|Home | About | Journals | Submit | Contact Us | Français|
This multi-center phase II trial was conducted by the Gynecologic Oncology Group to evaluate the activity and safety of irofulven in patients with recurrent epithelial ovarian cancer.
Eligible patients had documented recurrent ovarian cancer 6-12 months after receiving a front-line platinum-based regimen and no other chemotherapy. Patients were required to have measurable disease, performance status of 0-2, and adequate bone marrow, hepatic, and renal function prior to study entry. The dose of irofulven was 0.45 mg/kg IV on days 1 and 8 every 21 days. Responses were defined by RECIST.
Fifty-five of 61 enrolled patients were evaluable for response and toxicity. There were seven partial responses (12.7%) and 30 patients (54.6%) had stable disease. Median progression-free and overall survival were 6.4 mo (1.3-37.5) and 22.1+ mo (2.8-57.8+), respectively. Patients received a median of three cycles (range 1-21) of protocol therapy. Grade 4 hematologic toxicity was limited to reversible neutropenia and thrombocytopenia. Grade 4 non-hematologic toxicity was limited to one patient with anorexia and another with hypomagnesemia.
Irofulven administered at this dose and schedule was well tolerated but had modest activity as a single agent.
Epithelial ovarian cancer remains the leading cause of gynecologic cancer death among American women. Approximately 21,550 women will be diagnosed in 2009 and an estimated 14,600 will die of this disease . More than 70% of the patients are diagnosed with advanced stage disease . Initial management includes cytoreductive surgery followed by platinum and taxane-based chemotherapy. Although approximately 80% of women will have no detectable disease at the end of front-line treatment, most of them will recur and require further therapy . The clinical efficacy of re-treatment with platinum-based chemotherapy is limited by the development of drug resistance. Therefore, new chemotherapeutic agents with novel mechanisms of action are needed in the treatment of this disease.
Irofulven (MGI-114, 6-hydroxymethylacylfulvene, HMAF) is a unique cytotoxic agent that is related to the Jack O’Lantern (Omphalotus illudens) mushroom-derived iludin toxins. It is a semisynthetic derivative of the sesquiterpene, illudin S, which binds to DNA, produces DNA single strand breaks leading to cell cycle arrest in S phase and cell death through caspase mediated apoptosis . Irofulven has a more favorable therapeutic index than its parent illudin S. Resultant apoptosis is independent of p53 status, multidrug resistance (MDR), bcl-2 expression, and mismatch repair enzymes . It also binds to RNA and proteins. Radio labeled irofulven localizes primarily in the nuclear compartment followed by cytosolic and membranous compartments . In vitro, irofulven was active against numerous cell lines known to be resistant to alkylating agents, cisplatin, doxorubicin, topoisomerase inhibition, and taxanes [7-9]. In preclinical models, the drug works synergistically with topotecan, platinum compounds, taxanes, and radiotherapy.
Early clinical trials with irofulven revealed that the agent has significant toxicity on certain schedules. The daily × 5 day schedule demonstrated antitumor activity in a number of tumors including ovarian and endometrial cancer but was associated with severe gastrointestinal, renal, metabolic, hematologic, and ocular toxicities [10-12]. Exploration of alternative doses and schedules demonstrated improved tolerability while maintaining the antitumor activity of the study drug. In a phase I trial exploring weekly and biweekly schedules of administration of irofulven, a patient with heavily pretreated recurrent ovarian cancer who received irofulven on the day 1 and 8 schedule every 21 days, had a complete response which lasted 11 months . The maximum tolerated dose was determined to be 18 mg/m2 for a day 1 and 8 every 21 day dosing schedule not to exceed 0.55 mg/kg/infusion and 50 mg total dose per infusion. Based on these data, the dose of 0.45 mg/kg/infusion was chosen for further evaluation in this phase II trial.
