Sunwaybio.com.cn

P h a s e I T r i a l o f I n t r a p e r i t o n e a l I n j e c t i o n o f t h e E 1 B - 5 5 -
k d - G e n e – D e l e t e d A d e n o v i r u s O N Y X - 0 1 5 ( d l 1 5 2 0 ) G i v e n o n
D a y s 1 T h r o u g h 5 E v e r y 3 W e e k s i n P a t i e n t s W i t h
R e c u r r e n t / R e f r a c t o r y E p i t h e l i a l O v a r i a n C a n c e r
By P.A. Vasey, L.N. Shulman, S. Campos, J. Davis, M. Gore, S. Johnston, D.H. Kirn, V. O’Neill, N. Siddiqui, Purpose: Resistance to chemotherapy in ovarian
1010 pfu developed common toxicity criteria grade 3
cancer is frequently associated with mutations in the
abdominal pain and diarrhea, which was dose-limit-
p53 gene. The adenovirus dl1520 (ONYX-015) with the
ing. The maximum-tolerated dose was not reached at
E1B 55-kd gene deleted, allowing selective replication
1011 pfu, and at this dose level patients did not expe-
in and lysis of p53-deficient tumor cells, has shown
rience significant toxicity. There was no clear-cut evi-
preclinical efficacy against p53-deficient nude mouse-
dence of clinical or radiologic response in any patient.
human ovarian carcinomatosis xenografts.
Blood samples were taken for adenovirus DNA and
Patients and Methods: We undertook a phase I trial
neutralizing antibodies. Polymerase chain reaction
of intraperitoneal dl1520 in patients with recurrent
data indicating presence of virus up to 10 days after the
ovarian cancer. Sixteen women with recurrent/refrac-
final (day 5) infusion of dl1520 are suggestive of con-
tory ovarian cancer received 35 cycles (median, two
tinuing viral replication.
cycles) of dl1520 delivered on days 1 through 5 in four
Conclusion: This article therefore describes the first
dose cohorts: 1 ؋ 109 plaque forming units (pfu), 1 ؋
clinical experience with the intraperitoneal delivery of
1010 pfu, 3 ؋ 1010 pfu, and 1 ؋ 1011 pfu.
any replication-competent/-selective virus in cancer
Results: The most common significant toxicities re-
patients.
lated to virus administration were flu-like symptoms,
J Clin Oncol 20:1562-1569. 2002 by American
emesis, and abdominal pain. One patient receiving 1 ؋
Society of Clinical Oncology.
EPITHELIAL OVARIAN cancer is the fourth most though other factors, such as tumor burden and histology, frequent cause of cancer death among women in both are also important.3 Patients who relapse after 6 months are the United States and the United Kingdom, and it is the most termed “platinum-sensitive” and are usually treated with frequent cause of death from gynecologic cancer in the platinum-based salvage therapy. Patients who either developed world.1 It is considered one of the most chemo- progress on first-line therapy or relapse within 6 months of sensitive cancers, with response rates to platinum-contain- completion are termed “platinum-resistant,” have a poor ing regimens of greater than 60%.2 However, fewer than prognosis, and are suitable candidates for experimental 25% of patients with advanced disease survive 5 years, and treatment of recurrent disease is complicated by the emer- In normal cells, the tumor suppressor gene p53 mediates gence of drug resistance. The probability of response to cell cycle arrest and apoptosis in response to DNA damage second-line or salvage chemotherapy is related to the induced by radiotherapy, DNA alkylating chemotherapy interval between primary chemotherapy and relapse, al- agents, or foreign DNA synthesis.4-6 Correlations have beenestablished in laboratory studies between the emergence ofcisplatin-resistant ovarian cancer cells and the presence ofmutations in the p53 gene, and it is probable that a From the Beatson Oncology Centre and Stobhill Hospital, Glasgow, substantial number of drug-resistant tumor cells at the time Royal Marsden Hospital, Surrey, and Imperial College and ImperialCancer Research Fund, London, United Kingdom; and Brigham and of clinical relapse in ovarian cancer lack functional p53.7 Women’s Hospital, Dana-Farber Cancer Institute, and Massachusetts dl1520 (ONYX-015) is an adenovirus made from human group C adenovirus (serotypes 2 and 5) that has several Submitted June 7, 2001; accepted November 6, 2001. small mutations within the serotype 5 portion of the virus. It Supported by the Cancer Research Campaign, which funds the has been attenuated by deletion of the E1B 55-kd gene Address reprint requests to Paul Vasey, MD, Cancer Research region, the protein product of which is known to bind and Campaign Department of Medical Oncology, Cancer Research Cam- inactivate p53 and allow continued DNA synthesis and viral paign Beatson Laboratories, Garscube Estate, Switchback Rd, Glas- replication. Mutants such as dl1520 that lack this early gene gow G61 1BD, United Kingdom; email: pav1y@clinmed.gla.ac.uk. product are severely deficient in their ability to replicate in 2002 by American Society of Clinical Oncology.
