This is an old version of the article... For the newest version, always go to:
http://www.virtualtrials.com/williams.cfm

Treatment Options for Glioblastoma and other Gliomas

Prepared by Ben A. Williams

Glioblastoma Diagnosis, March, 1995

Last Updated: September 9, 2004

Copyright 2004, Ben Williams

Disclaimer: the information presented here is the opinion of Ben Williams.
It is for informational purposes only, do not consider it medical advice. Discuss the ideas presented here with your own doctors.

---

Since my own diagnosis of glioblastoma (GBM) I have spent considerable time researching the literature for treatment options, and the following discussion summarizes what I have learned. Most of the information is from medical journals. Some is from information that has been contributed by others to various online brain tumor patient support groups which I have followed up on, and some is from direct communications by phone or e-mail with various physicians conducting the treatments that are described. References are presented at the end for those who would like their physicians to take this information seriously. Although this discussion is intended to be primarily descriptive of the recent development of new treatment options, it is motivated by my belief that the development of new agents, per se, is likely to fall short of providing effective treatment. What is needed, in addition, is a new approach to treatment that recognizes the power of evolution as the enemy of victims of cancer.

A more extensive account of my philosophy of treatment, and the reasons for it, are provided in my (2002) book, 'Surviving "Terminal" Cancer: Clinical Trials, Drug Cocktails, and Other Treatments Your Doctor Won't Tell You About'. It can be ordered elsewhere on this website, from Amazon.com, from your local bookstore, or directly from the publisher:

Fairview Press 2450 Riverside Ave. Minneapolis, MN 55454 1-800-544-8207 FAX: 612.672.4980 www.fairviewpress.org

Treatment for GBMs and other high-grade gliomas is changing rapidly. Until the last five years there was a standard treatment in the USA, including surgery, radiation, and chemotherapy with a nitrosourea, either BCNU alone or CCNU combined with procarbazine and vincristine (known as the PCV combination). While this treatment has worked for a small minority of people, its 5-year survival rate has been only 2-5%. Alternatives to this traditional treatment regimen are imperative if GBM patients are to have any realistic hope of surviving their disease. Fortunately, new treatments are now being introduced at a rapid rate. But unfortunately, most still are not available outside of clinical trials and there is no consensus about what is the best treatment for this deadly disease.

There are two general premises to the approach to treatment that will be described. The first of these is borrowed from the treatment approach that has evolved in the treatment of AIDS. Both HIV and cancer involve biological entities that mutate at high rates. This implies that unless a treatment is immediately effective the dynamics of evolution will create new forms that are resistant to whatever the treatment may be. However, if several different treatments are used simultaneously (instead of sequentially, which is typically the case), any given mutation has a much smaller chance of being successful.

A second general principle is that any successful treatment will need to be systemic in nature because it is impossible to identify all of the extensions of the tumor into normal tissue. Moreover, cancer cells are typically evident in locations in the brain distant from the main tumor, indicating that metastases within the brain can occur, although the great majority of tumor recurrences are within or proximal to the original tumor site. Localized treatments such as radiosurgery may be beneficial in terms of buying time, but they are unlikely to provide a cure. Even if the localized treatment eradicates 99.9% of the tumor, the small amount of residual tumor will expand geometrically and soon will cause significant clinical problems.

Until recently, the only systemic treatment available has been chemotherapy, which historically has been ineffective except for a small percentage of patients. An important issue, therefore, is whether chemotherapy can be made to work substantially better than it typically does. Agents that facilitate or augment its effects are critically important. Such agents are available but not widely used. Also becoming available are new systemic treatments that are much less toxic than traditional chemotherapy. The availability of these treatments raises the possibility that some combination of these new agents can be packaged that is substantially less toxic and yet provides effective treatment based on several different independent principles. Thus, the AIDS-type of combination approach is now a genuine possibility whereas it would not have been ten years ago. Because oncologists have been slow to appreciate the significance of the increased availability of these relatively nontoxic treatments, patients learn about them piecemeal if at all. Thus, patients themselves need to become familiar with these new agents and the evidence available regarding their clinical effectiveness. It is possible, although by no means proven, that some combination of these new agents offers the best possibility for survival.

Patients may or may not learn about the treatments that will be described from their physicians. To appreciate why this may be, it is important to understand how American medicine has been institutionalized. For most medical problems there is an accepted standard of what is the best available treatment. Ideally this is based on phase III clinical trials. Treatments that have been studied only in nonrandomized phase II trials will rarely be offered as a treatment option, even if the accepted "best available treatment" is generally ineffective. What happens instead is that patients are encouraged to participate in clinical trials. The problem with this approach is that most medical centers offer few options for an individual patient. Thus, even though a given trial for a new treatment may seem very promising, patients can participate only if that trial is offered by their medical facility. An even more serious problem is that clinical trials with new treatment agents almost always study that agent in isolation, usually with patients with recurrent tumors who have the worst prognoses. For newly diagnosed patients this is at best a last resort. What is needed instead is access to the most promising new treatments, in the optimum combinations, at the time of initial diagnosis.

Physicians rarely will inform a patient about clinical trials being conducted elsewhere. Moreover, the idea that several different agents from separate phase II clinical trials might be combined will be met with great resistance. Patients themselves will therefore need to become informed about what options are available, and which kinds of combinations seem most promising. In addition to the information that will be presented here, other information, especially about which new clinical trials are available, are available elsewhere on this website (address: www.virtualtrials.com).

The Role of Chemotherapy

A continuing debate in neuro-oncology is whether chemotherapy is appropriate for patients with glioblastoma as the diagnosis. A number of different clinical trials have failed to show a significant survival advantage from the addition of chemotherapy to radiation, although a meta-analysis (a somewhat controversial statistical procedure) of the entire corpus of clinical trials has shown an increase in the median survival time of 2-3 months. Given such a small increase, the rigors of undergoing chemotherapy, for many, seem to outweigh the minimal increase in survival. This is especially true in Europe. A recent large phase III clinical trial conducted in Europe has created even more misgivings about the efficacy of chemotherapy, as it reported that the use of the PCV chemotherapy regimen produced no statistically significant difference in median survival rate (1). However, that study has been significantly criticized on several grounds (2).

The debate over the use of chemotherapy has unfortunately focused on the median survival time as the measure of chemotherapy's effects. Chemotherapy, at least that which has been investigated up until the present time, benefits only a minority of patients, but those who do benefit may gain a great deal. Approximately 20-30%of patients receiving traditional chemotherapy show some shrinkage in their tumors, and such shrinkage is associated with increased survival times. The problem with using medians as the primary statistic is that they are insensitive to changes in the distribution of effects that occur in only one tail of the distribution. A much better statistic, in my opinion, is the two-year survival rate. Patients who receive only radiation have a two-year survival rate of 0-10%, while those also receiving chemotherapy have a two-year survival rate of 15-25% (3). Thus, the use of chemotherapy increases two-year survival by a significant amount. On the other hand, it will have little benefit for the majority of patients. But given the poor prognosis from all forms of GBM treatment, a patient cannot afford to forego the possibility that he/she will be among the minority for whom chemotherapy provides a significant benefit.

One approach to determining whether an individual patient will benefit from chemotherapy is simply to try 1-2 rounds to see if there is any tumor regression. The debilitating effects of chemotherapy typically occur in later rounds, at which point there is a cumulative decline in blood counts. The extreme nausea and vomiting most associated with chemotherapy in the mind of the lay public is now almost completely preventable by the new anti-nausea agents, Zofran and Kytril. Marijuana also can be very effective in controlling such effects, and has an appetite-boosting effect as well. Thus, for those patients who are relatively robust after surgery and radiation, some amount of chemotherapy experimentation should be possible without major difficulties.

An alternative way to ascertain the value of chemotherapy for an individual patient is the use of chemo-sensitivity testing for the various drugs that are possible treatments. Such testing requires a live sample of the tumor and thus must be planned for in advance of surgery. Culturing the live cells is often problematic, but at least a half-dozen private companies across the country offer this service. Cost ranges from $1000-$2000 depending on the scope of drugs that are tested. Recent evidence has shown that chemosensitivity testing can significantly enhance treatment effectiveness for a variety of different types of cancer, including a recent Japanese study using chemosensitivity testing with glioblastoma patients (4). In general, when chemosensitivity testing indicates an agent has no effect on a patient's tumor the drug is unlikely to have any clinical benefit. On the other hand, tests indicating that a tumor culture is sensitive to a particular agent do not guarantee clinical effectiveness, but it substantially increases the likelihood that the agent will be beneficial. More information about chemosensitivity testing is presented in a separate article listed in the "noteworthy treatments" section that includes the present paper.

The poor success rate of chemotherapy for glioblastomas is not necessarily true of other types of glioma. For both anaplastic astrocytoma (AAIII) and oligodendrogliomas, the PCV chemotherapy regimen has been shown to have substantial effectiveness. In one of the most recent studies, for example, PCV combined with the radiation sensitizer, DFMO, produced median survival times for AAIII patients of 5-6 years (5).

Which Chemotherapy?

Until the past 3-5 years the standard chemotherapy for gliomas was either BCNU as a single agent or the PCV combination, which consists of the sequential presentation of CCNU ( an oral chemical sibling of BCNU), procarbazine, and vincristine. . A major change in the chemotherapy used for gliomas has occurred following the introduction of a new drug called temozolomide (trade name: temodal in Europe, temodar in the USA). This is an interesting drug in part because its development (in England) was based on chemosensitivity assays using a wide variety of cultures of brain-tumor cells. Surprisingly, the other common chemotherapy agents used for brain cancer were developed not from their effect in the laboratory on cultured brain tumor cells but on cell cultures from other types of cancer. Perhaps as a consequence temozolomide appears to produce a notably higher response rate. At the time of this writing, it has received approval by the FDA for the treatment of grade-3 gliomas (anaplastic astrocytomas) but has been denied approval for the treatment of glioblastomas. However, the European equivalent of the FDA has approved the drug as a treatment for all forms of glioma, including glioblastoma. The basis of the FDA denial of approval for glioblastomas was a phase III clinical trial involving patients with recurrent tumors, in which temozolomide was compared with procarbazine, a commonly used second-line chemotherapy for recurrent glioblastomas (6). Three different measures of clinical efficacy were assessed: 1. the percentage of patients who had no tumor progression for at least six months after initiation of treatment: (21% for temozolomide, 9% for procarbazine); 2. the median time after treatment initiation before tumor progression was evident: (2.9 months vs. 1.9 months); 3. the median survival time: (7.3 vs. 5.8 months). The first two of these differences were statistically reliable using the standard .05 criterion. The difference in median survival time was not statistically significant. The FDA apparently focused on the median survival time, and for that reason declined to approve temozolomide as a new treatment for glioblastomas. (As I have discussed in my book, cited above, this decision, along with numerous others, demonstrates how the FDA's decisions have frequently worked against the best interests of cancer patients, often denying them access to promising new treatments because of arbitrary criteria).

In actual fact the FDA's refusal to approve temodar for glioblastomas has not prevented its use, and it is rapidly becoming the preferred drug for recurrent gbms and increasingly as the initial treatment after radiation. The reason this is possible is that an oncologist can prescribe a drug for any purpose as long as the drug has FDA approval for some purpose (known as off-label usage). Part of the reason for temodar's increasing usage is that it has significantly less toxicity than the drugs traditionally used, especially with respect to the cumulative suppression of the stem cells in the bone marrow that make blood cells. Moreover, it is important to keep in mind that none of the conventional chemotherapy drugs used for glioblastomas has evidence supporting their use in terms of the clinical trial outcomes that the FDA has required for temodar.

