|Home | About | Journals | Submit | Contact Us | Français|
Refractory anaplastic oligodendroglioma (AO) and oligoastrocytoma (OA) tumors are challenging to treat. This trial primarily evaluated toxicity and estimated the maximum tolerated dose (MTD) of intra-arterial (IA) melphalan, IA carboplatin and intravenous (IV) etoposide phosphate in conjunction with blood-brain barrier disruption (BBBD) in these tumors. The secondary measure was efficacy.
Thirteen subjects with temozolomide (TMZ) - refractory AO (11) or OA (2) underwent BBBD with carboplatin (IA, 200 mg/m2/day), etoposide phosphate (IV, 200 mg/m2/day), and melphalan (IA, dose escalation) every 4 weeks, for up to 1 year. Subjects underwent melphalan dose escalation (4, 8, 12, 16, and 20 mg/m2/day) until the MTD (one level below that producing grade 4 toxicity) was determined. Toxicity and efficacy were assessed.
Two of four subjects receiving IA melphalan at 8 mg/m2/day developed grade 4 thrombocytopenia, thus the melphalan MTD was 4 mg/m2/day. Adverse events included asymptomatic subintimal tear (1 subject) and grade 4 thrombocytopenia (3 subjects). Two subjects demonstrated complete response, 3 had partial responses, 5 demonstrated stable disease and 3 progressed. Median overall PFS was 11 months. Subjects with complete or partial response demonstrated deletion of chromosomes 1p and 19q. In the 5 subjects with stable disease, 2 demonstrated 1p and 19q deletion and 3 demonstrated 19q deletion only.
In these patients with AO or OA tumors who failed TMZ, osmotic BBBD with IA carboplatin, IV etoposide phosphate, and IA melphalan (4mg/m2/day for 2 days) shows acceptable toxicity and encouraging efficacy, especially in subjects demonstrating 1p and/or 19q deletion.
Patients who harbor anaplastic oligodendroglioma (AO) and oligoastrocytoma (OA) tumors are difficult to treat. Although extensive resection is associated with increased survival (11, 85), further adjuvant treatment (radiation therapy or chemotherapy) is recommended in patients with these aggressive tumors. (84) The responsiveness of AO to chemotherapy was established in trials of recurrent tumors, initially with the procarbazine, lomustine and vincristine (PCV) regimen and more recently with temozolomide (TMZ). Approximately two thirds of patients with recurrent AO after prior RT demonstrate either a complete response (CR) or partial response (PR) to PCV chemotherapy, with a time to progression of 12 to 18 months.(12, 80) Because this regimen is associated with significant hematologic and GI toxicity, most patients do not tolerate the six cycles of PCV intended. The reported response rate of recurrent oligodendroglioma to TMZ after failure of RT is up to 50% with a median progression-free survival (PFS) of 10 to 12 months.(9, 87, 90) TMZ is better tolerated with modest myelosuppression and easily controlled nausea and vomiting. Although no formal comparison between PCV and TMZ in recurrent oligodendroglial tumors is available, TMZ is the drug of choice at most institutions because of the better tolerability and the ease of administration.(84)
In patients with TMZ-refractory AO or OA, treatment options are lacking. Many patients are not good candidates for additional surgery due to tumor size and location.(72, 73) A few agents have been investigated for second-line chemotherapy for recurrent AO, including CPT-11, carboplatin, cis-retinoic acid and others with mixed results.(8, 10, 13–15, 38, 46, 79, 83, 90)
Chemotherapy administration by the IA route in conjunction with BBBD is an approach that can improve chemotherapy delivery to brain and tumor while avoiding systemic toxicities (2, 18, 24, 40, 50, 55, 59, 62, 63, 66). Our group has obtained encouraging preliminary results using IA chemotherapy in conjunction with BBBD in patients with many tumor types including aggressive oligodendroglioma. (23, 27, 34, 39, 40, 43, 55, 58, 66, 89). Rationale and justification for use of this therapy for aggressive oligodendroglial tumors gains further support from our most recent report, summarizing the multi-institutional experience of 149 newly diagnosed patients with primary central nervous system lymphoma (PCNSL). Although PCNSL is chemoresponsive, like AO and OA, chemotherapeutic agents are often ineffective due to limited ability to cross the BBB. Use of radiation therapy is often associated with significant neurotoxicity. Delivery using BBBD and IA administration in 149 consecutive PCNSL patients who had not previously been treated with whole-brain radiotherapy has yielded exciting results, with 25% of patients being alive at 8 years and, in some patients, survival extending beyond 20 years.(2) Importantly, survival occurred without the cognitive loss associated with up-front radiation. These encouraging results, in addition to prior studies that span more than 20 years (23, 27, 34, 39, 40, 43, 55, 58, 66, 89), support investigation of this method of improved delivery to other chemoresponsive tumors, such as AO and OA.