0732-183X/02/2006-1562/$20.00 normal cells.8,9 In vitro studies have demonstrated that Journal of Clinical Oncology, Vol 20, No 6 (March 15), 2002: pp 1562-1569 Downloaded from www.jco.org on August 17, 2006 . For personal use only. No other uses without permission. Copyright 2002 by the American Society of Clinical Oncology. All rights reserved. INTRAPERITONEAL ONYX-015 IN RECURRENT OVARIAN CANCER dl1520 is capable of efficient, selective replication and tolerated dose (MTD). Secondary objectives included the cytopathogenicity via cytolysis in p53-deficient human determination of dl1520 propensity to replicate in ovarian tumor cells, whereas replication was abrogated in cells carcinoma cells within the ascitic fluid, and the evaluation containing wild-type p53.10 In addition, laboratory studies of the humoral immune response (antibody development) to in Glasgow have demonstrated that selective replication of dl1520 in both blood and ascitic fluid.
dl1520 in cells with nonfunctioning p53 is due to theinduction of apoptosis in cells with wild-type p53.11 Cells with nonfunctioning p53 that are therefore resistant to apoptosis are permissive for replication, leading to virusspread and subsequent cytolysis of the cell population.
This study was conducted at four cancer centers in the United States There is some controversy because p53 gene sequence alone and the United Kingdom. Eligible patients had histologically orcytologically confirmed primary epithelial ovarian cancer, with recur- does not predict for the antitumor efficacy of dl152012-14; rent disease suspected either radiologically or by rising CA125 levels.
however, p53 function can be lost via many mechanisms.
All patients had received previous chemotherapy with platinum- Furthermore, in vivo correlation of p53 function and dl1520 containing regimens and had relapsed within the previous 6 months.
efficacy was demonstrated in three nude mouse-human Karnofsky performance status of Ն 70%, age of at least 18 years, and ovarian carcinomatosis xenograft models, A2780/Cp70, life expectancy of at least 3 months were required, as was writteninformed consent. Patients were, in the opinion of the investigators, OVCAR3 (mutant p53), and A2780 (wild-type p53).15 In suitable candidates for either a laparotomy or laparoscopy. Patients in these experiments, both antitumor efficacy and improved whom laparoscopy was contraindicated were those with known dense survival were demonstrated for the mice carrying mutant adhesions, periumbilical infection, peritonitis, intestinal obstruction, p53 xenografts, whereas no such effects were evident for the ileostomy, or gross obesity. Normal organ function was required, as wild-type p53 xenografts. Clinical studies of dl1520 have evidenced by the following: neutrophil count more than 2 ϫ 109/L,hemoglobin concentration more than 10 g/L, serum creatinine level less been undertaken in patients with recurrent head and neck than 150 ␮mol/L, aminotransferase levels less than 2.5 times the upper cancer, a disease in which p53 gene mutations or deletions limit of normal, prothrombin time or international normalized ratio Յ are present in up to 70% of patients.16 Significant activity 2, and partial thromboplastin time within normal limits. Patients were has been documented in a dose-finding phase I study, with excluded from study entry if they had known chronic liver dysfunction evidence on magnetic resonance imaging of tumor necrosis before the development of ovarian cancer or had either cirrhosisevident on gross examination during laparotomy/laparoscopy or more at the site of viral injection in five of 32 patients (four of than 50% liver replaced by tumor. Other exclusion criteria included whom had mutant p53 tumors).17 Dose-limiting toxicity ongoing, active infection including human immunodeficiency virus, (DLT) was not reached at 1010 plaque-forming units (pfu)/d recent viral syndrome, recent chemotherapy or radiotherapy, concom- by five consecutive daily doses every 4 weeks.
itant hematologic malignancy, ongoing requirement for immunosup- Direct administration of cytotoxic drugs into the perito- pressive medication including glucocorticoids or cyclosporine, andpregnancy/lactation. Patients treated on any other investigational pro- neal cavity is a strategy designed to enhance locoregional gram within the previous 6 weeks were also excluded from participa- drug delivery while abrogating systemic toxicities. Further- tion, as were patients previously treated on a research protocol more, this route allows concentrations to be attained at the involving the administration of adenovirus-based therapies. The proto- site of the tumor that are many times higher than would be col was reviewed and approved by the institutional review boards, the tolerated in the systemic circulation; these can easily exceed biosafety committees, and in the United States by the Food and DrugAdministration before patient accrual.
