| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Cancer Therapy: Clinical |
Authors' Affiliations: 1 Division of Medical Oncology, Juravinski Cancer Centre, Hamilton, Canada, and 2 Wellstat Biologics Corporation, Gaithersburg, Maryland
Requests for reprints: Pierre P. Major, The Margaret and Charles Juravinski Cancer Centre, 699 Concession Street, Hamilton, Ontario, Canada L8V 5C2. Phone: 905-387-9495, ext. 64603; Fax: 905-575-6326; E-mail: pierre.major{at}hrcc.on.ca.
 |
Abstract |
|---|
 Top Abstract Patients and MethodsResultsDiscussionReferences |
|---|
Experimental Design: For the first dose, a 3-h infusion was used compared with the 10- and 30-min infusions administered in the two previous trials. Subsequent doses were infused over 1 h. Six doses were given per 3-week cycle. Escalation of the first dose was done separately from the escalation of doses 2 to 6. Viral clearance was determined using whole blood reverse transcription-PCR.
Results: Eighteen patients with advanced chemorefractory cancer were enrolled. The first dose was safely escalated to 24 x 109 plaque-forming units/m2 and doses 2 to 6 were safely escalated to 120 x 109 plaque-forming units/m2. Tolerability was improved compared with the rapid bolus dosing used previously with the elimination of severe flu-like symptoms. Furthermore, infusion reactions were markedly decreased in this trial compared with previous PV701 trials. The presence of neutralizing antibodies did not significantly affect PV701 clearance. Four major and two minor tumor responses were observed.
Conclusions: Using slow infusion, patient tolerability was improved, while the first dose was safely escalated relative to two previous PV701 trials. Based on improved tolerability and encouraging signs of activity, this slow infusion regimen was selected for further PV701 clinical development.
Importantly, the mechanisms for efficacy and toxicity seem to be distinct. In preclinical studies, tumor regressions following i.v. administration required live virus and followed viral replication and tumor cell lysis, whereas toxicity was indirect, due to the release of proinflammatory cytokines, and was observed with both live and UV-killed virus (17).
In the first phase 1 trial of PV701, 79 patients were treated with i.v. bolus dosing to determine the maximum tolerated dose for single-dose and multiple-dose regimens (6). Flu-like symptoms were the most common adverse events, consistent with cytokine-induced toxicity. Most symptoms were associated with the first dose and subsequent doses were better tolerated. This improved tolerability of repeat doses was due to a desensitization associated with the reduced induction of proinflammatory cytokines and occurred before the development of patient antibodies to PV701. There was no cumulative toxicity. Signs of efficacy included two major responses and 4-year survival in a patient with advanced mesothelioma.
Although comparing favorably with the safety of many chemotherapeutics, the previous bolus regimen resulted in one third of patients having reversible grade 3 fatigue, almost exclusively following the first dose of cycle 1 (6). Before beginning phase 2 trials, means to improve tolerability were sought. One approach used a dose de-escalation methodology that incorporated an additional lower desensitizing step and was referred to as "two-step desensitization." Although this regimen was successful at reducing the severity of flu-like symptoms compared with the first phase 1 trial, this approach required a significant 12-fold reduction of the first dose (7).
An alternative method to improve tolerability was addressed by adjusting the infusion rate. The basis for this approach stems from preclinical observations of an i.v. toxicity profile for UV-inactivated virus which was similar to the toxicity profile for live virus, indicating that toxicity could result from the exposure to virus particles rather than requiring virus replication. Thus, slowing the infusion rate, and consequently, the peak virus levels in the blood, was expected to decrease the release of proinflammatory cytokines and improve tolerability. When the slow PV701 infusion rate was tested in mice, cytokine-related toxicity was greatly reduced without affecting efficacy. Independently, in preclinical experiments using vesicular stomatitis virus, another enveloped negative strand RNA virus, Bell and colleagues showed that slowing the i.v. infusion rate allows for more efficient delivery of the virus to tumors.3 In the phase 1 trial of PV701 reported here, slow infusion was coupled with desensitization to improve patient tolerability, especially regarding cytokine-related adverse events. Additionally, this trial was the first study to measure PV701 clearance from whole blood. Marked improvements in patient tolerability to PV701 compared with bolus dosing were observed, including the elimination of severe flu-like symptoms. Furthermore, signs of clinical activity, including six responses, were noted, and this slow infusion regimen was selected for phase 2 testing of PV701.
