Phase II study of daily oral perifosine in patients with advanced soft tissue sarcoma


  • Informed consent for participation in this study was obtained from the subject(s) and/or guardian(s).



A multicenter Phase II study was performed to evaluate the clinical activity of an initial loading (150 mg every 6 hours × 4 doses) dose followed by continuous daily oral dosing (100 mg/day) of perifosine in patients with advanced soft tissue sarcomas (STSs).


Patients with measurable metastatic STS received perifosine as first-, second-, or third-line treatment and underwent disease assessment every 8 weeks until disease progression, excessive toxicity, or patient refusal.


Twenty-three patients received 66 cycles (1 cycle = 4 weeks) of perifosine. One partial response of 9 months duration was observed. The overall 3 and 6 month progression-free survival was 22% and 9%. NCI CTC (v2.0) Grade 1 to 2 gastrointestinal toxicity or fatigue were the most common (>50% of subjects) toxicities observed. The steady-state plasma perifosine levels (Css) were similar to prior experience (mean 6 μg/mL). Patients with Css levels >6 μg/mL appeared more likely to remain on study past 2 months than those with levels <6 μg/mL.


Despite not achieving the primary objective of ≥40% 6-month progression-free survival rate, optimism remains for this agent in STS patients. Prolonged responses in heavily pretreated STS patients continue to be observed with perifosine treatment. Continued assessment of perifosine in STS appears warranted, with special attention to specific histologies or tumor characteristics that might identify a more sensitive population and achieving perifosine Css levels >6 μg/mL. Cancer 2006 © 2006 American Cancer Society.

Soft tissue sarcomas (STSs) are a heterogeneous group of tumors with an annual US incidence of 9000 cases.1 The primary treatment of STS is surgery with or without radiation when feasible. Once STS has metastasized or recurred, the only chance of long-term survival is with metastatectomy, but only a few patients are candidates for this surgery.

Chemotherapy for patients with advanced STS is largely palliative. Doxorubicin and ifosfamide are the most active agents against STS, with response rates near 25%; a third agent is dacarbazine, with a response rate of 15%.2 Many different combination regimens of standard cytotoxic agents have been explored with observations of improved response rates, but little to no improvement in survival.2

As the understanding of the various processes critical to cellular oncogenesis has increased, anticancer agents specifically targeted against these processes have been developed. An example relative to STS is imatinib, a potent inhibitor of various protein tyrosine kinases including c-kit. Imatinib consistently causes prolonged regression or tumor stabilization in patients with gastrointestinal stromal tumors (GIST) that overexpress c-kit.3

Alkylphospholipids (ALKs) are a class of antineoplastics that display potent antitumor effects against a wide array of human tumor models and the lead compound, miltefosine, is approved in Europe for topical treatment of cutaneous metastases.4 Perifosine is a more bioavailable analog of miltefosine with equal or greater antitumor potency as compared with other alkylphospholipids. The mechanism of their antitumor effect is unclear. Possible mechanisms have focused on ALKs direct absorption into the cells with accumulation in cell membranes.5, 6 Observed cellular effects of ALKs include cell cycle arrest at G1-S and G2/M by up-regulation of p21waf1/cip1 protein levels7 and dose-dependent inhibition of Akt/Protein Kinase B phosphorylation.8, 9

Phase I trials of perifosine have focused on daily administration, either continuously or with brief interruptions.10, 11 The predominant toxicities observed have been gastrointestinal toxicity and fatigue, and pharmacokinetic studies supported a loading dose usually administered in divided doses due to dose-dependent gastrointestinal toxicity.11 During the performance of the Phase I trial at the University of Wisconsin (UW), a prolonged partial response (>12 months) was observed in a heavily pretreated patient with advanced leiomyosarcoma treated at the daily dose level of 50 mg. This led to a Phase II study of perifosine in patients with advanced STS employing the recommended Phase II dose and schedule from the UW trial.