concentrations shown in vitro to be required to overcomeclinical drug resistance.18 This method of administration is particularly relevant in ovarian cancer, since the disease at Before treatment, patients had routine hematologic and biochemical presentation is confined to the abdominal cavity in approx- analyses performed, in addition to undergoing chest radiography, imately three quarters of patients and subsequent relapses urinalysis, ECG, CA125 marker level assessment, and a complete also tend to remain thus compartmentalized. This model physical examination. A baseline computed tomography scan of the system of metastatic ovarian cancer growing within a abdomen with documentation of any measurable tumor was carried out.
clearly defined anatomic space bathed in free fluid has many In addition, blood was drawn for serum antibody to group C adenovi-rus, adenovirus DNA (by polymerase chain reaction [PCR]), immune potential advantages in delivery, safety, and efficacy of function evaluation (CD3, CD4, CD8, total lymphocyte count, delayed- novel therapies. This study therefore seeks to exploit the type hypersensitivity skin test). Finally, a test for pregnancy was molecular differences between normal and malignant cells, utilizing the specificity of dysfunctional p53-dependentcytotoxicity. The primary objective was therefore to deter- mine the safety of intraperitoneal administration with For placement of the intraperitoneal catheter, all patients underwent dl1520 daily for 5 days and to determine the maximum- either a laparoscopy or laparotomy, the choice depending on the Downloaded from www.jco.org on August 17, 2006 . For personal use only. No other uses without permission. Copyright 2002 by the American Society of Clinical Oncology. All rights reserved. presence of ascites and the size and site of recurrent tumor masses.
cell pellet was smeared and fixed onto slides before in situ hybridiza- Generally, laparoscopy was performed (following paracentesis in the tion (ISH) for adenovirus DNA to determine the ability of dl1520 to presence of ascites) if there were no visible or resectable lesions, replicate in ovarian cancer cells within the ascitic fluid in vivo.
whereas laparotomy was preferred if there was an option to resect Certified pathologists examined slides to characterize any infected disease. In the case of laparoscopy, contingency plans were made to cells. Finally, determination of neutralizing antibody titers was carried proceed to laparotomy if there were complications or to confirm relapse if laparoscopy results were negative. The catheter (Tenckhoff orPort-a-Cath [Pharmacia Deltec, St Paul, MN]) was placed intraopera- tively, and a tumor biopsy was performed to assess p53 status by gene DLT was defined as National Cancer Institute of Canada common sequencing. Any adhesions in the peritoneum were divided if technically toxicity criteria (CTC) grade 4 flu-like symptoms (eg, fever, fatigue, possible. To prevent catheter leakage, a minimum 7-day period was myalgia) or any other grade 3 toxicity attributed to ONYX-015 allowed to elapse before usage, and therefore heparinized saline was administration. The MTD was defined as the dosage at which two of six instilled into the tubing during this period.
patients experienced a DLT after the first treatment cycle with dl1520 was grown and “titered” on the human embryonic kidney cell ONYX-015. Patients were enrolled sequentially onto treatment cohorts line HEK293,19 formulated in a sterile viral solution in TRIS buffer (10 of increasing dose level as follows. Three patients were entered onto mmol/L [pH 7.4], MgCl 1 mmol/L, NaCl 150 mmol/L, and 10% cohort 1, and if no DLT occurred in the first two patients over the first glycerol), and supplied frozen in single-use plastic screw-cap vials 3 weeks and no DLT occurred in the first week after infusion for the containing 0.5 mL of virus at a specified concentration. dl1520 was third patient, recruitment to the next dose level proceeded. If one of produced and supplied by Magenta Corp (Rockville, MD). These vials three patients in a cohort experienced a DLT, up to three additional were stored below Ϫ20°C and were thawed and initially diluted to the patients were enrolled at the same dose. If one or more of the additional appropriate titer for a particular dose level. Thawed virus was main- patients had a DLT (ie, Ն two of six), the MTD was defined and no tained at 2°C to 8°C during dilution and handling, before being warmed further escalation took place. Patients were eligible for repeat dosage to room temperature for intraperitoneal administration. Dilutions to a cycles of dl1520 at the same dose if they experienced no DLT or no final volume of 500 mL were performed immediately before infusion.
disease progression. A maximum of six cycles was allowed.
dl1520 adenovirus infusion through the catheter proceeded over 15minutes suspended in 500 mL of physiologic saline. After infusion, the patient was rotated on all sides over a 30-minute period to maximizeperitoneal exposure. Vital signs were taken every 15 minutes before Sixteen patients, all white women, received 35 cycles of and at the start of treatment and at 60, 90, and 120 minutes afterward.
dl1520 in four dose cohorts: 1 ϫ 109 pfu, 1 ϫ 1010 pfu, 3 Repeat infusions were carried out daily for 5 days.