|
Patients and Methods |
|---|
|
TopAbstractPatients and MethodsResultsDiscussionReferences |
|---|
18 years) were eligible for this trial if they had pathologically confirmed advanced, incurable solid tumors, and at least 30 days had passed since any prior surgery, systemic therapy (6 weeks for mitomycin or nitrosourea), radiotherapy, or systemic corticosteroid use. Other eligibility criteria included Karnofsky performance status of 70, the presence of at least one bidimensionally measurable lesion, and normal organ function including a serum creatinine level <1.5 times the upper limit of normal, bilirubin level <1.5 times the upper limit of normal, serum transaminase levels <2.5 times the upper limit of normal or <5 times the upper limit of normal in the presence of hepatic metastases, leukocyte count of >3,000/µL, neutrophil levels of >1,500/µL, platelet levels of >100,000/µL, and hemoglobin levels of >9 g/dL. Patients were not eligible if they had hypersensitivity to eggs, were receiving systemic corticosteroids, had an autoimmune disease, HIV or hepatitis B or C infection, uncontrolled cardiac disease, or had any uncontrolled bacterial infection. Given the known occurrence of tumor site–specific inflammation, patients with brain metastases, abdominal tumors which had previously caused bowel obstruction, pulmonary metastases >5 cm or causing impaired pulmonary function (FEV1 of <75% of that predicted for age and height, and/or a resting pulse oximetry of <90% oxygen saturation), or tumor at any site which was judged to be dangerous if tumor inflammation or peritumoral edema occurred, were excluded. Pregnant or nursing women were excluded, as were patients with a past or current history of working with poultry, as these latter patients may have preexisting anti–Newcastle disease virus antibodies. Although no human-to-human transmission has been documented with Newcastle disease virus, as a precaution, patients were instructed to avoid infants <1 month of age, pregnant women, and those with immunodeficiency until 3 weeks following discontinuation of study therapy. This trial was approved by Health Canada and the research ethics boards of the Hamilton Health Sciences Corporation. All patients provided written informed consent. PV701. PV701 was grown to high titer in specific pathogen–free embryonated chicken eggs (SPAFAS, Inc., Preston, CT) and purified from allantoic fluid. PV701 doses were expressed as the amount of infectious virus in plaque-forming units (PFU) as determined by plaque assay using HT1080 human fibrosarcoma cells (6). The virus was formulated in a mannitol-lysine solution, and stored below –65°C. For administration, vials were thawed at room temperature, diluted into i.v. saline bags, and infused along with 500 mL of saline within 4 h of preparation.
Treatment. Each 21-day PV701 treatment cycle consisted of six i.v. infusions (days 1, 4, 8, 10, 12, and 15) over a 15-day period followed by 6 days of rest. Dose 1 of the first cycle was administered over 3 h. All other doses were administered over 1 h.
As before (6), antipyretics were administered prophylactically and antidiarrheals were used as needed. Ondansetron (i.v., 8 mg x3) was given on day 1. Additionally, patients received i.v. hydration with 
4 L of saline on day 1. Because the first dose may have been escalated higher than in any previous study, patients were kept overnight as a precaution for the first dose of cycle 1. All other doses were administered on an outpatient basis.
Patients were monitored extensively by physical and laboratory examinations. Radiographic assessments were done after two cycles and any responses were confirmed no earlier than 30 days later.
