Treatment of recurrent malignant gliomas
with chronic oral high-dose tamoxifen
William T. Couldwell, David R. Hinton, Amy A. Surnock, Christopher M. DeGiorgio, Leslie P. Weiner, Michael L.J. Apuzzo, Lena Masri, Ronald E. Law and Martin H. Weiss
From the Departments of Neurological Surgery [WTC, DRH, AAS, CMD, MLJA, REL, MHW], Neurology [CMD, LPW], Pathology [DRH],
and Preventative Medicine, Division of Biometry [LM],
University of Southern California School of Medicine,
Los Angeles, California
Supported by a grant from the National Institutes of Health (K08 NS01672-01) (WTC).
Please address correspondence to:
William T. Couldwell, M.D., Ph.D.,
Department of Neurological Surgery
New York Medical College
Munger pavilion, Room 329
Valhalla, NY 10595
Tel: (914) 493-8392
Fax: (914) 594-3641
High dose tamoxifen treatment for recurrent gliomas
Brain Neoplasm; Chemotherapy; Protein Kinase C; Tamoxifen
The present clinical trial was undertaken to assess the clinical safety and possible efficacy of
administering tamoxifen to patients with recurrent malignant glial tumors at dosages calculated to
achieve levels sufficient to inhibit protein kinase C within the tumor cells. Chronic oral tamoxifen
was administered in very high dosages to 32 patients (20 males:12 females, age range 26-75, mean
49 years) with histologically verified malignant glioma [anaplastic astrocytoma (AA; 12 patients) or
glioblastoma multiforme (GBM; 20 patients)], who had demonstrated clinical and radiographic
progression or recurrence following external beam radiation therapy (and additional chemotherapy
in 11; immunotherapy in 2). The dosage of tamoxifen administered was 200 mg/day to males and
160 mg/day to females given in a b.i.d. schedule. Clinical and radiographic (defined as a greater
than 50% decrease in volume of the enhancing lesion volume on MRI and a decrease in metabolic
activity on serial positron emission tomographic (PET) scans) response was noted in 8 patients
(25%; 4/12 with AA and 4/20 GBM), with an additional 6 patients (19%) exhibiting stabilization
of disease with minimal side effects. Median survival from the time of diagnosis for the entire
cohort was 24 months (104 weeks), anaplastic astrocytoma group 42.5 months (185 weeks), and
glioblastoma group 17.4 months (75.5 weeks). From the initiation of tamoxifen, median survival
for the entire cohort was 10.1 months (44 weeks), anaplastic astrocytoma group 16 months (69
weeks), and glioblastoma group 7.2 months (31 weeks). The mean length of followup of all
patients after initiating tamoxifen was 16 months (69 weeks), while the mean length of followup of
alive patients is 22.6 months (98 weeks) (range up to 51 months). These data suggest that a
subgroup of patients with malignant gliomas respond or stabilize with chronic high-dose tamoxifen
therapy. This therapy may represent an alternative or adjuvant to existing chemotherapies for these
tumors; further clinical trials are warranted.
Malignant gliomas (glioblastoma multiforme and anaplastic astrocytoma), the most common
primary tumors arising in the human brain, represent a formidable clinical challenge. Despite
advances which have been made in conventional surgical, radiotherapeutic, and chemotherapeutic
modalities, the prognosis of such patients remains poor. The median survival of patients harboring
glioblastoma, the most aggressive grade of these tumors, remains less than 12 months from the time
Previous work has demonstrated that the proliferation rates of malignant gliomas are
sensitive to inhibitors of the Protein Kinase C (PKC) intracellular signal transduction system in vitro
(1-4). Malignant human gliomas express very high PKC activity when compared to non-
transformed glial cells (2-3 orders of magnitude increase), and this high activity correlates strongly
with the proliferation rates of these tumors in vitro. These observations have supported an
important role of the PKC system in regulating glioma growth and have led to the speculation that
PKC inhibitors may be utilized as adjuvants in the therapy of patients harboring malignant gliomas.