On the other hand, it remains unclear what the long-term outcomes of using temodar will be. A recent review of clinical trials using temodar concludes that while it appears to produce a higher response rates than previously-used chemotherapies, there is little evidence that it increases overall survival (7). A recent phase II study using temozolomide as a single agent, but starting the chemotherapy before the standard radiation treatment, resulted in tumor shinkage in 42% of patients with glioblastomas, but a median survival time of only 13.2 months, only slightly longer than the typical 11-12 month survival seen with previously used chemotherapies (8). However, at the most recent meeting of the American Society of Clinical Oncology (ASCO), a large European study was reported establishing that temozolomide clearly is beneficial compared to radiation treatment alone, although it remains unclear how much better it is than the previously used chemotherapy protocols. The protocol for the European study was to present low-dosage temodal during the standard six weeks of radiation, then continue the drug after radiation was completed with a higher dosage on a schedule of 5 consecutive days of drug per 30 day cycle.(9). Median survival was 15 months, compared to a median survival of 12 months for patients receiving radiation only, a difference that was statistically significant. More impressive was the difference in two-year survival rate, which was 26% for the patients receiving temodal but only 8% for those receiving only radiation. Very similar outcome results were obtained in a smaller nonrandomized study in Germany (10) As a result of these new findings, , the protocol of temodar presented during radiation is rapidly becoming the most popular first line treatment for glioblastoma patients.

The best results from using temodar as the initial treatment comes from a small phase II study done in Switzerland, which combined temodal with thalidomide, starting after the standard radiation treatment (11). Subjects received either thalidomide alone or thalidomide + temodal. The median survival time for the thalidomide-alone group was 63 weeks, while that for the group with thalidomide + temodar was 103 weeks. But the latter group involved only 25 patients, so it is obviously important to replicate these results.

Because temodar has substantially less toxicity than previously used chemotherapy drugs, there are now numerous clinical trials underway that combine temodar with other agents. These trials are almost all phase II trials conducted with patients with recurrent tumors, i.e., patients for whom initial treatment has not prevented regrowth of the tumor. The clinical history of such patients is highly varied, which greatly complicates the evaluation of the outcomes of the trials. Previous phase II trials have been evaluated in terms of the percentage of patients who have tumor shrinkage as a result of receiving the treatment. But this has become problematic because many of the new biological agents, which will be discussed later, are not directly cytotoxic but work instead by inhibiting tumor growth. New criteria of evaluation have thus been adopted, the most popular being the median time before tumor progression, and increasingly, the percentage of patients who have no regrowth of tumors six months after treatment initiation (known as "6-month progression-free survival" {PFS-6}). Because phase-II trials have only the outcomes of previous trials as historical controls, the new temozolomide combination trials are being compared to those previous outcomes. A recent compilation of results from eight different phase-2 clinical trials has reported that the average PFS-6 for all trials was 15% (12). The results for temodar as a single agent are slightly better (21%), but hardly enough to be remarkable (6). But when temodar is combined with accutane (also known as 13-cis-retinoic acid, to be discussed later), the PFS-6 improved to 35% (13) When combined with a new drug called marimastat (14), PFS-6 was 39%. Marimastat is one of the new cytostatic drugs which stops tumor growth by an inhibiting the enzyme process whereby the tumor digests the extracellular matrix of surrounding cells, allowing the tumor to invade the adjacent tissue.. But marimastat has the unfortunate side effect of severe arthralgia and also is not available outside of clinical trials.

Temodar has also been combined with several conventional chemotherapies. When combined with CPT-11, drug developed for colon cancer but now being intensively studied in its own right as a treatment for brain cancer, the PFS-6 was 38-39% (15, 16). When combined with VP-16, a drug often used for gliomas in children, the PFS was 67%, although this trial involved only 12 patients (17). The combination of temodar with BCNU is also being studied, but has been complicated by issues of toxicity and the optimal schedule of dose administration for the two drugs. However, a recent published report failed to show any benefit of combining BCNU with temodar, compared to temodar alone, as the PFS-6 for the combination was only 11 weeks, accompanied by considerable toxicity (18). Better results have been obtained when temodar has been combined with cisplatin, In a pair of clinical studies performed in Italy (19, 20), the PFS-6 was 34% and 35%. Temodar has also been combined with procarbazine (21). While the report of that study did not include the PFS-6 statistic, it did report an unusually high percentage of tumor regressions, suggesting that this combination might be effective.

There are also several clinical trials underway combining temodar with a variety of new biological agents that hold great promise of improving outcomes without increasing treatment toxicity. These include drugs that target the signaling pathways involved in cell division, and agents that inhibit the growth of new blood vessels. In the latter category is a trial conducted jointly by several hospitals in New York which combined temodar with celebrex, the anti-inflammatory drug that is now widely used for arthritis (22). For the 29 patients in the study, the PFS-6 was 35%.. I will discuss several of these new agents in greater detail in later sections.

It is important to recognize the limitations of the PFS-6 measure of treatment efficacy. While it does provide a rough means of comparing different treatments, it says very little about whether the various treatment protocols will improve overall survival. It is entirely possible that treatments with low PFS-6 values produce a greater percentage of long-term survivors than those with higher PFS-6 values. Nevertheless, one major conclusion allowed by the above comparisons is that combinations of treatments are notably superior to single-agent treatments, and that the combinations can include agents of relatively mild toxicity (e.g.,accutane, celebrex). It is entirely feasible that the use of such lower-toxicity agents will allow combinations involving 3 and 4 different agents, which presumably should improve treatment outcome still further.

While temodar is becoming the drug of choice for the initial treatment of glioblastoma, it is clear from the above discussion that the majority of patients will receive minimal benefit. Unlike a generation ago, it is now common for patients who have failed one chemotherapy to proceed to other chemotherapy drugs. These include the nitrosoureas, BCNU and CCNU (and ACNU in Europe and Japan), but also the platinum drugs, and irinotecan, a new drug developed for colon cancer known as (also known as CPT-11).

An important variation in the use of BCNU as the chemotherapy agent has been the development of polymer wafers known as gliadel. A number of such wafers are implanted throughout the tumor site at the time of surgery. The BCNU then gradually diffuses from the wafers into the surrounding brain. A possible problem with the treatment is that the drug will diffuse only a small distance from the implant sites, so that significant portions of the tumor will not make contact with the drug. A phase III clinical trial has demonstrated that survival time for recurrent GBM is significantly increased by the gliadel wafers relative to control subjects receiving wafers without BCNU, although the increase in survival time, while statistically significant, was relatively modest (23). The median survival time from the time of re-operation for the recurrent tumor was 31 weeks, while that for the placebo control group was 23 weeks. Survival rates six months after the treatment were 56% for the gliadel group while 36% for the placebo group. On the other hand, the differences in survival between the two groups was near zero when measured one year after treatment, indicating that the beneficial effects of gliadel were relatively short-term in nature. A second small randomized clinical trial was conducted in Europe, but involving patients who received gliadel at the time of initial surgery as a primary treatment, rather than as treatment for recurrent tumors (24). Here the survival rate after one year was 63% but only 19% for those receiving the placebo. The two-year survival rate was 31% of the gliadel patients compared to only 6% for the placebo patients. While these differences were statistically significant, it is important to recognize that the comparison condition was a placebo control. It is unclear whether gliadel is superior to other control conditions such as the IV BCNU that is commonly administered (although the two-year survival rate does seem unusually high). Indeed, there is reason to believe that the difference, if any, would be minimal, in that the median survival time for glioblastoma is approximately 1 year. Moreover, both gliadel clinical trials involved patient populations that included approximately 1/3 of the patients with diagnoses other than glioblastomas, so the survival times that were obtained are somewhat inflated from what they would have been if only glioblastoma patients had been included. Probably the best estimate of the benefit of gliadel as an initial treatment comes from a. third much larger randomized clinical, also done in Europe (25) which reported a median survival of 13.9 months for patients receiving gliadel compared to a median survival of 11.6 months for patients implanted with placebo wafers. As with other forms of chemotherapy, however, larger differences are evident for long-term survival. After a follow-up period of 3-4 years, 9 of 120 patients who received gliadel were alive, compared to only 2 of 120 of those receiving the placebo. Such results are not notably different from the historical results with BCNU given intravenously. But gliadel has the major advantages over intravenous BCNU that it avoids the systemic side effects of IV BCNU, which can be considerable, not only in terms of low blood counts but also in terms of a significant risk of major pulmonary problems. However, treatment with gliadel produces its own side effects, including an elevated risk of intracranial infections and seizures (26). However, the lack of systemic toxicity for gliadel makes it a candidate for various drug combinations (e.g., temodar + gliadel). Clinical trials involving such combinations are now underway.

Although gliadel has FDA approval, medicare, and some insurance companies, have until recently refused coverage for its usage. But thanks to a letter writing campaign organized by the Musella Foundation, Medicare recently has now agreed to its coverage.

Many of the standard forms of chemotherapy for other types of cancer have also been tested against glioblastomas. Most of these tests have involved patients with recurrent tumors, for whom the prognosis is especially grim. The historical norm for patients with recurrent glioblastomas who are given some form of additional chemotherapy has been a survival time of 3-6 months, although there is enormous variability. Some of the better results have occurred with the platinum drugs cisplatin and carboplatin, with carboplatin now the preferred drug because it has considerably less toxicity. In a representative study of carboplatin (27), of 29 patients with recurrent glioma, 4 achieved partial tumor regressions, and another 10 achieved stable disease, for a response rate of 48%. Of those responding to carboplatin, the median time to tumor progression was 26 weeks. However, other treatment studies using the platinum drugs have produced highly variable results, with the source of the variability not clearly identifiable. Considerable attention has been given to improving the effectiveness of these drugs by combining them with other agents. One recent study of carboplatin has used intra-arterial infusion in combination with RMP-7 (Cereport), an agent that disrupts the blood-brain barrier. A clinical trial presented at the 1998 meeting of the American Society of Clinical Oncology reported a median survival time of 37 weeks for 37 patients with recurrent GBM (28). However a recent randomized clinical trial compared IV carboplatin with or without RMP-7 and found no advantage to adding RMP-7 (29).

More impressive results using cisplatin has come from its implantation directly into the tumor bed in polymer wafers similar to gliadel. A study in Belarus reported that patients receiving the cisplatin wafers at the time of initial surgery had a median survival time of 428 days, compared to 211 days for patients who received only radiation (30). Positive results using carboplatin in combination with VP-16 (also known as etoposide) have also been recently reported (31). Patients with recurrent gliomas had a median survival time of 14.5 months, with a median time to progression of 9.6 months. This trial included both glioblastoma and grade III gliomas, and the statistics for the different diagnoses were not separately reported.

Yet another drug commonly used for other forms of cancer, taxol, has also been investigated as a treatment for gliomas but without notable success for glioblastomas (32). However, this probably was due to taxol not crossing the blood-brain barrier. When taxol is directly into the brain via catheters, in combination with a new convection system designed to cause it to diffuse thoughout the brain,11 of 15 glioblastoma patients with recurrent tumors had tumor regressions, with 5 complete regressions (33). This is the best result of chemotherapy reported in the literature that I have seen. However, significant toxicity accompanied the treatment and it is unclear whether it is still being pursued.

