A phase 1/2 study of orally administered synthetic hypericin for treatment of recurrent malignant gliomas
Abstract
BACKGROUND:
Hypericin is a potent inhibitor of glioma growth in vitro. To examine whether synthetic oral hypericin can be tolerated by patients with recurrent malignant gliomas (anaplastic astrocytoma and glioblastoma) and to investigate its efficacy against these tumors, the authors undertook an open-label, sequential dose escalation/de-escalation tolerance study.
METHODS:
Patients with documented recurrent or progressive malignant gliomas who had received standard radiation therapy with or without chemotherapy were included. Patients were excluded for previous treatment with agents known to contain hypericin or treatment within 30 days with medications known to cause photosensitivity. Enrolled patients were given gradually increasing dosages of oral synthetic hypericin (0.05-0.50 mg/kg) for up to 3 months if no toxicity was observed, and patient response to treatment was noted. The patients were examined each month and underwent magnetic resonance imaging to evaluate tumor status at 3 months.
RESULTS:
Synthetic hypericin administered orally appeared to provide stabilization or a slight (<50%) decrease in tumor volume (coded as stable disease) at 3 months for 7 of 42 patients (17%) and produced a tumor reduction >50% (partial response) in 2 patients (5%). Seventeen patients (40%) survived for 3 months on daily synthetic hypericin at dose levels of 0.33 ± 0.070 mg/kg daily. The mean maximum tolerated dose was 0.40 ± 0.098 mg/kg daily. Twelve patients continued on hypericin therapy beyond 3 months. The median survival was 26 weeks (Kaplan-Meier method).
CONCLUSIONS:
The results of this study indicated that synthetic, oral hypericin is well tolerated in this patient group. The response results were comparable to those reported from other studies of salvage therapies for recurrent malignant brain tumors. Cancer 2011;. © 2011 American Cancer Society.
Hypericin is a natural compound found in the stems and petals of plants of the genus Hypericum, which includes the common St. John's wort plant, Hypericum perforatum. In contrast to the St. John's wort plant extract of hypericin, which contains a variety of chemical constituents, synthetic hypericin (used in the current trial), a polycyclic naphthodianthrone, is synthesized chemically in 3 steps. Early clinical trials evaluated synthetic hypericin (VIMRxyn; Hy BioPharma, Jamison, Pa) as an antiviral agent in the treatment of human immunodeficiency virus1 and hepatitis C.2 In addition, synthetic hypericin is known as a potent protein kinase C inhibitor.3 To evaluate the potential of hypericin as an inhibitor of glioma growth, previous investigations were conducted on glioma cells in vitro.3 It has been demonstrated that hypericin inhibits glioma growth in a dose-related manner in vitro, with the marked inhibition of growth, motility, and invasion in the low-micromolar concentration range.4 Because the reported inhibitory effects of protein kinase C are enhanced by visible light, [3H]thymidine uptake has been measured in both the presence and the absence of visible light. In glioma line A172, the presence of light slightly increased the inhibitory effect of hypericin. Moreover, in an apoptosis assay to determine whether the treatment of glioma cells with hypericin was cytostatic or cytocidal, DNA from cells that were treated with hypericin for 48 hours exhibited a classic “ladder” pattern of oligonucleosome-sized fragments characteristic of apoptosis. These data suggest that hypericin may have potential as an antiglioma agent.4
Studies on its mechanism of anticancer activity indicate that hypericin binds to heat shock protein 90 (Hsp 90), leading to its ubiquitinylation. This disrupts several critical client proteins that regulate cell growth pathways, resulting in destabilization, rapid degradation, and elimination from the cells of a variety of Hsp 90 client proteins, ultimately leading to cell death.5 Additional research indicates that photoactivated hypericin possesses significant antiproliferative effects on activated normal human lymphoid cells and malignant T-cells purified from the blood of patients with Sezary syndrome or the leukemic phase of cutaneous T-cell lymphoma. The mechanism of the antiproliferative activity may be related to the high rate of apoptotic death in these lymphoid cells.6 Experiments have demonstrated that hypericin also has antiangiogenic activity in several rat ocular models.7 Because hypericin has been used in clinical trials for other tumor types,8-10 the objective of the current study was to determine the tolerability and any potential antitumor activity of orally administered, synthetic hypericin in patients harboring malignant gliomas (anaplastic astrocytoma and glioblastoma multiforme) that had recurred after standard therapy.