This current prospective trial was undertaken primarily to evaluate toxicity and estimate the maximum tolerated dose (MTD) of IA melphalan administered with IA carboplatin and intravenous (IV) etoposide phosphate in conjunction with osmotic BBBD. The results of this completed Phase I study are reported here. As a secondary endpoint was to examine response to this therapy, we also report the preliminary results of the ongoing Phase II component of the study.
Adult patients with AO or OA brain tumors who had failed prior treatment with TMZ were enrolled in this prospective Phase I/II clinical trial. Patients were treated with IA melphalan, IA carboplatin and IV etoposide phosphate in conjunction with osmotic BBBD, on two consecutive days, every four weeks for up to one year, from October 2005 to January 2008. All subjects were treated at Oregon Health & Science University (OHSU). Data were extracted from password-protected institutional databases and verified using the patients’ records. Original pathology reports, computed tomography (CT), and magnetic resonance imaging (MRI) scans were reviewed to verify histopathologic diagnoses and radiographic tumor characteristics. This study was approved by the Institutional Review Board (IRB) at OHSU.
Data were prospectively collected at study entry and at every visit during the study. Eligibility criteria included pathologic evidence of AO or OA according to World Health Organization classification criteria; prior treatment with TMZ with at least 28 days elapsed since completion of chemotherapy; prior consultation with a radiation oncologist; age 18 to 75 years; KPS greater than or equal to 50; Eastern Cooperative Oncology Group (ECOG) Performance Status less than or equal to 2; White blood cell count > 2.5 × 103/mm3; Absolute granulocyte count > 1.5 × 103/mm3; Platelets > 100 × 103/mm3; Serum Creatinine < 1.5 × upper limit of normal; Bilirubin < 1.5 × upper limit of normal; baseline SGOT/SGPT < 2.5 × upper limit of normal, negative pregnancy test and adequate birth control methods in women with childbearing potential; no uncontrolled significant medical conditions, no radiographic signs of intracranial hypertension and/or spinal block; no significant risk of general anesthesia; no contraindication to carboplatin, melphalan, etoposide phosphate or sodium thiosulfate; and informed consent. Prior carboplatin/etoposide treatment was not an ineligibility criterion. No other anti-tumor treatments were allowed while on study.
The IA/BBBD technique and the concomitant supportive care have been previously described in detail. (24, 26, 50) Briefly, BBBD treatment was completed on two consecutive days, every 4 weeks, for up to one year, for a maximum of 24 treatments per patient. One cycle of therapy was equal to 4 weeks. After induction of general anesthesia, IV etoposide phosphate (200 mg/m2/day; total dose 400 mg/m2) was infused. A femoral artery was accessed, and either an internal carotid or vertebral artery was catheterized and confirmed by fluoroscopy. In the case of multifocal disease, both internal carotid arteries and one vertebral artery were catheterized in an alternating monthly fashion (one vessel per treatment) to ensure homogeneous drug delivery throughout the affected brain regions. Warmed (37C) 25% mannitol was delivered at a predetermined flow rate of 4–12 mL/s into the intracranial artery for 30 seconds. Flow rate was determined by iodinated contrast injection and fluoroscopy, as the lowest infusion rate in which there was retrograde flow from the arterial catheter. After administration of mannitol, IA melphalan (dose escalation) was infused over 10 minutes. Following melphalan, carboplatin (200 mg/m2/day; total dose 400 mg/m2) was infused IA over 10 minutes. Following administration of IA carboplatin, the trans-femoral catheter was removed and pressure was manually applied to the femoral artery puncture site for 10 minutes. Sodium thiosulfate (STS), IV, was administered 4 and 8 hours after carboplatin at 20 and 16 gm/m2 respectively for hearing protection. Because high-dose STS causes transient hypernatremia, hypertension and controllable grade II nausea and vomiting, patients were pre-medicated with antiemetics prior to STS infusion. Subjects treated with IA carboplatin were given dexamethasone taper after the second day of BBBD. Granulocyte colony stimulating factor (G-CSF) was given 48 hours after last dose of carboplatin, based on patient’s body weight.