Toxicity assessment and dose escalation. Toxicity was assessed using the National Cancer Institute Common Toxicity Criteria version 2.0. Dose-limiting toxicity was defined as any clinically significant grade 4 hematologic toxicity (except for lymphopenia and leukopenia or neutropenia lasting <5 days), or nonhematologic toxicity of at least grade 3 (except for flu-like adverse events, and reversible elevations of liver enzymes in those with hepatic metastases). At least three patients were enrolled per cohort. There were two sets of dose escalations between cohorts: first for dose 1 then for doses 2 to 6.
Anti-PV701 antibodies. Serum samples were analyzed for neutralizing antibody (6) and for anti-PV701 IgG antibodies. For the latter analysis, serial dilutions of heat-inactivated serum samples were added to PV701-coated microplates. Following antibody absorption, and reaction with peroxidase-conjugated anti-human IgG, antibody concentration was determined by colorimetric assay using 3,3',5,5'-tetramethylbenzidine and a standard curve. This immunoassay had a limit of detection of 0.03 mg/mL and a lower limit of quantification of 0.09 mg/mL for anti-PV701 IgG.
Viruria. Urine samples were collected, diluted 1:1 with viral transport/preservation medium (M4RT; Remel, Lenexa, KS) and frozen at 
–65°C. After thawing, samples were passed through a desalting column to exchange the urine matrix with the assay medium. Serial dilutions of the sample were done across a microtiter plate followed by the addition of HT1080 fibrosarcoma cells. Following 68 h at 37°C, cell viability was determined by colorimetric assay and the virus concentration was determined by comparison to a standard. This quantitative assay had a limit of detection of 70 PFU/mL and a lower limit of quantification of 600 PFU/mL. However, patient urine output was not measured, and therefore, the results represent a semiquantitative assessment.
Viremia and clearance by reverse transcription-PCR. Whole blood for viremia was collected during the first two cycles prior to each dose, and prior to the first dose of each subsequent cycle. Whole blood for clearance was collected from six patients (three each from the first and last cohorts) following each of the first three doses of cycles 1 and 2, and stored at <–65°C. As a control, a low concentration of heterologous RNA (internal positive control) was spiked into each sample prior to extraction. Total RNA was extracted and amplified in multiplex reactions using the ABI Prism 7700 Sequence Detection System (Applied Biosystems, Foster City, CA) and PV701 and internal positive control–specific primer/probe combinations. PV701 copy number, expressed as PFU-equivalents per milliliter, was calculated following interpolation of the specimen cycle threshold to a reference curve of PV701 extracted and amplified in parallel. The assay limit of quantification was 1,000 PFU-eq/mL of whole blood.
|
Results |
|---|
|
TopAbstractPatients and MethodsResultsDiscussionReferences |
|---|
|
One or more flu-like symptoms were observed in all patients after the first dose of cycle 1 (Table 2 ). These were mild-to-moderate and decreased in intensity and frequency with subsequent dosing (Fig. 1 ). Mild-to-moderate gastrointestinal symptoms such as nausea and vomiting were also commonly observed during the first cycle. Hematologic effects were also transient and mainly mild-to-moderate. Although some grade 3 leukopenia, neutropenia, and thrombocytopenia events were seen (Table 2), none were clinically relevant and all resolved rapidly despite continued treatment with higher doses. Here, as in previous clinical trials of PV701, common adverse events, including flu-like symptoms and hematologic effects, were consistent with the production of proinflammatory cytokines (see Cytokine levels) and were predicted from preclinical studies.
|
|
Following PV701 dosing, >80% of patients had transient liver transaminase elevations (mainly mild-to-moderate). Two patients had grade 3 transaminase elevations that returned to grade 1 or within normal limits even during continued dosing.
Adverse events associated with the location and size of tumor(s) were common. Examples of these tumor site–specific adverse events include: inflammation, pain, and swelling at tumor sites; obstructive jaundice in a patient with bile duct metastasis that improved with corticosteroids; and abdominal pain and bowel obstruction in two patients with extensive intra-abdominal tumors.