Eligibility Criteria/Human Subjects

Patients eligible for the study had histologically confirmed advanced STS and were at least 18 years of age. Measurable lesions were required and if the measured lesions had been previously irradiated, they must have progressed >25% since completion of radiation therapy. All patients must have had laboratory values obtained ≤7 days before registration satisfying the following: absolute neutrophil count (ANC) ≥1,500/μL; platelet count (PLT) ≥100,000/μL; total bilirubin ≤upper normal limit (UNL); aspartate aminotransferase (AST) ≤2.5 × UNL; and creatinine ≤UNL or calculated creatinine clearance ≥60 mL/min. Patients also were expected to live at least 12 weeks, have an Eastern Cooperative Oncology Group (ECOG) performance score of 0 or 1, and were capable of understanding the investigational nature, potential risks and benefits of the study, and able to provide valid informed consent per federal guidelines and institutional approval. Prospective patients who were pregnant or breastfeeding women, or anyone of childbearing potential who was unwilling to employ adequate contraception were all ineligible. Other exclusion criteria included: ≥3 prior cytotoxic chemotherapy regimens for metastatic sarcoma; nitrosoureas or mitomycin C ≤6 weeks before study entry; chemotherapy, radiotherapy, immunotherapy, or biologic therapy ≤4 weeks before study entry; failure to recover from acute, reversible effects of prior therapy regardless of interval since last treatment; other concurrent therapies considered investigational; uncontrolled brain metastases; history of allergic reactions attributed to compounds of similar chemical or biologic composition to perifosine; uncontrolled intercurrent illness; HIV-positive patients receiving combination antiretroviral therapy; and prior malignancy (except if disease free for ≥5 years, or adequately treated basal cell, squamous cell skin cancer, or noninvasive carcinoma).

Before study performance at each site, the study was reviewed and approved per local and federal requirements by the institutional review board for human subjects safety.

Perifosine Administration

Perifosine (NSC #639966) was supplied as film-coated tablets, containing 50 mg active ingredient, by the Division of Cancer Treatment and Diagnosis, National Cancer Institute.

Cycle 1 of study treatment consisted of 100 mg of perifosine, taken orally, once a day except a loading dose of 150 mg every 6 hours (×4) on the first day. Thereafter, Day 1 was administered at the same dose as the remainder of the cycle (ie, 100 mg of perifosine). Treatment was given on a continuous daily basis and a treatment cycle was 28 days. The perifosine dose was reduced according to prestudy-defined adverse event criteria to 50 mg daily or, if a second reduction was needed, every other day. Reductions lower than 50 mg every other day were not permitted. Unresolved adverse events would result in a hold of study drug for up to 2 weeks, or until the grade resolved to 1 or lower, at which time perifosine would be administered at a reduced dose. Perifosine dose reescalation was not allowed after dose reduction.

Disease Assessment

Objective response to therapy was assessed using the Response Evaluation Criteria in Solid Tumors (RECIST).12 Measurable disease was defined as at least 1 lesion whose longest diameter could be accurately measured (by physical or radiologic exam) as ≥2.0 cm. Patients with global deterioration of health status requiring discontinuation of treatment without objective evidence of disease progression at that time, and not related to study treatment or other medical conditions, were considered to have progressive disease (PD) due to symptomatic deterioration.

Determination of duration of stable disease or response and time to progression or survival was calculated from the time of subject study registration until progression or death. The 6-month progression-free rate was estimated by dividing the number of patients who have not progressed within 6 months by the total number of evaluable patients. Patients were followed until death or a maximum of 5 years postregistration.

Perifosine Plasma Concentrations

Steady-state plasma concentrations of perifosine were determined by drawing predose blood samples (10 mL drawn into a plastic heparinized green top Vacutainer) in the morning on Day 15 of Cycle 1 and Day 1 of Cycles 2–6. Prior pharmacokinetic studies in humans observed steady-state concentrations by Day 8.11 Blood samples were immediately centrifuged (1200g for 15 minutes) and plasma aliquots of 2 mL each were frozen (−70°C or colder) in polypropylene cryotubes until shipment. Samples were shipped on dry ice to the 3P Analytical Laboratory, University of Wisconsin Comprehensive Cancer Center (Madison, WI) for analysis. Perifosine in plasma was measured by a validated reversed-phase liquid chromatography /electrospray mass spectrometry (LC/MS) method developed by Woo et al.13 Specific methods employed are detailed by Van Ummerson et al.11