Patients underwent 2 sequential BBBD procedures 24 hours apart via the right internal, left internal or vertebral arteries with each treatment course. The phase I study consisted of dose escalation of melphalan while carboplatin and etoposide phosphate doses remained fixed. MTD was defined as one dose level below the dose level that produced grade 4 non-hematologic or grade 4 hematologic toxicity in 33% of subjects. The first 3 patients were treated at melphalan dose 4 mg/m2 each consecutive day (dose level 1). Dose escalation proceeded as follows: Dose Level 1, 4 mg/m2 daily X 2 (8 mg/m2/course); Dose Level 2, 8 mg/m2 daily X 2 (16 mg/m2/course); Dose Level 3, 12 mg/m2 daily X 2 (24 mg/m2/course); Dose Level 4, 16 mg/m2 daily X 2 (32 mg/m2/course); Dose Level 5, 20 mg/m2 daily X 2 (40 mg/m2/course).
The dose escalation algorithm was as follows: A group of 3 subjects entered at dose level 1. If none of 3 subjects experienced toxicity as defined above, the entry dose level was escalated for the next 3 subjects. If one of 3 subjects experienced toxicity, the same dose was repeated for the next 3 subjects. If two or more subjects experienced toxicity, the dose escalation stopped, and the MTD was one dose level lower. Subjects remained at the same dose level during their entire course of treatment. The dose escalation was based on the first course of treatment. If a subject went off study prior to receiving one full course (two consecutive days) of treatment, the subject was deemed inevaluable for the dose escalation, however the subject was followed and toxicities continued to be monitored. An additional subject was entered at that dose level. Once the melphalan MTD was determined, the phase II portion of the study began with a fixed dose of melphalan (the MTD).
The carboplatin (200 mg/m2/day; total dose 400 mg/m2) and etoposide phosphate (200 mg/m2/day; total dose 400 mg/m2) remained at fixed doses for the duration of the study.
Common Toxicity Criteria for Adverse Events (CTCAE, version 3.0) was used to grade toxicities involving all systems. Subjects may have had a subsequent treatment course at full dose of carboplatin, melphalan, and etoposide phosphate if admission absolute granulocyte count was greater than or equal to 1.2 × 103/mm3; platelet nadir from previous course of carboplatin, melphalan and etoposide phosphate was greater than 20,000; and there was no documented sepsis. The chemotherapy dose was reduced for hematologic toxicity as follows: (1) first event platelet nadir less than 20 × 103/mm3, subject underwent 25% dose reduction of carboplatin on subsequent courses. (2) second event platelet nadir less than 20 × 103/mm3, subject underwent 25% dose reduction of melphalan on subsequent courses. (3) third event platelet nadir less than 20 × 103/mm3, no melphalan administered. (4) subject was allowed to go off protocol at any time, per the treating physician’s discretion. If the subject developed febrile neutropenia while on G-CSF, the subject underwent 25% dose reduction of carboplatin on subsequent carboplatin courses.
If the subject developed febrile neutropenia after having undergone 25% dose reduction of carboplatin further dose reduction followed. For patients with a calculated creatinine clearance (CrCl) less than 60 and greater than or equal to 40, the carboplatin dose was decreased to 75%. For patients with a calculated CrCl less than 40 and greater than or equal to 30, the carboplatin dose was decreased to 50%. Patients with CrCl less than 30 were not eligible for this protocol.