Patients with persistent or recurrent platinum-sensitive epithelial ovarian or primary peritoneal carcinoma were eligible for this trial. Patients must have received a platinum/taxane-based chemotherapeutic regimen for the management of primary disease, which may have included consolation or extended therapy after surgical or non-surgical assessment. Patients may not have received chemotherapy for recurrent disease. Patients were considered platinum-sensitive and eligible if they had progressed between 6 and 12 months of their last platinum dose. Patients who progressed greater than 12 months from their last platinum dose were not eligible as retreatment with platinum-based therapy is routinely used in this setting. Measurable disease defined by Response Evaluation Criteria in Solid Tumors (RECIST) was required . Other eligibility criteria included: (1) Gynecologic Oncology Group (GOG) performance status of 0-2; (2) adequate bone marrow (Absolute Neutrophil Count (ANC) ≥ 1500/mm3, platelets > 100,000/mm3), hepatic (transaminases and alkylating phosphatase levels ≤ 2.5 × the upper limit of institutional normal (ULIN) total bilirubin ≤ 1.5 × ULIN) and renal function (serum creatinine ≤ ULIN); (3) no ≥ grade 2 peripheral neuropathy; (4) no radiation to more than 25% of marrow bearing areas; (5) no myocardial infarction, cerebrovascular events, transient ischemic attacks or congestive heart failure in the past six months prior to enrolling on the trial; (6) no electrocardiogram (ECG) evidence of acute ischemia or significant conduction abnormality (bifascicular block, defined as left anterior hemiblock in the presence of right bundle branch block; second or third degree atrioventricular (AV) blocks); (7) no other invasive malignancies with the exception of non-melanomatous skin cancer in the past five years; (8) negative serum pregnancy test if still of childbearing potential; (9) no history of retinopathy and/or macular degeneration; and (10) no prior therapy with irofulven. All women provided written informed consent and participating institutions obtained annual institutional review board approval in accordance with Federal, State and local institutional requirements and guidelines.
Patients received Irofulven at a dose of 0.45 mg/kg intravenously over 30 minutes on days 1 and 8 every 21 days. Doses were not to exceed 50 mg for each infusion. Cycles were administered every 21 days until disease progression or unacceptable toxicity. Progression was determined radiologically. Toxicities were graded according to the National Cancer Institute Common Toxicity Criteria (CTC) Version 2.0.
Pretreatment evaluation consisted of history and physical exam, assessment of GOG performance status, chest x-ray, ECG, complete blood count (CBC), serum chemistries (lactate dehydrogenase (LDH), blood urea nitrogen (BUN), creatinine, magnesium, calcium phosphate, transaminases, alkaline phosphatase, and total bilirubin), urinalysis, CA-125, and documentation of measurable disease by computerized tomography (CT) scan. During the study, interval history, physical examination, toxicity assessment, CBC, serum chemistries, and CA-125 were obtained at the start of each cycle. CT scan to evaluate response was performed every two cycles. Evaluation of response was by RECIST . Subsequent cycles were not administered until the absolute neutrophil count was ≥ 1500/mcl and the platelet count was ≥ 100,000/mcl. Therapy was allowed to be delayed up to a maximum of two weeks. Patients who did not recover blood counts to the aforementioned parameters or did not recover non-hematologic toxicity to grade <1 by the end of this two week period would be removed from study. Doses would be reduced to 0.35 mg/kg 1 level reduction) or 0.25 mg/kg (second dose level reduction) for febrile neutropenia, grade 4 neutropenia lasting > seven days, grade 4 thrombocytopenia, grade ≥ 2 renal toxicity, grade ≥ 3 hepatic toxicity, grade ≥ 3 gastrointestinal (GI) toxicity unresponsive to medical treatment or ≥ 2 nonhematologic toxicity with an impact on vital organ function. Patients who experienced any degree of decrease in visual acuity would have irofulven dosing discontinued until the condition had fully reversed. If treatment was restarted, the dose of irofulven was reduced by 25%. Continuation of therapy after experiencing visual symptoms, grade 1 or 2, without decrease in visual acuity, was left up to the patient and treating investigator; a dose decrease of 25% was permitted. There were no dose re-escalations on this trial.