One of the newer chemotherapy agents is CPT-11 (also known as irinotecan), which has been FDA-approved for the treatment of colon cancer. Its application to gliomas has been pioneered by Dr. Henry Friedman at Duke University and is now undergoing clinical trials at a number of other medical centers as well. The initial results from the early trial were that 9 of 60 patients with recurrent gliomas had a confirmed partial response, while an additional 33 patients had stable disease lasting more than 12 weeks (34). However, results from other reported studies have been less positive (35, 36). Part of the reason for the discrepant outcomes appears to be that CPT-11 interacts pharmacologically with anti-seizure medications, causing its serum concentration to be decreased.

Like temodar , CPT-11 is now being studied in various combinations with other chemotherapy regimens, notably gliadel, intravenous BCNU, and temodar, but whether these combinations are more effective than CPT-11 alone has not been definitively determined. One interesting sidelight about CPT-11 is that the gastro-intestinal toxicity that it produces, which can be severe, is substantially attenuated by low dosages of thalidomide (see below for further discussion of thalidomide as a treatment agent in its own right). However, there is some concern thalidomide might also interact pharmacologically with CPT-11 metabolism to decrease its concentration in the body.

One of the most promising new chemotherapy regimens, apparently being neglected in this country, involves mitoxantrone, a drug which in cell culture studies has been shown to be among the most toxic to glioblastoma cells. In a study done in Italy (37) patients with recurrent glioblastomas were treated either with the PCV protocol alone, or in combination with repeated administration of mitoxantrone to the tumor cavity via a catheter placed there during surgery. Median survival, measured from the time of treatment initiation for tumor recurrence, was 6 months for the PCV-only subjects but 12 months for those receiving PCV + mitoxantrone. and 16.8 months for those receiving PCV + mitoxantrone and also surgery to remove the recurrent tumor. Two-year survival was 30% for those receiving mitoxantrone, compared to only 5% for those receiving PCV alone.

Increasing Chemotherapy Efficacy

As noted earlier, an important issue is whether chemotherapy can be made more effective by a variety of agents which themselves are not cytotoxic in nature. Brain tumors, and other types of cancer, possess active pump-like mechanisms by which the chemotherapy agent is extruded from the cell body. One of these pump mechanisms utilizes calcium channels, so that calcium channel blockers can interfere with its action, thus allowing the chemotherapy agent longer time to be effective. This is important because chemotherapy is effective only when cells are dividing, and only a fraction of the cell population is dividing at any given time. The longer the chemotherapy remains in the cell, the more likely it will be there at the time of cell division. If the extrusion of the chemotherapy drug could be inhibited, chemotherapy should in principle become more effective. Calcium channel blockers, which include commonly used medications for hypertension such as verapamil, have thus been studied for that purpose (38). Unfortunately, these agents have potent effects on the cardiovascular system, so that dosages sufficiently high to produce clinical benefits usually have not been achievable. However, a recent study (39) did report a substantial clinical benefit for patients with breast cancer with a relatively low dosage (240 mg/day). In addition, the combination of verapamil with tamoxifen (which itself blocks the extrusion pump by a somewhat different mechanism) may possibly increase the clinical benefit (40). In laboratory studies other calcium channel blockers, especially, nicardipine and nimodipine (41, 42) have also been shown to effectively increase chemotherapy effectiveness, and may have direct effects on tumor growth themselves.

Most recently, a new facilitator of chemotherapy, 06-benzylquanine, has been the subject of clinical trials, but as yet no data have been published that allow an evaluation of whether chemotherapy is more efficacious as a result of addition of this new drug to the treatment regimen. The statin drugs used for the treatment of high cholesterol levels, such as simvastin, have also been shown to augment the effects of BCNU in laboratory studies (43), but have not yet been combined with chemotherapy in any reported clinical study. Most recently, a common drug used in the treatment of alcoholism, Antabuse (also known as disulfuram), has been shown in laboratory studies to be a powerful inhibitor of the extrusion pump mechanism, although as yet this has not been studied clinically (44). One potential concern about increasing chemotherapy effectiveness via any of the foregoing methods is that toxicity to normal cells that are in the process of dividing (notably the bone marrow cells producing blood cells) may also be increased. That is, keeping the chemotherapy agent in the cell for a longer period of time may be functionally similar to higher dosages of the chemotherapy agent. As yet it is unclear whether this in fact will be the case. A promising method of dealing with this problem is to combine 06-BG with gliadel or other agents delivered directly to the brain. Presumably this allow 06-BG to increase the effectiveness of the chemotherapy but avoid additional systemic toxicity.

As will be discussed in a later section perhaps the most promising methods for improving the efficacy of chemotherapy involve its combination with anti-angiogenic agents and with agents that block the signaling channels that stimulate tumor growth, although the mechanisms underlying such synergies are not now clearly understood.

Timing of Chemotherapy

The usual practice of administering chemotherapy has been to start it soon after the regular external beam radiation has been completed. Alternative schedules are to present radiation and chemotherapy together or to present chemotherapy prior to radiation. After reviewing the different outcomes of these various schedules, no meaningful conclusion is possible. In some cases chemotherapy prior to radiation seems to improve outcome, in others it has produced a worse outcome. Similar variability in outcomes has been reported when chemotherapy is presented concurrently with external-beam radiation, although the recent study described earlier (9) that presented low-dosage temodar during radiation seems to produce a better outcome than other protocols. It is also the case that the effect of scheduling may depend on the type of chemotherapy that is used.

A second issue regarding the timing of chemotherapy has been whether large bolus dosages of chemotherapy every 4-6 weeks should be replaced by continuous low-dosage chemotherapy. Several prominent oncologists have argued that the rationale for periodic administration of the maximum tolerated dosage is based on inadequate experimental data and needs to be reconsidered. They have also reported experimental studies showing that rodents that have become resistant to chemotherapy administered with the usual bolus injections will nevertheless show a clinical response when the same chemotherapy is administered continuously at low dosages (45, 46). Moreover, unlike the bolus dosage, continuous low dosages have minimal toxicity. If this finding turns out to be generally true, it will constitute a major revolution in how chemotherapy is used, The first clinical study using this new approach to chemotherapy has recently been reported with breast cancer patients (47). These patients, who had failed several previous chemotherapy regimens for metastatic cancer, received low dosage cyclophosphamide daily and low dosage methotrexate on days 1 and 2 of each week. Of 63 patients, two had complete remissions, ten had partial remissions, and eight had stable disease for more than six months. Because this was a phase 2 study, there was no control group, but given that the patients had previously failed several other therapies, a plausible interpretation is that the new continuous regimen for chemotherapy improved its efficacy. It remains to be seen whether other types of cancer will show similar effects. It is noteworthy that one of the early trials using temozolomide (48) used a schedule approximating continuous dosing, as patients received temozolomide each day for 7 weeks, with a dosage about 1/3 of that used when the drug is presented for 5 days out of each 28-day block, which is the most commonly used schedule at the present time.. The results were that 7 of 17 glioma patients had tumor shrinkage while another 6 of the 17 had stable disease. However, a subsequent study using a similar protocol(49) failed to show a similar improvement, as its PFS-6 was 19%, essentially similar to the usual schedule of temodar. It is important to recognize, however, that animal models using the low-dosage protocol have typically combined it with various other agents to obtain the maximum improvement.

Yet another modification of the usual protocol for temodar has been a schedule of one week on, one week off (i.e., days 1-7 and 15-21 of a 28 day cycle (50). Here the PFS-6 was 48%, compared to the historical norm of 21% when temodar is used on its usual 5 of 28 day schedule. Although his study involved only 21 patients, it seems likely that this alternative schedule may be a substantial improvement.

The Role of Radiation

The initial approach to using radiation to treat gliomas was whole-head radiation, but this was abandoned because of the substantial neurological deficits that resulted, sometimes appearing a considerable time after treatment. Current clinical practice uses a more focused radiation field that includes only 2-3 cm beyond the periphery of the tumor site. Even this more localized use of radiation can produce significant neurological damage, depending on the original site of the tumor. Because of the potential for such damage, the currently accepted level of radiation that is considered safe is limited to 55-60 Gy. Even at this level, significant deficits may occur, often appearing several years after treatment. Attempts to minimize this damage are now being developed in the form of the Peacock system, which uses many more individual radiation beams, thus allowing the tissue other than at the tumor site to receive substantially less radiation exposure. Unfortunately, only a few treatment centers have such a system, and there is no published evidence that shows that the new system really does produce a clinical benefit beyond the traditional radiation treatment.

The major development in the use of radiation in the treatment of gliomas has been the addition of localized radiation to the tumor field, after the external-beam radiation treatment is finished (or sometimes concurrently), either by use of implanted radiation seeds (typically radioactive iodine), a procedure known as brachytherapy, or by the use of radiosurgery.

Brachytherapy

Beginning in the early 1980's, patients with recurrent GBM's were treated by implants of high-intensity radioactive iodine pellets at the site of tumor recurrence, which were then removed after 4-5 days of exposure. Survival time after this procedure was approximately a year, in contrast to only 3-6 months for patients receiving only additional chemotherapy for their recurrent tumors. Brachytherapy was then extended to initial treatment, usually occurring soon after the completion of the standard radiation, followed sometimes by standard chemotherapy and sometimes by no further treatment. Median survival time for patients receiving brachytherapy typically has been 18-20 months. In one representative study (51), survival rates at 1, 2, and 3 years were 83%, 34%, and 27%. Survival rates for a comparable group of control subjects not receiving brachytherapy were 40%, 12.5%, and 9%. Thus, brachytherapy appeared to more than double the survival rates at all points after treatment and as a result became a widely recommended procedure. Unfortunately, only 25-40% of patients are eligible to receive the treatment because of restrictions due to tumor size and location. The use of such selection criteria potentially biases the outcome results because patients who meet brachytherapy criteria but do not receive the procedure have survival times substantially longer than those not meeting the criteria (14 months vs. 6 months). Some significant portion of the apparent benefit of brachytherapy is therefore due to this selection bias (52). Moreover, two separate phase III trials that used random assignment of eligible patients to brachytherapy versus a control have failed to show a statistically significant benefit of brachytherapy, although there was a numerical advantage in both cases for the patients receiving brachytherapy. (53, 54)

A major disadvantage of the use of the standard brachytherapy procedure is that it produces very significant radiation necrosis that later causes major neurological symptoms. Initially this is treated with steroids, but ultimately results in the need for surgical removal. Approximately one-half of all patients undergoing brachytherapy require re-operation within the first year post-treatment, and this value underestimates the true rate of re-operation because many patients do not survive long enough for re-operation to become an issue. For those subjects who do survive longer, the re-operation rate is 60-70%

Because of the substantial radiation necrosis, and the negative evidence from the recent phase III clinical trials, most treatment centers have abandoned the short-term high-intensity implants that were in favor a decade ago. However, some centers continue to use brachytherapy with permanent low-intensity implants, a procedure that has been used widely in Europe for lower-grade glioma tumors, especially with children. The original pioneers of high-intensity implants also attempted to use the low-intensity implants but reported they were ineffective. Now, however, the low-intensity implants appear to be enjoying a revival. In a study (55) of 22 adult patients with recurrent gliomas (18 GBM, 4 AA-III), surgery for removing the recurrent tumors included placement of permanent low-intensity radiation seeds. For the GBM patients, the one-year survival rate was 59% and the median survival time was 64 weeks. No patients required re-operation to remove radiation necrosis. Unfortunately, these survival data are difficult to compare with the earlier results using high-intensity implants because the surgery to remove the recurrent tumors was apparently more successful than is usually the case, in that the majority of patients had post-surgery MRI's with no evidence of residual tumor. Whether this reflects the differences in the skill of the surgeon, or the degree of aggressiveness in removing the tumor, is unclear. A second study with essentially the same protocol but at a different treatment center produced a median survival time of 47 weeks (56). A subsequent retrospective comparison of the two protocols has compiled the results across several different clinical studies( 57) with the result that both produce a median survival time of approximately 52 weeks. However, the need for re-operation due to radiation necrosis was significantly less for the permanent low-intensity implants.