Criteria for early termination were as follows: (1) progression of disease, (2) voluntary patient withdrawal, (3) investigators decision that is in the patients best interest to withdrawal, (4) if the patient became pregnant, (5) if, in the opinion of the Investigator, laboratory values became abnormal to a clinically significant degree during the course of the study, (6) noncompliance, (7) significant protocol violation, and (8) for any reason, at the Sponsor or Investigators discretion.
Toxicity assessment included audiologic, opthalmologic and neuropsychologic evaluations. Audiologic assessments were conducted at baseline, monthly prior to each BBBD treatment, and within 30 days after the final course of BBBD. Ototoxicity was defined by the ASHA significant change criteria. (1) Opthalmologic assessment included visual acuity, intraocular pressure, color vision, fundoscopy, visual fields, extra-ocular movement, accommodation and convergence. Neuropsychological and Quality of Life assessments, conducted at baseline and after the final course of BBBD, included: Rey Complex Figure Test; Controlled Oral Word Association Test; Auditory Verbal Learning Test; Trail Making Test A and B; Wechsler Adult Intelligence Scale – III; Reading subtest from the Wide Range Achievenemt Test – 3; Finger Tapping; Symptom Checklist 90-R; and Quality of Life Assessment using the European Organisation for Research and Treatment of Cancer (EORTC) QLQ C – 30.
Subjects were followed for tumor response, duration of response, and survival. Neurological and MRI evaluations were performed monthly during protocol treatment. After the final course of the BBBD treatment regimen, evaluation was every 3 months for one year, every 6 months for 2 years, then annually. According to protocol, two sets of response data were followed. T1-weighted MRI with Gd was used for evaluation of tumors with enhancing component and T2-weighted MRI and FLAIR was utilized to assess non-enhancing tumors (adapted from Radiation Therapy Oncology Group – RTOG-98-02). Complete response (CR) refers to disappearance of all tumor on consecutive MRI images (contrast-enhances T1, T2, or FLAIR) with stable or improved KPS. Partial response (PR) refers to 50% or greater decrease in tumor on consecutive MRI images with stable or improved KPS. Progressive disease (PD) refers to 25% or greater increase in tumor on consecutive MRI images or clinical deterioration in the setting of stable MRI. Stable disease (SD) includes all other situations.
Demographic, baseline, and treatment characteristics were summarized using descriptive statistics. Adverse effects were summarized as both proportion of patients with each complication, total number of episodes and rate of episodes per number of procedures. Overall survival (OS) and progression-free survival (PFS) were measured from the first BBBD treatment date to date of first relapse (for PFS), death or last follow-up. The Kaplan-Meier method was used to estimate OS and PFS. Descriptive summaries include OS, time to best response, and response rate.
Patient and clinical characteristics are summarized in Table 1. Thirteen subjects (7 males and 6 females), aged 20 to 63 years (median 38 years), with AO (11) or OA (2) were enrolled. All patients were evaluable for toxicity and tumor response. All patients had a KPS greater than or equal to 50 (median 90.0, mean 85.4; range 50 to 100). KPS values were as follows: 100 (5 patients), 90 (3 patients), 70 (1 patient), 60 (2 patients) and 50 (1 patient). At initial diagnosis prior to entry in this study, all subjects had undergone surgery. Five patients (38.4%) had undergone a single surgery, 6 (46.2%) had undergone two surgeries, 1 (7.7%) had undergone 3 surgeries and 1 (7.7%) had undergone 4 surgeries. Five subjects had been treated with RT prior to study entry. In 9 subjects, surgery was followed by chemotherapy as initial treatment, in 3 subjects, RT followed surgery and one subject had surgery alone as initial treatment. All 13 subjects had failed TMZ either initially or at first or second recurrence, prior to study entry. Prior to study entry, 2 subjects (15.4%) failed 3 chemotherapy regimens, 4 subjects (30.8%) failed 2 chemotherapy regimens, and 7 subjects (54%) failed only TMZ. Of the patients who received only TMZ chemotherapy prior to study entry, 3 of these had received RT and 4 had not (4/13 patients had failed only with surgery and TMZ prior to study entry). To date, the 13 subjects have undergone a total of 147 BBBD treatments (median 10; range 1–24). Chromosomes 1p and 19q were both deleted in 8 subjects, 19q alone was deleted in 3 subjects, and 2 subjects had no 1p or 19q chromosome deletion. No subject was seen with 1p alone deletion.