Tamoxifen inhibits PKC activity and growth in some malignant glioma cell lines within the
micromolar concentration range in vitro, a property distinct from its estrogen receptor blockade
effect (1,5,6). Treatment of patients with recurrent malignant gliomas with low-dose oral
tamoxifen (40 mg/day) failed to demonstrate significant increased survival (7). Since the growth
and PKC inhibitory response to tamoxifen is dose-dependent (1,8,9), the present clinical trial was
undertaken to assess the clinical safety and possible efficacy of administering very high dosages of
tamoxifen to patients with recurrent malignant gliomas. The dosages chosen were calculated to
achieve target in vivo levels sufficient to inhibit the PKC signal transduction system in these cells.
Preliminary results with short followup in a subset of the patients in this study have been reported
in a brief communication (10).
Thirty-five adult patients with histologically verified malignant glioma [anaplastic astrocytoma
(WHO grade III) or glioblastoma multiforme (WHO grade IV)], whom had demonstrated clinical
and radiographic progression or recurrence following radiation were enrolled in an open-label
prospective study administering daily high-dose oral tamoxifen. In addition, 11 patients had
previously been administered cytotoxic chemotherapy, and 2 patients had received prior
immunotherapy; tamoxifen was initiated following demonstrated failure and discontinuation of
these treatments. To be eligible for enrollment in the study, all patients had demonstrated
increasing volume of gadolinium-enhancing lesion(s) on serial post-radiotherapy Magnetic
Resonance Images (MRI) and either high metabolically active lesion compatible with recurrent
tumor as measured by Positron Emission Tomography (PET; 18FdG uptake; 13 patients), or
recurrent malignant glioma verified histologically by open surgical resection (12 patients) or
stereotactic biopsy (10 patients). Biopsy was performed in all cases in which the PET did not
indicate significantly increased metabolic activity compatible with gross tumor. Informed consent
was obtained in all patients in accordance with the Institutional Review Board of the University of
Southern California. The patients had no intercurrent illness such as other malignancy, history of
previous malignancy, blood dyscrasias, gynecological, ocular or gastrointestinal disease. All
pathology was reviewed by a neuropathologist (Dr. Hinton). Baseline blood work (Complete
Blood Counts [CBC] and Serum Chemistry) was performed prior to initiation of therapy.
Tamoxifen (ICI Pharmaceutical, Wilmington, Delaware) was first administered for 4 days at
standard antiestrogen doses (20 mg orally b.i.d.) to observe for any side effects. If tolerated, the
dose was increased weekly to achieve target dose over a 1 month period (80 mg b.i.d. in females,
100 mg b.i.d. in males). CBC and Serum Chemistry panel were performed after every 2 weeks
while escalating the dose and every 2 months thereafter while on the drug. Patients included for
evaluation included those who tolerated the drug at maximal dosages for a period of 2 months or
longer, survived for at least one month following attainment of maximal therapy (to enable
significant steady-state tissue levels of tamoxifen to be obtained), and were clinically and
radiographically followed during the treatment period. Thirty-two of 35 patients were evaluable
(Table 1; 20 males:12 females, 12 anaplastic astrocytomas and 20 glioblastomas, age range 26-75,
mean 49 years). Two patients were non-compliant and one patient expired from progressive
disease within 3 weeks of the start of treatment.