Low-intensity radiation seeds have also been implanted at the time of initial surgery, so that they were in effect at the time of the normal external-field radiation treatment (58). Of the 14 GBM patients receiving this treatment, median survival time exceeded 23 months. It was also the case that several of these patients would not have been eligible for traditional high-intensity brachytherapy because of the location of their tumors which otherwise indicated poor prognoses. Thus, the use of low-intensity implants at the time of external radiation seems a very promising procedure. Unfortunately, the procedure is not widely used, so that patients needing initial surgery are in no position to arrange for the treatment. Two of the major brain tumor treatment centers that do use low-intensity brachytherapy in combination with the initial external-beam radiation are Wayne State University (in collaboration with the Henry Ford Hospital) in Detroit and the University of California, San Francisco.

Radiosurgery

An increasingly popular alternative to brachytherapy is radiosurgery, using either a linear accelerator (the LINAC) or the Gamma Knife. Unlike brachytherapy, radiosurgery does not require surgery and can be conducted in a single session of treatment rather than requiring the 5-7 days typical of brachytherapy. Radiotherapy also produces a substantially lower rate of re-operation for radiation necrosis.

A retrospective comparison of the two procedures (59) has indicated that their effects are generally similar, at least for recurrent GBM. Survival rates 12 and 24 months after treatment were 45 and 19%, respectively, for radiosurgery, and 44 and 17%, respectively, for brachytherapy. Thus, given that radiosurgery is easier to administer and has a lower rate of radiation necrosis, it increasingly has become the preferred procedure. However, commentators on this comparison have noted that the tumor volumes for brachytherapy patients were notably larger, and that brachytherapy might be the preferred procedure when larger tumors are involved. However, several centers are now employing fractionated radiosurgery to treat larger tumors, typically consisting of one radiation session per week for four weeks. Initial reports with this procedure with a small number of subjects were very positive (60), but a subsequent report with an expanded number of patients produced results more similar to previous findings with radiosurgery with a single radiation session. (61)

An important feature of both radiosurgery and brachytherapy is that younger patients seem to receive substantially more benefit from the procedure than older patients. For example, in a recent compilation of 78 patients who underwent radiosurgery as part of their initial treatment, the median survival for all patients was 20 months, while patients who were younger than 40 years had a median survival time of 49 months.

As yet there are no direct comparisons of the LINAC and Gamma-Knife methods of administering radiosurgery. In principle there is no reason to believe that there should be a difference in outcome due to the two pieces of machinery used to produce the radiation. However, two recent studies have reported impressive results with the gamma knife. The first of these, from the University of Pittsburgh (62) used the gamma knife on 64 GBM patients either at the time of initial treatment or after tumor recurrence. Median survival time, measured from initial diagnosis, was 26 months; median survival time measured from the time of radiosurgery was 16 months. The 2-year survival rate was 51%. Such outcome results are among the best reported for any form of radiosurgery, although the authors note that the patients involved were selected partly on the basis of the size of their tumors. In the second study (63), conducted at the medical center of the University of Maryland (in Baltimore) subjects receiving the gamma knife after the usual external beam radiation were compared to those receiving only the external beam radiation. The median survival for those receiving the gamma knife was 25 months; median survival for those receiving only the external beam radiation was 13 months. Although the two groups were not randomly assigned, the authors note there were no significant differences between the groups in age, Karnofsky performance status, extent of resection, and tumor volume, which are the known to be the major prognostic variables independent of treatment.

Radiation via Monoclonal Antibodies

An alternative for providing a radiation boost beyond that from the standard external field radiation involves attaching radioactive iodine 131 to a monoclonal antibody that targets a specific antigen, tenascin, which occurs on almost all high-grade glioma tumors. The monoclonal antibodies are infused directly into the tumor cavity over a period of several days, and reportedly have substantially less toxicity than either brachytherapy or radiosurgery. The median survival time from a phase 1 clinical trial of this treatment for recurrent GBM tumors was 56 weeks (64). One important feature of this treatment is that it apparently does not preclude later treatment with either radiosurgery or brachytherapy. Moreover, it also apparently produces considerably less necrotic tissue. However, it can produce significant suppression of the stem cells in the bone marrow that produce blood cells.. In the first study that reported using this approach as initial treatment (65) patients received the monoclonal antibodies, followed by the standard external-beam radiation and then a year of chemotherapy. Of 33 patients only one required re-operation for necrotic tissue caused by the radiation. Median survival time was 79 weeks for the patients with glioblastoma (27 of 33 of total patients) and 87 weeks for all patients. Estimated two-year survival rate for GBM patients was 35%.

Gliasite

Yet another new radiation delivery system is "Gliasite". This consists of an inflatable balloon catheter that is placed in the resection cavity at the time of tumor debulking. Low-dose radiation is then delivered with a solution of radioactive iodine 1-2 weeks after surgery, and remains there for 3-6 days after which the balloon is removed. In the initial clinical trial (66), 21 patients who received the treatment after tumor recurrence had a median survival of 12.7 months. However, this was a mixture of glioblastoma and grade III tumors. For only the glioblastoma patients, median survival after treatment was 8 months. Most remarkably, the treatment produced no symptomatic radiation necrosis.

New Treatment Agents Currently Available

In this next section, all of agents described are FDA-approved and thus can be obtained by prescription, despite the fact that their approvals have been for diseases other than brain tumors. This unfortunately causes some oncologists to be unwilling to prescribe them, although there is no legal basis for that reluctance. The drugs that will be described differ from conventional chemotherapy in that they do not kill all dividing cells, and as a result have little of the traditional toxicity for the bone marrow that causes weakening of the immune system and anemia. This makes them ideal candidates for drug cocktails, including combinations with chemotherapy. Several of these combinations were described in the earlier section on chemotherapy, and these leave little doubt that clinical outcomes are greatly improved when these new "biological agents" are added to traditional treatments.

Tamoxifen.

This drug is well known for its usage in the treatment of breast cancer. Its mode of action there is to compete with estrogen for attachment to the estrogen receptors of breast cells, thus reducing estrogen's ability to serve as a growth factor for carcinogenesis. This mode of action has little to do with tamoxifen's ability to serve as a therapeutic agent for gliomas. Effects on glioma are instead due to tamoxifen being an inhibitor of protein kinase C activity - an intracellular enzymatic reaction that is involved in glioma cell proliferation. To obtain inhibition of PKC activity, and thus slow or stop the growth of the cancer cells, very high doses of tamoxifen are used, in contrast to its usage for breast cancer. The typical dosage for breast cancer is 10-20 mg daily, while for gliomas the dosage used has ranged from 160-240 mg per day. This high dosage is potentially problematic and does indeed have side effects. The most important is an increased risk of blood clots. For women, there is also an increase in the risk for uterine cancer, and for men, impotence and loss of libido are frequent problems. Weight gain is another significant side effect. Overall, however, such side effects are mild in comparison to traditional chemotherapy.

A stage II clinical trial (67, 68) evaluating the effects of tamoxifen for patients with recurrent gliomas has reported that it produced tumor regression in 25% of patients and stabilization of tumor growth for an additional 20% of patients. The percentage of patients with responses to treatment was greater with Grade III Astrocytomas than for patients with GBMs. The median survival time from the initiation of tamoxifen treatment was 16 months for Grade III tumors and 7.2 months for glioblastomas. This perhaps seems to be a minimal benefit (survival time for recurrent glioblastomas typically ranges from 3-6 months when second-line chemotherapy is used) but it should also be noted that a significant percentage of those who had either regression or stabilization had survival times greater than two years. Thus, for those "responders" tamoxifen produced a major benefit.

Tamoxifen has also been used in combination with traditional chemotherapy, in part because it should in principle reduce the level of chemo-resistance in addition to having its own direct effects on tumor growth. A European clinical trial combined tamoxifen with carboplatin as the initial treatment after radiation (69). Dosages of tamoxifen ranged from 40 to 120 mg/day, all of which were smaller than that used when tamoxifen has been used alone (160-240 mg/day). Combined over all dosages, the 12-month and 24-month survival rates were 52 and 32 %, respectively. For the patients receiving the highest dosage of tamoxifen, 12-month survival rate was 78%. In comparison, a matched set of subjects who received carboplatin alone after radiation had 12- and 24-month survival rates of 30% and 0%. However, a second similar study combining tamoxifen with carboplatin (70) reported a median survival time of only 55 weeks, which was not statistically superior to historical controls using carboplatin alone (48 weeks). However, the latter study noted that a minority of patients did have unusually long survival times, which was not reflected in the median survival times.

Tamoxifen with a dosage of 240 mg/day has also been studied in combination with BCNU as the initial treatment after radiation (71). Median survival time was 66.1 weeks, while the 1-year, 2-year, and 3-year survival rates 65%, 45% and 24%, respectively. It should be noted that while the 1-year survival rate and median survival time are only marginally greater than those obtained with chemotherapy alone, the 2-year and 3-year survival times are substantially greater. This benefit of a notable increase in the number of longer-term survivors again reflects the fact that tamoxifen is effective only for a minority of patients, but for those its benefits can be very substantial. The fact that tamoxifen benefits only a minority of patients is relevant to the negative results of a phase III trial conducted in France (72). Patients received BCNU alone or BCNU in combination with 40-100 mg/day of tamoxifen (note that these dosages are substantially below that used in the other studies). No increase in median survival time was found, whereas the addition of tamoxifen did significantly increase the frequency of serious blood clots.

Most recent has been the preliminary report of a trial combining tamoxifen with temodar (73). While details of this preliminary report are sketchy, its notable feature is that the combination treatment, presented as the initial treatment after standard radiation, resulted in all of the patients being alive at 12 months after diagnosis. More details are clearly needed, but the results as described are unusually promising.

An important recent development with respect to tamoxifen has been the report (74) that it may be possible to predict which patients will be among the minority that benefit from tamoxifen. This Canadian study compared patients who responded to tamoxifen with those who did not and reported that there was a systematic difference in the metabolites from tamoxifen. This potentially allows a decision very early in treatment about whether tamoxifen is worth continuing. A second major development is that tamoxifen's efficacy may be increased by suppressing thyroid function (75). Thyroid hormones maintain the level of the insulin-like growth factor (IGF), which is now known to play an important role in causing resistance to several different kinds of cancer treatments (to be discussed further in a later section). Eleven of 22 patients with recurrent tumors became hypothyroid as a result of a drug treatment. Their median survival time was 10.1 months, versus 3.1 months for patients whose thyroid function was not effectively suppressed.