A GOG Phase II trial of cisplatin in this disease yielded a response rate of 24%. As a result, this agent has been used extensively in combination chemotherapy. A more recent study of taxol demonstrated a response rate of approximately 36%. In effect, these are the standards against which other Phase II trials must be measured. Twenty-two additional agents have been examined in this setting. Response rates ranged from 0-18% with only six agents having a response rate in excess of 10%. Consequently, if a new agent has a response rate of 15% or less, it will be of no clinical significance. Conversely, if the true response rate is at least 30%, further study is clearly indicated [15,16].
The study employed a two-stage accrual design  with an early stopping rule in the event that treatment demonstrated insufficient activity. During the first stage of accrual, 22-29 patients would be entered and evaluated. If at least four responses were observed among the first 22-24 patients, or at least five responses out of 25-29 patients, a second phase of accrual was to be initiated that would increase accrual to 53-60 patients. The regimen would be considered active if at least 11 responses were observed among 53 patients, at least 12 responses were observed among 54-57 patients, or at least 13 responses among 58-60 patients. If the true response rate is 15%, the average probability of designating the treatment as active is limited to 10%. Conversely, if the true response rate was 30%, then the probability of correctly classifying the treatment as active was 90%.
Sixty-one women were enrolled onto the trial. Two patients were ineligible due to incorrect cell type (1) and a platinum-sensitive interval > 12 months (1). Four patients were inevaluable as they never were treated, leaving 55 evaluable patients. Patient characteristics are provided in Table 1. All patients had prior platinum-based chemotherapy, and one patient had prior radiotherapy (to liver).
Patients received a median of 3 cycles (1-21) of protocol therapy. There were no treatment related deaths. Grade 4 hematologic toxicity was limited to reversible neutropenia and thrombocytopenia. One gastrointestinal (anorexia) and one metabolic (hypomagnesemia) grade 4 toxicities were observed (Table 2). Other significant (grade 3) nonhematologic toxicities included metabolic and neurological side effects. There were five grade 3 ocular toxicities. Two patients had photophobia, one had flashing lights and floaters, and the last two patients had blurred vision.
The overall response rate was 12.7% (7/55). There were no complete responsesThirty patients had a best response of stable disease (54.6%) with a median duration of 8.2 months (2.6-37.5) (Table 3). The median progression-free survival is 6.4 months (1.3-37.5) with a median overall survival of 22.1+ months (2.8-57.8) (Figure 1). The response rate was 16.7% (4/24) in the first stage of accrual.
Initial phase II trials of irofulven in patients with recurrent gynecologic malignancies were based on consecutive daily schedules for four or five days repeated every 28 days [10, 11]. While the drug was shown to have activity, these trials were fraught with major toxicities. These included significant neutropenia, thrombocytopenia, nausea with emesis, and severe electrolyte abnormalities with renal tubular acidosis that occasionally culminated in treatment-related deaths.
Subsequently, more intermittent dosing schedules were explored that demonstrated more favorable toxicity profiles . Two dosing schedules were recommended for further phase II testing: days 1 and 8 every 21 days at 18 mg/m2/infusion or days 1 and 15 every 28 days at 24 mg/m2/infusion. There were no treatment related deaths with either of these regimens. Significant GI toxicity was rare with the days 1 and 8 every 21 day schedule. Hematologic toxicity was slightly more pronounced with this schedule compared with the day 1 and 15 every 28 day schedule. The incidence of visual toxicity did not appear to depend on duration of infusion. There was an association between dose per infusion and occurrence of visual toxicity. The mean dose per infusion in patients with visual toxicity was 23.6 ± 3.9 mg/m2 compared with 20.2 ± 4.1 mg/m2 in patients without visual toxicity. The risk of visual toxicities appeared to be even more accurately predicted by dosing based on body weight than body surface area. Patients with visual toxicities received a mean dose of 0.65 ± 0.09 mg/kg compared with 0.53 ± 0.12 mg/kg in patients without visual toxicities. Sophisticated ophthalmologic examination and pathology evaluations demonstrated that visual toxicity is based on selective cone loss in the retina [18,19]. The incidence of visual toxicity was much lower in patients who received less than 0.55 mg/kg and a total maximum dose of 50 mg per infusion, hence, the basis for dosing at 0.45 mg/kg for this trial with a limit of 50 mg per infusion.