All patients underwent initial and serial MRI studies with and without gadolinium
enhancement every three months during tamoxifen treatment. Initial tumor volume was estimated
by measuring the cross-sectional diameters at the level of the largest contrast-enhancing tumor
extent on axial images. These measures were multiplied, and the product was multiplied by the
extent of maximal enhancement on coronal studies. Estimation of the tumor volume excluded areas
of cystic change or edema. Subsequent comparison studies chose comparable MRI slices with
tumor measurement by the same technique, including contrast dosages. In addition, serial PET
scans were performed on 28 of 32 patients during treatment. Treatment response was defined as a
greater than 50% decrease in volume of the enhancing lesion volume on MRI and a decrease in
metabolic activity (18FdG uptake) on PET scans with clinical neurological improvement (including
Karnofsky scores). Stabilization or no change was defined as <50% reduction in tumor volume
radiographically and no clinical progression. Progressive disease included all other patients
(worsening clinically or radiographically). Corticosteroid dosages in all patients were either
maintained or decreased on comparison radiographic studies.
Among the twenty patients harboring glioblastoma, clinical and radiographic response was noted in
4 (20%; transient in 2 patients with recurrence occurring at 5 and 22 months of therapy), and
stabilization occurred in 4 (20%) other patients (subsequent progression and death occurred in 2 of
these patients after discontinuing tamoxifen for compassionate reasons because of low Karnofsky
scores after stabilization). The responding patients have been followed up to 51 months (1 achieved
complete radiographic remission after 24 months and has been discontinued from tamoxifen
therapy). Among the 12 patients with anaplastic astrocytoma, clinical and radiographical response
was noted in 4 (33%; transient in one patient with recurrence following 14 months of therapy), and
stabilization of disease occurred in an additional two patients (17%). Median survival from the
time of diagnosis for the entire cohort was 24 months (104 weeks), for those patients harboring
glioblastoma 17.4 months (75.5 weeks), while those patients with anaplastic astrocytoma achieved a
median survival of 42.5 months (185 weeks). From the initiation of tamoxifen, median survival
for the entire cohort was 10.1 months (44 weeks), for the glioblastoma group 7.2 months (31
weeks), and for the anaplastic astrocytoma group 16 months (69 weeks). The mean length of
followup of all patients in the study from initiation of tamoxifen was 16 months (69 weeks). The
mean length of followup of alive patients is 22.6 months (98 weeks). Kaplan-Meier survival plots
from the time of initiation of tamoxifen therapy for the two histological subtypes are shown in
A major complication encountered was deep venous thrombosis occurring in 2 elderly
males on maximal therapy, which was managed by anticoagulation following discontinuation of
therapy (both of these patients had not demonstrated response). Four other patients exhibited side
effects during the course of their treatment; one patient developed nausea following dose escalation
of tamoxifen which prompted temporary withdrawal of therapy, with no resolution of the symptom.
The drug was restarted and the nausea slowly resolved. Other minor complications included "hot
flashes" experienced by a young female just after initiation of therapy, and fatigue with maximal
therapy in 2 other patients. No evidence of tamoxifen-associated retinopathy was noted among the
These data indicate that a subgroup of patients with malignant gliomas respond or stabilize with
chronic oral high-dose tamoxifen therapy. Furthermore, the results indicate that high-dose oral
tamoxifen appears to be well tolerated in most patients. In addition, 6 patients whom had
previously failed standard nitrosourea chemotherapy either stabilized or responded, indicating that
the therapy may play a role as salvage therapy following standard chemotherapy in some patients.
Prediction of the patients that may respond to chronic high-dose tamoxifen is unknown at
present, but appears to be similar to the percentage of low-passage glioma lines that are sensitive in
vitro (1) suggesting that in vitro-sensitivity testing may be feasible. The mechanism for the clinical
antitumor effect in these patients is unknown. The observations of the limited clinical efficacy of
the therapy when given in dosages calculated to block the estrogen receptor (7), and the fact that the
concentration of tamoxifen necessary for growth inhibition in these cells lies within the micromolar
range in vitro (greater than necessary for blockage of the estrogen receptor; Ref. 1) suggest that
alternative mechanisms may be responsible. The mechanism may involve inhibition of PKC
activity in the glioma cell, since tamoxifen has been demonstrated to inhibit PKC activity and
growth as well as induce apoptosis in these cells in vitro (1,6,8,9). However, alternative
pleiotropic effects of the drug on these cells may be responsible for the clinical efficacy noted.
