey RT et al. XI International Conference on AIDS. July 7-11, 1996; (abstract We.B.290). 9 Carr A et al. XI International Conference on AIDS. July 7-11, 1996; (abstract We.B.292). 10 Jacobson EL et al. Procceedings of the National Academy of Science USA. Sept., 1996; 93:10405-10. 11 Hale P. Searchlight. Spring, 1996; pages 10-15. Responding to CMV Neurologic Infections by Theo Smart CMV infections of the central nervous system (brain and spinal cord) could be the most under-recognized neurologic complication of AIDS, according to researchers such as J. Allen McCutchan, M.D., of the University of California, San Diego. Usually by the time CMV neurologic disease is diagnosed, most cases are unresponsive to treatment, and death comes within weeks. CMV CNS disease may be a common cause of death in patients with CMV retinitis, particularly, says Dr. McCutchan, in those who "go blind, give up, and withdraw without being aggressively diagnosed." Prospects for earlier diagnosis and improved treatment are increasing, though, as tests that measure CMV DNA in the cerebrospinal fluid become more widely used, and new treatments for CMV become available. CMV can cause a number of distinct neurologic syndromes (in addition to retinal infections). These include mononeuropathy multiplex (inflammation or disease involving individual nerves in unrelated portions of the body), myelitis/polyradiculopathy (inflammation of a large number of spinal nerves accompanied by pain, muscle wasting and paralysis) and two forms of encephalitis (or brain disease): diffuse micronodular encephalitis (DME) and ventriculoencephalitis (VE). Mononeuropathy multiplex occurs when CMV infects the cranial and peripheral nerves resulting in face, wrist and foot palsies (for example, drooping muscles in the face similar to what occurs in people who have suffered a stroke). Mild forms that resolve spontaneously often strike patients fairly early in the course of HIV disease. Polyradiculopathy is strictly a late-stage AIDS event. It starts out as weakness or pain in the legs or back, but within a couple of weeks can progress to paralysis of the lower extremities and loss of bladder control leading to an inability to urinate or loss of anal- sphincter control. Polyradiculopathy or myelopathy may accompany or lead to CMV encephalitis. The two CMV-encephalitic syndromes are distinguished primarily by the site of infection within the CNS. For both syndromes, symptoms of dementia, delirium, confusion, apathy and lethargy are common.1,2 Some symptoms of DME depend upon the location of the nodular lesions (as is often the case in toxoplasmosis, CNS lymphoma and PML). Patients with VE, meanwhile, often experience cranial nerve palsies and nystagmus (rhythmic involuntary eye movement). Both conditions are rapidly fatal. In one study of patients with VE, the median survival was only five weeks,3 while in another study those diagnosed with DME had a slightly longer median survival of 8.5 weeks.4 The actual rate of CMV infection of CNS is unclear. One two- year study reported that CMV encephalitis was found in only 1% of patients with advanced AIDS.5 This figure may accurately reflect the rate of diagnosis, but not the rate of incidence. In neurological autopsy studies, CMV infection of the CNS is found in between 20% and 28% of the brains studied.6,7 Some researchers believe that the rate of CMV neurologic disease is increasing as people live longer in an immune-compromised state. Others, such as David Clifford of Washington University and the Neurologic Subcommittee of the AIDS Clinical Trial Group, believe that there merely may be an increase in diagnosis: "Some of the cases we believed were HIV encephalopathy may actually have been caused by CMV -- there has been misclassification over the years. We've only known how to diagnose CMV encephalopathy over the last few years." Other rates of incidence have been measured by the University of California, San Diego -- the site of the HIV Neurobehavioral Research Center funded by the National Institute of Mental Health. According to Dr. McCutchan, the site has been able to prospectively conduct neurocognitive tests on AIDS patients that come to the university's AIDS clinic, and then frequently perform neurological autopsy in the event of death. In one recent study, researchers from the center calculated that autopsy-confirmed CMV encephalitis occurs in about 42% of patients with CMV retinitis.8 The rate was almost 75% in more advanced cases of retinitis. Meanwhile, it was very uncommon in people without CMV retinitis. Not all of the CMV neurologic disease detected upon