Immunogenicity and shedding. No patient had measurable baseline levels of anti–Newcastle disease virus IgG. All patients developed antibodies by day 7. During the first few treatment months, neutralizing antibody levels increased with repeated dosing, reaching a plateau after cycle 4 (Fig. 2A ; median of 1:2,560; range, 1:640-1:5,120). Antibody levels from the six patients whose whole blood samples were measured for clearance of PV701 genomes (see below) were typical of other patients in this study, with levels at the start of the second cycle ranging from 1:320 in one patient to 1:640 in the other five patients, compared to the median of 1:640 for all patients. In general, the development of neutralizing antibody to PV701 paralleled the development of total anti–Newcastle disease virus IgG levels (Fig. 2A).
|
Pharmacokinetic results. All patients were tested for viremia in whole blood by quantitative reverse transcription-PCR prior to each dose. As expected, patients were negative prior to dosing and became positive after treatment. Seventeen of 18 patients remained positive for viremia throughout the first two cycles. Among these patients, the median viremia was 3.0 x 104 PFU-eq/mL with a range of 1.0 x 103 to 2.3 x 106 PFU-eq/mL. Distribution/elimination of PV701 particles from blood followed a biphasic pattern (Fig. 2B), with the initial log-linear phase noted immediately post-infusion having a relatively short half-life of <1 h. The terminal log-linear phase began after 8 h and had an average half-life of >10 h. The concentration-time data for repeat doses indicated that the exposure to virus-related material was roughly proportional (no significant differences noted in the log-transformed, dose-normalized AUC, and Cmax values; data not shown). No accumulation of the virus-related material was noted on repeat dosing. Additionally, there was no significant difference in clearance between cycles 1 and 2.
Cytokine levels. Plasma samples were obtained from three patients in the first cohort and three patients in the last cohort, and analyzed by immunoassay for serum IFN-
and tumor necrosis factor (6). As observed previously (6), IFN- was detectable at 6 h post-dosing and peaked at 20 h post-dosing (median of 836 pg/mL), returning to baseline by 72 h. This median peak level, however, was >10-fold lower than that observed in the earlier study with bolus dosing (6). This result is consistent with the hypothesis that the improved tolerability of slow infusion is due to a reduction in the release of proinflammatory cytokines. Tumor necrosis factor was detectable in only half of the patients at 20 h post-dosing.
Antitumor activity. All 18 patients were evaluated for response after two cycles. Fifteen patients had at least stable disease and received further treatment. Eleven of 18 patients (54%) had a time-to-progression of at least 4 months. In addition, six objective tumor responses were noted in diverse tumor types (Table 3 ; Fig. 3 ). Tumor biopsies were obtained from two responding patients and extensive inflammation and/or fibrosis was noted in each instance (see, for example, Fig. 3D). One of the two patients with colorectal cancer who responded also had elevated serum levels of carcinoembryonic antigen, which spiked shortly after dosing followed by a decline to 70% below the baseline (Fig. 3C). A similar carcinoembryonic antigen pattern was seen in other patients with colorectal cancer previously treated with PV701, including a patient with a major radiographic response (6).
