       GMHC Treatment Issues, Volune 10, Number 5 - May 1995
       ----------------------------------------------------------
       Published by Gay Men's Health Crisis, Treatment Education,
       129 West 20th Street, New York, NY 10011.
       ----------------------------------------------------------

       Viral Entry Discovery Suggests New Treatments
       by Theo Smart

       A central mystery surrounding the infection of CD4 cells by HIV
appears to have been solved by Edward Berger, M.D., and colleagues at
the National Institute of Allergy and Infectious Diseases (NIAID).
Scientists have long known that the envelope protein of HIV binds to
the CD4 receptor on certain types of white blood cells. But that
action by itself was not adequate to fully explain how the virus fuses
with and inserts its genetic material into a cell. In fact, non- human
cells altered to bear the human CD4 receptor were not infectable with
the virus. There had to be an unidentified co-factor on human cells to
which the virus clung, and investigators have spent a decade searching
for it.

       The NIAID investigators now have identified the co-factor as a
cell surface protein they call fusin due to its pivotal role in
helping HIV fuse to CD4 cells. Their discovery reportedly rapidly has
been confirmed by a number of other labs. The discovery of fusin is
considered a breakthrough not only because it reveals a new target for
anti-HIV drugs but because it may lead to the first true animal model
for HIV infection.

       Gene Libraries

       Dr. Berger's team already had a model for HIV fusion using
mouse cells modified to carry CD4 receptors. But, again, merely having
a CD4 cell receptor did not make the mouse cell susceptible to HIV
infection. To find the fusion cofactor, the researchers introduced a
series of genes from a human CD4 cell line into the genetically
engineered mouse CD4 cells and fusion occurred. Certain that one of
these human genes was making a protein that allowed fusion, the
researchers used a process of elimination to isolate the gene
responsible for the fusin protein.

       The fusin gene proved to be a member of the gene family that
produces "G protein-coupled" cell receptors, often exploited by other
viruses when entering cells. In contrast to the protruding CD4
receptor, G protein-coupled receptors weave in and out of the cell
membrane. They form a depression that usually contain these receptors'
critical binding region. If HIV binds to such a receptor on CD4 cells,
it is already partially inside the cell.

       Fusin as an Antiviral Target

       To see whether fusin protein was required for infection of
human CD4 cells, the researchers mixed CD4 cells with antibodies to
fusin and effectively blocked infection of the CD4 cells. Antibodies
to the V3 loop on the HIV envelope protein molecule gp120 also blocked
fusion, suggesting that HIV binds to fusin via this highly conserved
loop, although this has yet to be demonstrated. A number of other
researchers have also observed that antibodies to the V3 loop
neutralize the virus and block fusion.

       Besides underscoring that fusion could not occur without fusin,
the antibodies' antiviral activity raised the prospect that compounds
interfering with this protein may block infection of CD4 cells.

       But if this cofactor plays a role in the infection of CD4
cells, it would seem that simple drug screening tests would have
picked up fusin blockers long ago, even if no one understood exactly
how the drug worked. Indeed, other researchers screening potential
anti-HIV drugs have found a few compounds believed to inhibit viral
entry other than by blocking gp120 binding to the CD4 receptor. These
chemicals include Rhne-Poulenc Rorer's betulinic acid derivatives,
distamycin analogs from Pharmitalia, and the bicyclam compounds owned
by Johnson Mathey.

       "It is possible that some of these compounds actually interfere
with fusion," says William Rice, Ph.D., head of the Laboratory of
Antiviral Drug Mechanisms at the National Cancer Institute (NCI). It
is conceivable, too, that these compounds bind to the part of the
gp120 V3 loop that may attach to the fusin receptor. A recent
laboratory study of the bicyclams found that HIV reacted to these
compounds precisely by acquiring resistance-conferring mutations in
the V3 loop.

       None of these substances have made it into human studies, and
Rhne-Poulenc says it has stopped research into betulinic derivatives
due to poor "pharmacodynamic properties." Rice thinks that many
pharmaceutical companies have a bias against drugs that act at this
stage of the infection process: "It's very hard to push forward
surface-active agents [drugs that prevent the virus from entering the
cell]," he says, "because of the problems with dextran sulfate and
some of the other sulfated polysaccharides." These drugs blocked
infection by coating the surface of the cell, but they also caused
hemorrhaging in clinical trial participants.

       According to Dr. Rice, the NCI currently is planning clinical
trials of one of the Pharmitalia compounds, which has recently shown
anti-HIV activity in an animal model. Dr. Rice believes that the
discovery of the fusin protein will lead to the development of more
precise laboratory tests more capable of detecting more
fusion-specific inhibitors. Meanwhile, Johnson Mathey wants to launch
phase I studies of JM3100 within the year. This company's biomedical
group is small, though, so the company is trying to get help to run
the studies from the AIDS Clinical Trials Group.

       One problem with developing drugs that directly block fusin, is
that no one yet knows what human protein normally binds to it.
Antivirals that work by blocking fusin may interfere with its natural
function. The part of the natural ligand that binds to fusin also may
share homology with sequences from the V3 loop of HIV. Some
researchers have found similarities between segments of the V3 loop
and parts of MHC-II proteins, involved in antigen presentation to CD4
cells.

       The Chemokine Connection

       Another potential shortcoming of anti-fusin drugs is that the
researchers found that although the protein seemed necessary for the
infection of CD4 cells, anti-fusin antibodies had no effect on the
infection of macrophages. Drugs that are too specific to fusin may not
block infection of macrophages. [Last minute note: Anne Bousseau,
Rhne-Poulenc's AIDS Program Director, told Treatment Issues that
betulinic derivatives also block HIV entry into macrophages. These
compounds interact with gp120 outside the V3 loop and produce an
effect more general than mere inhibition of V3 loop-fusin would
produce.]