MATERIALS AND METHODS
To determine whether hypericin can be tolerated by patients with recurrent malignant gliomas and to investigate its efficacy against these tumors, we undertook an open-label, sequential dose escalation/de-escalation tolerance study in 42 patients (7 with anaplastic astrocytomas and 35 with glioblastomas) who had a diagnosis of recurrent malignant glioma. The study was conducted between 1996 and 1998 at 6 clinical centers in the United States and Canada (Westchester Medical Center, New York Medical College, Valhalla, NY; Trinity Medical Center, Minot, ND; Tom Baker Cancer Center, University of Calgary, Calgary, Alberta, Canada; University of Washington, Seattle, Wash; University of Southern California University Hospital, Los Angeles, Calif; and Walt Disney Memorial Cancer Institute at Florida Hospital, Orlando, Fla). The study proceeded sequentially in groups of 4 study participants, with each group receiving a step-wise higher dosage of hypericin if the previous group experienced no severe phototoxicity (mild phototoxicity was accepted) or other toxicities, including neurotoxicity, gastrointestinal toxicity, or myelotoxicity, after 14 days.
Study participants were treated as outpatients. Each patient returned to the clinic once daily during the initial 2 weeks or the first month (depending on dose escalation) to receive the dose of hypericin and, when indicated, to have a blood sample drawn. Thereafter, patients were provided their medication as a “mail prescription” prepared by the hospital pharmacist assigned to the study. Unless they were withdrawn early because of toxicity or intolerance (or lack of effectiveness indicated by clinical symptom progression or tumor progression on imaging), each study participant continued in the study for 3 months of hypericin treatment and remained under observation as noted until the follow-up visit. Each patient underwent a physical examination 4 times during the first month and then once monthly in Months 2 and 3. Volumetric magnetic resonance imaging (MRI) was done for tumor evaluation at 3 months.
Inclusion/Exclusion Criteria
Men or nonpregnant women aged ≥18 years with documented (by MRI supported by biopsy or [18F]fluorodeoxyglucose positron emission tomography screening) recurrent or progressive malignant gliomas were considered for the study. All tumor pathology at initial or repeat surgical resection was reviewed by a neuropathologist (D.R.H.). Included patients had previously received standard external-beam radiation therapy with or without adjunctive chemotherapy, had a Karnofsky performance status ≥70, had adequate baseline organ function documented by clinical laboratory testing, and had no evidence of intercurrent illness.
Patients were excluded from the study if they had received previous treatment at any time with agents known to contain hypericin (eg, Hyperforate 500 mg, Psychotonin M alcohol extract, St. John's wort), treatment within 30 days before study entry with any medications known to cause phototoxicity or photosensitization, or treatment within 2 weeks before study entry with any investigational drug(s). Other exclusion criteria included the presence of significant cardiovascular, renal, gastrointestinal, hepatic, or central nervous system disorders; concurrent substance abuse; or generalized skin rash or other disorders that potentially could interfere with the assessment of phototoxicity.
All patients were adequately informed of the nature and risks of study and provided written informed consent before the first study procedures. Institutional review board approval was obtained at each of the treating sites.
