NEW YORK Three researchers at the Chemotherapy Foundation
Symposium XVII reported efforts to improve survival in the brain
cancers glioblastoma multiforme and astrocytoma with new
Robert L. Fine, MD, director of the Experimental Therapeutics
Program, Columbia University, is concentrating on a factor that may
block chemotherapeutic agents from reaching their targets.
Since paclitaxel (Taxol) exhibits substantial activity against human
glioblastoma cell lines in vitro but has proved only minimally
effective clinically, Dr. Fine reasoned that the blood-brain barrier
might be part of the problem. In laboratory studies, he found that
tam-oxifen (Nolvadex) was able to bind and inhibit P-glycoprotein, a
fundamental part of the blood-brain barrier.
Encouraged by rat studies combining tamoxifen and paclitaxel, Dr.
Fine is conducting a clinical trial that has enrolled 12 patients
with recurrent gliomas and 5 with metastatic brain tumors. In all
cases, he said, surgery was deemed to be the next appropriate
Patients were randomized to receive either paclitaxel alone at the
standard dose of 175 mg/m² or paclitaxel at the same dose
preceded by tamoxifen, 150 mg/m² twice daily for 6 days.
Patients had surgery 3 to 18 hours after the paclitaxel infusion
ended. Concentrations of paclitaxel in the resected tissue were
Interim results, Dr. Fine reported, showed that the combination of
tam-oxifen and paclitaxel dramatically increased deposition of the
latter drug in and around gliomas. At the center of the tumor, the
concentration doubled (mean, 2.02). At the interface of tumor and
normal tissue, it tripled. In adjacent normal tissue, it increased by
The increase in the adjacent tissue, Dr. Fine emphasized, may be
particularly relevant, since most relapses in glioma occur within 1
to 2 cm of the original resection. Paclitaxels
pharmacokinetics, he added, were not altered by the combination with tamoxifen.
For metastatic tumors, unfortunately, we did not see much
difference, Dr. Fine said. In fact, metastatic tumors, I
think, are biologically different from gliomas in the sense that
drugs penetrate better into metastatic tumors than they do into
gliomas, at least in our studies. One reason tamoxifen may not
boost pacli-taxel concentration in metastatic tumors, he suggested,
may be that these tumors do not have as much P-glycoprotein as gliomas.
In fact, Dr. Fine found that paclitaxel penetrated all areas of
metastatic brain tumors equally well, suggesting there is no blood-brain
barrier in metastatic disease. Thus, he believes that the best
way to treat metastatic brain tumors with chemotherapy is to use the
agents that have been reported to be of clinical efficacy instead of
using agents that may not be clinically useful in that tumor type but
are able to penetrate the blood-brain barrier.
Dr. Fine is also studying whether the combination of tamoxifen and
paclitaxel can improve drug penetration into the cerebrospinal fluid
(CSF). In rat studies, high-dose tamoxifen increased penetration of
paclitaxel into the CSF by 40-fold. To date, he has tried the regimen
in two breast cancer patients with carcinomatous meningitis, first
giving paclitaxel alone and then the combination. With tamoxifen
pretreatment, he reported, the levels of paclitaxel in the CSF
quadrupled at 30 minutes.
Despite his results with the tamoxifen-paclitaxel combination, Dr.
Fine remains cautious. A single tumor cell can have multiple
mechanisms of drug resistance, he said, and thus
inhibiting P-glycoprotein is not going to cure cancer because other
mechanisms of resistance are operative at the same time.
Thalidomide Plus Carboplatin
Michael L. Gruber, MD, director of neuro-oncology and clinical
associate professor of neurology, New York University Medical Center,
is investigating whether suppressing tumor angiogenesis can improve
chemotherapeutic effectiveness. He is combining thalidomide
(Thalomid) with carboplatin (Paraplatin) in patients with recurrent
glioblastoma multiforme or malignant astrocytoma.
The phase I trial to define the maximum tolerated dose of thalidomide
enrolled 19 patients, 17 of whom had glioblastomas. The researchers
dose-escalated beginning at 100 mg/m², and determined that the
maximum tolerated dose was 300 mg/m². In this trial, patients
were to receive six cycles of carboplatin if possible and to continue
on thalidomide until disease recurrence.
Among the trials 10 evaluable patients, one had a complete
response, Dr. Gruber reported, and is alive 84 weeks after initiation
of the therapy. Two other patients had a partial response, and four
had stable disease.
Of the 71 patients in the phase II trial, 40 previously had
nitrosourea-based therapy, 11 had received other forms of
chemotherapy, and 20 had only surgery and radiation.
Among the 53 patients evaluable for response, 2 had a partial
response, 35 had stable disease, and 18 had disease progression. The
median time to tumor progression was 22 weeks, and median survival
was 39 weeks. Interestingly, when we looked at survival, it had
quite a nice tail, Dr. Gruber said. About 30% of patients
were alive at 1 year.
Plans for a phase III trial, he said, call for patients in one arm to
be treated with thalidomide plus carboplatin and those in the other
arm to receive carboplatin only. The aim of the study is to determine
whether thalidomide is indeed the active agent, Dr.
In a current study in which thalidomide is the only agent being given
to patients with recurrent brain metastases, 2 of the 13 patients
enrolled have had a complete response.
Temozolomide (Temodar), recently approved by the FDA for anaplastic
astrocytoma, is being investigated for potential usefulness in
glioblastomas as well, alone and in combination with other agents,
said Victor A. Levin, MD, professor of neuro-oncology, M.D. Anderson
In a randomized trial in glioblastoma multiforme, 46% of patients had
a partial response or stable disease with temozolomide, compared with
33% of those given procarbazine (Matulane). Median progression-free
survival at 6 months was 21% with temozolomide and 8% with procarbazine.
An oral agent, temozolomide crosses the blood-brain barrier and
achieves good distribution in the CSF. Dr. Levin noted that toxicity
is much more predictable than with the nitrosoureas. With the
nitrosoureas, you sometimes have to wait 6 to 8 to 10 weeks after the
stem cell effects are over before you can re-treat, he said.
Temozolomide can be given like a clock, almost every 4 weeks.
In recurrent glioblastoma, Dr. Levin and his colleagues are
evaluating the usefulness of a combination of temozolo-mide, at a
dose of 150 mg/m2, and interferon-alfa (Intron A) given three times a
Preliminary results, he reported, show a 22% partial response or
stable disease rate in glioblastoma, and a median time to failure of
14 weeks. In anaplastic tumors, the partial response and stable
disease rate was 68%, and the median time to failure has not yet been
reached. If the combination is going to work, it will work best
in the anaplastic astrocytoma patients, Dr. Levin said.
Also underway is research combining temozolomide with carmustine
(BiCNU) in patients with glioblastoma and astrocytoma.
In another study, patients with newly diagnosed glioblastoma are
being randomized to receive either temozolomide alone or temozolomide
with prinomastat (an investigational agent from Agouron).
In patients with recurrent brain tumors, temozolomide is being given
in combination with cis-retinoic acid, which alone is one of the most
active agents against glioblastoma multiforme, Dr. Levin said.
In yet another study, temozolomide is being combined with marimastat,
an investigational matrix metalloproteinase inhibitor licensed to
Schering Plough by its British developers. In this limited
phase II study, Dr. Levin said, were asking, Do we
really have a killer therapy, something worth testing in a large
randomized, meaningful phase II study? Were not looking for a
me-too therapy. Were looking for something that will change the
direction of treatment.