There were 5 proposed IA melphalan DLs: 4, 8, 12, 16, and 20 mg/m2/day for 2 days. Four patients received DL 2 (8 mg/m2/day); and 2 of these patients developed grade 4 thrombocytopenia, thus the melphalan MTD was DL 1 (4 mg/m2/day). The first 3 patients entered at DL 1 (4 mg/m2). For each patient, there were no grade 4 toxicities attributable to the first course of treatment. Of the second group of 3 patients, entered at DL 2 (8 mg/m2), patient #4 did not receive a full course of treatment (only one day). The patient was not included in the dose escalation assessment. Of note, the patient did not experience any grade 4 toxicity. Patient #5 did not experience grade 4 toxicity attributable to the first course of treatment. Patient #6 experienced grade 4 thrombocytopenia during the first course of treatment; thus additional patients were entered at DL 2. Patient #7 developed grade 4 thrombocytopenia during first course of DL 2 treatment. Thus 2 patients entered at DL 2 experienced grade 4 toxicity. Per the protocol, the dose escalation was stopped, and the MTD was determined to be DL 1 (4mg/m2). Two patients entered at DL 2 underwent 25% dose reduction of melphalan after experiencing grade 4 thrombocytopenia. Toxicity data from the MTD dose escalation study is summarized in Table 2.
Toxicity data was recorded for all patients, for all grades, by type using the CTCAE v.3.0. Table 3 lists all Grade 3–5 toxicity events observed and the number of patients experiencing each event. From 147 procedures, there were a total of 65 grade 3 AEs, and 32 grade 4 AEs. The most common grade 3 AEs were thrombocytopenia (29 events, 20% of procedures), neutropenia (14 events,10% of procedures) and leukopenia (14 events, 10% of procedures). The most common grade 4 AEs were thrombocytopenia (17 events, 12% of procedures) and neutropenia (10 events, 7% of procedures). Five subjects underwent a total of 11 platelet transfusions, and 3 subjects underwent a total of 4 PRBC transfusions. One subject developed transient speech impairment that resolved within 72 hours, and one developed a grade 3 pseudoaneurysm of the left internal carotid artery secondary to intra-arterial catheter placement after experiencing two asymptomatic grade 1 subintimal tears.
Responses to study treatment for each patient are presented in Table 4. Figures 1–3 demonstrate representative subjects. As all 13 subjects demonstrated an enhancing lesion, T1-weighted MRI with Gd was used to measure response. Two subjects demonstrated CR (Figures 1 and and2),2), 3 demonstrated PR (Figure 3), 5 demonstrated SD and 3 developed PD. In the 10 subjects demonstrating a response or stable disease, the median PFS using Kaplan-Meier estimation was 19 months. Overall PFS in all 13 subjects was 11 months. Survival from the date of first BBBD protocol treatment ranged from 1 to 27+ months and 7 patients were still alive at the time of this report. Each of the 3 patients with PD progressed within one month from first BBBD treatment. The 5 subjects with CR or PR as best response demonstrated deletion of both 1p and 19q. Of the 5 subjects with SD as best response, 2 demonstrated both 1p and 19q deletion, and 3 demonstrated 19q deletion only. The 2 patients in whom both 1p and 19q were intact developed PD within 1 month and died at 1 and 8 months respectively following first BBBD treatment. The 2 patients achieving CR did not receive RT either prior to or after BBBD treatment. Of the 3 patients achieving PR, 2 received RT prior to, and none received RT following BBBD treatment. Both patients achieving CR had a KPS of 100, and those with PR had KPS of 50, 70, and 90.