Treatments Administered
Hypericin was administered as an oral solution at doses ranging from 0.05 to 0.50 mg/kg once each morning for up to 3 months. Patients in the first group (Group I, 4 patients) received an oral dose of 0.05 mg/kg hypericin once daily for 14 days. Then, because no severe phototoxicity or other toxicity was observed at the initial dosage, the Group I patients were given higher dose levels of the drug. The next treatment group (Group II, 4 patients) received 0.1 mg/kg daily as the first dose. Because little (if any) phototoxicity or other toxicity was observed at that dosage level, dose levels for these patients were escalated, and all subsequent patients (Group III, 34 patients) received 0.20 mg/kg daily as the initial dosage. In each group, if no toxicity was observed after 14 days of dosing at 1 level, then the dosages were increased step-wise from 0.25 mg/kg to 0.5 mg/kg, at the discretion of the investigators if the drug was tolerated. The sponsor was notified of negative effects experienced by the patients. The investigators and the sponsor confirmed the dose escalation and de-escalation. The dosing escalation criteria were intended to avoid or minimize potential harm to the patients.
Hypericin dosing was calculated according to body weight. Patients were weighed within 48 hours before administration of the first dose of study medication, and that weight was used to calculate the dosage for the duration of the patient's participation in the trial. The maximum tolerated effective dose was defined as the dose that caused antiglioma activity with minimal phototoxicity or other toxicities, including neurotoxicity, gastrointestinal toxicity, or myelotoxicity.
The objective of the dose adjustment scheme was to have a minimum of 16 total patients from any group complete 3 months of treatment at the maximum tolerated dose. If severe phototoxicity or other grade III or greater toxicities were observed at any dose level, then no further dose increases were made. In the event that either of these conditions was met, the next dose group received a dose that was a half of the log10 dose lower than the dose that produced toxicity. Whenever severe toxicity was observed in a patient, that patient was discontinued from the study and/or the dosage was adjusted downward at the discretion of the physician (with the consent of the patient). In the event that moderate (tolerable grade II) toxicity was observed in any patient, dose de-escalation (a half-log10 dosage level) was undertaken in that patient. In all instances, the patients were monitored closely for latent toxicity. If patients developed mild, moderate, or severe phototoxicity, then a dermatologist was consulted on the same day, and appropriate photographs were taken of the affected skin with the patient number and day of treatment recorded.
Stock solutions of chemically synthesized hypericin (not the plant extract) were prepared from hypericin powder by VimRx Pharmaceuticals (Wilmington, Del; data on file at Hy BioPharma Inc.). All dosage preparation steps were documented by the study personnel, and appropriate quality-assurance measures were applied to assure the accuracy of the dosage calculations, weighing of drug, dilutions, and other dosage preparation steps.
Drug Concentration Measurements and Pharmacokinetics
Blood was drawn for pharmacokinetic analyses into heparinized tubes at predetermined times of 0 hours and 6 hours (1 day weekly) during Weeks 0, 1, 2, and 3; an additional set of blood samples was obtained in Week 12 at 0 hours, 2 hours, 4 hours, 8 hours, 10 hours, 12 hours, 24 hours, and, in 2 cases, 48 hours. The plasma was centrifuged at ×1000g for 10 minutes, aliquoted into cryovials, stored at −20°C, and shipped in batches on dry ice to the Clinical Core Laboratory at New York University Medical Center.
Hypericin-spiked standard solutions were prepared from a 2-mg/mL stock in ethanol to working dilution of 1 μg/mL. Aliquots of these working standards were used to provide plasma standards in the range from 20 ng/mL to 1280 ng/mL as well as precision samples in low, medium, and high range values of the standard curve (40 ng/mL, 120 ng/mL, and 640 ng/mL). Aliquots of 0.5 mL were removed into 12 × 75-mm glass centrifuge tubes and stored frozen at −20°C until they were used. For a given assay, a 6-point standard curve with each concentration extracted in duplicate was used along with triplicate samples of the 3 precision standards.
Each plasma sample was analyzed twice at 590 nM using a validated high-performance liquid chromatography method previously described11 with the following modifications: Two 0.5-mL aliquots of the plasma sample were processed through 2 consecutive extraction cycles with 0.125 mL of dimethyl sulfoxide. This was followed by the addition of 0.625 mL of a mixture consisting of acetonitrile/2-butoxyethanol, 90/10 (volume/volume), to the samples, which were then centrifuged. Supernatant was collected and transferred to a 2-mL volumetric flask; then, the volume was brought up to 2.0 mL.