       Dr. Berger's team observed that fusin closely resembles a
receptor for IL-8, one of the chemokines involved in directing
lymphocytes to infected tissue. The group believes that a different
but possibly related cofactor allows the HIV to fuse to macrophages.
This suspicion is strengthened by reports that the V3 loop of gp120
determines cell tropism -- strains of HIV that infect macrophages more
efficiently have slightly different V3 loops than the HIV that
concentrates on CD4 lymphocytes. Given the similarity between fusin
and the IL-8 chemokine receptor, the fusin-like protein on macrophages
may be the receptor for Dr. Robert Gallo's anti- HIV chemokines:
Rantes, MIP-1a and MIP-1b (see January's Treatment Issues, pages
10-12). The evidence for this association is circumstantial at
present.

       Animal Models

       Even if this breakthrough in basic science does not lead to any
new therapies, it may have one other practical implication: the
development of a small animal model for HIV. The lack of the fusion
co-factor had made this impossible. Scientists could alter mouse or
rabbit cells to carry human CD4 cell receptors, but these cells could
not be infected with HIV. Now, with the same gene altering techniques
that the team used to isolate the fusin gene, they could genetically
modify animals to have lymphocytes with both human receptors. Such an
animal model would speed the identification and development of the
most promising anti-HIV compounds.

       ***************************
       Protease Inhibitor New Math
       by Dave Gilden

       Two months have passed since three protease inhibitors became
commercially available in the United States. In this period, the three
corporate sponsors, Abbott, Merck and Roche, have jockeyed for market
share, with varying degrees of success. Each of the companies'
products, respectively ritonavir (brand name: Norvir), indinavir
(Crixivan) and saquinavir (Invirase), have encountered various
controversies stemming from the conditions under which each compound
is manufactured and administered. In each case, a peculiar brand of
mathematics has been employed to resolve seeming contradictions. We
review below these new developments, both pharmaceutical and
mathematical.

       Zero plus Zero Equals One

       On May 7, Hoffmann-La Roche released the final clinical data on
its North American trial NV14256. Preliminary surrogate marker (viral
load and CD4 count) from this 1,000-person trial were presented to the
FDA last fall as part of the package supporting marketing approval of
the drug (see Treatment Issues, November 1995, page 2 and December
1995, page 2). The trial tested saquinavir plus the nucleoside analog
ddC (also a Roche drug) against either drug alone in volunteers
previously treated with AZT. The initial surrogate marker data made
the combination look only slightly better than saquinavir or ddC
monotherapy. This observation was a big disappointment considering
that protease inhibitors are supposed to be a powerful new drug class
whereas tolerable doses of ddC have weak antiviral activity even
compared to other nucleoside analogs. At the time, Treatment Issues
concluded, "Saquinavir's main advantage is that it is a relatively
benign way of slightly boosting the mediocre performance of nucleoside
analogs but the retail price is $7,200 per year!"

       The new data on progression to AIDS and deaths (final surrogate
marker data have yet to be released) document the extent to which
saquinavir supplements nucleoside analog therapy. Once again, ddC and
saquinavir monotherapies came out nearly the same, but the combination
therapy results exhibited greater superiority over monotherapy than
one would expect from the available surrogate marker data (see table).

       The ddC/saquinavir combination produced an immediate, sustained
median viral load reduction of about 0.6 log (75 percent), compared to
little to no long-term reduction for either drug alone. As was the
case in a previous trial of AZT plus delavirdine (see Treatment
Issues, January 1996, page 6), a threshold leading to significant
clinical approval may have been crossed when the induced drop in HIV
levels exceeded 0.5 log.

       A major limitation of the NV14256 trial is that one still
cannot tell how saquinavir plus ddC stacks up against the more potent
nucleoside analog combinations (for example, d4T plus 3TC or ddI or
AZT/3TC) in the advanced, AZT-experienced population under
examination. The ongoing international trial comparing AZT/saquinavir
to AZT/ddC to the triple combination in treatment-naive volunteers
will partly fill this gap. Roche has developed a new "soft gel
capsule" formulation of saquinavir that is much better absorbed by the
stomach. It is hoped that this improved version will help the compound
fulfill the anti-HIV potential it displayed in the lab. A 400 person,
48-week trial of the soft gel capsule plus participants' choice of
nucleoside analogs is now fully recruited. The saquinavir dose used in
this trial (1,200 mg three times daily, twice the current recommended
dose of the present formulation) achieves peak blood levels eight
times that attained by the current regimen with the marketed
formulation.

       A trial combining saquinavir with ritonavir is also in progress
(see last month's Treatment Issues). Adding ritonavir inhibits
breakdown of saquinavir by the liver and can raise saquinavir blood
levels ten-fold. Combining the two protease inhibitors probably would
allow for reductions in the recommended dosing of one or both drugs
while enhancing their anti-HIV efficacy.

       Both Gay Men's Health Crisis and Project Inform have written
Roche asking the company to stop promoting and pricing the saquinavir
now on the market as if it were equivalent to other protease
inhibitors instead of just a "kinder, gentler ddC." At the same time,
they asked that the company accelerate its efforts to bring its
protease inhibitor up to the original expectations for it.

       Roche officials profess great optimism about saquinavir's
future. Franz Humer, the head of Roche's pharmaceuticals division,
told Financial Times (London) that saquinavir would become a
blockbuster drug for the company, according to the paper's May 7
edition. He predicted that annual sales of saquinavir would reach $160
to $400 million within five years and that total sales of all Roche
HIV-related treatments would be as much as $800 million at that time.

       The Incredible Shrinking Markup

       One of the fears about widespread use of suboptimal saquinavir
is that such exposure will start HIV on the evolutionary pathway to
resistance to both higher doses of saquinavir and to other, more
potent protease inhibitors, specifically Merck's indinavir (Crixivan).
Merck, which just started selling indinavir at the end of March,
reported that 13,000 people were taking the drug by mid-May. Rapid
emergence of indinavir-resistant HIV would threaten what is becoming a
very popular product.