autopsy in this study was rapidly progressive or caused the patients' death, but there is substantial evidence suggesting that patients are adversely affected by these undiagnosed infections. One study by Dr. McCutchan notes that even less advanced CMV infections can affect neurocognitive function. His team compared 16 patients with newly diagnosed CMV retinitis (but no evidence of outright dementia) to 322 matched controls with similar CD4 cell counts, age and education. They found neuropsychologic impairment (most commonly attention and verbal deficits) in a third of the patients with newly diagnosed CMV retinitis "caused by CMV rather than HIV." Evidence of physical harm to the central nervous system was detectable by MRI brain scans approximately twice as often in the patients with retinitis than in the matched controls.9 Dr. Clifford, however, does not agree with the methodology used in the report and says that there is a need for a "more focused and conservative study. I think this was a seriously flawed study because patients were sicker than the control group since they had an opportunistic infection and these were very mild neurocognitive deficits that could have been affected by the fever associated with CMV infection as well as visual problems associated with CMV retinitis." In another study from San Diego's Neurobehavioral Center, though, researchers reported that the risk of cognitive impairment associated with CMV encephalopathy was seven-fold higher than that associated with HIV encephalopathy.10 The authors concluded that "unrecognized CMV encephalitis may be an important cause of both mild and severe neurocognitive impairment in patients with advanced HIV infection." Recognizing the Infection In aggressive cases, the onset of symptoms is so abrupt, and the progression so rapid, that the encephalopathy may kill the patient within the time it takes clinicians to distinguish it from other AIDS-related neurologic conditions. There are some distinctive features of CMV neurologic disease, though. For example, CMV encephalitis occurs very late in the course of HIV disease, while signs of the most common alternate diagnosis, HIV encephalopathy, appear earlier and more gradually. Other hallmarks of CMV CNS infection, such as electrolyte abnormalities, pleocytosis (an abnormally high level of white blood cells in the CSF) and contrast enhancement of the brain's ventricles on MRI brain scans suggest CMV infection, particularly in the presence of CMV retinitis. These markers allow for a presumptive, but not a definitive diagnosis, since they also may occur in other neurologic conditions as well. Until recently, finding proof of active CMV infection in the CNS has been difficult without conducting a brain biopsy. Even though the CSF is believed to be the primary route through which CMV enters the CNS (particularly in the case of VE and polyradiculopathy), the virus can only rarely be cultured from the CSF.11 "The one thing that has helped is the CMV DNA PCR test in the [cerebral] spinal fluid," notes Dr. McCutchan. Some studies report that such assays are dramatically more sensitive than culturing, and that the virus is usually only detected in cases of active CMV CNS infection.12 One study by Achim et al. found that the test was not as specific as reported by other groups, since CMV DNA was detected by PCR in the CSF of 58% of the patients without CMV-associated neurological conditions.13 Lack of experience using the new tests may explain some of these contrasting results. Also, this was a study of autopsy specimens, and McCutchan points out that there is a strong possibility that the virus could have entered the CSF post-mortem. "I am very suspicious that things can change after death. I think we need to look at the CSF during life." It stands to reason that even if CMV DNA can be found in the CSF of people who do not have clinical CMV CNS disease, quantitative assays may provide a better indication of who is at risk for disease. A German study suggests that CMV DNA loads in the CNS were ten- to 1,000-fold higher in patients with CMV neurologic disease than in those without.14 Researchers from Roche Molecular Systems and University of California San Francisco reported at September's ICAAC that CMV DNA was detected in 24 of 26 patients with polyradiculopathy, and that levels were consistently above 1,000 CMV DNA copies per ml (except in one case of a patient on anti-CMV therapy).15 CMV DNA levels were generally under 100 copies per ml of CSF in patients with other CMV neurologic conditions, though. Another group of researchers recently published data correlating