Accutane

This drug, which is FDA-approved for the treatment of severe acne, is an acid form of vitamin A chemically known as 13-cis-retinoic acid (also known as isotretinoin). Acid forms of Vitamin A are not stored in the liver; so unlike regular Vitamin A, high dosages may be used with much less risk of liver toxicity. Its presumed mechanisms of action are to activate genes that cause cancer cells to differentiate into normal cells and to block the receptor for the epidermal growth factor (EGFR). High levels of expression of that receptor cause cell division to occur at a rapid rate. A variety of other anti-proliferative effects have been identified as well (76)

A stage II clinical trial evaluating accutane for recurrent gliomas has been conducted at the M. D. Anderson Brain Tumor Center (77). The median survival time was 58 weeks for glioblastoma patients and 34 weeks for grade III gliomas. This difference is survival time is opposite in direction than that obtained with other treatments. However, there was wide variability in both tumor types, so that the difference was not statistically reliable. Aggregated over all tumor types (43 evaluable patients) 3 achieved a partial tumor regression, 7 had minor regressions, and 13 had tumor stabilization, for a total response rate of approximately 50%. A recent more complete report of using accutane with 86 glioblastoma patients with recurrent tumors was somewhat less impressive. (76). Median survival time from the onset of treatment was 25 weeks and PFS-6 was 19%. Accutane now is used at M. D. Anderson as a "maintenance therapy" for patients after initial treatment with radiation or traditional chemotherapy. It also has been used in Germany for patients who have had a complete response to other treatment modalities as a maintenance therapy (78) The major side effects have been dry skin, cracked lips, and headaches, although occasional liver toxicity has also occurred. Increases in blood lipid levels frequently occur, often requiring anti- cholesterol medication such as Lipitor. Accutane also may produce severe birth defects if taken during pregnancy.

Because accutane's toxicity is very different from that of chemotherapy, it is now often used in combination with chemotherapy, notably temodar. When temodar is used alone for recurrent glioblastomas, the percentage of patients who have are alive without tumor progression six months after the start of treatment is 21%. When accutane is used in combination with temodar, the corresponding number is 32% (14).

There is also experimental evidence that accutane is synergistic with other drugs that are known to cause cell differentiation (79). This approach to cancer treatment will be discussed more fully in a later section.

Currently under investigation in clinical trials is a second form of retinoic acid called fenretimide, which laboratory studies has indicated is more active against gliomas than is accutane. However, this new drug has not yet been studied in combination with chemotherapy, and is not available outside of clinical trials.

The similar pattern of treatment outcomes for tamoxifen and accutane raises an important question. Given that both seem to significantly benefit a minority of patients, the issue is the overlap in the populations helped by the different treatments. That is, would the patients helped by tamoxifen also be those most likely to be helped by accutane, and vice-versa. If so, this suggests that the two agents should be synergistic in their effects, so that a patient receiving both agents should have a very high likelihood of a positive clinical outcome. If not, they should be additive, not necessarily in terms of benefit for individual patients, but instead in terms of the percentage of the total population who respond to one or the other treatment.

Thalidomide

This drug became infamous during the 1950's because it produced a large number of birth defects involving abnormal or completely missing limbs. It is now believed that this was due to its effects on inhibiting new blood vessels because limb buds are especially dependent on the growth of new blood vessels for normal development. Thalidomide has been approved by the FDA for the treatment of leprosy, but it also can be obtained for off-label uses such as the treatment of cancer. Unfortunately, a considerable amount of paperwork is necessary, both by the pharmacist who supplies it and the physician who prescribes it, so obtaining it for off-label uses is not as simple as having your physician write a prescription. These bureaucratic restrictions have been imposed despite the fact that the majority of potential users of the drug, males, and females past the age of menopause, are in no way affected by the drug's teratological potential.

Thalidomide's utility as a cancer treatment comes from it being the first anti-angiogenic drug that has been FDA approved. Exactly how thalidomide retards the growth of new blood vessels is not entirely understood, and it seems likely that new anti-angiogenic drugs will soon replace it because of their greater effectiveness. To date, however, it has been shown to have significant clinical effectiveness for both Kaposi's sarcoma and multiple myeloma, and is currently being evaluated for other forms of cancer as well.

In the first clinical trial using thalidomide as a single agent for the treatment of recurrent GBM (80), involving 36 patients with GBM or AA-III tumors, there were two partial regressions, two minor regressions, and 12 patients with stable disease for a minimum of 8 weeks. Median survival times were 74 weeks for those with tumor regression, 30 weeks for those with stable disease, and 22 weeks for those classified as nonresponders. However, PFS-6 was only 4%. The major side effects were somnolence (thalidomide was originally introduced for its sedative purposes; presumably such effects could be counteracted by various stimulants) neuropathy of various sorts, and constipation. Because such side effects are greater with higher dosages, it is of interest to note that results very comparable to the preceding study have been obtained in Australia using substantially lower dosages. Whereas the American studies have used a maximum dose of 1200 mg/day, the Australian study use a maximum dose of 500 mg/day (81).The best results using thalidomide as a single agent comes from a recently published study performed in Switzerland (82). Nineteen glioblastoma patients with glioblastomas received 200 mg/day of thalidomide, starting after radiaation, escalating to 600 mg/day if tolerated. The actual median dose used was 200 mg/day. Median survival time was 63 weeks. Median progression-free survival was 17 weeks. Some patients had surgery for recurrent tumors so it is difficult to know how much of the survival time was due to the additional surgery.

The same study also reported the results of 25 patients who received the same regimen of thalidomide but in combination with temozolomide. Here the median survival time was 103 weeks and he median progression-free survival was 36 weeks.

Other trials have combined thalidomide with chemotherapy agents other than temozolomide. A clinical trial involving the combination of thalidomide with carboplatin for recurrent glioblastomas was reported at the 1999 meeting of the American Society for Clinical Oncology (83). Of 46 patients assessable for efficacy, 5 had a partial regression, 28 had stable disease and 13 had progressive disease. Estimated median survival for all patients was 40 weeks.

Thalidomide has also been studied in combination with BCNU (84) with patients with recurrent high-grade gliomas. Although the PFS-6 for all patients was only slightly better than temodar alone (27% vs. 21%), 9 of 40 patients had major tumor regressions while an additional 9 had stable disease. Both of these are higher than when temodar is used as a single agent in a similar population. Because of the disparity in the two different measures of treatment efficacy, any evaluation of the combination still remains unclear.

Comparison of the above results suggests an important point to highlight. Thalidomide appears to be more effective as a treatment when given as initial treatment rather than for tumors that have recurred. This appears to be true for anti-angiogenic treatment generally, the rationale being that mature tumors have a more developed vasculature so that preventing the growth of new blood vessels is less effective in starving the tumor.

Many other anti-angiogenic agents are currently being investigated in clinical trials, although few with brain tumor patients as the clinical populations. It is important to recognize that these agents may not cause tumor regressions in themselves because they are not cytotoxic in the usual sense. Their target is the blood vessels feeding the tumor, not the tumor itself, so tumor stabilization may be the most likely outcome. However, it is believed that by keeping the tumor from expanding, the chances of traditional treatment modalities being effective will be greatly increased. A more extensive discussion of how to maximize the effects of anti-angiogenic agents is presented in a later section.

STI-571 (Gleevec)

This new drug treatment that targets a specific receptor involved in the growth of a form of leukemia recently received a great deal of publicity because of its unprecedented effectiveness. As will be discussed later, this general strategy of identifying the growth signals for tumor growth and then targeting those signals, or their receptors, is one of the major new hot areas in cancer research. Such growth signaling channels often, but not always, are specific to the individual type of cancer. Although Gleevec was developed specifically for chronic myelogenous leukemia, the receptor involved has biochemical similarities to those for a more general type of growth signal, platelet-derived growth factor (PDGF), which is also involved in the growth of gliomas and other forms of cancer (e.g., small-cell lung cancer). Laboratory research has supported the importance of this similarity in that gleevac has been shown to strongly inhibit glioma growth (85), with the result that gleevac currently is being studied in clinical trials involving gliomas. Because it has approval for its usage for leukemia, the drug is also available outside of clinical trials. Results from the first clinical trial using the drug for brain tumors was reported at the 2002 meeting of the American Society of Clinical Oncology (86). Only the abstract was available at the time of this writing. Of 31 patients who were evaluable, 14 had stable disease, the reminder had progressive disease. A second study of gleevec as a single agent was reported at the 2004 ASCO meeting (87) Of 51 heavily pretreated patients, three had a partial response and five others had prolonged periods of disease stabilization. Thus, gleevec does have some activity against glioblastomas but it appears limited. However, better results have been obtained in a subsequent clinical trial (88) performed in Germany which combined gleevec with hydroxyurea, an older drug that at one time was believed to be a radiation sensitizer among other functions. This more recent trial was notably more successful, as 5 of 14 patients with recurrent glioblastomas had tumor regressions, another 5 had stable disease and 4 had disease progression. Thus, while Gleevec has only a small degree of activity as a single agent, it is a promising candidate for part of a cocktail treatment with other treatment agents, especially considering it has minimal toxicity.

Celebrex (and other NSAIDs)

Carcinogenesis of several types involves an inflammatory process. When anti-inflammatory drugs such as ibuprofen are taken on a regular basis the incidence of colon cancer is reduced as much as 50%. This astonishing effectiveness has motivated investigation of the mechanisms of these benefits. One component of the inflammatory process is angiogenesis, which is now believed to be a critical component of cancer growth. COX-2 enzymes are believed to play an important role in inflammation, so that COX-2 inhibitors should reduce angiogenesis and inhibit tumor growth (89, 90). Many nonsteroidal anti-inflammatory drugs (NSAIDs) are known to be COX-2 inhibitors, but most (e.g., ibuprofen) also inhibit COX-1 enzymes, which are necessary for healthy maintenance of the stomach lining, which is why many users of NSAIDs eventually develop intolerance to them. Thus, much recent attention has been given to the new COX-2 inhibitors such as Celebrex and Vioxx, which were developed to avoid COX-1 inhibition for the purposes of arthritis treatment. Because inhibition of angiogenesis is one of the major new approaches to the treatment of cancer (see discussion in a later section) many oncologists have begun adding Celebrex or Vioxx to their regular treatment protocols, based on laboratory findings that Cox-2 inhibitors inhibit tumor growth. In the last meeting of American Society for Clinical Oncology (ASCO), there were scores of new clinical trials reported that combined one or another Cox-2 inhibitor with conventional radiation, chemotherapy, and new targeted treatments. The great majority of these were phase 2 clinical trials which had only historical controls with the conventional treatment alone to assess the value of the added Cox-2 inhibitors, but almost all concluded there appeared to be a significant benefit, including one clinical trial using such a combination with glioblastomas (22). In one representative study with patients with non-small cell lung cancer, 16 patients received 800 mg/day of celebrex between rounds of chemotherapy. Four had a complete response to treatment, 12 had partial responses, and 4 had stable disease (91). While these results were from a phase II clinical trial without a control group, the outcome was better than the historical norms for comparable patients.

In a second study, prostate cancer patients with rising PSAs after initial treatment received 400 mg/day of celebrex (92). Of 11 patients followed for 6 months, 2/11 had a decrease in PSA, and 3/11 had stabilization in the PSA rise. The remainder had a decrease in the rate of PSA rise. The authors concluded that celebrex significantly slowed the progression of prostate cancer, although how long that suppression would continue remains tentative because of the early-stage nature of the clinical trial.

The only clinical trial reported to date that has used celebrex in the treatment of gliomas combined it with temodar, and was described earlier (22) in the section on chemotherapy.

Because of the mild toxicity of NSAIDS, considerable recent research has investigated the mechanisms of its clinical benefit. Whereas initial research focused on the anti-angiogenic properties of this class of drugs, several other mechanisms have been identified, including the enhancement of various aspects of the immune system, and inhibition of the genes that prevent damaged cells from undergoing apoptosis.(93). Not all NSAIDS are equal in their anti-proliferative effects, as there is some evidence that one of them, celebrex, is considerably more potent than others in directly inhibiting tumor growth by down-regulating the cyclin proteins regulating the different stages of cell division (94). It is critical to note that many of the mechanisms by which NSAIDS work are strongly involved in the growth of high-grade gliomas, and that the expression of the cyclogenase enzyme that is the target of COX-2 inhibitors correlates strongly with the proliferation rate of glioblastoma tumors and correlates inversely with survival time (95, 96).