The discovery that aggressive oligodendroglial tumors were sensitive to PCV chemotherapy (12, 86) increased interest in optimizing adjuvant chemotherapy regimens for these tumors. Most prospective uncontrolled single arm studies evaluating chemotherapy in recurrent oligodendroglial tumors involved PCV or TMZ chemotherapy. Although grouping of oligodendroglial tumors with other gliomas in clinical trials precludes conclusions to be drawn regarding oligodendroglial tumors specifically, two trials directly evaluated oligodendroglial tumors. A Radiation Therapy Oncology Group (RTOG) Phase III trial for AO randomized 289 patients to receive either neoadjuvant dose intense PCV chemotherapy followed by RT (PCV+RT) or postoperative RT alone. With 3-year follow-up on most patients, the median survival times were similar (4.9 years for PCV+RT versus 4.7 years for RT), however PFS time favored PCV+RT (2.6 years versus 1.7 years for RT alone, P = 0.004). Unfortunately, 65% of patients experienced grade 3 or 4 toxicity and one patient died. Patients with tumors lacking 1p and 19q (46%) had longer median survival times (>7 years) compared with tumors not lacking 1p and 19q (2.8 years, P<0.001).(11) In another similar multicenter randomized controlled trial aimed to determine if PCV chemotherapy improved survival, 368 patients with AO were randomized to either RT alone, or to RT followed by standard PCV chemotherapy. Like the RTOG trial, RT+PCV increased PFS compared with RT alone (23 versus 13.2 months, P = 0.0018) but there was no difference in OS. Again, combined loss of 1p and 19q identified a favorable subgroup.(85) These studies suggest that, although PFS is longer with chemotherapy in addition to RT, it appears to be associated with significant toxicity that limits dose escalation.
Two principles strongly influence the clinical effectiveness of antitumor agents. First, the agent must reach the tumor in therapeutic concentrations. Second, effective drugs must demonstrate consistent and clear dose-responsiveness. (76) One simple and common strategy to improve antitumor toxicity involves increasing the dose of chemotherapeutic agents until toxicity occurs. In this study, IA melpahlan dose levels tested were 4 and 8 mg/m2/day. The melphalan MTD in this Phase I study was determined to be 4 mg/m2/day. However, we might have achieved a higher MTD of melphalan had there been a median dose (6 mg/m2/day). This dose, under investigation in further IA/BBBD melphalan clinical studies from our group, may allow further dose escalation.
Unfortunately, simple dose escalations are often limited by systemic toxicities. Therefore, innovative approaches are required to achieve increased chemotherapy dose delivered to the tumor, while protecting non-tumor tissues. Methods to increase dose intensity include high-dose myeloablative chemotherapy followed by autologous peripheral blood progenitor cell rescue (APBPCR), osmotic BBBD combined with intra-arterial chemotherapy delivery, use of chemoprotective agents that protect against systemic toxicities while allowing increased tumor dose intensity, and tumor sensitizers which cause tumor cells to be more susceptible to chemotherapeutic agents (for review, see Kraemer et al., 2002).(43)
A primary hindrance to systemic dose escalation of most chemotherapeutic agents is hematopoietic toxicity. Myeloablative therapy with APBPCR is a strategy developed to achieve higher dose intensity and avoid cranial radiation and its associated toxicities. Mohile et al. (52) reported 20 patients with aggressive oligodendroglioma treated with 4 cycles of PCV chemotherapy every six weeks. Fifteen patients who demonstrated a response were treated with myeloablative therapy using busulfan and thiotepa followed by APBPCR. Fourteen patients underwent transplant, with median disease-free and OS of at least 36 weeks (not yet reached at time of report). This strategy allowed deferral of radiation for at least 3 years. A major limitation of this approach, however, is the acute toxicity associated with both the induction and consolidation regimens. A potential approach to improve effectiveness of this therapy could be to deliver the chemotherapeutic agent via the IA route following BBBD, followed by APBPCR. This may allow a higher dose to be delivered to the tumor, while avoiding many acute toxicities associated with myeloablative therapy.
IA administration of chemotherapy for the treatment of brain tumors in both animal studies and clinical trials provides a higher concentration of drug to the tumor and surrounding brain (17, 28, 29, 36, 75) with less associated systemic toxicity compared with IV administration (6, 75, 82). Several investigators have assessed the toxicity and/or efficacy of specific antineoplastic agents administered IA to treat brain tumors (3, 6, 28, 36, 37, 47, 53, 57, 77, 81, 88).