Pharmacokinetic analyses of plasma levels were undertaken using WinNonLin Pro Node (version 3.3; Pharsight Corp., Mountain View, Calif), and the dose proportionality of minimal plasma concentration (Cmin) and maximal plasma concentration (Cmax) values were assessed by regression analyses using Sigma Plot (version 8.0.2; SPSS Inc., Chicago, Ill).
Statistical Analysis
A Kaplan-Meier test was used to determine the median survival of all patients in the intent-to-treat population. Survival was calculated from the first day of hypericin dosing to the date of death for 9 patients who were uncensored. For 33 patients who were censored, survival was calculated conservatively from the first day of dosing to the last available date patients were known to be alive, which was the date they received their last dose of medication.
RESULTS
Demographics and Monitoring of the Study Population
Demographic and other baseline characteristics for all patients entered into the study are summarized in Table 1. The mean weight (±standard deviation) of the patients at enrollment was 81.6 ± 16.61 kg. The median follow-up for all patients was 3 months (12 weeks), and follow-up ranged from <1 month (<1 week) to 36 months (155 weeks).
| Age | |
|---|---|
| Mean±SD | 50.0±12.62 y |
| Median | 49.5 y |
| Range | No. of Patients |
| 18-40 | 11 |
| 41-64 | 25 |
| 65-75 | 4 |
| >75 | 2 |
| Sex | No. of Patients |
| Men | 25 (60) |
| Women | 17 (40) |
| Race | No. of Patients |
| Caucasian | 35 (83) |
| Hispanic | 3 (7) |
| American Indian | 2 (5) |
| Not defined | 2 (5) |
| Diagnosis | No. of Patients |
| GBM | 35 (83) |
| AA | 7 (17) |
| Karnofsky score | Score |
| Mean±SD | 78.6±13.36 |
| Median | 80.0 |
| Range | 50-100 |
- Abbreviations: AA, anaplastic astrocytoma; GBM, glioblastoma; SD, standard deviation.
Pharmacokinetics
Pharmacokinetic studies were performed in 6 patients. Dose proportionality assessed by regression analyses of Cmin and Cmax values as a function of dose indicated the interpatient correlation between dose and steady-state plasma concentration levels. Figure 1 contains plots of the mean ± standard deviation data for the steady-state Cmin and Cmax, which clearly indicate dose proportionality (P ≤ .0003). A linear regression analysis of these parameters indicated correlations (r2) of 0.95 and 0.83 for Cmin and Cmax, respectively.
Mean minimal serum concentrations (Cmin) and peak serum concentrations (Cmax) of hypericin are illustrated as a function of dose at steady state.
Figure 2 illustrates mean and log plasma profiles of hypericin levels in patients who received a mean dose of 0.40 ± 0.07 mg/kg. By using a “2-compartment open-model” with first-order absorption and elimination rates, a correlation coefficient r2 = 0.941 was achieved. At steady state, a long half-life was manifested with both the mean and the geometric mean levels.
Mean and semi-log plasma hypericin profiles are shown from patients with glioma after daily doses over a 3-month period (N = 6; dose = 0.40 ± 0.07 mg/kg). V2/F indicates volume of distribution in the tissue compartment; t1/2, half-life.
Table 2 summarizes the pharmacokinetic parameters obtained by fitting the mean and/or geometric mean plasma levels in 6 patients. It is noteworthy that a long half-life of 68.6 to 72.1 hours was observed at steady state, and apparent extensive tissue distribution was observed with a volume of distribution in the tissue compartment between 713 and 756 mL/kg predicted by the pharmacokinetic model. Also noteworthy were the magnitude of rate constants K12 versus K21, which indicate that the rate of distribution of hypericin into the tissue compartment was >2-fold that of the distribution back into the “plasma compartment.” These findings strongly suggest that hypericin is distributed rapidly into the intracellular tissue compartment, where it behaves as a “tissue reservoir.”