the presence of CMV DNA in the CSF of all but one of 13 patients with CMV CNS disease, while no CMV DNA was found in control subjects. High levels of CMV DNA (1,000 copies per 8 ml) were associated with severe disease, VE in particular, and uniformly shorter survival.16 Treatment Options There have been no prospectively controlled studies of the treatment of CMV CNS disease, and the case reports are scant. Thus, there are no clear data on what is the best drug or dose to use or the duration of treatment. There are case reports of clinical responses to ganciclovir or foscarnet, but there is reason to believe that standard treatments of CMV retinitis may be inadequate as therapy for CMV infections in the brain. Dr. McCutchan compared case studies of treated patients with polyradiculopathy to those of untreated patients, and he found that half of the treated patients survived for a median of eleven weeks while all seven untreated patients died within four weeks. It is hard to construe less than two months additional survival as a great success, though. Also, there have been a number of reports that CMV encephalitis developing in patients while receiving maintenance therapy for CMV retinitis. Researchers from Australia have reported on the development of VE in one patient on ganciclovir who then failed to respond to treatment with foscarnet.17 Another group described two cases of CMV encephalitis emerging in patients while on ganciclovir, who then failed to respond to the higher ganciclovir doses used for induction therapy.18 Yet another team noted five cases of CMV encephalitis that failed to respond to ganciclovir therapy, even though retinitis was stable in four of these same patients.19 Some of the treatment failure may be due to drug resistance. One team reported on culturing ganciclovir-resistant virus from a patient who subsequently developed polyradiculopathy while on ganciclovir therapy.20 Other researchers published a case of polyradiculopathy that was caused by a ganciclovir- resistant CMV, though the patient eventually responded to foscarnet and ganciclovir combination therapy.21 In addition, there are little data to suggest that either ganciclovir or foscarnet is absorbed into the central nervous system. In early pharmacokinetic studies, the level of ganciclovir in the CSF was found to be substantially lower than in the plasma.22 On the other hand, the blood-brain- barrier may be compromised by CMV infection of the CSF. Lawrence Drew, M.D., of the University of California San Francisco, presented an abstract at the Second National Conference on Human Retroviruses with data demonstrating that ganciclovir, with or without foscarnet, can lower CMV load in the CSF. Using the CMV bDNA assay, his team reported that treatment had antiviral activity in the CSF in seven out of eight patients with CMV neurologic disease.23 Dr. Drew concluded that the failure of treatment in these patients must have been due to some other factor such as the "advanced degree of disease." Dr. McCutchan has another possible explanation for treatment failure, "The pathology suggests that some of the damage is irreversible. When infection in the nervous system is diffuse (as in DME), there may be recovery on treatment, but when the infection is in the retina or the nerve roots (as in polyradiculopathy or VE), there's no recovery of necrotized [dead] nerves. Early neuropathological autopsy studies by Dr. Clayton Wiley show massive brain destruction, as CMV has burnt past [the layers of the central nervous system] until you reach normal looking brain, full of viral antigens. There's not much active CMV until you reach the active burning at the edges of tissue damage and heating in the next level full of virus. In retinitis studies, even after treatment is initiated, this 'fire' always moves just a little further." In this scenario, only earlier detection and treatment may advert permanent nerve damage. More hopeful data were presented by French researchers at the Third Conference on Human Retroviruses last January.24 The team reported rapid clinical improvement (within 10 to 13 days) in all of four patients with CMV neurological disorders (three cases of encephalopathy and one polyradiculopathy) on a combination regimen of ganciclovir at 5 mg/kg of body weight every 12 hours, and foscarnet 60 mg/kg every eight hours as induction therapy, followed by ganciclovir at 5 mg/kg, and foscarnet 90 mg/kg every day for maintenance. CMV DNA became undetectable in both of two patients within whom this was evaluated. No one stopped treatment due to toxicity, although one case of anemia and one