Glitazones (Avandia, Actos)

All of the drugs discussed above in this section have been developed for medical purposes other than the treatment of brain cancer, and only subsequently were their anti-cancer properties discovered. The latest example of such "accidental" anti-cancer drugs is a family of drugs ("thiazolidinediones", also know as "glitazones") developed for type II diabetes that now are used by hundreds of thousands of patients. The two drugs of this category available in the USA are rosiglitazone (trade name Avandia) and pioglitazone (trade name Actos). Their mechanism of action for the treatment of diabetes is to increase cellular insulin sensitivity. But laboratory research has identified multiple other mechanisms of action as well that potentially have major benefit for cancer patients (97, 98), including inhibiting various steps in the cell cycle, induction of cell differentiation, induction of apoptosis (programmed cell death), and inhibition of angiogenesis. Among the experimental studies are those targeting glioblastoma tumors, both "in vitro" studies of cell cultures (99) and "in vivo" studies of implanted tumors in rodents (100). Of special interest is the finding that this class of drugs is synergistic with retinoids such as accutane in regulating the genes necessary for differentiation and apoptosis (101). As yet no clinical trials using these drugs for the treatment of brain cancer have been reported. One small clinical trial with advanced "vascular tumors" was reported using one of these drugs (Actos) as part of a drug cocktail, also involving chemotherapy on a daily low-dosage schedule (metronomic chemotherapy) in combination with Vioxx, a Cox-2 inhibitor (102). Of six patients, all of whom had been heavily pretreated without success, three had complete remission of their disease and a fourth had a partial remission. While such results are difficult to evaluate in the absence of more knowledge about the clinical outcomes of this class of patients, the results at least encourage further investigation. Of special note is that the concentrations of the drug necessary to produce major tumor regressions in animal models (100) are similar to those routinely used by diabetics, which are known to have minimal toxicity.

"Supplements" with Demonstrated Efficacy

Melatonin

This is a naturally occurring hormone secreted by the pineal gland that regulates the body's diurnal rhythm. It is commonly used for the treatment of jet lag and for insomnia. It is readily available in any health food store and most drug stores. Its role in cancer treatment has been predicated on the assumption that it boosts the immune system, with the current hypothesis being that it augments the activity of T-helper cells. It recently also has been shown to inhibit angiogenesis (103). It may also have direct cytotoxic effects on some types of cancer cells, notably melanoma cells. It has no known toxic side effects. Clinical research on the use of melatonin for cancer treatment has been done primarily in Italy, where it has been used either as a single agent after radiation treatments, or in combination with various chemotherapy or immunotherapy regimens, most frequently interleukin-2. Part of the rationale for such combinations is that it decreases the side effects of the chemotherapy, especially with respect to blood counts. One of the clinical studies that has been reported (104) randomly assigned GBM patients either to radiation-alone or to radiation concomitant with 20 mg/day of melatonin. Melatonin was continued after completion of the radiation. Survival time was significantly longer for subjects receiving the melatonin. In terms of one-year survival rates, 6/14 patients receiving melatonin were alive, while only 1/16 patients without melatonin was alive. This GBM study involved a relatively small number of patients, so that the effects might be considered tentative until a larger study is conducted. However, the effect of melatonin was statistically reliable even with the small number of subjects. Moreover, comparable effects have been reported in a similar design for the use of melatonin with advanced lung cancer (105). Like the GBM study, a substantial increase in survival rate occurred for the patients receiving melatonin.

To date there have been at least a dozen phase-2 clinical trials using melatonin either alone or in combination with other agents and five phase-3 trials involving random assignment of subjects to melatonin versus some type of control group. The majority of these have been relatively small and have involved patients in the terminal stages of their disease, which is perhaps why American oncologists have largely ignored them. However, several recent trials have been much larger and seem to leave little doubt that melatonin significantly increases the efficacy of chemotherapy. The most extensive randomized clinical trial involved 250 patients with advanced metastatic cancer of various types (106). Patients were randomly assigned to chemotherapy alone (using different chemotherapies for different types of cancer) or chemotherapy plus 20 mg of melatonin per day. Objective tumor regression occurred in 42 (including 6 complete regressions) of 124 patients receiving melatonin but in only 19/126 (with zero complete regressions) of the control patients. A comparable difference occurred for survival rate: 63/124 of those receiving melatonin were alive after one year while only 29/126 were alive of those receiving chemotherapy alone. A different trial, involving 100 patients with metastatic nonsmall-cell lung cancer (107), compared chemotherapy alone or chemotherapy in combination with melatonin. For the chemotherapy alone patients there were 9 of 51 who had a partial tumor regression, while 17 of the 49 chemo + melatonin patients had either a complete (2) or partial (15) regression. Twenty percent of the chemo-alone patients survived for one year and zero for two years, while the corresponding numbers for chemo + melatonin were 40% and 30%. Melatonin not only increased the efficacy of chemotherapy, but also significantly reduced its toxicity. These trials leave little doubt that the effects of melatonin are robust and of major clinical significance. Moreover, a recent study has shown that using multiple components of the pineal gland secretions instead of melatonin alone enhances clinical effectiveness still further (108).

PSK and other polysaccharides

PSK is the abbreviation for polysaccharide krestin (sometimes known simply as krestin), which is an extract from the mushroom, Coriolus Versicolor. It has become a standard component of cancer treatment protocols in Japan (a Chinese version of the same extract is known as PSP) for many different kinds of cancer predicated on the assumption that it is an immune-system enhancer. Among the effects on the immune system that have been identified are gamma-interferon production, interleukin-2 production, and in increase in T-cell activity. Other effects include inhibition of matrix-degrading enzymes that underlie tumor invasion of adjacent tissue, and the inhibition of angiogenesis.

In one representative study, with non-small cell lung cancer (109), stage I patients receiving PSK (3 g/day) had a five-year survival rate of 39% compared to 22% for patients not receiving PSK. For stage III patients, the 5-year survival rate with PSK was 16% versus only 5% for those not receiving PSK. Both differences were statistically significant. Other studies involving colon cancer and stomach cancer have also shown that PSK substantially increased survival rates. I have found only one study that used PSK in the treatment of glioma, in combination with ACNU (a chemical cousin of BCNU) and vincristine (110). The survival rate for high-grade patients after one, two, and three years was 77%, 49%, and 47%, respectively. No control condition was studied that did not receive PSK, so exactly what its effect was is unclear. Note, however, that the two-year and three-year survival rates are substantially greater than that typically seen for GBM following traditional treatment with chemotherapy alone. However, the abstract of the study (the study was in an inaccessible Japanese journal) did not report the results separately for glioblastomas versus grade III gliomas.

PSK is not easily obtained in this country. The only source I have found is JHS Natural Products in Eugene, Oregon (phone # 541-344-1396 or 888-330-4691; website:www.jhsnp.com). A month's supply costs $125. Other mushroom extracts that also have the long-chain polysaccharides (beta-glucans) that appear to be the active ingredient in PSK are more readily available. These include maitake, reisha, and shitake mushrooms. However, none of these has the same level of scientific evidence for treatment efficacy in human clinical trials. Maitake D-fraction seems an especially promising mushroom extract based on a recent laboratory study of chemically-induced tumors in mice (111). Tumor growth was inhibited 90% when the mushroom extract was combined with chemotherapy versus an inhibition of only 50% when chemotherapy was used alone for control subjects.

Gamma-Linolenic Acid (GLA) and Fish Oil

GLA is an essential fatty acid found in evening primrose oil, borage seed oil, and black currant seed oil. At least 100 laboratory studies have shown it to be highly cytotoxic to many different kinds of cancer cells, with the presumed mechanism that metabolism of GLA by the cancer cells creates high levels of free radicals that are lethal to the cells. Iron and zinc potentiate this cytotoxic effect; Vitamin E (and perhaps other anti-oxidants) counteracts it. GLA is harmless to normal cells and has been shown to have clinical utility for a variety of disorders, notably rheumatoid arthritis. It also has been shown to lower LDL cholesterol and increase insulin sensitivity. GLA is also known to change the structure of cell membranes, which is believed to underlie the finding that it increases the effectiveness of both chemotherapy and radiation. At the same time GLA has been shown to protect normal cells from radiation damage.

Evidence that GLA is effective against gliomas comes from a study conducted in India (112, 113) in which GLA was infused directly into the tumor bed. Of the 15 patients treated, most had major tumor regressions, and 12 of the 15 were alive at the time of the report's publication (1-2 years later). The three who died were all quite elderly and probably would not have received any conventional treatment beyond radiation in this country. A subsequent study (114) involving patients with very advanced disease had notably less success but here too there were notable tumor regressions attributable to the treatment.

A critical question is whether oral ingestion of GLA has any clinical effects. A recent clinical trial involving its use for breast cancer substantiates that it does (115). Advanced breast cancer patients received the standard treatment of tamoxifen alone or tamoxifen in combination with GLA, in the form of 2.8 g of GLA/day. The source of GLA was borage seed oil, which is approximately 20% GLA, which meant that the patients were taking 15 g of borage seed oil per day. Borage seed oil is available in any health food store, usually in the form of 1000 mg capsules, although supposedly it can also be obtained in liquid oil form and makes tasty salad dressings. It is important to find a reliable source, because some sources have high alkaloid levels that are poisonous. The measure of treatment effectiveness in the breast cancer clinical trial was the status of patients three months after the initiation of treatment. With tamoxifen alone none of the patients had a complete response to treatment, 13% had partial regression of their tumors, while 81% had stable disease. For tamoxifen + GLA the corresponding percentages were 5, 37, and 55%, a significant improvement.

The use of GLA as a cancer treatment is controversial because one of its major metabolites is arachnidonic acid, which is the precursor to both the lipoxygenase and cyclogenase inflammatory pathways. These inflammatory pathways are believed to stimulate the growth of cancer cells, which seems to contraindicate using GLA. However, it should be noted that GLA has been used successfully as a treatment for rheumatoid arthritis because of its anti-inflammatory effects, so obviously the story is more complicated. Part of the source of confusion is that the effects of GLA are dose-dependent. In laboratory studies low dosages have been shown to stimulate tumor growth, while at higher dosages the effect is clearly cytotoxic. (116, 117). A second important factor is the presence of n-3 fatty acids (fish oil being the most common). When fish oil is also present, its metabolic pathway competes for enzymes that also are involved in GLA metabolism, thus preventing the formation of arachnidonic acid. The optimal use of GLA may therefore be in combination with fish oil, not as a single agent.

The major fatty acids found in fish oil, eicosapentenoic acid (EPA) and docosahexanoic acid (DHA), have also been demonstrated to have potent cytotoxic effects on cancer cells in numerous laboratory experiments. Part of their mechanism of action is similar to that of GLA, in that the metabolism of these fatty acids creates high levels of free radicals. In addition, a recent laboratory study has shown that EPA-treated tumors showed a significant arrest of cell division due to inhibition of cyclins at the G1 phase of cell division, which resulted in an increased rate of programmed cell death known as apoptosis (118).