| Parameter | Units | Mean Estimate | GeoMean Estimate |
|---|---|---|---|
| K10 | 1/h | 0.040 | 0.034 |
| K12 | 1/h | 0.181 | 0.180 |
| K21 | 1/h | 0.071 | 0.080 |
| V1/F | mL/kg | 280.46 | 334.84 |
| Tmax | h | 4.45 | 4.635863 |
| Cmax | ng/mL | 621.53 | 522.41 |
| AUC | h*ng/mL | 35,586.12 | 34,356.99 |
| K01 t1/2 | h | 2.43 | 2.45 |
| K10 t1/2 | h | 17.30 | 20.20 |
| Alpha | 1/h | 0.282 | 0.284 |
| Beta | 1/h | 0.010 | 0.0096 |
| Alpha t1/2 | h | 2.46 | 2.44 |
| Beta t1/2 | h | 68.58 | 72.09 |
| CL/F | mL/h/kg | 11.24 | 11.49 |
| V2/F | mL/kg | 713.46 | 756.32 |
- Abbreviations: Alpha, absorption/distribution rate constant; Alpha t1/2, absorption/distribution half-life; AUC, area under the plasma concentration–time curve; Beta, terminal elimination rate; Beta t1/2, terminal plasma half-life; CL/F, total body clearance; Cmax, peak serum concentrations; h, hour; K01 t1/2, absorption half-life; K10 t1/2, elimination half-life; K10, elimination rate; K12, distribution rate in tissues; K21, distribution from tissues; Tmax, time of peak concentrations; V2/F, volume of distribution in tissue compartment.
Tolerability
In total, 17 patients completed a 3-month treatment regimen, and 12 patients elected to continue therapy under a compassionate-use exemption. For the latter patients, daily hypericin at levels were 0.33 ± 0.070 mg/kg. The mean maximum tolerated dose was 0.40 ± 0.098 mg/kg daily.
Adverse Events
Toxicity was graded in the standard manner using National Cancer Institute Common Toxicity Criteria.12 Of 42 patients who were enrolled in the intent-to-treat cohort, 31 patients (73.8%) experienced adverse events attributed to study medication. According to what might be expected from the pharmacology of hypericin, the body system/organ class with the highest incidence of adverse events attributed to study medication was skin and subcutaneous tissue disorders (30 of 42 patients; 71.4%), including photosensitivity reaction (18 of 42 patients; 42.9%), erythema (11 of 42 patients; 26.2%), and skin burning sensation (6 of 42 patients; 14.3%). No patient had a severe enough adverse effect from skin reaction that hospitalization was indicated. Nervous system disorders (35 of 42 patients) included convulsions, hyperesthesia, and parasthesia, although the cause of these disorders most likely was disease progression rather than study medication. Finally, gastrointestinal side effects, including abdominal distension, vomiting, and diarrhea, occurred in 10 of 42 patients (23.8%). No myelotoxicity occurred during treatment with hypericin, although 1 patient had a mild case of thrombocytopenia. No significant alopecia occurred in any of the patients; 1 patient had alopecia listed as an adverse event.
Response to Treatment and Survival
Twenty percent of patients with glioblastoma multiforme responded to the treatment, and 29% of patients with anaplastic astrocytoma responded (Table 3). For the 25 patients who failed to respond during the initial 3 months, the dose level ranged from 0.20 to 0.50 mg/kg, and 60% of those patients received doses ≥0.40 mg/kg. All of these patients failed within 30 to 60 days of initiating hypericin treatment. Among the patients who failed to complete 3 months of treatment, none demonstrated stable disease or partial response. For the patients who did complete 3 months of treatment, final dose and response data are listed in Table 4.
| Diagnosis | Total No. of Patients | Nonresponders | SD+PRa | SD | PR |
|---|---|---|---|---|---|
| Glioblastoma multiforme | 35 | 28 | 7 | 6 | 1 |
| Anaplastic astrocytoma | 7 | 5 | 2 | 1 | 1 |
| Total | 42 | 33 | 9 | 7 | 2 |
- Abbreviations: PR, partial response; SD, stable disease.