|
|
|
Discussion |
|---|
|
TopAbstractPatients and MethodsResultsDiscussionReferences |
|---|
This study is the third phase 1 trial to show that i.v. administration of PV701 is well tolerated by most patients. As in the first trial (6), a desensitizing first dose was used, allowing for the administration of higher subsequent doses. The key difference in the current study was the use of a slower infusion rate to which we attributed the improvement in patient tolerability. In the current trial, the first dose (24 x 109 PFU/m2) was administered over 3 h compared with 10 min in the first trial (which used a dose of 12 x 109 PFU/m2). Thus, the maximum rate of virus administration for the first dose was 9-fold slower than in the first study (8 x 109 versus 72 x 109 PFU/m2/h). Similarly, the maximum rate of virus administration for repeat doses was nearly 5-fold slower (120 x 109 versus 576 x 109 PFU/m2/h). In contrast to the original phase 1 study in which 33% of patients had grade 3 fatigue and 12% had grade 3 fever, no patient in this trial had grade 3 flu-like adverse events. Compared with the second PV701 trial in which grade 3 flu-like symptoms were similarly eliminated (7), the improvement in tolerability in the current trial was achieved using a first dose that was 24-fold higher. The doses in the current study were also markedly higher (e.g., 400-fold for the first dose) than that used in the clinical study of Newcastle disease virus strain HUJ by Freeman et al. (9). The higher doses used here may be particularly helpful from a therapeutic perspective because higher doses have been consistently associated with greater tumor responses preclinically. As additional prophylaxis, patients in this trial received aggressive fluids after the first dose. This may have also contributed to the improved tolerability, especially in terms of decreased fatigue.
As noted earlier (6, 7), there were three classes of adverse events due to PV701: (a) flu-like symptoms, (b) tumor site–specific adverse events, and (c) infusion reactions. The main adverse events in cycle 1 were flu-like symptoms, which were never above grade 2 in intensity, and were attributed to the observed induction of proinflammatory cytokines. Most other cycle 1 side effects were mild-to-moderate and no patients had major toxicity, indicating that this regimen is safe for outpatient administration. Although a maximum tolerated dose was not reached, the first dose was not escalated to >24 x 109 PFU/m2 following the observation of dose-dependent mild hypotension and moderate fever. With repeat dosing, there was a reduction in the incidence of flu-like adverse events, even at doses five times higher than the first dose. Most patients experienced little, if any, toxicity following cycle 1. As noted in the previous PV701 trials, prolonged use of PV701 seems safe with no evidence of cumulative toxicity.
Tumor site–specific adverse events, together with signs of tumor inflammation, were commonly observed. These reactions are important to understand in terms of both safety and the potential effect on response. Regarding safety, various adverse events occurred which depended on the location of each individual's tumor. For example, patients with s.c. metastases had redness, pain, and/or swelling of their lesions with subsequent tumor regression. Patients with liver metastases experienced acute, reversible right upper quadrant pain. A patient with a bile duct tumor experienced obstructive jaundice, which was rapidly reversible with steroids. Two patients with large intra-abdominal tumors experienced various symptoms of bowel obstruction; none of which required surgical intervention. For future studies, patients with extensive intra-abdominal tumors (e.g., tumors 5 cm) should be followed closely. In contrast to the first phase 1 study in which severe dyspnea occurred in nine patients with extensive lung involvement (6), severe dyspnea was not an issue in this trial. This is likely explained by the fact that in the current trial, patients with large lung tumors (5 cm) and those with baseline pulmonary compromise of at least segmental size were excluded, and all patients with pulmonary metastases were required to pass pulmonary function tests prior to enrollment.
Tumor inflammation in response to PV701 therapy must also be kept in mind from a response perspective. Shortly after dosing and before some of the tumor responses, lesions became inflamed or swollen. Histologically, extensive inflammation and/or fibrosis were observed in both of the tumor biopsies examined, and clinical signs of inflammation were documented in three others following PV701 treatment. The most obvious example of an inflammatory reaction occurred in a woman with in-transit melanoma. Many of her lesions became visibly inflamed after dosing before undergoing tumor regression. If such lesions had been internal, they could have been easily confused with tumor progression on CT scan. Sze et al. has taken this concept further by arguing that tumor swelling and/or inflammation is likely a desirable consequence of virus administration indicating activity at the tumor site (19).