A clinical trial comparing fish-oil supplements versus a placebo has also been reported, involving patients with several different types of advanced cancer (119). Thirty malnourished patients suffering from cachexia were randomly assigned to receive 18 g of fish oil per day, in combination with 200 mg of Vitamin E, or a placebo sugar pill. An additional thirty subjects, adequately nourished, received a similar random assignment. For both groups the fish oil significantly increased survival. For the malnourished patients the median survival times, as estimated from their survivor functions, were 110 days for the patients receiving placebo and 210 days for patients in the fish oil group. For the adequately nourished patients, the corresponding numbers were 350 versus 500 days. In laboratory studies (120) fish oil has also been shown to significantly increase the effectiveness of chemotherapy.

The evidence that the above three agents all have significant clinical efficacy for many types of cancer is compelling It is therefore remarkable that all of them continue to be ignored as treatment options/supplements by American oncologists, especially given that none of the three have shown any type of toxicity even with very high dosages.

Vitamin D

Numerous laboratory studies have shown that Vitamin D is highly cytotoxic to cancer cells, due to several different mechanisms (although it is labeled a vitamin it more properly should be considered a hormone). While most research has focused on its ability to upregulate genes that cause cancer cells to differentiate into mature cells, other effects have also been identified, including cell cycle regulation, inhibition of the insulin-like growth factor, and the inhibition of angiogenesis (121). However the form of Vitamin D most commonly available is not readily usable for cancer treatments because the dosages producing anti-cancer effects also cause hypercalcemia, which can be life threatening (the major function of Vitamin D is to regulate calcium absorption and resorption from the bones and teeth). But like many vitamins/hormones, the generic designation refers not to a specific chemical structure but to a family of related molecules that may have different properties of various sorts. For Vitamin D several of these variants (commonly referred to as analogues) have been shown to effectively inhibit cancer cell growth but without the same degree of toxic hypercalcemia. In a recent paper in the Journal of Neuro-oncology (122), 10 patients with glioblastoma and one with grade III AA tumors received a form of Vitamin D called alfacalcidol in a dosage of .04 micrograms/kg each day, a dosage which produced no significant hypercalcemia. The median survival was 21 months, and three of the 11 were long-term survivors (greater than 5 years. Although the number of patients who responded to the treatment was not a high percentage , the fact that any relatively non-toxic treatment can produce that number of long-term survivors is remarkable. This is an especially interesting finding because there is strong reason to believe that Vitamin D is synergistic with retinoids such as accutane (123). Its effectiveness is also increased in the presence of dexamethesome (124) and a variety of anti-oxidants, notably carnosic acid, but also lycopene, curcumin, silibinin, and selenium (125).

Unfortunately, alfacalcidol is not available in the USA. But it is available in Europe and Canada. For those in the USA it is possible obtain it from various online marketers. One source that several members of the brain tumor community have used is Masters Marketing. Its web address is http://www.mastersmarketing.com. Undoubtedly there a number of other possible suppliers. It also should be noted that several other Vitamin D analogues are available, which also have much reduced hypercalcemic effects. One of these, paricalcitol, was developed for treatment of a disorder of the parathyroid gland, and recently has been the subject of several experimental studies (126, 127, 128) that have shown it to be highly cytotoxic to many different type of cancer. Given that other forms of Vitamin D have been shown to be highly cytotoxic to to glioblastoma cells, and that glioma cells are known to have receptors for Vitamin D, it seems likely that paricalcitol should have efficacy for glioblastoma as well. Unfortunately, its routine use is complicated by the fact it is available only in a form that requires intravenous injection.

It is important to note that even these new forms of Vitamin D can occasionally produce dangerous serum calcium levels, in part because there is a great deal of variability in their effects across individuals. It is thus important that blood calcium levels be monitored, especially while a nontoxic dosage is being established.

Supplements With Potential Efficacy But Not Yet Clinically Tested

Genistein

This is an isoflavone derived from soy products (it is also found in red clover extract) that has been shown in the laboratory to be highly cytotoxic to many different types of cancer, including glioma cells. In addition to the laboratory evidence, there is also substantial epidemiological evidence that high dietary intakes of soy products decrease cancer mortality by at approximately 50%. Only recently has it begun to be studied in clinical trials, mainly with respect to the treatment of prostate cancer.

Soy extracts containing genistein are available in most health-food stores. The concentration of genistein is often not well specified, so it is unclear what is actually in the extract. Most importantly, the listed amounts of genistein are so low that they are unlikely to provide much clinical benefit. The highest concentration (about 10 times greater than the others that I have found) appears to be in a product called "mega soy" made by the Life Extension Foundation (phone: 800-841-5433; website: lef.org). It can be ordered from them or from L&H Vitamins, a discount mail-order company that is a good source for many types of products otherwise found in health-food stores (phone #: 800-221-1152).

Although there is as of yet no direct evidence of the clinical effectiveness of genistein, the laboratory studies that are available make a strong case for its potential efficacy. In one representative laboratory experiment mice received different concentrations of genistein added to their regular diet (129). The measure of its effect was the number of lung metastases caused by melanoma cells injected into the mice. The number of lung tumors was reduced by 50-75% depending upon the amount of genistein added to the diet. Interestingly, even greater inhibition of tumor growth was observed in another study when whole soy extracts were added to the diet, rather than genistein alone (soy contains numerous isoflavones other than genistein).

Recent experimental studies have examined the mechanisms whereby genistein produces its anti-cancer effects (130). The consensus is that this results from its ability to inhibit tyrosine kinase activity. This is a general class of chemical signals that strongly stimulate cell division. The epidermal growth factor, discussed earlier with respect to the mechanism of accutane's effect, is one member of this class of signals, and some investigators believe that genistein works by blocking the EGF receptor. Genistein also appears to produce inhibition of protein kinase C (discussed earlier with respect to the mechanisms of tamoxifen. This in turn suggests that a combination of genistein and tamoxifen might be especially effective. Finally there is increasing evidence that genistein is an inhibitor of angiogenesis.

Of special interest to brain cancer patients is a recent laboratory study in which glioblastomas cells were treated with a combination of genistein with BCNU (131). The result was a highly synergistic suppression of the rate of growth. This observation is important because genistein has much in common with new drugs being developed to block the EGF signaling channel, which themselves seem to be more effective when used in combination with conventional treatment modalities.

Selenium

This is a trace element commonly found in the soil, which is absorbed into various foods, most commonly onions and garlic. Its potency as a an anti-cancer agent was discovered almost by accident in a randomized placebo-controlled trial in which selenium was being tested as a possible preventative agent for skin cancer (132). While selenium had no effect on the incidence of skin cancer, it had substantial effects on the incidence of other types of cancer, including lung, colorectal, prostate, and the total of all cancers. The most dramatic effect occurred for prostate cancer, for which the incidence was reduced by 63% for those receiving selenium relative to the rate in the placebo controls. The incidence of brain cancer was not recorded in this study. An important question is whether selenium is effective as a treatment for existing cancers in addition to being useful as a cancer preventative. Laboratory research suggests that it should indeed be effective, as it has been shown to inhibit tumor growth in a dose-dependent manner in vitro, and its use as a dietary supplement significantly inhibits the growth of pulmonary metastases after injection of melanoma cells into mice (133). Laboratory studies also have show I t to inhibit the growth of glioma cells (134). Recent studies have identified two of its mechanism of action, inhibition of protein kinase C (135), known to be important in the growth of gliomas, and inhibition of angiogenesis (136). It is important to note that selenium can be highly toxic at high dosages, and that the degree of toxicity varies with the compound in which it comes. Selenomethionine is the preferred form because it is the least toxic. The most common dosage used is 200 micrograms/day, although dosages to 400-800 mcg/day have been used without evident toxicity. There is some evidence that its effects may be synergistic with Vitamin D.

Green Tea

Green tea has been consumed in both China and Japan for 5000 years based on its medicinal properties. It is now believed that its primary anti-cancer ingredients are polyphenolic catechins, the most prominent of which is epigallocatechin-3-galate (EGCG). A recent review has summarized its anti-cancer effects in several different animal models using both mice and rats (including major inhibition of glioblastoma cell lines), both when human tumors have been implanted and when they have been induced by various chemical carcinogens (137). In a representative study of chemically-induced tumors in mice (138), green tea was provided as the sole source of fluid, at a concentration of 6% (6 g of tea per liter of water), the incidence of lung tumors was reduced by 30%. The same study identified several different mechanisms of action, the most prominent of which was the inhibition of angiogenesis.

A recent review by the new Division of Alternative Medicine of the National Institutes of Health has identified green tea as the most promising of treatments advocated by proponents of alternative medicine. Accordingly, several clinical trials investigating its efficacy are ongoing. The only one reported to date used green tea in the treatment of patients with androgen independent metastatic prostate cancer (139). Dosage was 6 g of green tea per day. Only limited clinical benefit was reported. It is important to recognize that anti-angiogenic agents generally take a long time to produce clinical regressions, work better with less advanced stages of disease, and also work better in combination with other treatment agents.

Quercetin

This is a member of the class of flavonoids found in fruits and related plant products. Its most abundant sources are onions and apples. Like genistein it appears to be an inhibitor of tyrosine kinase activity, and appears to be synergistic with genistein when the two have been combined in laboratory studies involving both ovarian and breast cancer cell lines. It currently is being investigated in phase-1 clinical trials. Given that apples are one of its major sources, it is interesting that a recent paper in Nature (June 22, 2000) has reported that material extracted from fresh apples inhibited in a dose-dependent manner the growth of both colon and liver cancer cell lines.

Curcumin

This is an ingredient in the Indian cooking spice, turmeric. It has been shown to inhibit the growth of cancer cells of various types in laboratory studies (140). Like genistein and quercetin it inhibits the tyrosine kinase signaling and also inhibits angiogenesis. When the three supplements have been directly compared curcumin was the more powerful inhibitor, but it also should be noted that it bioavailability from oral intake is limited. However, bioavailability appears to be significantly increased when curcumin is combined with piperine (the main ingredient in black pepper).

Silibinin (an ingredient of Silymarin)

Silymarin is an extract from the milk thistle plant that has been used extensively in Europe as an antidote for liver toxicity, due to mushroom poisoning and overdoses of tylenol. Its active ingredient is a molecule called silibinin. Recently a great deal of laboratory research has shown it to have anti-cancer effects as well. Like genistein and quercetin it is a tyrosine kinase inhibitor, but it appears to have multiple other effects, including the inhibition of the insulin-like growth factor (IGF) that contributes to the development of chemoresistance (141) (see the section on tamoxifen), and the inhibition of angiogenesis (142). It also inhibits the 5-lipoxygenase inflammatory pathway and suppresses nuclear factor kappa B, which is known to be antagonistic to apoptosis (143) It also appears to protect against common chemotherapy toxicities (144), while at the same time increasing the effectiveness of chemotherapy (145).

Bromelain

This is an extract from pineapple that contains proteinases that have a variety of anti-cancer effects (146. While it has been shown to be cytotoxic to a variety of different cancer cell lines, most germane to the present discussion is its ability to inhibit the growth of glioma cells (147) by a variety of different mechanisms.

Lycopene

This is a carotenoid that is found most abundantly in tomatoes but occurs in various other red-colored vegetables as well (including watermelon).Unlike the most well-known carotenoid, beta-carotene, it does not get transformed into Vitamin A, and thus has no hepatic toxicity. In a small clinical trial involving prostate cancer patients about to undergo surgery (148), for those who consumed lycopene for several seeks before surgery both the size and malignancy of their tumors were significantly reduced relative to those not receiving lycopene. In an experimental study involving both cell cultures and implanted glioma tumors in rats (149), lycopene (and beta-carotene) were found to substantially inhibit tumor growth in both experimental preparations, and in fact had a greater inhibitory effect than did a collection of retinoids commonly used clinically. Of further relevance to gliomas is that one of lycopene's mechanisms of action is to inhibit the insulin-like growth factor, which as noted above is involved in the development of resistance to a variety of different treatment agents. (150). Also of interest is evidence that it synergizes with Vitamin D (151).