- a PR was defined as a reduction >50% in tumor area.
| No. of Patients | ||||
|---|---|---|---|---|
| Final Hypericin Dose, mg/kg | Total | PRa | SD | No Response |
| 0.2 | 1 | 0 | 1 | 0 |
| 0.3 | 3 | 0 | 1 | 2 |
| 0.35 | 2 | 1 | 0 | 1 |
| 0.4 | 6 | 1 | 3 | 2 |
| 0.45 | 1 | 0 | 1 | 0 |
| 0.5 | 4 | 0 | 1 | 3 |
- Abbreviations: PR, partial response; SD, stable disease.
- a PR was defined as a reduction >50% in tumor area.
For purposes of efficacy analysis, patients were divided into 3 populations based on their length of treatment, as presented in Table 5. No patient had a complete response to treatment. Twenty-two percent of patients (9 of 42) in the intent-to-treat population achieved either stable disease (17%; 7 patients) or a partial response (5%; 2 patients) during hypericin treatment. The objective of the study was for at least 16 patients to complete treatment with hypericin for 3 months (per protocol population). Of the 17 patients who completed 3 months of treatment with hypericin, 53% (9 of 17 patients) achieved either stable disease (41%; 7 patients) or a partial response (12%; 2 patients). One additional patient completed at least 2 months of treatment. Thus, for the 18 patients who completed >60 days of treatment, 50% achieved either stable disease (39%) or a partial response (11%), similar to what was achieved by patients who successfully completed only 3 months of treatment.
| No. of Patients (%) | |||||
|---|---|---|---|---|---|
| Analysis Population | All Patients | PRa | SDb | PR+SD | Total Nonresponders |
| Intent-to-treat: Patients who initiated dosing | 42 | 2 (5) | 7 (17)c | 9 (21) | 33 (79) |
| Per protocol: Patients who achieved 3 mo of therapyd | 17 | 2 (12) | 7 (41)c | 9 (53) | 8 (47) |
| Evaluable: Patients who achieved >60 d of therapye | 18 | 2 (11) | 7 (39)c | 9 (50) | 9 (50) |
- Abbreviations: PR, partial response; SD, stable disease.
- a PR was defined as a reduction >50% in tumor area.
- b SD was defined as no definitive progression in tumor size as determined by magnetic resonance imaging.
- c One patient was stable at 3 months but had achieved a PR at 6 months.
- d Per protocol patients were defined as those who completed 12 seven-day weeks (84 days of dosing).
- e For a patient to be evaluable and allowed to continue in the study, the patient must have received a minimum of 5 consecutive doses in the first week and must have missed only 1 dose per week in the first 2 weeks of therapy. Beyond Day 14, the patient, although deemed evaluable, would be allowed to continue provided they did not skip more than 2 consecutive doses in any given week and a total of 4 skipped days in the 3-month treatment period.
Of the 42 patients who were enrolled, 17 patients (40%) survived to 3 months after enrollment. The median survival for all patients in the intent-to-treat analysis, as determined by the Kaplan-Meier method, independent of their response to treatment, was 26 weeks (Fig. 3). In addition, 8 of 42 patients survived beyond 6 months (mean survival, 20.8 ± 11.35 months). Further analysis indicated that the mean progression-free survival of all patients in the intent-to-treat group was 20 weeks (Fig. 4).
This Kaplan-Meier plot illustrates the median survival of patients with glioma in the intent-to treat population. The median survival was 26 weeks.
This Kaplan-Meier plot illustrates the median progression-free survival of patients with glioma in the intent-to-treat population. The median time to progression was 20 weeks.
Partial radiographic reduction in tumor size (>50% but <100%) was observed in 2 patients who had received hypericin for >3 months. Both of those patients had recurrent disease, 1 with recurrent anaplastic astrocytoma and the other with recurrent glioblastoma multiforme. Figure 5 illustrates the response obtained in the patient who had a recurrent anaplastic astrocytoma along with the documented radiographic reduction in tumor size. That patient was stable at last follow-up (36 months post-treatment). The other patient, who had a recurrent glioblastoma, had undergone repeat craniotomy, resulting in postoperative bone flap infection, necessitating its removal. This patient underwent hypericin treatment and achieved a partial radiographic reduction in tumor volume.