       To gauge the extent of this threat, Merck is planning a trial
of indinavir's efficacy in individuals with at least one year's prior
saquinavir therapy. This six-month open-label study will test
indinavir versus both old and new formulations of saquinavir. Merck is
also planning dose- ranging and safety trials of indinavir and
saquinavir combined. Other upcoming studies include the effect of
taking indinavir with meals (currently not recommended), twice daily
dosing versus thrice daily (the present recommended schedule),
maternal-fetal HIV transmission and indinavir for early HIV infection
(CD4 count over 500). Trials in infants are being held up by Merck's
inability to produce a palatable liquid formulation.

       The rapid growth of indinavir sales has raised the threat that
supplies will run out for new patients attempting to obtain the drug.
Until full-scale production facilities build up an adequate inventory
(expected by the end of this year), Merck will fill prescriptions for
only about 30,000 Americans, to whom it will guarantee continuing drug
availability. The final 2,500 of these slots are reserved for people
with CD4 counts below 50, on the assumption that this is the
population with the most urgent need for a new and potent therapy.

       To ensure a steady supply of indinavir to its clientele, nearly
all prescriptions are being filled through Stadtlanders' pharmacy, a
mail-order firm based in Pittsburgh, Pennsylvania. In addition to
filling orders, Stadtlanders' is supposed to monitor supply, make sure
customers receive refills in a timely fashion, and arrange with
insurance companies and other third-party payers for direct
reimbursement.

       After some start-up troubles, the system is operating fairly
smoothly, most people report, although the turnaround time for orders
is at least four days, and can be several weeks, not the 48 hours
Merck originally envisioned. The biggest controversy has been the
price Stadtlanders is asking. Merck's wholesale price for a year's
supply of indinavir is $4,380, and Stadtlanders originally wanted
$6,022.50 retail for the same amount of drug. This exceptionally high
37.5 percent markup was a disappointment considering that Merck seemed
to be intentionally limiting its per unit profit in order to create a
larger total market for indinavir. The retail price sparked wide
protests and calls for a boycott especially from ACT UP chapters in
New York, Philadelphia and San Francisco (ACT UP/Golden Gate) as well
as the PWA Health Group, also in New York City. Stadtlanders
immediately argued that 80 percent of patients on indinavir were
receiving a substantial discount due to negotiations with their
reimbursers (HMOs, Medicaid, etc.).

       Stadtlanders eventually agreed to provide similar discounts to
cash customers who hold a discount card from the Community
Prescription Service -- another mail-order pharmacy. (Call
800/842-0502 to obtain the card, which costs $18 annually.) With the
addition of this measure, most people will be charged a retail price
representing a more normal fifteen percent markup. The main exception
is purchasers with traditional medical insurance whose carriers pay
Stadtlanders directly. Those insurance companies will still pay full
price. (If the purchasers pay Stadtlanders and then apply on their own
for insurance reimbursement, they could receive the discount.)

       Stadtlanders maintains it has to preserve the high official
price as a benchmark for the discounts many third-party payers
customarily receive. While Stadtlanders made sure it received an
adequate return for the temporary distribution program, it lost
considerable good will in the process.

       There remains the issue of side effects experienced by the
general population, rather than the more restricted cohorts in
clinical trials. In a posting on the Internet's "Crixivan E-mail
Discussion List," Kiyoshi Kuromiya, director of the Critical Path AIDS
Project in Philadelphia, commented on his own personal experience as
well as that of others. He wrote, "I think. . .  that the incidence of
'flank pain,' crystalluria, and kidney stones is much higher than the
three percent that Merck predicts. . .  It can happen within hours of
taking the Crixivan with insufficient water. It can lead to pain (that
resembles acute appendicitis or severe gas pains) which may resolve
after a few hours (ten in my case) on its own. Or it can lead to
problems over a few days or weeks or months." To prevent the
precipitation of indinavir in the urine, Mr. Kuromiya recommends
ingesting the medication with several glasses to a quart of water.

       Merck officials profess to be mystified by such remarks, which
they consider exaggerated. (It remains, though, that Merck advises
people taking indinavir to drink an extra two quarts of water a day to
preclude indinavir deposition in the kidneys, which can cause
considerable pain.)

       Ritonavir Plus Other Drugs

       Abbott, with the most expensive protease inhibitor of them all
($6,500 a year wholesale) claims to have a similar number of protease
inhibitor clients as Merck does. Despite speculation that a price cut
is imminent, the company says no such move is planned. The only hope
Abbott publicly holds out for lessening ritonavir's cost is that it
might eventually be possible to combine it with saquinavir. This still
highly experimental combination might allow some reduction in the
standard ritonavir dose, but saquinavir is nearly as expensive as
ritonavir. The final cost of this combination cannot be predicted at
this time and could well be more than ritonavir alone (though the
benefits might be greater, too). Some data on ritonavir/saquinavir
will be presented at the International Conference on AIDS this summer.

       Officially, about twenty percent of those who start ritonavir
therapy discontinue it. The main reason is nausea and diarrhea. In
addition, the problem of ritonavir's interactions with other drugs has
confused and troubled many care providers and patients alike. Abbott
has sent at least two different cards to practitioners outlining
drug-drug interaction information. It also has a 27-page document on
the subject for investigators taking part in ritonavir clinical
trials.

       The table on page 6 summarizes the feared interactions. Of
particular concern is ritonavir's effect on most antidepressant drugs,
whose blood levels ritonavir probably increases, and on
estrogen-containing birth control pills, whose blood level is reduced
by 40 percent (all other steroid hormones, including
progesterone-based birth control, testosterone, anabolic steroids and
corticosteroids, are probably increased).