Infusion reactions, consisting principally of mild-to-moderate back and chest pain, were more common in the two previous PV701 trials. In the current trial, infusion reactions occurred infrequently, were milder, and were only seen at the two highest dose levels. In all three trials, infusion reactions were never associated with the first PV701 dose. There was a 17-fold reduction in the per dose incidence of infusion reactions during the first two cycles in this trial compared with the first phase 1 trial using rapid bolus dosing (2% versus 35%; P < 2 x 10–17, Fisher's exact test). There was also a highly significant reduction in the incidence of infusion reactions when comparing this trial (that used a 1-h infusion of repeat doses) to the second trial that used 30-min infusions (2% versus 12%; P < 0.002, Fisher's exact test). Although the cause of these reactions remain unclear, they were managed effectively by further slowing the i.v. infusion rate and, in all instances, a full dose was given.
As expected (10), all patients were seronegative at the start of the current study. After treatment, patients developed neutralizing antibodies, reaching a modest plateau titer of 1:2,560. This level is within 2-fold of that seen previously (6, 7), a difference within the variability of the assay. Interestingly, the development of neutralizing antibody titers did not affect the clearance of PV701, and viremia was maintained during the first two cycles of treatment, although there was no accumulation of PV701 in whole blood upon repeat dosing. These studies measured the presence of PV701 nucleic acid in whole blood by reverse transcription-PCR. In future studies, it will be interesting to see how the pharmacokinetics of infectious virus compares to the pharmacokinetics of viral nucleic acid. As in previous PV701 trials, tumor regressions were observed to occur long after antibody formation.
In this trial, as in the other phase 1 PV701 trials (6, 7), signs of antitumor activity were observed. Half of the patients had a time-to-progression of at least 4 months with one third of all patients surviving for >2 years. These observations were notable in patients with advanced chemorefractory disease and need to be tested rigorously in future randomized controlled studies in well-defined study populations. Together with the earlier trials, tumor responses have now been observed in diverse cancers including mesothelioma, melanoma, and carcinomas of the colon, rectum, cervix, anus, head and neck, and pancreas.
There are several strategies that could be used in the future to improve therapeutic outcome using PV701. One potential strategy would be to treat patients earlier in the natural history of their disease, such as in the adjuvant setting, or when the disease burden is minimal. The ongoing benefit at 3.5 years after enrollment, noted in the melanoma patient who received localized radiation therapy shortly after finishing PV701 treatment, suggests a strategy of combining PV701 with other modalities such as chemotherapy and/or radiation. Indeed, in preclinical models, significant antitumor effects have been noted following the combination of chemotherapeutics or radiation with PV701.4
In conclusion, PV701 has shown a predictable and easily manageable toxicity profile. The predominant toxicity profile is characterized by flu-like symptoms which are manageable on an outpatient basis with patient monitoring and symptom control measures, including prophylactic antipyretics and hydration. The slow i.v. infusion approach employed in this trial yielded important improvements in patient tolerability despite using higher initial levels than in the previous trials. Given these improvements in tolerability and the encouraging signs of activity noted in this heavily treated patient population, further investigation of PV701 is planned in phase 2 studies using the dosing regimen developed in this trial.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Note: Presented in part at the 39th and 40th American Society of Clinical Oncology Annual Meetings in New Orleans, Louisiana, and Chicago, Illinois in 2002 and 2003, respectively.
3 Bell J.C., personal communication. 
4 Lorence R.M. and Roberts M.S., unpublished results.
Received 7/25/06; revised 9/20/06; accepted 10/ 6/06.
|
References |
|---|
|
TopAbstractPatients and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  E. Carnero, W. Li, A. V. Borderia, B. Moltedo, T. Moran, and A. Garcia-Sastre Optimization of Human Immunodeficiency Virus Gag Expression by Newcastle Disease Virus Vectors for the Induction of Potent Immune Responses J. Virol., January 15, 2009; 83(2): 584 - 597. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Cancer Research | Clinical Cancer Research |
| Cancer Epidemiology Biomarkers & Prevention | Molecular Cancer Therapeutics |
| Molecular Cancer Research | Cancer Prevention Research |
| Cancer Prevention Journals Portal | Cancer Reviews Online |
| Annual Meeting Education Book | Meeting Abstracts Online |