Statistical Considerations

Due to the potential that agents like perifosine could manifest clinical benefit in the form of prolonged stable disease or tumor shrinkage, the primary endpoint of this study was to evaluate the 6-month progression-free rate (6MPFR). All patients meeting the eligibility criteria who signed a consent form and began on treatment were considered evaluable. Patients discontinuing treatment before 6 months after study entry were replaced in accrual and classified as not evaluable for the 6MPFR if: 1) surgical resection of disease was performed, 2) the patient died of causes unrelated to the study treatment or disease, or 3) the patient was lost to follow-up. Patients alive 6 months after registration, with either stable disease, a partial or complete response to treatment were considered a “success” unless specified otherwise. The criteria for the final efficacy decision, with respect to increasing the 6MPFR, was based on a novel statistical design, named the “three-outcome Phase II design.”14 The largest success proportion where the proposed treatment regimen would be considered ineffective in this population was 20%, and the smallest success proportion that would warrant subsequent studies with the proposed regimen in this patient population was 40%. The goal of a 6MPFR >20% was based on a study by Okuno et al.15 that observed a 11% 6MPFR rate with a minimally effective single agent (gemcitabine) in patients with previously treated advanced STS. If at most 3 successes were observed in the initial 15 evaluable patients, this would result in termination of accrual for lack of efficacy with respect to the 6MPFR in this patient population. Otherwise, we were to accrue an additional 27 patients. Of the 42 evaluable patients, this treatment regimen was to be considered “ineffective” if 10 or fewer successes were observed, “inconclusive” if 11 or 12 successes were observed, and “promising” if at least 13 successes were observed. The confidence interval (CI) reported for the estimate of the 6MPFR is calculated via the method of Duffy and Santner.16

Secondary endpoints included survival time, time to disease progression, duration of response, time to treatment failure, and summaries of the pharmacokinetics correlative study. All adverse events were evaluated using the NCI Common Toxicity Criteria, version 2.0 (CTC v2.0); thereafter, summarized in a tabular manner as a maximum grade per patient for a given type of event. Kaplan-Meier methodology17 was used to describe the distribution of time to disease progression and survival. All analyses were performed using SAS v. 8.0 (Cary, NC).


Patient Characteristics

A total of 23 patients from 5 P2C treating institutions (Mayo Clinic Rochester, Mayo Clinic Jacksonville, Johns Hopkins, Washington University, and University of Wisconsin) were enrolled between June and October 2003. Patient characteristics for these patients are summarized in Table 1. At study entry, patients ranged in age from 24 to 77 (median, 53); 65% were female, and 13% were African American. Six (26%) patients presented with liver metastases when they enrolled and 20 (87%) had received prior chemotherapy for metastatic disease (3 patients, 0; 9 patients, 1; 11 patients, 2).

Table 1. Patient Characteristics (N = 23)
CharacteristicFrequency (%)
  1. NOS indicates not otherwise significant.

Age, median, range53, 24–77
Sex, women15 (65)
 Leiomyosarcoma7 (30)
 NOS liposarcoma4 (17)
 NOS stromal sarcoma3 (13)
 NOS sarcoma2 (9)
 NOS fibrous histiocytoma2 (9)
 Synovial sarcoma2 (9)
 Chondrosarcoma2 (9)
 Desmoid1 (4)
Liver metastases6 (26)
Number of prior chemotherapy regimens for metastatic disease 
 03 (13)
 19 (39)
 211 (48)

Perifosine Administration

Overall, patients received a total of 66 cycles of treatment (median of 2 cycles per patient; range, 1–10). Seventy-four percent of patients received only 1 or 2 cycles of treatment. The mean percentage of targeted dose administered per patient and cycle was 96% (range, 40%–112%). Only 2 patients had dose reductions. One patient misunderstood the drug instructions, taking only 50 mg of perifosine per day during Cycle 1. This patient escalated to 100 mg per day for Cycle 2 and subsequently progressed at evaluation for the conclusion of that cycle. The second patient had a dose reduction to 50 mg of perifosine during Cycle 2 of treatment, due to increased tumor pain, and subsequently progressed thereafter. Reasons for discontinuation of treatment were disease progression (96%) and refusal of treatment due to Grade 3 diarrhea (4%).