       Most of these interactions are theoretical, as actual studies
have been conducted with only a handful of compounds. The information
in the table is largely derived from predictions based on ritonavir's
mostly inhibitory effect on different metabolic pathways in the liver
and on other drugs' breakdown via these pathways. In particular, drugs
broken down by the CYP3A or CYP2D6 pathways, which ritonavir blocks,
have increased concentrations in the body, requiring careful
monitoring for toxic effects. Conversely, blood levels of some
medications are raised somewhat because they are broken down by liver
enzymes whose activity ritonavir induces. Such drugs may require dose
increases to maintain their effectiveness.

       Coming Up: Does One plus One Equal One?

       The problems with ritonavir promise to become a less serious
issue upon the arrival of the second generation protease inhibitor
Abbott now has under development. Known as ABT-378, the compound is
said to be more powerful than any of the current protease inhibitors
and to be relatively unaffected by the HIV mutations conferring
resistance to the ones currently on sale. Human testing is scheduled
for later this year, with the first information on any in vivo
improvements in toxicity and efficacy available in December. (Merck,
too, has an active second generation protease inhibitor program.)

       Finally, more early clinical trial data were released in May on
Agouron's experimental protease inhibitor, nelfinavir (brand name:
Viracept). After 16 weeks on nelfinavir monotherapy, twenty volunteers
experienced average viral load drops of 81 to 99%, depending on dose
(the reduction at four weeks was 95 to 99%). CD4 counts went up by an
average of 115 to 170. In a 36-person study of nelfinavir plus d4T,
average viral load drops over 98% were observed for each of the three
nelfinavir doses at 60 days, while CD4 count increases averaged 105 to
130. These data may sound impressive, but numbers of trial
participants and duration of follow-up are small. The current larger
trials eventually will provide sounder information. Although the
present figures allow no definitive conclusion, it is curious that
nelfinavir with or without d4T yielded substantially the same results.
One plus one seems to equal one in this case.

       A marketing application for nelfinavir is expected to be filed
with the FDA early next year. With a variety of other protease
inhibitors also on the way, we can expect that our lessons in protease
inhibitor arithmetic will continue.

              Final Clinical Data from NV 14256
                     AIDS-Defining Events
       Regimen                 or Death  Death Alone
       ddC                         85            28
       Saquinavir                  77            34
       Saquinavir plus ddC         46*            9*

       *statistically significant compared to the two monotherapies
        characteristics of participants:

       number: 978
       CD4 counts: 50-300 (median of 160-180 depending on the
         treatment)
       prior AZT: at least 16 weeks (median of 18 months)
       time on assigned therapy: ddC -- 43 weeks; other two -- 56
         weeks
       median follow-up: 73-74 weeks

       *********************************************************
       How Ritonavir Affects The Body's Processing of Concurrent
       Medication*

       Except for the last table, the interactions listed here are all
theoretical predictions based on the way the liver is thought to
metabolize the listed drugs and ritonavir's known inhibiting and
inducing effects on that metabolism. The simple two-drug interactions
in the table do not take into account other drugs a patient may be
receiving (interactions between three or more drugs), nor does the
table allow for interpersonal biochemical variations, especially
abnormal liver function. The adequacy of the proposed drug
substitutions has not been established, nor is their safety assured.
Their levels may be altered by ritonavir, and individuals could
experience undesirable changes in these drugs' toxicity and activity.
(This issue has been raised in particular about Prozac, whose blood
levels ritonavir may increase.) Direct monitoring of blood
concentrations is frequently advisable. Drugs whose blood levels can
be tested readily are marked with a dagger ().

       *************************************
       Never use concurrently with ritonavir:
       (narrow therapeutic window)

       Omitted from ascii electronic version
       *************************************

       **********************************
       Contracting Out the FDA: An Update
       by Derek Link

       Over the last six months, "FDA reform" bills have progressed at
a slow but steady pace through both houses of Congress (see January's
Treatment Issues, pages 1-5). The Senate Labor Committee passed in
late March an amended version of S.1477, Sen. Nancy Kassebaum's (R-KS)
FDA bill, S.1477, now awaits consideration on the Senate floor, a
debate which has not yet been publicly scheduled. The House Commerce
Committee introduced its FDA reform proposal in late March too. No
action has been taken on the House FDA plan, although a subcommittee
vote could occur soon.

       The Senate bill passed the Labor Committee with nine
amendments. Several of these, particularly the ones offered by Sen.
Dan Coats (R-IN), came as a surprise assault. The bill had gradually
evolved into a moderate proposal, but the Coats amendments undid much
of the consensus building. They would force the FDA to immediately
contract with private consultants for the review and approval of all
medical devices, broaden the grounds for privatizing FDA's other key
responsibilities, and dramatically lower safety, purity and
effectiveness standards for drugs, devices and foods.

       Patient advocates opposed Coats' amendments on the grounds that
the regulatory system must be built on sound public health principles
and rigorous, objective clinical research. Privatization should
therefore proceed cautiously, in these advocates' view. There is no
private sector industry immediately capable of taking on FDA's tasks.
Even if Congress rushes ahead with farming out FDA activities, it will
take time for this new industry to develop. Immediate privatization
risks slowing drug development by throwing the entire system into
chaos.

       Also, drug and device makers could hire a private concern to
review their products under the Coats plan. This builds conflict of
interest into the regulatory system. Reviewers would have an incentive
to approve products based on potential future business rather than on
an objective examination.

       The House Commerce Committee, after keeping its FDA proposals
secret for months, released them in late March with much fanfare. At a
bizarre media spectacle held in its ornate chambers, the Commerce
Committee released three FDA bills -- a bill for drugs and
biotechnology products, a bill for foods and a bill for medical
devices. Several enthusiastic private citizens, whom the committee
brought to the ceremony and described as victims of the FDA
bureaucracy, provided a facade of popular support for the measures.
Committee members even brought out a cake and sang a sweet song
through teary eyes for a heart disease patient from New Jersey. The
Committee also announced it had formed an FDA task force of four to
guide the bills through the legislative process.