case of leucopenia were observed. These results provide additional justification for ACTG 305, a six month multicenter study of high dose, aggressive combination foscarnet/ganciclovir therapy in at least 30 patients with recently diagnosed encephalopathy or polyradiculopathy. Participants will be treated for 28 days with foscarnet 90 mg/kg and ganciclovir 5 mg/kg both twice daily followed by maintenance therapy with the same doses, but on a once-a-day basis. If they can tolerate it, those with prior experience on ganciclovir will receive higher ganciclovir doses (7.5 mg/kg twice a day for induction therapy, and 10 mg/kg once a day for maintenance therapy). Up to ten patients who cannot tolerate either agent will be treated with the alternate drug as a monotherapy. If patients fail on maintenance therapy, higher induction therapy doses will be reinitiated. ACTG 305 seeks to answer many of the nagging questions surrounding CMV CNS disease. The study will evaluate both quantitative CMV DNA PCR and CMV bDNA as assays for disease severity, progression and response to therapy. The study also will investigate the role of drug resistance in the evolution of CMV CNS disease and in treatment failure. Finally, researchers will try to perform as many brain autopsies as possible on those patients who die during the study to determine the level of drug that actually penetrates the brain. According to Dr. Clifford, the study should be open by the time Treatment Issues goes to press. The Next Frontier It is hard to predict what effect the recent improvement in the treatment of CMV-retinitis will have upon the development of CMV-related neurological syndromes. The addition of cidofovir to the approved anti-CMV armamentarium may increase the options of treatment. Lobucavir, adefovir, and other new compounds (from Gilead and Glaxo Wellcome) are waiting in the wings. It is unclear whether the use of oral ganciclovir will decrease the rate of extra-ocular disease or merely engender resistance by supplying suboptimal levels of the drug. Finally, the growing use of localized therapies for CMV- retinitis that do not treat extra-ocular CMV disease, such as the ganciclovir implants, may allow the infection to spread to the CNS more often. "Local therapy worries me the most," says Dr. McCutchan. "We are doing ganciclovir implants and now cidofovir intraocular injections at our site, but we don't have the funding to prospectively follow these patients to see whether there is an increase in the rate of CMV disease in the CNS." Even if the use of localized therapy does increase the incidence of CMV encephalopathy or polyradiculopathy, Dr. McCutchan believes that with viral load monitoring of the CSF and an increased awareness of the conditions, the prospects for treating CMV CNS disease are improving. "One of the reasons to look at these endstage diseases is that we're going to get better and better at bringing people back from the brink. Of course, what we'd really like to do is to keep people from ever reaching that brink." 1 Holland NR et al. Neurology. March, 1994; 44(3 Pt 1):507-14. 2 Kalayjian RC et al. Medicine (Baltimore). March, 1993; 72(2):67-77. 3 Kalayjian, RC et al. Ibid. 4 Holland NR et al. Op cit. 5 Gallant JE et al. Journal of Infectious Diseases. December, 1992; 166(6):1223-7. 6 Morgello S et al. Human Pathology. March, 1987; 18(3):289-97. 7 Cinque P et al. Scandinavian Journal of Infectious Diseases-Supplementum. 1995; 99:92-4. 8 Bylsma SS et al. Archives of Ophthalmology. January, 1995; 113(1):89-95. 9 McCutchan JA et al. Second National Conference on Human Retroviruses and Related Infections. January 29-February 2, 1995; 107(abstract 289). 10 Ellis R et al. Second National Conference on Human Retroviruses and Related Infections. January 29-February 2, 1995; 107(abstract 288). 11 So YT, Olney RK. Annals of Neurology. January, 1994; 35(1):53-8. 12 Cinque P et al. Op cit. 13 Achim CL et al. Journal of Infectious Diseases. March, 1994; 169(3):623-7. 14 Kuhn JE et al. Journal of Medical Virology. September, 1995; 47(1):70-82. 15 Long CM et al. Thirty-Sixth Interscience Conference on Antimicrobial Agents and Chemotherapy, September 15-18, 1996; (abstract H033). 16 Arribas JR et al. Journal of Infectious Diseases. August, 1995; 172(2):527-31. 17 Paterson DL et al. Seventh Annual Conference of the Australian Society of HIV Medicine. November 16-19, 1995; 81(abstract 91). 18 Mastroianni CM et al. Journal of Infection. November, 1994; 29(3):331-7. 19 Berman SM, and Kim RC. American Journal of Medicine. May, 1994; 96(5):415-9. 20 Smith IL et al. Journal of Infectious Diseases. June 1996; 173(6):1481-4. 