Osmotic BBBD has been shown to further intensify drug delivery to the tumor and surrounding brain (2, 5, 7, 24, 43–45, 49) particulary in smaller, less permeable tumors and when using higher-molecular weight agents. As the normal BBB prevents passage of ionized water-soluble compounds with molecular weights greater than 180 Dalton (D), and most chemotherapeutic agents have molecular weights between 200 and 1200 D, this barrier is a major impediment to chemotherapy access.(44) Transient osmotic disruption of the blood-brain, blood-CSF and blood-tumor barriers is achieved throughout a vascular circulation by IA infusion of a hyperosmotic agent, usually mannitol. Pharmacokinetics in animals showed that vascular permeability to methotrexate was maximal by 15 minutes after infusion of mannitol and returned to pre-infusion levels within 2 hours (60). A 10- to 100-fold increase in delivery was measured in intracerebral tumors and tumor-infiltrated brain comparing IV administration with BBBD (55, 59, 68, 69). In humans BBB permeability remained elevated at 2 hours after BBBD and returned to baseline levels by 4 hours (78). A correlation of dose intensity and survival was demonstrated in CNS lymphoma patients treated with IA/BBBD therapy (42) and none of the assessed patients suffered neurocognitive decline (50). Osmotic BBBD is used clinically to enhance chemotherapy delivery in brain tumor patients at nine centers across the United States, Canada and Israel. With over 6000 procedures performed, this procedure has shown low morbidity and mortality, with acceptable levels of toxicity (19, 24, 59).
Response data reported here are encouraging with 5 of 13 patients demonstrating a response, 5 remaining stable and only three developing progression. Each of the 3 patients with PD progressed within one month from first BBBD treatment, suggesting severe advanced disease. In the 10 subjects demonstrating a response or stable disease, the median PFS using Kaplan-Meier estimation was 19 months while overall PFS in all 13 subjects was 11 months. Survival from the date of first BBBD protocol treatment ranged from 1 to greater than 27 months and 7 patients were still alive at the time of this report, suggesting there is a subgroup of patients who respond well to this therapy. The 5 subjects with CR or PR demonstrated deletion of both 1p and 19q, consistent with other studies showing improved chemotherapy response in patients harboring tumors with these deletions. The 2 patients in whom both 1p and 19q were intact developed progressive disease within 1 month and died at 1 and 8 months following first BBBD treatment. Interestingly, the 2 patients achieving CR did not receive RT either prior to or after BBBD treatment.
The rationale leading to testing of IA melphalan with BBBD comes from several lines of evidence. Melphalan (L-Phenylalanine Mustard), originally studied for use in multiple myeloma, has demonstrated activity in brain tumors including medulloblastoma, pinealblastoma and gliomas (31, 48, 65, 74). Melphalan has a steep dose response curve, has possible synergism, is non-cross-resistant, and has differing non-hematologic toxicities when used in combination with cyclophosphamide (48). It was chosen to be administered along with carboplatin and etoposide phosphate following BBBD for the above reasons and because it is well tolerated when given IA following BBBD. The permeability of melphalan across the BBB has been studied (55). Like other water-soluble agents such as methotrexate, osmotic BBBD combined with IA delivery can significantly increase the brain levels of water soluble chemotherapeutic agents such as this (4, 60, 61).
It is important to emphasize that results reported here come from the completed Phase I portion of a Phase I/II design. Although important information concerning safety and toxicity of IA/BBBD therapy for AO and OA can be drawn, inherent limitations on response assessment due to small numbers of patients and incomplete Phase II data should be noted.