       The House bills go far beyond S.1477, even with the Coats
amendments. They privatize virtually all of FDA's functions, ranging
from the inspection of drug-making facilities and blood banks to the
review and approval of all drugs, devices, blood and biotechnology
products. The House bill lowers safety and effectiveness even more
than the Senate bill, allowing drugs to be sold with no clinical
research whatsoever. The bill also allows higher levels of pesticides
in food and restricts the FDA's ability to inform the American public
about the dangers of products on the market.

       AIDS organizations, and GMHC in particular, have helped lead
the public opposition to the Republican FDA proposals. A growing array
of groups -- patients, seniors, unions, women, minorities,
environmentalists and consumers -- have now joined to defeat these
bills. But the coalition is coming together in a frenzied and ad hoc
manner that is limited both organizationally and financially.
Discussions are now taking place as to how to form an aggressive,
independent and organized public health lobby for the FDA.

       Whether Congressional leaders can pass the FDA bills in both
chambers, reconcile the substantial differences between House and
Senate proposals, and secure assent from President Clinton remains a
matter of much speculation. The growing opposition to the FDA bills, a
dwindling legislative year and an intensifying electoral campaign make
the likelihood of swift, easy passage decline from day to day. This
year may prove to be just a dress rehearsal for future showdowns.

       *********************
       Making Inroads on PML
       by Theo Smart

       Progressive multifocal leukoencephalopathy (PML) can paralyze,
blind, mute and kill people with AIDS within weeks. According to some
estimates, the incidence of the condition has doubled in the last
several years. This disease of the central nervous system presently
occurs in around six percent of AIDS cases and the rate is expected to
increase as people with AIDS live longer with low CD4 cell counts
(partly due to improved prophylaxis and treatment of other
opportunistic infections).1 Although there is still no treatment for
PML, a number of compounds that have shown potential in the
laboratory, such as topotecan and cidofovir, are closer to being
studied in people with PML. There have also been advances in diagnosis
and basic science research which may lead to improved treatment and
ultimately prevention.

       The JC Virus

       PML is caused by the JC virus, which infects and kills
oligodendrocytes. These brain cells produce the myelin that surrounds
and protects nerve cells. The virus, named after the first patient
with PML in whom it was isolated, belongs to a family of viruses that
also includes wart viruses. The JC virus is common. By middle age, 80
percent of adults, have been exposed to it. The primary illness that
it causes is unknown, but a case of meningoencephalitis has been
observed in an otherwise healthy teenage girl with rising JC antibody
titers.2

       After acute infection, the virus can remain latent in the
kidney, lymphoid organs, bone marrow and circulating B lymphocytes,
but most researchers believe that the virus is not latent in the brain
or central nervous system (CNS). A number of factors such as immune
suppressive chemotherapies (most commonly, cyclosporine and
methotrexate), and other viruses, including HIV and CMV can stimulate
the virus out of latency. After reactivation, the researchers think
that virus is transported to the brain by circulating B-cells. If this
indeed is the case, early treatment of people with reactivated JC
virus -- which has not yet taken up residence in the CNS -- possibly
could prevent PML, should an effective anti-JC drug be discovered.

       Once the virus enters the brain, treating it may become more
difficult since the parts of the JC virus genome that drive
reproduction are prone to mutation. A number of studies report that JC
virus associated with PML lesions has unique genetic sequences in the
part of the virus' genome that drives replication. These genes
frequently differ from the genes of the JC virus found in the kidneys
of the same patients.3 The genes in the "neurotropic" JC virus bear a
strong resemblance to genes involved in making myelin and other neural
proteins.4 Exactly how the virus adopts these changes is unknown, but
these mutations may allow natural processes in the brain to accelerate
spread of the virus there. This suggests yet another reason to treat
before the virus invades the CNS. It also suggests that drugs being
considered as possible treatments for PML should show activity against
the neurotropic strain of the virus.

       Improvements in Diagnosis

       Although a brain biopsy is still considered the standard for
definitive diagnosis in some cases, PML can be diagnosed without
undergoing this invasive procedure. Many clinicians rely on axial
computed tomography (CT) or magnetic resonance imaging (MRI). The MRI
is more sensitive, and also can reveal the number and extent of
lesions. With either technique, the pattern and enhancement of the PML
lesions usually can be easily distinguished from toxoplasmosis,
lymphoma and HIV encephalopathy.

       Viremia (detectable virus) in the cerebral spinal fluid (CSF)
is increasingly being used in diagnosis. A recent study by Dawn
McGuire, M.D., and colleagues in San Francisco looked for JC virus in
the CSF of 26 patients with PML plus 114 HIV- positive controls and
sixteen HIV-negative controls.5 Using a very sensitive PCR test that
could detect one copy of JC virus in 50 microliters of cerebrospinal
fluid, the researchers found the virus in 24 of the 26 patients with
PML, ten of the HIV-positive patients (who will be closely monitored
for development of the condition) and one HIV- negative participant
with neurologic abnormalities. An ongoing ACTG study of the treatment
of PML (see below) should confirm the utility of JC virus PCR for
diagnosis.

       Viremia also may prove to be a prognostic tool. A team of
French investigators has tested over 165 people with HIV along with 65
HIV-negative immune compromised persons and found that the incidence
of detectable JC virus in peripheral blood leukocytes was almost twice
as high in people with HIV (28.9 percent versus 16.4 percent in the
HIV-negative patients).6 The individuals with both HIV and JC virus
are being followed to ascertain the predictive value of the tests.
Clearly, 28.9 percent of people with HIV do not develop PML, however,
the study may determine whether a higher JC viral load or certain
strain of the virus distinguishes those people more likely to develop
the condition.