These coronal magnetic resonance images from Patient C-05, who had recurrent anaplastic astrocytoma after treatment with radiotherapy and lomustine/vincristine, reveal tumor regression. The patient achieved a partial response after treatment with hypericin. Extended treatment (36 months) was administered; and, at last follow-up (36 months after treatment), the patient still was alive and functional.
DISCUSSION
The prognosis for patients with malignant gliomas remains poor despite advances in surgery, radiation, and chemotherapy. The rationale for postulating an antitumor effect for hypericin originally related to its ability to inhibit protein kinase C, an enzyme that has been demonstrated to correlate with glioma cell growth in vitro.3, 13, 14 However, other antitumor mechanisms, including ubiquitinylation of Hsp 90 and mediation by an O2-dependent degradation domain through the ubiquitin-proteasome pathway, also probably are involved.5, 15
Hypericin readily crosses the blood-brain barrier and may be administered parenterally or orally.16 Pharmacokinetic studies have demonstrated that hypericin has a long half-life at steady state with extensive intracellular tissue distribution. Quantifiable levels of hypericin have been documented in the glioma tissue of a patient with recurrent glioblastoma multiforme who underwent repeat surgery and previously had received hypericin. Intratumoral hypericin levels were determined from the tumor extractions and were measured in the range from 596 to 678 ng/g.
The current protocol evaluated the safety of daily oral administration of hypericin for patients with recurrent malignant gliomas in a phase 1/2 dose-escalation study. The maximum tolerated dose was reached at 0.40 ± 0.098 mg/kg daily, and 40% of patients enrolled survived to 3 months on hypericin at daily levels of 0.33 ± 0.070 mg/kg. This provides us with a guideline of dosage for patients with recurrent gliomas. Twelve patients in our study continued on hypericin therapy beyond 3 months. These results indicate that hypericin is well tolerated orally in this patient group. The dose-limiting toxicity in these patients was photosensitivity. No significant myelotoxicity or alopecia was observed in these patients.
The response results from this study are comparable to those from other published studies on salvage therapies for recurrent malignant brain gliomas that have been suggested as promising. Treatment with temozolomide for recurrent glioblastoma using the standard 3-week-on and 1-week-off regimen in 33 patients with chemotherapy-naïve, recurrent glioblastoma resulted in a 6-month progression-free survival rate of 30.3% and an overall response rate of 9%.17 Implantation of Gliadel wafers (carmustine) placed surgically for the treatment of recurrent malignant gliomas resulted in an overall median survival of 46 weeks after placement.18 Bevacizumab, the latest addition to the US Food and Drug Administration-approved armamentarium for recurrent malignant gliomas, improved the 6-month progression-free survival rate to 42.6% of patients who received bevacizumab alone.19 In our study, hypericin administered orally appeared to provide stabilization and even produced a slight (<50%) decrease in tumor volume (coded as stable disease) at 3 months for 7 of 42 enrolled patients (17%) and a tumor reduction >50% (but <100%), which was coded as a partial response, for an additional 2 patients. Forty percent of the patients enrolled in our study survived to 3 months after enrollment on daily hypericin. Eight of 42 patients (19%) survived beyond 6 months (mean survival, 20.8 months).
When the treatment effect of hypericin was examined for each tumor type independently (Table 3), because of the small sample size, hypericin did not demonstrate a significant treatment effect for recurrent anaplastic astrocytoma. When combined with the glioblastoma response data, however, the astrocytoma data were suggestive of a treatment effect on astrocytoma as well. A larger sample size will be required to further evaluate the response to hypericin in patients with recurrent gliomas.