Finally, previous work has demonstrated that PKC inhibitors, including tamoxifen, may
also potentiate the effect of radiation therapy on malignant glioma cells (11). In this regard, the
radiosensitivity of intrinsically radioresistant glioma cells may be enhanced by depletion of PKC in
vitro (12). These data, taken together with the clinical data from the present study suggest that
high-dose tamoxifen or other more potent PKC inhibitors may become useful adjuvant treatment
for the patients with malignant glioma both during and following radiation therapy.
1. Baltuch, G., Couldwell, W.T., Villemure, J.G., Yong, V.W. Protein Kinase C inhibitors
suppress cell growth in established and low passage glioma cell lines. A comparison
between staurosporine and tamoxifen. Neurosurgery, 33: 495-501, 1993.
2. Couldwell, W.T., Antel, J.P., Yong, V.W. Enhanced protein kinase C activity correlates
with the growth rate of malignant gliomas: Part II. Effects of glioma mitogens and
modulators of PKC. Neurosurgery, 31: 717-724, 1992.
3. Couldwell, W.T., Uhm, J., Antel, J.P., Yong, V.W. Enhanced protein kinase C activity
correlates with the growth rate of malignant human gliomas. Neurosurgery, 29: 880-887,
4. Nishizuka, Y. Intracellular signalling by hydrolysis of phospholipids and activation of
protein kinase C. Science, 258: 607-614, 1992.
5. O'Brian, C.A., Liskamp, R.M., Solomon, D.H., Weinstein, I.B. Inhibition of protein
kinase C by Tamoxifen. Cancer Res., 45: 1462-2465, 1985.
6. Pollack, I.F., Randall, M.S., Kristofik, M.P., Kelly, R.H., Selker, R.G., Vertosik, F.T.
Effect of Tamoxifen on DNA synthesis and proliferation of human malignant glioma lines
in vitro. Cancer Res., 50: 7134-7138, 1990.
7. Vertosik, F.T., Selker, R.G., Pollack, I.F., Arena, V. The treatment of intracranial
malignant gliomas using orally administered tamoxifen therapy: preliminary results in a
series of "failed" patients. Neurosurgery, 30: 897-903, 1992.
8. Couldwell, W.T., Hinton, D.R., He, S., Chen, T., Sabit, I., Weiss, M.H., Law, R.
Protein Kinase C inhibitors induce apoptosis in human malignant glioma lines. FEBS Lett.,
345: 43-46, 1994.
9. Pollack, I.F., Kawecki, S. The efficacy of tamoxifen as an antiproliferative agent in vitro
for benign and malignant pediatric glial tumors. Pediatr. Neurosurg., 22: 281-288, 1995.
10. Couldwell, W.T., Weiss, M.H., Weiner, L.P., DeGiorgio, C.M., Hinton, D.R.,
Ehresmann, G.R., Conti, P.S., Apuzzo, M.L.J. Clinical and radiographic response in a
minority of patients with recurrent malignant gliomas treated with high dose tamoxifen.
Preliminary report. Neurosurgery 32: 485-490, 1993.
11. Zhang, W., Yamada, H., Sakai, N., Niikawa, S., Nozawa, Y. Enhancement of
radiosensitivity by tamoxifen in C6 glioma cells. Neurosurgery, 31: 725-730, 1992.
12. Zhang, W., Law, R.E., Hinton, D.R., Anker, L., Weiss, M.H., Couldwell, W.T.
Radiosensitization of malignant human glioma cells by hypericin in vitro. Clin Cancer Res
2: 843-846, 1996.
Figure 1. Kaplan-Meier survival plots for the two histological subtypes anaplastic astrocytoma
(upper graph) and glioblastoma multiforme (lower graph). The time period of the
horizontal axis indicates the time from initiation of tamoxifen therapy.
Couldwell et al.