       Limitations of Present Therapies

       As there is no accepted treatment, patients with PML frequently
are subjected to experimentation with toxic agents. These experiments
have yielded a number of case reports claiming improvements in
response to treatments such as cytarabine, alpha interferon,
vidarabine (adenine arabinoside) and idoxuridine. Treatment with AZT
has been reported to lead to stabilization, but it is unclear whether
improvement was due to an anti-JC viral effect or treatment of the
underlying HIV infection.7 In another study, three months of
intrathecal treatment with AZT failed to improve the status of two
patients with PML.8

       Perhaps the most influential PML study was by Carolyn Britton,
M.D., which reported clinical improvement in eight of thirteen
patients (mean CD4 cell count of 106) treated with intrathecal
cytarabine (ARA-C).9 One must view such clinical improvement with
skepticism -- spontaneous remission of PML can occur in the absence of
treatment, particularly in patients with higher CD4 cell counts. A
number of other case reports and open label studies recount the
failure of cytarabine,10 as well as every other drug that has shown
activity in other case reports. One review of the published case
reports even noted that people treated with cytarabine died faster.
There is at least one published case of cytarabine causing PML in a
person with cancer (by suppressing the immune system).11 On top of
that, the drug is extremely neurotoxic. Nevertheless, because of
Britton's open label studies, and because they have no other
treatments to offer, many doctors and clinics administer cytarabine to
patients with PML.

       Cytarabine is being put to more rigorous challenge in ACTG 243,
the first well controlled clinical trial for PML. The trial compares
high-dose antiretroviral drugs alone to high doses of both intravenous
and intrathecal cytarabine. The endpoints will be survival and changes
in neurological condition. The mixed reports for these therapeutic
approaches give few researchers faith that this study will advance
treatment, though it will yield data on using JC viremia as a
diagnostic tool.

       New Leads

       Camptothecin derivatives -- in particular, camptothecin,
topotecan and irinotecan -- can suppress the neurotropic JC virus in
glial cultures.12 These compounds are chemotherapeutic agents that
work by inhibiting topoisomerase I, an enzyme that uncoils strands of
DNA, a necessary step in transcription, and ultimately cell
replication. These drugs are known to cause bone marrow suppression,
diarrhea, nausea and other side effects.

       A number of patients now have been treated with topotecan.13
Two end-stage patients failed to respond. Another patient experienced
lesion remission and symptomatic improvement. Despite the bone marrow
suppression that topotecan caused (in response to which the patient
was given the growth-factors G- CSF and EPO), the patient's CD4 cell
count increased. This patient had initiated therapy with a protease
inhibitor six weeks earlier, though, which suggests that this response
could have been due to the antiretroviral.

       Nevertheless, the case has led SmithKline Beecham to design a
study that should start next month in Los Angeles, San Francisco,
Miami, and possibly New York City or Baltimore and a site in France.
The study will enroll at least fifty volunteers with biopsy-proven PML
who have been on stable antiviral therapy for six weeks. Trial
participants will be randomized to treatment with topotecan or
observation (no treatment). All in the latter study arm will be
crossed over to the treatment arm after eight weeks -- or after only
four weeks if there is progression in disease, defined as growth in
PML lesions of at least 25% detected on MRI. For information on study
enrollment call Sharyn Arnold at 215/751-7074.

       A paper by a team of Belgian researchers, led by Erik de
Clercq, M.D., at the Thirty-fifth Interscience Conference on
Antimicrobial Agents and Chemotherapies reported that cidofovir, a
drug soon to be approved for the treatment of CMV retinitis, has
activity against a murine (mouse) polyomavirus, which is very closely
related to the JC virus.14 At a Japanese conference in May, the same
group reported even greater activity against SV40, a more closely
related simian polyomavirus that causes a condition similar to PML in
immunosuppressed monkeys with SIV.15

       They then treated one PML patient with 3 mg per kg of cidofovir
once every ten days.16 Although after the second infusion there was a
reduction in the size of the lesion, there was continued neurologic
deterioration. Treatment was discontinued after the fourth infusion,
and the patient died two months later. As there was change in the size
of the lesion, the Belgian team concluded that the treatment had
potential but that further studies should use more frequent dosing.

       Researchers in London are planning a controlled clinical study
that should begin by August. The dosing regimen in this study will be
the same as used for the treatment of CMV retinitis -- 5 mg per kg per
week for the first two weeks, and then every other week.

       Treating PML by Treating HIV

       Topotecan and cidofovir may prove to be more effective
treatments than cytarabine, which again, has been shown to cause PML.
Still, the most significant advance against PML may turn out to be the
new treatments for AIDS. There are numerous anecdotes of people with
PML experiencing lesion remission after starting treatment with
protease inhibitors. These drugs may do a much better job of salvaging
the immune system than treatments of the past, and improvement of PML
is fairly common in people with higher CD4 cell counts. Although
anti-HIV therapy may only work temporarily, people with PML should
certainly be offered the best available antiretroviral agents before
turning to unproven toxic chemotherapies that may actually aggravate
their condition.