21 Jokela JA et al. Thirty-fourth Interscience Conference of Antimicrobrial Agents Chemotherapies. October 4-7, 1994; 209(abstract H97). 22 Shepp D et al. Annals of Internal Medicine. September, 1985; 103(3):368-73. 23 Drew L et al. Second National Conference on Human Retroviruses and Related Infections. January 29-February 2, 1995; 117(abstract 338). 24 Couderc L-J et al. Third National Conference on Human Retroviruses and Related Infections. January 28-February 1, 1996; 84(abstract 173). Tuberculosis Therapy and the New Protease Inhibitors by Gabriel Torres, M.D. Aggressive treatment for tuberculosis (TB) requires combination therapy, usually with four drugs (isoniazid, rifampin, ethambutol and pyrazinamide) administered over nine to 12 months. TB treatment regimens that contain rifampin are far superior to those that do not contain rifampin. Rifampin- containing regimens are usually much shorter (six to nine versus 18 to 24 months) and have faster clearance of the tubercle bacterium from the sputum, higher cure rates and fewer relapses. Rifamycins, the class of drugs that includes rifampin and rifabutin unfortunately pose a significant problem for TB patients who are also HIV-positive (one-third the total TB caseload in New York City). Rifampin and rifabutin heighten the activity of the liver's drug-metabolizing cytochrome P450 enzyme system, leading to subtherapeutic blood levels of the anti-HIV protease inhibitors (saquinavir, indinavir and ritonavir). At the same time, the protease inhibitors inhibit the P450 enzymes, causing higher levels of rifampin and rifabutin with an accompanying increased risk of serious side effects (including bone marrow suppression and inflammation of various tissues and organs). The New York City Bureau of Tuberculosis Control of the Department of Health has recently issued a series of recommendations for HIV-positive patients with tuberculosis who also require treatment with protease inhibitors. The agency recommends that: 1. HIV-positive patients with tuberculosis who have already begun therapy with saquinavir or ritonavir should switch to indinavir and change from rifampin to rifabutin since rifabutin does not induce the cytochrome P450 system as strongly as rifampin. Saquinavir levels are reduced by 40% in the presence of rifabutin whereas ritonavir increases the levels of rifabutin, causing excess toxicity. Rifabutin is probably equivalent to rifampin for the treatment of tuberculosis and should be substituted for rifampin in a nine-month TB regimen. In recent studies, the standard 300 mg per day of rifabutin taken with 2,400 mg/day of indinavir caused indinavir blood levels to decrease by 30% and rifabutin levels to double. The advised regimen is 150 mg per day rifabutin taken along with the usual indinavir dose of 800 mg three times a day. 2. HIV-positive patients with newly diagnosed tuberculosis, CD4 counts still over 200 and no prior protease inhibitor therapy should hold off on protease inhibitor therapy until they have completed at least two, and preferably six, months of a rifampin-containing regimen. Other antiretroviral agents (nucleoside analogs and non-nucleoside reverse transcriptase inhibitors) can be used during the six months of TB therapy. Patients with lower CD4 counts have one of three options: a) initial two months with a four-drug, rifampin-containing regimen, followed by four to seven months of isoniazid and rifabutin, with indinavir added after two months; b) begin indinavir along with the TB regimen, but use rifabutin instead of rifampin for the first two months along with three other drugs, and continue with rifabutin and isoniazid for the last four to seven months; c) initial two months with a four-drug, rifampin-containing regimen, followed by a ten- month course of an non-rifampin three-drug regimen, with indinavir begun after the first two months. 3. If rifampin or rifabutin cannot be included in the TB regimen (because of patient intolerance, say), the regimen should include an injectable anti-TB drug for the first two to six months of treatment, in combination with other anti-TB agents. This TB regimen should be continued for at least 18 months. Patient mortality is heightened if rifampin or rifabutin cannot be included. 