ϫ 1010 pfu, and 1 ϫ 1011 pfu. Baseline characteristics ofthese patients are shown in Table 1. Most patients had platinum-resistant ovarian cancer and bulky residual tumor On day 5, blood was taken to assess liver function and to measure masses (defined as disease volume Ͼ 2 cm). The median virus DNA by PCR. Before dl1520 infusion, peritoneal fluid was drawn number of prior therapies was four, and all patients had for cytologic evaluation. Patients were then reviewed in the clinic ondays 8, 15, and 22 for each cycle of dl1520 treatment. At each visit, general status and toxicity assessment was carried out by examination, (62.5%) had been optimally cytoreduced at their initial adverse event reporting, and scoring of Karnofsky performance status.
operation for ovarian cancer. The three patients in cohort 4 Blood was also drawn for hematology, serum chemistry, CA125, and were selected specifically for nonbulky disease volume and adenovirus DNA and antibody tests, and a urinalysis was performed. In minimal prior therapy. One patient in cohort 2 had further addition, on day 15, peritoneal fluid was withdrawn, if present. If thiswas not possible, 1,000 mL of 0.9% saline was infused through the surgical cytoreduction to less than 2 cm residuum at the time catheter, allowed to distribute throughout the peritoneal cavity for 15 to20 minutes, and then drained. On day 22, repeat hematologic, biochem- Table 1. Baseline Patient Characteristics (N ؍ 16)
ical, and CA125 analyses were performed and patients received chestradiograms and ECGs. Every second cycle, patients had a repeat Tumor biopsies were performed at laparotomy when possible, and immunohistochemistry was performed on formalin-fixed paraffin- embedded tumors for p53 status. In addition, Oncormed Corp (Gaith- ersburg, MD) sequenced exons 5 to 9 of the p53 gene on pretreatment tumor biopsy samples. This analysis was carried out retrospectively and was not a prerequisite for entry onto the study. Ascitic fluid samples were obtained from eight patients at the time of day 5 and 15 samples/washes. The fluid was spun to pellet any cells, and the supernatant was frozen. The presence of adenovirus in plasma samplesor the cell-free fraction of patients’ peritoneal fluid was determined by *Defined as relapsing within 6 months or progressing on therapy.
PCR using primers of the E1A region of the adenoviral genome. The †Of seven assessable samples obtained.
Downloaded from www.jco.org on August 17, 2006 . For personal use only. No other uses without permission. Copyright 2002 by the American Society of Clinical Oncology. All rights reserved. INTRAPERITONEAL ONYX-015 IN RECURRENT OVARIAN CANCER Table 2. Treatment Delivery
with peritonism and was associated with diarrhea and/or increased stoma effluence in some patients. Other associ-ated symptoms included heartburn and vomiting. One pa- tient in cohort 2 experienced grade 3 abdominal pain with grade 3 diarrhea after 1 cycle of dl1520. The diarrhea and pain developed during the 5 days of viral administration and *Median number of cycles was two for all cohorts.
eventually required the administration of loperamide hydro- chloride and parenteral opiates. She was admitted forsymptom control and required a nasogastric tube insertionfor palliation. This toxicity profile was considered to be of catheter insertion. Treatment delivery is shown in Table 2.
dose-limiting, and three additional patients were recruited at Most patients received more than one cycle of dl1520, and this dose level. Four of the other five patients in this cohort three patients received four cycles (median, two cycles).
experienced grade 2 abdominal pain but no associated severe diarrhea or other features that would have stoppeddose escalation. There were no other dose-limiting toxicities Side effects from the administration of dl1520 were described in any of the other cohorts.
common, but CTC grade 4 toxicity was not reported. The During the study, there had been an impression that most common toxicities were related to the acute adminis- patients in cohorts 1 to 3 with known bulky intra-abdominal tration of dl1520 and are readily grouped together as disease were experiencing more severe symptoms of vire- “viremic” or “flu-like.” These toxicities, CTC grades 2 and mia and/or abdominal pain. As a consequence, cohort 4 3, are shown in Table 3. Other toxicities considered to be at patients were recruited specifically with nonbulky residual least possibly related to the administration of ONYX-015 tumor volume and good performance status and were less and reaching CTC grades 2 and 3 are shown in Table 4.