All patients were evaluable for toxicity. The most common toxicities were Grade 1–2 gastrointestinal toxicity (anorexia 10/23, diarrhea 13/23, nausea 20/23, vomiting 10/23) and fatigue (14/23). Gastrointestinal toxicities were usually more pronounced during the loading doses on Day 1. Twelve (52%) patients experienced Grade 3 or higher nonhematologic adverse events. The vast majority of Grade 3–4 toxicities were of questionable relation to study drug. There was 1 patient who experienced a Grade 4 adverse event considered possibly attributable to treatment. This patient experienced a Grade 4 ileus during Cycle 1, having daily nausea, discontinued perifosine, and subsequently progressed upon disease evaluation. One patient experienced Grade 4 cerebral ischemia (considered unrelated to study treatment) during Cycle 8, again showing progression upon the subsequent disease assessment. Three subjects each had worsened abdominal pain and dyspnea (Grade 3), considered clinical evidence of PD. One patient experienced Grade 3 anorexia and another experienced Grade 3 dehydration. See Table 2 for a more inclusive listing of observed toxicities considered possibly, probably, or definitely related to study treatment. No fatalities were observed during study treatment.

Table 2. Maximum Severity of Adverse Events Attributed to Perifosine (N = 23)*
Body systemToxicityGrade 1 No.Grade 2 No.Grade 3 No.Grade 4 No.
  • *

    These have an attribution of possible, probable, or definite.

Infection/febrile neutropeniaInfection0010
Constitutional symptomsFatigue6810


All patients were considered evaluable for the primary endpoint of the estimated 6 month progression-free survival rate. We note that 1 patient refusing treatment during Cycle 1 was never evaluated for response during treatment; however, further follow-up indicated that this patient progressed 60 days postregistration. Two (9%) patients were progression-free after 6 months (95% CI: 1%–28%). At the time of this analysis, all patients have experienced disease progression, with 57% having multiple progression sites. Ten (43%) patients had pulmonary involvement and 6 (26%) patients had hepatic involvement upon progression.

Twenty-two (96%) patients were evaluable for objective tumor response, having at least 1 postbaseline tumor assessment. One patient with a prior resected extraskeletal chondrosarcoma in the soft tissue of the pelvis, who had not responded to prior chemotherapy for biopsy-confirmed metastatic disease and had progressive pulmonary metastases (≤2 cm in diameter) at study entry, had a partial response that lasted 9 months before observed progression of pulmonary disease. Additionally, 5 patients had stable disease 8 weeks postregistration. Overall, 15 patients (65%) have died, with a median follow-up of 15.6 months (range, 2–19) for surviving patients. Estimated median survival is 9.4 months (Fig. 1). See Table 3 for further details with regard to patient outcome.

Figure 1.

Kaplan-Meier patient outcome curves.

Table 3. Patient Outcomes (N = 23)
EndpointEstimate (95% CI)
  • CI indicates confidence interval; m, months; NE, not evaluable.

  • *

    Kaplan-Meier method.

  • Not estimable.

6-Month progression-free survival9% (1%–28%)
Response rate n = 225% (0%–23%)
Stable disease n = 2223% (8%–45%)
Duration of stable disease, median, range n = 53.4 m, 2.8–7.6
Time to disease progression,* median1.8 m (1.7–2.0)
 3-month rate22% (10%–47%)
 6-month rate9% (2%–33%)
 9 month rate4% (1%–30%)
Overall survival,* median9.4 m (4.8-NE)
 6-month rate64% (47%–87%)
 9-month rate55% (37%–80%)
 12-month rate36% (21%–63%)

Steady-State Perifosine Plasma Concentrations

Steady-state plasma concentrations were available for 19 of 23 patients and are as follows (mean ± SD, μg/mL): Course 1, Day 15, 5.8 ± 2.0, n = 19; Course 2, Day 1, 6.0 ± 2.6, n = 18; Course 3, Day 1, 6.3 ± 2.5, n = 9; Course 4, Day 1, 8.6 ± 1.6, n = 3; Course 5, Day 1, 8.1, n = 2; Course 6, Day 1, 8.5, n = 3. A mean steady-state concentration for each patient was determined and these values were compiled as well (mean ± SD, 6.0 ± 2.3 μg/mL, n = 19). Further examination of individual steady-state levels revealed that 6 of 9 patients with Course 2, Day 1 concentrations ≥6 μg/mL continued on study past Course 2, whereas 3 of 9 patients with Course 2, Day 1 steady-state concentrations <6 μg/mL continued on study for >2 courses. Nine patients had steady-state concentrations through 3 courses of treatment, and 4 of 9 had slightly lower steady-state levels on Course 3, Day 1 as compared with Course 1, Day 15; the other 5 patients had slightly greater concentrations on Course 3, Day 1 as compared with Course 1, Day 15. The patient with a partial response had a mean steady-state concentration through 6 courses of 8.7 μg/mL (n = 5; range, 7.5–9.8).