4. Therapy with protease inhibitors should not be interrupted because patients can develop HIV that may be resistant to both the protease inhibitor they were taking as well as others they might take in the future. Similarly, therapy for TB should not be interrupted because the tuberculosis bacterium may develop multi-drug resistance and become impossible to eradicate. Patients with TB who are also receiving protease inhibitors should therefore receive directly observed therapy (DOT), in which caseworkers record their compliance with the treatment regimen, the side effects and drug-drug interactions. DOT will help ensure the success of therapy and avoid the development of resistance and cross- resistance to either class of drugs. 5. Patients who are beginning treatment with a protease inhibitor should be carefully screened for latent TB infection and disease and given complete preventive therapy with isoniazid, if necessary. Note: The Centers for Disease Control have just issued similar, although somewhat more conservative, guidelines as the NYC DOH. It recommends completion of a six-month TB regimen containing rifampin for all HIV positive patients with active TB regardless of CD4 cell count. For TB patients already on protease inhibitor therapy, the CDC believes one option is discontinuing therapy with protease inhibitors and completing a six-month course with a rifampin-containing TB regimen. The agency concurs with the second New York DOH option of switching from rifampin to rifabutin (150 mg per day) and to indinavir from other protease inhibitors but suggests treating patients with a four-drug rifabutin- containing TB regimen for nine months. The CDC also recommends measuring rifabutin plasma concentrations -- such tests are available only at specialized centers in the US. A third additional option proposed by the CDC is a four-drug rifampin-containing regimen for two months or until the sputum culture has converted to negative, followed by 16 months of continued TB treatment with isoniazid (INH) at 15 mg/kg of body weight and ethambutol at 50 mg/kg, given twice weekly. Some experts also recommend adding a third drug such as streptomycin for the continuation phase. This regimen allows for reintroduction of protease inhibitor therapy after the second month of TB therapy. This third option is not recommended for patients with INH-resistant TB. This advice is based on the best guesses of experts in the fields of HIV, TB and drug metabolism. The regimens have not been formally tested since most trials of protease inhibitors excluded patients requiring rifampin or rifabutin. The recommendations are a good starting point that will encourage practitioners who care for people with TB not to withhold potent antiretroviral therapy from patients who also have HIV. As TB is known to accelerate the rate of HIV disease, intervention with protease inhibitor therapy in this patient population is crucial. Treatment Briefs NTZ for Cryptosporidiosis The Food and Drug Administration (FDA) has given Unimed Pharmaceuticals permission to increase the size of the company's expanded access program for nitazoxanide (NTZ), a new treatment for cryptosporidiosis. There now is no limit to the number of people with HIV-related cryptosporidiosis who can access the drug through the open label program. There are currently no approved treatments for cryptosporidiosis, and numerous agents have failed to show a significant effect on this parasite in clinical trials involving people with HIV. The data on NTZ are much more encouraging, though. A Mexican study presented at the 1996 International Conference on AIDS claimed that NTZ reduced the symptoms of cryptosporidiosis and cleared the parasite from the intestines in most of the 30 patients treated. Those who failed NTZ at doses of 500 mg or 1,000 mg per day, responded when the dose was increased to 4,000 mg per day (abstract TH.B.4213). Results have not been quite as dramatic in U.S. dose-ranging study. Data on the first 22 patients to complete the study were reported at this year's Interscience Conference on Antimicrobial Agents and Chemotherapies (abstract LM50). Participants received either 500, 1,000, 1,500 or 2,000 mg per day. Overall, after four weeks of treatment, at least half of the participants experienced at least a 25% reduction in number of daily bowel movements. 