There are several potential reasons for the discrepancy in overall survival for these patients. First, in the current study, there were no exclusion criteria based on tumor size. Many of our patients had tumors as large as 4 × 4 × 4 cm, often with adjacent mass effect, yet these patients often had Karnofsky scores that were better than would be expected from tumors of that size. Because hypericin is administered orally with the goal of inhibiting protein kinase C, the time that was required for hypericin to be effective may not have be long enough for the drug to have an effect in some patients who had larger tumors. Future trials will need to include tumor size as an exclusion criterion to control for the effect of slower drug response. Finally, the effect of photosensitivity experienced by patients may be a potential attribute of long-term effects. For example, 1 patient who had an enduring effect from hypericin had undergone previous bone flap removal secondary to an infection. The possibility of hypericin photosensitization of the tumor bed to increase activity should be investigated further.
Since this hypericin study was performed, additional protein kinase C inhibitors have come up for clinical trial; however, to date, their use has not been without concern. Enzastaurin is a protein kinase Cβ selective inhibitor that has been used in investigational studies for a wide range of cancers, including gliomas, lung cancer, and non-Hodgkin lymphoma.20 In a phase 1 trial in patients with recurrent gliomas, Kreisl et al21 reported that increasing the dosing of enzastaurin from daily to twice daily dosing resulted in improved drug exposure but with unacceptable toxicities of thrombocytopenia and prolonged QTc at dose-limiting toxicities. These toxicities were not observed at the dose-limiting toxicity for hypericin. Bryostatin 1 is a protein kinase C partial agonist that reportedly has both antineoplastic and immune-stimulatory properties. In a phase 2 study in patients with recurrent metastatic melanoma, the major toxicity of this treatment was myalgia, but investigators reported that bryostatin 1 was not an effective treatment in these patients.22
The future of hypericin in the treatment of malignant gliomas most likely will focus on 4 potential areas of treatment: 1) first-line therapy in conjunction with temozolomide and radiation therapy; 2) combination therapy with other chemotherapeutic agents, including bevacizumab, irinotecan, and temozolomide; 3) use in temozolomide-resistant patients; and 4) local intratumoral application. First-line therapy with hypericin in conjunction with temozolomide and radiation therapy in newly diagnosed patients with malignant gliomas has the merit that hypericin has been established as a good radiation sensitizer and has an additive/synergistic effect with temozolomide in in vitro and in vivo models.23, 24 This combination model will optimize the chemotherapeutic and radiation-sensitizing model of hypericin. Hypericin also may be used as combination therapy with other chemotherapeutic agents like temozolomide instead of temozolomide alone, which is used currently after radiation therapy. Moreover, hypericin may be used in combination with drugs like irinotecan.25 The use of hypericin in salvage therapy for temozolomide-resistant patients is a possibility, and in vitro and in vivo experiments are pending to evaluate this application. Hypericin also reportedly has antiangiogenesis effects. Its use as an adjunct or in place of bevacizumab for temozolomide-resistant patients will need to be evaluated in the future.7 Finally, the use of hypericin for local photodynamic therapy of the tumor bed has been evaluated by other investigators for other cancers, including esophageal, nasopharyngeal, and bladder malignancies.26, 27 Whether it may be used as a local photodynamic agent for the treatment of malignant gliomas will need to be evaluated in the future.
In conclusion, the results from this phase 1/2 trial of orally administered, synthetic hypericin in patients with recurrent malignant gliomas indicate that it is well tolerated and exhibits modest efficacy in this patient group. Further trials of hypericin as adjuvant therapy for patients with malignant glioma are warranted.
Acknowledgements
We thank Leonard Liebes, PhD, for assistance with blood level analysis and Kristin Kraus, MSc, for editorial assistance.
FUNDING SOURCES
The synthetic hypericin used in this study was provided by Hy BioPharma (Jamison, PA).
CONFLICT OF INTEREST DISCLOSURES
Dr. Tobia is the President of Hy BioPharma and has stock or other ownership interests in the company. Dr. Cabana is the Vice President of Clinical Pharmacology and Regulatory at Hy BioPharma and has stock or other ownership interests in the company. Dr. Cabana is a consultant for NDS International, Inc. Dr. Chen was reimbursed by Hy Biopharma for travel expenses. The other authors had no financial support other than for expenses related to clinical trial implementation.