Boswellic Acid

This an extract from Indian folk medicine used for its anti-inflammatory effects. Laboratory studies have shown that its mechanism of action is the inhibition of the lipoxygenase inflammatory pathway, which is the source of inflammatory leukotrienes (152). This inflammatory pathway is distinct from the cyclogenase pathway that was discussed earlier in the section on Celebrex and other NSAIDs. Boswellic acid is now used in Germany as a substitute for steroids as a method of reducing the edema associated with gliomas. There have also been reports (153, 154) from in vivo animal laboratory experiments that it has direct anti-cancer effects . It seems plausible that its combination with celebrex or other COX-2 inhibitors might be synergistic.

Broccoli Sprouts

Brassica vegetables such as broccoli, cauliflower, brussel sprouts, and cabbage have long been believed to have anti-cancer properties, with the prevailing theory of the basis of that effect being that they contain a substance known as sulphoraphane. Recently it has been discovered that the 3-4 day-old sprouts of these vegetables contain 10-100 times the concentration of sulphoraphane as do the full-grown vegetables. To test whether the oral ingestion of sprouts has anti-cancer effects, dried broccoli sprouts were included in the diet of rats with chemically-induced cancers, with the result that considerable regression of the tumors were observed (155). Broccoli sprouts are also very tasty additions to salads.

Ellagic Acid

This is a phenolic compound present in fruits and nuts, including raspberries, blueberries, strawberries, and walnuts. In laboratory experiments it has been shown to potently inhibit the growth of various chemical-induced cancers, with the basis of the effect being an arrest of cell division in the G stage of cell division, thus producing the programmed cell death known as apoptosis. While there have been no trials to assess its clinical effects with human patients, it should be obvious that quantities of berries and nuts are among the more enjoyable dietary components, and even the possibility that they may have anti-cancer effects should encourage their usage.

Berberine

This is an alkaloid extract from Coptides Rhizoma commonly used in China as an herbal medicine. It is also found in high concentration in the widely-used supplement, goldenseal. In one laboratory study of using both various kinds of glioma cell cultures and implanted tumors in rodents (156), the cytotoxic effects of berberine were compared to those of BCNU and to the combination of berberine and BCNU. Berberine used alone produced a 91% kill rate in cell cultures, compared to 43% for BCNU. The combination produced a kill rate of 97%. Comparable results were obtained with the in vivo implanted tumors. Such results suggest that berberine is among the most promising treatment agents, but to date very little research using it has been reported.

Resveratrol

This is a naturally occurring polyphenol found most abundantly in grapes and mulberries. Red wine is among the sources. Numerous experimental studies have shown that it inhibits proliferation of various kinds of cancer, including leukemia, prostate, breast, and colon cancer. Among its mechanisms of action are activation of the P53 gene, inhibition of protein kinase C, and the inhibition of new blood vessel growth. In the one recent study of its use with glioma tumors (157), rats received either sub-cutaneous injections or intra-cerebral injections of tumor cells which in control animals rapidly grew and became fatal. With sub-cutaneous tumors a dose of resveratrol of 40mg/kg produced major growth inhibition with 70% of the rats becoming long-term survivors. A higher dosage (100 mg/kg) was necessary to inhibit the growth of the intracranial tumors, and even it was only marginally effective. The difference in outcome for the two preparations suggest that resveratrol may be impeded by the blood-brain barrier. However, the authors note that it had significant anti-angiogenic effects, which are not affected by the blood-brain barrier. Whether resveratrol has clinical utility for brain cancer is unclear, although it is known that anti-angiogenic agents of various sorts synergize with various kinds of conventional treatment.

Cannabis

After years of governmental discouragement of research on Cannabis (the plant from which marijuana is derived), the last few years has seen a proliferation of research on its mechanisms of action. One result of this research is that it has been shown to inhibit the growth of various kinds of cancer cells, including gliomas (158). In the most recent paper (159), cannabinoids were shown to significantly inhibit angiogenesis in gliomas implanted in mice, which was accompanied by significant inhibition of glioma growth. The result is noteworthy because cannabis is among the more potent anti-nausea agents for controlling the side effects of chemotherapy.

Skeptics of supplements/dietary components such as those discussed above have argued that the laboratory studies providing evidence for their anti-cancer effects have used dosages that can never be achieved in human patients, and therefore the supplements are unlikely to be useful clinically. Without a study of the dose-effect relations in clinical settings there is no easy way to evaluate this concern. However, in several cases investigators of the various substances have noted that their effects in the laboratory were obtained with dosages comparable to what easily can be realized by dietary supplementation. In any event, for most of what has been discussed there is little if any risk to using the supplements, with the only cost being financial in nature. It is important to keep in mind that cancer treatment of all types is probabilistic in its outcome. Thus, any agent that adds even a small amount to the probability that a treatment program will be successful, and which also has no toxicity, is something that should be taken seriously as an additional component of a multi-faceted treatment program.

Noteworthy Clinical Trials

The earlier section on chemotherapy discussed the large number of new clinical trials that are combining different chemotherapy agents, now most commonly temodar, with other treatment agents such as tamoxifen, accutane, thalidomide, etc., as well with other chemotherapy agents (CPT-11, etoposide). It is increasingly clear that these combination treatments are more successful than chemotherapy used as a single agent. In this section I will concentrate on new treatment approaches that are not predicated on the traditional cytotoxic mechanism of killing all dividing cells in the body that come into contact with the chemotherapy agent. These newer treatment agents are the direct result of basic research on the processes of cell division, which have identified numerous new targets for intervention when cells become malignant and multiply wildly without differentiating into mature cells serving their various specialized functions. Only in the last two years have these new approaches been extended to primary brain tumors. The discussion will thus be at the level of the general rationale of the different treatment approaches, along with a report of the results of the initial early-stage clinical trials.

Anti-Angiogenesis

In order for tumors to grow they must recruit new blood vessels to meet the greatly increased energy demands. If the growth of new blood vessels could be prevented, the tumor's growth would necessarily stabilize or decrease, thus giving other treatments the opportunity to kill the cancerous cells. As noted in an earlier section, thalidomide is one drug that has such an inhibitory effect, and its use in clinical trials has had some success with several types of cancer. Its results as a single agent for brain tumors have not been impressive, although its combinations with other agents do seem more promising. The mechanisms of thalidomide's effect on blood vessel growth are not well understood. Celebrex also in now commonly used for it anti-angiogenic effects, although it has other mechanisms of action as well. Other existing drugs discussed in prior sections also have been shown to have anti-angiogenic effects, although it isn't clear whether these are secondary to other mechanisms of action.

Second-generation anti-angiogenesis agents have been developed based on specific hypotheses about how blood vessel growth is regulated. The majority of these new agents have targeted specific growth factors that a tumor sends out to stimulate blood vessel growth, or have targeted the receptors for those growth factors. At least a dozen such factors have been identified, the most important being fibroblast growth factor, platelet-derived growth factor, and vascular endothelial growth factor (VEGF), which is generally regarded as the most important. The multiplicity of growth factors is important to note because it implies there are redundant processes involved in stimulating blood vessel growth, which in turn suggests that targeting individual growth factors alone is unlikely to be successful.

Because anti-angiogenesis drugs are considered one of the most promising new approaches to cancer treatment, literally dozens of drug companies are in the process of developing their own approach to this new treatment modality. A number of these new drugs have been in clinical trials during the past two years and early results are just now being reported. One of them, Avastin (made by Genentech), recently was approved by the FDA, based on a successful clinical trial with colon cancer. It also has been shown to benefit renal cell cancer patients and lung cancer patients. Whether it will work for brain cancer is currently a matter of conjecture. Other anti-angiogenic drugs have been studied with gliomas with results now being reported from early-stage clinical trials. The first is an analog of thalidomide called CC-5103 (also known as revimid), which was engineered to have thalidomide's therapeutic effects without its side effects. In a phase I trial with recurrent high-grade gliomas (160), little toxicity was observed and several patients had stable disease, although the results were too early to evaluate meaningfully except for toxicity, which was minimal. A second drug, known as PTK787, which inhibits the VEGF signaling channel, has been studied as a single agent and in combination with temodar. Of 47 evaluable GBM patients receiving it as a single agent, there were 2 partial tumor regressions, and 31 with stable disease, along with clear evidence that blood vessel growth had been inhibited (161). When studied in combination with temodar (162), several partial tumor regressions and stabilization of disease were observed, but it is too early to determine whether this is an improvement over temodar used as a single agent. A third anti-angiogenic drug, which seems especially promising, currently being studied in early stage clinical trials at the National Cancer Institute, is LY31765, which targets a variant of protein kinase C that has been shown to be a critical part of the signaling pathway for VEGF. Of 32 patients reported on in the last meeting of ASCO (163), tumor regressions have been seen in five patients and stable disease in a significant number of others. In addition, the treatment seems to have minimal toxicity.

Given that brain tumor patients are unlikely to have access to these new treatments for some time to come, it is of interest to note that at least a half-dozen agents, already discussed in earlier contexts, possess significant degrees of anti-angiogenic activity. These include tamoxifen, accutane, gamma-linolenic acid, genistein, PSK, selenium, curcumin, and green tea. Silymarin, an extract of thistle plants used in Germany as a treatment for liver toxicity, has also been shown in laboratory studies to have anti-angiogenic properties. Vitamin D3 also has potent anti-angiogenic effects Also discussed in earlier sections was the anti-angiogenic effect of the new Cox-2 inhibitors used for arthritis. Many common analgesics, including aspirin and ibuprofen also have anti-angiogenic effects because new blood vessel growth is part of the inflammatory process. The advantages of the new COX-2 inhibitors are mainly that they have less stomach toxicity than more common, and much cheaper anti-inflammatory drugs. This supposedly should allow higher dosages to be used.

Perhaps the most interesting anti-angiogenic drug developed for other purposes is Rosigtilazone (trade name: Avandia), which was developed to increase insulin sensitivity for type II diabetics (164, 165). One of its properties is that it activates the PPAR-gamma gene which has been shown to inhibit cancer growth by encouraging apoptosis. Both gamma linolenic acid and fish oil, discussed above, has a similar effect on the PPAR-gamma gene.

A second class of existing drugs that have significant anti-angiogenic effects are members of the tetracycline antibiotic family, specifically minocycline and doxycycline (166). These drugs also inhibit metalloproteinases, which are enzymes that break down the cell matrix of the surrounding cells that allows cancer cells to invade that tissue (167). Reports of significant anti-angiogenic activity has been reported for a variety of other drugs,: methotrexate, a chemotherapy agent used in low dosages for the treatment of rheumatoid arthritis (168);, rapamycin (169), an immunosuppression drug used in organ transplants; and perindopril (170), one of the ACE inhibitors used in the treatment of high blood pressure.

The mechanisms underlying the anti-angiogenesis effects of each of these agents are largely unknown and possibly very different. Nevertheless, it seems feasible that a combination of these different agents might produce inhibition perhaps sufficient to be effective in its right, but also to substantially increase the effectiveness of traditional treatments, and that of other anti-angiogenic agents. For example, one recent laboratory study showed that the combination of thalidomide and sulindac (an anti-inflammatory analgesic used for arthritis) produced substantially greater inhibition of new blood vessel growth than did either agent in isolation (171). A number of other studies have also shown highly synergistic effects from combinations of different anti-angiogenic drugs.