heavily pretreated (although all three were considered to be There were no cumulative nonhematologic toxicities noted, platinum-resistant). Although dl1520-related pyrexias were and no evidence of hematologic toxicity was demonstrated.
documented, there were no significant flu-like symptoms Most patients reported flu-like symptoms after adminis- and only one patient described grade 2 abdominal pain.
tration of dl1520, and eight patients (50%) described at least There were no toxicities higher than grade 2 in these grade 2 severity. Symptoms generally started after the first patients, which suggests a better tolerance for patients with daily infusion of dl1520 and consisted of any or all of the low volumes of tumor. However, the two patients with non- following: generalized malaise, headaches, nausea, myal- bulky disease in cohorts 2 and 3 both experienced the flu-like gias, pyrexias, and rhinorrhea. Three patients, one in cohort syndrome and either abdominal pain or diarrhea; therefore, no 2 and two in cohort 3, reported CTC grade 3 viremic conclusions can be made regarding this association. The MTD symptoms, but they were not considered to be dose-limiting.
was therefore not reached in any of the four cohorts. Further Acetaminophen (up to 4 g daily) and antiemetics (eg, dose escalation was not pursued because the limit of virus cyclizine hydrochloride 50 mg three times daily) were given manufacturing capacity had been attained.
as prophylaxis and as treatment of these symptoms, withsome amelioration on subsequent cycles demonstrated. Two patients reported the nausea associated with other flu-likesymptoms to be precipitated by head movement in a similar In seven patients, the p53 status of pretreatment tumor manner to that seen with viral labyrinthitis.
biopsy specimens was evaluated retrospectively by gene Additionally, abdominal pain was commonly observed sequencing. Five patients (71%) were found to have muta- after dl1520 administration. This had features consistent tions in the p53 gene on sequencing. Twenty-eight perito- Table 3. Viremic/Flu-Like Toxicity
NOTE. Grade was based on National Cancer Institute of Canada common toxicity criteria.
Downloaded from www.jco.org on August 17, 2006 . For personal use only. No other uses without permission. Copyright 2002 by the American Society of Clinical Oncology. All rights reserved. Table 4. Other Significant Toxicities (NCIC-CTC)
NOTE. Grade was based on National Cancer Institute of Canada common toxicity criteria.
*Dose-limiting toxicity (both in same patient).
neal washings sampled at day 5 and day 15 (and later if before progressive disease developed. One patient did possible; timings were not specified in the protocol) after experience a fall in CA125 from 1,585 kilounits/L to 692 dl1520 administration were obtained from eight patients.
kilounits/L during two cycles of dl1520 but was not classi- Using PCR, the presence of dl1520 viral DNA in the fied as having a true CA125 response20 because of the cell-free fraction was assessed in 25 of 26 of these wash- transperitoneal fluid shifts induced by the day 5 and day 15 ings. Seven of these eight patients had evidence of viral peritoneal aspirations and washes. In addition, this patient presence at day 5, and five of these had additional evidence subsequently developed progressive disease with rising of viral presence on day 15. In one patient, dl1520 DNA markers and new sites of disease requiring radiotherapy was detected in the peritoneal fluid up to 354 days after the during her third cycle. The median survival time for all fourth cycle of treatment. Sixty-two percent (16 of 26) of patients was 165 days (range, 15 to 528 days). Of the 16 the cytology specimens prepared from the cell fraction of patients on study, 15 stopped ONYX-015 because they the peritoneal washings were not assessable due to either eventually developed progressive disease and one was taken low cellularity or no evidence of malignancy in the speci- mens. Of the remaining 10 cytology specimens, adenovirus DNA was only detected in one patient specimen by ISH.
However, the positive cells in this specimen did not appear This phase I study successfully achieved the primary aim to be malignant based primarily on size and nuclear mor- of defining safety and evaluating dose-related toxicity phology assessment. Because of the small number of associated with intraperitoneal delivery of dl1520. Proof of assessable specimens, it is not possible to conclude if principle for viral replication, antibody response, and evi- dl1520 is able to replicate in tumor cells in vivo. In addition, dence of an antitumor effect were secondary objectives, but eight patients from cohorts 1 to 3 had blood samples they were important nevertheless in delineating the possible collected after the first infusion of dl1520 at 15 minutes, 1 role of such novel therapies in the future. The viremic/flu- hour, 6 hours, 12 hours, and 24 hours in order to examine like symptoms seen after infusion could suggest that active plasma for adenovirus by quantitative PCR. Circulating viral replication is occurring, as shown in previous studies levels of dl1520 were not detected in any of these samples.