Novel agents that have anticancer mechanisms targeted against signal transduction pathways have been observed to be effective anticancer agents, especially against tumors that have genotypic or phenotypic changes in the targeted transduction pathway. That approach has led to the use of imatinib in GIST and a similar approach should be explored in other STSs.

Alkylphospholipids are a class of anticancer agents that perturb signal transduction pathways (inhibition of Akt phosphorylation) and have shown consistent clinical anticancer activity, but their systemic application has been limited by toxicity. The alkylphospholipid perifosine appears to have a better toxicity profile and greater promise as a systemic agent.

Daily oral perifosine was well tolerated in a cohort of predominantly previously treated advanced STS patients. Whereas there were observed Grade 3 and 4 toxicities, the vast majority were likely related to accompanying disease progression. Similar to the Phase I experience, most of the gastrointestinal toxicity occurred during the Day 1 loading doses. Once patients were on their once-a-day dosing, a minority of patients experienced Grade 1–2 fatigue and gastrointestinal toxicity (nausea, diarrhea, abdominal pain).

The primary endpoint of observing a ≥40%, 6-month progression-free survival rate in enrolled subjects was not met, because the study did not meet the initial criteria of a >20% 6-month progression-free survival rate after the initial 15 evaluable subjects. Two out of the first 15 evaluable subjects were progression-free at 6 months. However, similar to the Phase I experience, another prolonged partial response (>6 months) was observed in a patient with advanced STS (chondrosarcoma) who had been previously treated with multiagent chemotherapy. The choice of a primary endpoint in initial Phase II studies of a novel agent is open to debate from among the choices of objective tumor shrinkage, overall survival, progression-free survival, or quality of life. Each endpoint has advantages and disadvantages, but given the increasing evidence that signal transduction agents or even classic cytotoxic agents can manifest significant clinical benefit in the form of objective stable disease3, 18 in patients with advanced sarcomas, progression-free survival would appear to be a reasonable endpoint for an initial Phase II study. Our primary endpoint could be criticized based on the goal of a >20% to 40% 6MPFR being too ambitious for an initial Phase II study. A review, published after initiation of our study, of the European Organization for the Research and Treatment of Cancer (EORTC) clinical trials' experience in patients with advanced STS noted a 6MPFR of 14% for “active” agents and 8% for “inactive” agents.19 The highest 6MPFR observed to date in EORTC studies was 29% with ET-743.20

Perifosine steady-state levels observed during this study are very similar to the prior Phase I experience11 that observed a mean steady-state concentration of 6.3 ± 2.5 μg/mL (n = 24) at the 100 mg daily dose compared with our results of 6.0 ± 2.3 μg/mL (n = 19). Similar to the Phase I experience, we also did not observe evidence of drug accumulation over repeated dosing. Whereas the mean concentration for each course gradually increased, this appeared more related to patients with higher steady-state levels remaining on study longer. The relatively small dataset limits the reliability of our observation that patients with steady-state levels >6.0 μg/mL were more likely to be progression-free after 2 courses than patients with lower steady-state levels. These data coupled with the knowledge of the effectiveness of topical miltefosine in cutaneous malignancies4 imply that a drug concentration:response gradient exists.

Despite the early negative outcome of this study relative to our primary endpoint, optimism remains for this agent as monotherapy in STS. The infrequent, but prolonged, tumor shrinkage (>6–12 months) observed during early clinical development of perifosine and its potential effects on signal transduction raise the possibility that specific STS histologies or tumor cell characteristics (e.g., increased phosphorylation of Akt) might identify a more sensitive population of STS patients. Continued exploration of perifosine's antitumor effects, especially with attempts at achieving higher steady-state levels, will help pursue this line of study. Consistent with this, a follow-up multiinstitution Phase II trial of perifosine in advanced STS is under way.


We thank the University of Wisconsin 3P Laboratory for performance of perifosine assays and Joy Kurt for article preparation