18% of the patients had a complete resolution of diarrhea, and 41% had a reduction of cryptosporidia in their stool samples. There was a trend toward better results on the higher dose arms, and no dose- limiting toxicity was observed. (According to Andy Young, of the PWA Health Group, a New York City AIDS buyers' club that sells NTZ, the only side effect reported to his organization is a green tint to urine and other bodily secretions.) In the expanded access program, participants will be randomized to receive either 1,000 or 2,000 mg per day of NTZ for one month. If the treatment produces no response, the dose may be increased each month to a maximum of 3,000 mg per day. The program is open to people with concurrent microsporidiosis, though it excludes those who have concurrent intestinal CMV, MAC or KS. For more information, call 800/864-6330, extension 3032. Unimed plans to begin phase II/III studies, in collaboration with the AIDS Clinical Trials Group, by the end of the year. Meanwhile, despite much harassment from the Food and Drug Administration, including confiscated shipments and fines, the PWA Health Group in New York still offers NTZ to anyone with a prescription who does not qualify for the expanded access program or who does not want to participate in the phase II/III study. Call 212/255-0520 for further information. ddI and d4T Plus Protease Inhibitors Despite the popularity of AZT/3TC, there are a number of reasons not to use these two drugs plus a protease inhibitor as part of your first strike against HIV. First, if the regimen is not a complete success and resistance occurs, it is not clear whether people who have become resistant to 3TC will receive any benefit from switching to ddI or ddC, since the mutation that confers resistance to 3TC could make the virus resistant to the latter drugs as well. However, this mutation, at amino acid number 184 on HIV's reverse transcriptase enzyme, is only rarely observed to develop in patients using ddI or ddC, which could indicate that patients may be able to switch to 3TC after failure on either of these two drugs. Second, the peripheral neuropathy caused by ddI, ddC or d4T is most commonly reported in late-stage patients. This observation indicates that combination regimens including ddI, d4T or ddC probably should be administered earlier in the course of disease, with the relatively benign AZT/3TC combination reserved for more advanced patients. The ddI/d4T combination is a possible competitor to AZT/3TC, but there has been little information on how to use ddI/d4T in conjunction with protease inhibitors, indinavir in particular. The problem is that both ddI and indinavir must be taken on an empty or nearly empty stomach. Additionally, ddI's buffer interferes with the absorption of indinavir, so they should be taken at least one hour apart. Fortunately, d4T can be given with or without food. Bristol-Myers Squibb, manufacturer of ddI and d4T, recommends the following dosing schedule when combining ddI/d4T with indinavir: Indinavir should be taken just before a light breakfast (if you must eat something), d4T with your dry toast and black coffee, and ddI at least two hours afterwards and at least 30 minutes before lunch. Indinavir then should be taken in the afternoon (at least two hours after lunch and one before dinner). d4T can be taken with dinner, and ddI at least two hours after dinner but at least one hour before ending your evening with a dose of indinavir. Though, not formally recommending it, Bristol-Myers concedes that many clinicians are simply foregoing all this complication and prescribing ddI in one 400 mg dose each day. The dosing regimen would of course be simpler with ritonavir or ritonavir/saquinavir, but the gastrointestinal side effects of ddI combined with those of ritonavir may make it difficult to start using these two agents at the same time. Those who endure the first few weeks of ritonavir may be able to add ddI/d4T at this point since the nausea, vomiting and diarrhea associated with ritonavir have usually diminished by then. Nerve Growth Factor Study Opens ACTG 291, a study of recombinant human nerve growth factor (NGF) in people with HIV-associated neuropathy opened this past summer and currently has enrolled seventy patients. NGF is a chemically synthesized version of a human protein that specifically stimulates the growth and survival of nerve cells in the peripheral nervous system. The trial should be fully enrolled by the spring of 1997, which would be a record accrual rate for a neurological study conducted by the AIDS Clinical Trials Group. The trial will assign 180 people to receive either 0.1 or 0.3 mg per kg of body weight of NGF as a subcutaneous injection twice each week. It is open to people with HIV with peripheral neuropathy that is not inherited, or caused by diabetes, vitamin B12 deficiency, or treatment with neurotoxic drugs such as ddC or d4T. Pregnant women or those who will not consent to practice birth control are excluded. The study is being conducted at numerous sites in the US. Call 800/TRIALS-A for more information. Data from a six month placebo-controlled study of NGF in 250 patients with neuropathy caused by diabetes were presented at the American Neurology