with oncolytic viruses. One of the first such illustrative Finally, six (46%) of 13 patients had positive titers of studies was performed at the National Cancer Institute, neutralizing immunoglobulin G antibodies to adenovirus wherein 30 cervical cancer patients were treated with direct dl1520 at the start of treatment. Of these, all six developed tumoral infusion of wild-type human adenoviruses of 10 increased levels of antibody titers during treatment, and six different serotypes.21 In this trial, all patients developed a of the seven patients without neutralizing antibody titers at humoral response with neutralizing antibodies within 7 baseline developed high titers of neutralizing antibodies days, viral replication was documented for up to 17 days in after dl1520 treatment, indicating a significant humoral one patient, and a transient viral syndrome lasting 2 to 7 response. There was no clear correlation between the days was reported. In the study reported by Ganly et al,17 21 presence or absence of neutralizing antibody titers at base- of 22 patients with advanced head and neck cancer showed line and the severity of the flu-like symptoms.
an increase in neutralizing antibody after intratumoralinjection of dl1520 adenovirus, and viral replication was demonstrated by ISH in four patients, all known to have There was no clear-cut evidence of clinical or radiologic mutant p53 tumors. In the current study, PCR data indicat- response in any patient. Stable disease was demonstrated, ing that, generally, viral DNA can be detected up to 10 days with four patients receiving more than two cycles of dl1520 after the final (day 5) infusion of dl1520 (and much longer Downloaded from www.jco.org on August 17, 2006 . For personal use only. No other uses without permission. Copyright 2002 by the American Society of Clinical Oncology. All rights reserved. INTRAPERITONEAL ONYX-015 IN RECURRENT OVARIAN CANCER in one patient) provides evidence consistent with viral tion), when tumor masses were either microscopic or less infection and possibly replication. However, this is not than 2 mm in diameter, all treated mice were rendered conclusive proof of ongoing viral replication, because levels tumor-free and had significantly improved survival. In were generally lower than on day 5 and not enough is contrast, no abrogation of tumor growth was demonstrated known about the peritoneal clearance of dl1520. In addition, in mice treated with dl1520 at day ϩ31, ie, when tumor the relatively acute onset of the flu-like syndrome may also nodules were more than 2 mm in size. Such data strongly suggest that this is due merely to viral particles, as there suggest that unless penetration of dl1520 into tumor nodules may not have been enough time for replication to have taken can be improved, efficacy is likely to be limited to patients place. Furthermore, no conclusive evidence of viral repli- with low tumor burdens. Furthermore, adequate coverage of cation was demonstrated in cells distilled from the ascitic peritoneal surfaces by intraperitoneally administered fluids fluid of eight patients, although many samples were ob- is affected by adhesion formation, produced via an inflam- tained after a peritoneal wash; therefore, a dilutional effect matory reaction induced by both the initial surgery and is inevitable. For future studies, PCR for E4 or hexon some cytotoxics agents. Anticancer agents administered in protein mRNA expression could be used as other surrogate small fluid volumes are unlikely to achieve adequate intra- indicators for replication, as these studies may be more abdominal dispersion, even with multiple positional sensitive for detecting gene expression than immunohisto- changes as achieved in this study. Dilutions in 2 L of fluid chemistry for viral proteins or ISH. However, it should be have been advocated as the minimum volume required to noted that gene expression alone cannot be used as defini- achieve uniform coverage of all peritoneal surfaces.24 tive evidence for completed replication and release of As delivery of intraperitoneally administered anticancer agents is unlikely to penetrate bulky tumor nodules, vascu- Although in vivo studies have clearly demonstrated lar delivery to tumor centers may be important. Although efficacy and improved survival for mice bearing mutant p53 clearly not the planned primary modality of access to tumor peritoneal tumours,15 no clear-cut antitumor efficacy could in this study—the intention was to retain high concentra- be demonstrated in the current study. Tumor response was tions of dl1520 in the peritoneal cavity—this may have an not a primary objective of this study, but it is relevant to effect on efficacy. However, the reticuloendothelial cell discuss the factors potentially influencing the clinical activ- uptake of systemic virus particles is likely to abrogate ity of oncolytic viruses in patients with peritoneal carcino- distribution via the circulation to other metastatic tumor matosis. Despite the stated pharmacologic advantage