              1	Major E and Ault G. Current Opinion in Neurology.
       June, 1995; 8(3):184-190.

              2	Blake K et al. Archives of Disease in Childhood.
       Jul, 1992; 67(7):956-7.

              3	Loeber G and Dorries K. Journal of Virology. May,
       1988; 62(5):1730-1735.

              4	Amemiya K et al. Journal of Biological Chemistry.
       Jul 15, 1992; 267(20):14204-14211.

              5	Major E and Ault G. Current Opinion in Neurology.
       June, 1995; 8(3):184-190.

              6	Dubois V et al. AIDS. April, 1996; 10(4):353-358.

              7	Conway B et al. Reviews of Infectious Diseases.
       May- June, 1990; 12(3):479-82.

              8	Allegre T el al. 9th International Conference on
       AIDS. June 6-11, 1993; 9(1):423 (abstract PO-B16-1727).

              9	Britton CB et al. Eighth International Conference
       on AIDS. Jul 19-24, 1992; 8(1):Th67 (abstract ThB1512).

             10	Antinori T et al. AIDS. July, 1994; 8(7):1022-4.

             11	Hwang TL et al. Journal of Neuro-Oncology. 1986;
       3(4):335-9.

             12	Kerr DA et al. Virology. Oct, 1993;
             196(2):612-618.

             13	Broom, C. Personal Communication, April 22, 1996.

             14	Snoeck R et al. 35th Interscience Conference on
       Antimicrobial Agents and Chemotherapy. September 17-20, 1995,
       abstract H49, pg 188.

             15	Andrei G et al. Ninth International Conference on
       Antiviral Research, May 19-24, 1996.

             16	Snoeck et al. Ninth International Conference on
       Antiviral Research, May 19-24, 1996.

       ********************************************
       Free Introductory Offer for Viral Load Tests
       by Theo Smart

       As Treatment Issues went to press, FDA marketing approval was
imminent for Hoffmann-La Roche's Amplicor HIV-1 Monitor test kit.
Approval will make the test, based on polymerase chain reaction (PCR),
the first commercial method for measuring HIV in the blood, or viral
load. The Amplicor kit will receive the FDA go-ahead for use in
determining a patient's prognosis and as an aid in assessing response
to anti-HIV treatment. The latter indication is very important as it
will allow much more frequent use of the test. In recent weeks, AIDS
activists have put considerable pressure on the FDA to support such
treatment monitoring.

       Roche in partnership with a number of clinical laboratories is
offering to provide two free baseline tests to all comers for the
first 60 days Amplicor is on the market. To enroll in what is being
termed the Amplicor Access Program, call toll- free 888/AMPLICOR
(888/267-5426) This program will give Amplicor a head start over
Chiron's competing bDNA test, whose approval also should come soon.
Persons with baseline viral load by Amplicor should stay with that
brand since it is not completely interchangeable with bDNA.

       The company will use the 60-day period to convince Medicaid and
other third party reimbursers to pay for the test. Roche is not
committed to launching an indigent patient program for those who
cannot afford viral load testing after the first 60 days. The company
should be able to gauge the need for such assistance since those who
enroll in its baseline testing giveaway will be asked about their
insurance coverage. But any support Roche could give is complicated by
the fact that much of the testing fee -- expected to be $150 or more
initially -- goes to the commercial laboratories that perform it, not
Roche itself.

       ****************
       Treatment Briefs

       Glaxo, Merck and ADAP

       Responding to a campaign by Project Inform and the company's
own Community Advisory Board, Glaxo Wellcome has broadened the
financial assistance program for its new popular antiviral 3TC.
Previously, Glaxo refused to provide free or discounted 3TC to anyone
covered by a state ADAP (AIDS Drug Assistance Program) that refused to
include the drug in its formulary. (ADAPs, which serve 62,000 people
nationwide, pay medicine costs for uninsured people with HIV of low or
moderate income.) Glaxo refused to be the reimburser of last resort,
but that stance left people without access to 3TC in the eleven states
whose ADAPs did not include 3TC. With ADAPs across the country facing
a financial crunch (see Treatment Issues, February 1996, pages 11-12),
these people were unlikely to obtain 3TC any time soon, no matter how
much pressure Glaxo's policy put on the states.

       In announcing the change, Glaxo vice president Steve Skolsky
noted that Glaxo will not necessarily back up everyone eligible for
state ADAPs, only those who also meet Glaxo's criteria for financial
assistance.

       Merck & Co. has also come under sharp criticism for a similar
policy in regard to financial assistance for indinavir. Only three
ADAPs (Maryland, North Dakota and Washington) now pay for this
protease inhibitor, and only seven others were expected to add it
soon.

       Fortunately, as part of the Ryan White Care Act extension
passed by Congress in May, an extra $52 million will be available this
year for ADAP expenses. New York and California ADAPs have already
decided to cover protease inhibitors starting in July. Funding is
still short, though, and it remains to be seen how well the ADAPs can
keep up with the costs of the new anti-HIV treatments.

       DMP 266: Keep the Drug but Dump the Trial?

       "DMP 266" is a nonnucleoside reverse transcriptase inhibitor
(NNRTI) like nevirapine or delavirdine that was originally created by
Merck but languished when that company decided to concentrate on its
protease inhibitor program. It was recently bought by DuPont Merck, a
joint venture owned by DuPont and Merck. In its first human trials DMP
266 by itself reduced plasma HIV levels (viral load) in 20 volunteers
by a median of over 1.4 logs, or greater than 96 percent.

       Such figures make DMP 266 one of the most potent anti-HIV
agents ever tried in humans. But previous NNRTIs are notorious for the
speed with which HIV develops resistance to them, through a single
point mutation in the virus's reverse transcriptase enzyme. The
duration of DMP 266's effect remains unknown. Several mutations in
reverse transcriptase appear necessary for high-level resistance to
emerge. Researchers hope HIV strains resistant to DMP 266 are less
likely to exist prior to treatment and will be slower to evolve after
therapy commences. To ensure stability of response, though, future
trials will test the compound only in combination with other drugs
except for short two-week intervals.

       Two new trials are about to begin: One tests DMP 266 in
combination with indinavir while the other is a study of DMP 266 plus
AZT and 3TC. Details on the second trial have yet to be released, but
the first study is a fourteen-week, placebo- controlled trial taking
place at ten East Coast sites. The 360 participants must have CD4 cell
counts of between 100 and 500 and viral loads exceeding 20,000.