The response rate on this intermittent dosing schedule is 12.7%. This rate of response was in patients who were platinum-sensitive by the definition of recurrence within 6-12 months of their last platinum dose. The response rate was even lower in a similar study conducted by Seiden and colleagues in which patients had platinum-sensitive or resistant disease . The response rate was slightly higher (22%) in the study by Sarosy and associates who evaluated a daily × 5 schedule every 28 days  but with greater toxicity. The higher response rate in the latter trial, of course, may not be a function of schedule but an artifact of comparing trials with small numbers of patients with varying eligibility criteria.
While irofulven clearly induces some tumor shrinkage, its antitumor activity may ultimately be limited by the inability to increase the dose given its toxicities. As a DNA damaging agent, irofulven may combine well with other drugs with complementary mechanisms of action and toxicity profiles that will allow for reduction in dose and therefore toxicities. The combination of irofulven and irinotecan was found to be feasible in a phase I trial, although there were few responses . Irofulven produces DNA damage largely ignored by the nucleotide excision repair (NER) system The single strand DNA breaks induced by irofulven suggest that the drug maybe highly active in BRCA-deficient tumor cells analogous to poly (ADP-ribose) polymerase (PARP) inhibitors. This synergistic mechanism of action of DNA damage between irofulven and other DNA damaging agents may stem from the NER system being overwhelmed at two distinct points in the pathway leading to the prolonged presence of stalled polymerases eventually triggering apoptosis as a consequence of collisions with the replication fork resulting in additional lethal secondary DNA damage. DNA damage induced by irofulven is enhanced when combined with platinum or alkylating agents in preclinical models [8,9]. Irofulven at low nontoxic doses in combination with angiogenesis inhibitors, anginex or topomimetic 0118, was more effective at inhibiting tumor growth in mice bearing human xenografts than full doses of irofulven . Anginex is a 33mer designer peptide that targets galectin-1 which is unregulated in tumor activated endothelial cells. Galectin-1 is necessary for tumor cells to adhere to, and migrate on the extracellular matrix. Topomimetic, 0118 is a calyx arene-scaffold surface topomimetic that embodies the molecular dimensions, surface topology, and chemical composition of anginex but has greater biological activity.
Based on all of the data, including our results presented here, irofulven will not have a role in the treatment of ovarian cancer as a single agent. Its future may lie in combinations with newer agents such as angiogenesis antagonists or inhibitors of DNA repair, such as PARP inhibitors, smac mimetics or bcl-2 antagonists (such as BH3 mimetics).
This study was supported by National Cancer Institute grants CA 27469 (Gynecologic Oncology Group) and CA 37517 (Gynecologic Oncology Group Statistical and Data Center).
The following Gynecologic Oncology Group member institutions participated in this study: Abington Memorial Hospital; Walter Reed Army Medical Center; Northwestern University/Feinberg School of Medicine; University of Mississippi; Colorado Gynecologic Oncology Group; University of Pennsylvania Cancer Center; University of Texas Southwestern Medical Center at Dallas; University of California Medical Center at Irvine; Rush-Presbyterian-St. Luke’s Medical Center; SUNY Downstate Medical Center; University of New Mexico Health Sciences Center; Cooper Hospital/University Medical Center; Columbus Cancer Council/Ohio State; University of Massachusetts Memorial Health Care; University of Oklahoma; University of Virginia Health Sciences Center; University of Chicago; Case Western Reserve University; Tampa Bay Cancer Consortium; Gynecologic Oncology Network/Brody School of Medicine; University of Texas – Galveston; Community Clinical Oncology Program.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.