In an effort to improve survival, other novel approaches that increase dose intensity while avoiding toxicity will need to be investigated. Another such strategy currently under exploration is to protect organ systems damaged by chemotherapy during dose escalation. If the organ systems in which toxicity occurs could be protected, then dose may be escalated safely. Sodium thiosulfate (STS), utilized in this study, has been shown to significantly decrease development of chemotherapy-associated toxicities such as ototoxicity and myelosuppression.(25) Addition of a second thiol agent such as N-acetylcysteine (NAC), a cysteine analog with strong antioxidant activity which also induces the de novo synthesis of glutathione (GSH), may allow even further dose escalation, and a higher chemotherapeutic MTD. NAC has been well-studied for this purpose, does not cross the BBB, and thus may protect against ototoxicity without impacting the efficacy of tumor treatment. (20–22, 58)
Preclinical studies suggest chemoprotective effects can be optimized, without inhibiting tumor toxicity, by delayed administration of thiol chemoprotectant and/or by the use of two routes of administration (IA carboplatin; IV STS).(54) IA chemotherapy administration allows for higher concentration of chemotherapy within the arterial circulation supplying the tumor while a 4 and 8 hour delay in administration of STS by the IV route can neutralize the minimal residual plasma concentration of chemotherapeutic agent, thus reducing systemic toxicity. Using this simultaneous two-route administration of IV STS and platinum IA chemotherapy (without BBBD) in advanced head and neck cancer, this approach has allowed for a high concentration of cisplatin to be delivered via the arterial circulation supplying the tumor, while IV STS neutralized the plasma concentration of cisplatin, thus reducing systemic adverse effects.(67, 70, 71) These exciting results suggest that careful use of thiols may safely allow chemotherapy dose escalation to achieve specific tumor cytotoxicity, while avoiding nonspecific systemic toxicities.
A related approach aimed to improve chemotherapy efficacy even further involves increasing tumor cell susceptibility to cytotoxic effects. This can be accomplished by depletion of GSH, an endogenous cysteine-containing tripeptide that plays a major role in tumor cell resistance to cisplatin and its analogs.(51, 91) Depletion of GSH in tumor cells may improve chemotherapy tumor cytotoxicity. (16, 33, 35, 41, 56, 64) This can be achieved through inhibition of the rate-limiting enzyme in GSH biosynthesis using the highly specific agent buthionine sulfoximine (BSO).(32) Alternatively, acetaminophen at high doses generates a toxic reactive intermediate that can overwhelm GSH stores. (30) As alluded to above, it will be essential that directed therapies aimed to enhance chemotherapy efficacy via GSH depletion include measures to protect against nonspecific toxicities such as ototoxicity, peripheral neuropathy, and myelosuppression by administration of thiol agents.
In AO or OA patients who failed TMZ chemotherapy, BBBD with IA carboplatin, IV etoposide phosphate, and IA melphalan (4mg/m2/day for 2 days) shows acceptable toxicity and encouraging efficacy, especially in subjects demonstrating 1p and/or 19q deletion. Innovative approaches to increase dose intensity while avoiding toxicity will be needed to further improve treatment for these aggressive tumors.
Disclosure of funding received
This research was supported by a Veteran’s Administration Merit Review Grant and NIH grants CA137488, NS34608, and NS44687 from the National Cancer Institute and the National Institute of Neurological Disorders and Stroke to EAN
This work was presented as a select poster at the 13th annual meeting of the Society for Neuro-Oncology, Las Vegas, NV, November 21, 2008
We thank Cynthia A. Lacy, B.S.N., R.N.; Rose Marie Tyson, M.S.N., A.N.P.; and Marianne Haluska, M.S.N., A.C.N.P. for providing clinical information on patients enrolled in this study and for their helpful advice. We also thank Emily Hochhalter for help with manuscript preparation.
Conception and design: Edward A. Neuwelt, Nancy D. Doolittle
Provision of study patients: Johnny B. Delashaw, Edward A. Neuwelt
Collection and assembly of data: Nancy D. Doolittle, Seymur Gahramanov, Nancy A. Hedrick, Edward A. Neuwelt
Data analysis and interpretation: Nancy D. Doolittle, Daniel J. Guillaume, Seymur Gahramanov, Edward A. Neuwelt
Manuscript writing: Daniel J. Guillaume, Nancy D. Doolittle, Edward A. Neuwelt
Final approval of manuscript: Daniel J. Guillaume, Nancy D. Doolittle, Seymur Gahramanov, Nancy A. Hedrick, Johnny B. Delashaw, Edward A. Neuwelt