for the sites. Intraperitoneal administration of dl1520 is unlikely to delivery of intraperitoneal chemotherapy, proof is still escape this effect, as solutes vacate the peritoneal cavity into lacking as to the clinical relevance of such data, and little is the systemic circulation either by diffusing through the known about the biokinetics of distribution for intravenous parietal/visceral peritoneum or by absorption through the and intraperitoneal adenovirus. Delivery of any drug to the lymphatics. The plasma-peritoneal barrier has been de- innermost core of tumor nodules depends on vascular scribed as having unidirectional transport characteristics, delivery rather than regional administration. Using rat with intraperitoneally administered substances appearing peritoneal tumor nodules, Los et al22 compared the concen- rapidly in the systemic circulation, whereas intravenous tration of cisplatin at the periphery of the tumor (Ͻ 1.5 mm administration of the same substance causes it to appear from tumor surface) with the concentration at the center of the tumor, after both intraperitoneal and intravenous admin- The presence of neutralizing antibodies, indicating an istration. Increased concentration of cisplatin was demon- active antiviral immune response, is likely to be a major strated at the periphery with intraperitoneal administration, factor in limiting the efficacy of oncolytic viruses such as but there was no difference in the tumor core. This sug- dl1520. Binding of antibody to virus will affect infection,26 gested that any major therapeutic benefit was likely to be and the development of virus-specific cytotoxic T lympho- restricted to small tumor nodules, and therefore surgical cytes could potentially result in the lysis of infected cells tumor debulking is likely to facilitate activity. In concor- before successful virus replication.27 However, Khuri et al28 dance with this, Heise et al23 hypothesized that the tumor treated head and neck cancer patients with intratumoral mass at the time of treatment might be an important dl1520 combined with systemic chemotherapy and found determinant of the antitumoral efficacy of dl1520. Two that the presence of baseline neutralizing antibodies did not mutant p53 xenografts, OVCAR3 and A2780/Cp70, were prevent tumor responses from occurring. One postulated treated with intraperitoneal dl1520 at three time points reason for maintaining the efficacy of intratumoral dl1520 following intraperitoneal inoculation with tumor cells. At injections is the inefficient penetration of solid tumors by the earlier time points (day ϩ3 and day ϩ17 after inocula- Downloaded from www.jco.org on August 17, 2006 . For personal use only. No other uses without permission. Copyright 2002 by the American Society of Clinical Oncology. All rights reserved. In vitro studies using nude mice engrafted with the temic exposure of virus. Local toxicity (abdominal pain/ human tumor xenograft model HlaC strongly support the peritonism) secondary to an inflammatory process was also combination of dl1520 with chemotherapeutic agents, par- common and troublesome in this patient population and ticularly cisplatin, and these data suggest that additive or correlates with the injection site pain reported by over 50% potentially synergistic effects can be demonstrated.19 This of the head and neck cancer patients in the previous studies.
significant antitumor activity for intratumoral dl1520 in In the head and neck cancer trial by Khuri et al,28 dl1520 combination with cisplatin and fluorouracil was confirmed replication was demonstrated within tumor tissue in four of in a phase II trial in squamous carcinoma of the head and six posttreatment biopsy specimens, with necrotic tumor neck,28 and interestingly, response rates were not related to tissue present in an additional three. In the current study, it p53 status. Furthermore, the sequence of agents in combi- was not possible to conclusively demonstrate that intraperi- nation may have therapeutic relevance.23 The mechanism(s) toneally administered dl1520 can gain access to and repli- for this dl1520/cisplatin interaction is unclear, but these cate in mutant p53 tumor cells. One patient had material preclinical and clinical trials suggest that dl1520 may be suggestive of ISH-positive malignant cells, but this was not able to sensitize both infected and uninfected cells to killing confirmed and therefore not classified as a true-positive by chemotherapy. E1A gene expression has been shown to result because of the poor quality of the cellular material increase cellular sensitivity to chemotherapy in a p53- obtained and questionable assay specificity. However, one independent manner.6,31,32 As dl1520 expresses E1A when must remember that intratumoral injections of dl1520 in the infecting both p53 wild-type and p53 dysfunctional tumors, head and neck clinical trials consisted of only a few this may account for this chemosensitization. In addition, milliliters of 0.9% saline containing up to 1011 pfu of virus.