       The DMP 266/indinavir trial is a step toward creating simpler,
more "compact" drug combinations that strongly suppress HIV while
having fewer toxicities and simpler dose schedules than current
combination therapies. But this trial has already aroused some
criticism in the AIDS community: over 100 participants in the trial's
control arms will receive indinavir monotherapy (800 or 1,000 mg three
times a day) for twelve to fourteen weeks.

       Indinavir monotherapy is generally not advised for fear that
HIV resistant to it will easily evolve, especially in people with high
baseline viral loads. Volunteers assigned to this therapy risk not
being able to receive full benefit from indinavir and other protease
inhibitors that trigger similar resistance mutations -- ritonavir, for
example.

       PMPA Beats SIV Again

       Gilead Sciences' drug PMPA can prevent mucosal transmission of
simian immunodeficiency virus (SIV) and potently reduce SIV burden in
infected macaques, according to data presented at the Ninth
International Conference on Antiviral Research, held in Japan on May
19 to 24. These studies follow earlier data showing that administering
PMPA could protect uninfected macaques against injected SIV (see
Treatment Issues, December 1995, page 3).

       In the treatment study, ten macaques infected with SIV were
injected for four weeks with two different daily doses of PMPA. Levels
of the virus in the monkeys were reduced 100- to 1000-fold while viral
load increased in two untreated macaques. Viral load became
undetectable in eight out of ten of the treated monkeys, although SIV
DNA (a measure of infected cells rather than free virus) was still
detectable in most. After four weeks, treatment was discontinued and
SIV levels rebounded to baseline.

       There appeared to be no toxicity at the low dose, but the
higher dose caused slight reductions in red blood cells, hemoglobin
and phosphates, all of which reversed when treatment stopped. Studies
of PMPA's effect in humans are expected to begin sometime this year.

       Researchers in the transmission study treated four monkeys with
an intravaginal topical gel containing ten percent PMPA. After
intravaginal inoculation with SIV, the team performed weekly viral
detection tests to see whether the macaques had become infected. All
remained free from infection. After eight weeks of follow-up, the
monkeys were still SIV- negative. Similar studies with microbicides
such as the commercially available nonoxynol-9 have found that only
one- third of treated monkeys were protected from SIV exposure.

       Oxymetholone Adds Weight

       People with HIV experienced dramatic weight gains in a recently
published German study of the oral anabolic steroid oxymetholone and
the antihistamine ketotifen (see the British Journal of Nutrition,
January 1996; 75(1):129-38). Ketotifen is thought to increase weight
by lowering the body's production of tumor necrosis factor (see
Treatment Issues, May 1995, page 7), but it seemed to add nothing to
the observed benefit from oxymetholone. Marketed in Europe, both drugs
are available in the United States only through unofficial channels.

       The 60-person, 30-week study followed volunteers who had lost
more than ten percent of their body weight within the previous four
months. Persons who had had an active opportunistic infection in the
previous six months were excluded, as were those whose weight loss due
to chronic fever and diarrhea with no known cause.

       Thirty trial participants were randomized to two treatment
arms. One group received 50 mg of oxymetholone three times a day. The
other took the same dose of oxymetholone plus 50 mg twice a day of
ketotifen. The remaining 30 participants went untreated and served as
controls.

       Within four weeks, those on both treatments began to gain
weight. At thirty weeks, treated volunteers had increased their weight
by ten percent whereas untreated volunteers had lost four percent. By
the crude Body Mass Index, treatment led to an increase in lean body
mass. Appetite and quality of life improved in the treated patients.
Both agents were well tolerated.

       Anti-CMV Drug Trials Back On-Line

       Fomivirsen (formerly ISIS 2922) is an antisense RNA molecule
designed to attack CMV. It is under study as an intraocular injection
for CMV retinitis. Last year, Isis Pharmaceuticals placed fomivirsen
trials on hold when the dose under evaluation, 330 mg, was found to
cause a loss of peripheral vision in twenty percent of the patients
with early stage disease (see Treatment Issues, July/August 1995).
This side effect only occurred in one out of forty patients with more
advanced retinitis, though, so the company continued testing
fomivirsen in that population.

       The company then performed a dose escalation study in twelve
patients with early stage retinitis, starting with a 75 mg dose. There
was no toxicity and little treatment effect, so the researchers
increased the dose to 150 mg. This dose has produced none of the
undesirable effects, and the company claims it has produced prolonged
responses in a larger percentage of patients (although it is vague
about both the number of responders and the duration of benefit).

       Isis has now reinstated trials at the lower, safer dose in
persons with early, non-sight threatening (peripheral) CMV retinitis.
The study compares the time to progression in patients who are
immediately treated to that of patients who do not receive treatment
until after progression. Enrollment has not been rapid due to the
earlier problems. The same is true of a second trial that compares
fomivirsen plus oral ganciclovir to induction therapy with intravenous
ganciclovir followed by oral ganciclovir maintenance therapy. (Adding
oral ganciclovir to the intraocular fomivirsen injections addresses
the threat of CMV breakthrough in the other eye or elsewhere.)

       Two other studies have been opened. One is for patients with
advanced CMV retinitis and compares two different dosing regimens. The
second is for patients who have failed the other antiviral therapies
and have no remaining therapeutic options. One of fomivirsen's key
selling points is that there is no cross-resistance to other anti-CMV
drugs, so there is a possibility that the drug can offer benefit in
otherwise untreatable patients. For information about the location of
trial sites near you, call 800/679-4747.

       Copyright (c) 1996 - GMHC Treatment Issues.  Noncommercial
       reproduction encouraged.  Distributed by AEGIS, your online
       gateway to a world of people, knowledge, and resources.
       714.248.2836 (v.34+) * 8N1/Full Duplex // 714.248.0433 (ISDN).