Association conference in October. The study found that NGF improved neurologic function and produced an increase in the patients' ability to sense cooling or painful heat. Genentech, the company that makes NGF, is planning a follow-up phase III study in over 1,600 patients. The study is currently on hold though, because the company has a shortage of the drug. This also may prove to be a problem in the current ACTG study. Although, there is enough NGF to complete the trial, Genentech claims that it does not have enough of the drug to promise to continue treating study participants when the trial is over. Ironically, according to David Clifford, M.D., of Washington University, and a member of the ACTG's Neurology Subcommittee, "Ninety-nine percent of the drug is being thrown away." This happens because the drug, administered in very small doses, must be administered immediately after the manufactured concentrate is diluted. Dr. Clifford says that the vials in which Genentech supplies the NGF contain enough to treat around 40 patients -- if they all showed up at the same time. Since they do not, the rest must be thrown out. Less wasteful packaging and closer co-ordination of study participants' visits are possible solutions to the supply problem. According to Kathleen Rinehart, manager of Genentech's corporate communications department: "When we were making the multi-use vial, we didn't know what dose of NGF would end up being used or whether it would be used subcutaneously or intravenously, so we made vials large enough to cover the range of what might be needed. Right now, we are currently working on a reconfiguration of those vials." Cytarabine Nixed for PML The first well-controlled clinical trial to be performed in persons with Progressive Multifocal Leukoencephalopathy (PML), ACTG 243, closed last summer after finding that treatment with cytarabine (also known as ARA-C) offered no benefit. The trial compared high-dose antiretroviral drugs alone to high doses of both intravenous and intrathecal (spinal) cytarabine. The study was closed early, after the Protocol Safety Monitoring Board reviewed the data on the first 62 subjects in the study and found no statistical difference in survival either between the three arms individually or between the two cytarabine arms compared to placebo. Manageable, but slightly higher rates of bone marrow toxicity were observed on the IV cytarabine arms. A higher dropout rate occurred on the intrathecal arm, which involved the administration of the drug through a needle inserted into the membranes surrounding the spinal cord. The results of the study came as no surprise, since there were reports that people treated with cytarabine die faster than those who received no treatment and since there was at least one published case of cytarabine causing PML in a person with cancer by suppressing the immune system. (See TL Hwang et al. Journal of Neuro-Oncology. April, 1986; 3(4):335-9.) ACTG 243 may yet yield useful information on the natural history of PML. For one thing, it gathered useful data that should confirm whether a PCR (polymerase chain reaction) test that detects JC virus in the cerebrospinal fluid, can be used to diagnose PML. (JC virus is believed to cause this brain disease.) More Free Viral Loads Roche Diagnostic System's introductory offer of two free HIV viral load tests with its Amplicor PCR assay (see last May's Treatment Issues, page 12) has been a victim of its own huge success -- or poor planning, depending on how you look at it. At the end of October, some 25,000 plasma specimens had still to be assayed for HIV level, out of a total of 80,000 sent in. Some people have been waiting since July for the results, which are supposed provide two baseline values for gauging future disease progression and treatment effects. Roche now hopes to clean up the backlog by mid-December, but of what use are results from months-old test specimens? To make up for its program's failure to deliver in a timely fashion, Roche is now planning to send doctors a letter offering them two coupons for every patient who had not received the two test results by October 18. The coupons will be good for six months. Roche is trying to ensure that the labs performing the tests will report the viral load results within ten days of receiving a specimen. To prevent a new backup, it has signed up more medical laboratories and will sent out the coupons in staggered fashion. Coupon distribution is supposed to begin December 16. Patients who may qualify for these coupons should contact their doctor to make sure that he or she has applied for them.