the induction of apoptosis by platinum drugs is enhanced by It is therefore not surprising that comparable doses of virus wild-type p53 expression,33 and recent work has suggested diluted in 500 mL of saline within the peritoneal cavity that the therapeutic success of platinum and paclitaxel result in a much lower pickup rate for demonstrating active combinations (paclitaxel does not require the presence of viral replication. Furthermore, anatomic barriers to infectiv- functional p53 to induce apoptosis34) may reflect the effi- ity and spread (eg, adhesions) and the presence of neutral- cacy of the agents on different cellular populations with izing antibodies in the peritoneal fluid would further miti- different genetic backgrounds.35 It is relevant to note in this gate against viral replication capability and, consequently, context that two patients with platinum-resistant disease, our ability with currently available assays to demonstrate having developed progressive disease on dl1520 therapy, subsequently demonstrated falling CA125 after further car- In conclusion, replication-selective oncolytic adenovi- boplatin chemotherapy. One of these patients had a con- ruses such as dl1520 may offer a novel approach to the firmed CA125 response after three cycles, and although her treatment of ovarian cancer. Preclinical studies suggest that disease subsequently progressed, this supports the hypoth- they may be most effective when used in conjunction with esis that eradication or attenuation of a population of conventional cytotoxic agents, such as cisplatin or carbo- platinum-resistant, mutant p53-expressing clones by dl1520 platin. The optimal sequencing of these agents are being may have allowed further sensitivity to platinum.
examined in preclinical models, and further trials will be This article describes the first clinical experience with the required to carry these “proof of principle” experiments intraperitoneal delivery of any replication-competent/-selec- forward. The effect of antiviral immunity dl1520 antitumor tive virus in cancer patients. Multiple doses can be safely activity is still incompletely understood, and further inves- administered to patients after a mini-laparotomy, cytoreduc- tigation is needed in order to optimize treatment. Better tive surgery, and installation of a peritoneal catheter. The delivery of virus to intraperitoneal tumor nodules is required, main side effects were the flu-like syndrome, nausea/ particularly with respect to longer retention times within the vomiting, and abdominal pain/peritonitis. This latter symp- abdominal cavity. Modification of dl1520 and other second- tom was occasionally severe and reproducible, perhaps generation viral constructs will be necessary to enhance repli- signifying an inflammatory process initiated by viral infec- cation and virulence against the target tumor cell population; tion. The MTD, as defined, was not reached at 1011 pfu, and however, the systemic toxicities and abdominal pain associated at this dose level, patients with good performance status and with the intraperitoneal administration will need to be carefully nonbulky disease did not experience significant toxicity.
monitored. Highest efficacy may be seen dl1520 when admin- However, intraperitoneal administration of dl1520 was as- istered to patients with low tumor burden. Therefore, the ideal sociated with greater systemic toxicity, compared with that scenario for dl1520 therapy as a single agent may be as seen in the studies utilizing intratumoral injections in head consolidation treatment in patients with minimal residual and neck cancer.17,28 This is likely to reflect greater sys- disease following conventional chemotherapy. However, it Downloaded from www.jco.org on August 17, 2006 . For personal use only. No other uses without permission. Copyright 2002 by the American Society of Clinical Oncology. All rights reserved. INTRAPERITONEAL ONYX-015 IN RECURRENT OVARIAN CANCER will be important to observe evidence of clear-cut clinical efficacy or a biologic effect (eg, tumor cell infection) before The authors acknowledge the vital contribution of all the nurses, data proceeding with combination studies of intraperitoneal dl1520 managers, and hospital staff involved in this trial, without which this 1. Bray F, Sankila R, Ferlay J, et al: Estimates of cancer incidence and antitumoral efficacy that can be augmented by standard chemo- and mortality in Europe in 1995. Eur J Cancer 38:99-166, 2002 therapeutic agents. Nat Med 3:639-645, 1997 2. Conte PF, Cianci C, Gadducci A: Update in the management of 20. Rustin GJS, Nelstrop AE, McClean P, et al: Defining response advanced ovarian carcinoma. Crit Rev Oncol Hematol 32:49-58, 1999 of ovarian carcinoma to initial chemotherapy according to serum CA 3. Eisenhauer EA, Vermorken JB, Van Glabbeke M: Predictors of response to subsequent chemotherapy in platinum pre-treated ovarian 21. Smith R: Studies on the use of viruses in the treatment of cancer: A multivariate analysis of 704 patients. Ann Oncol 8:963-968, carcinoma of the cervix. Cancer 9:1211-1218, 1956 22. Los G, Mutsaers PHA, van der Vijgh WJF, et al: Direct 4. Debbas M, White E: Wild type p53 mediates apoptosis by E1A, diffusion of cis-diamminedichloroplatinum(II) in intraperitoneal rat which is inhibited by E1B. Genes Dev 7:546-554, 1993 tumours after intraperitoneal chemotherapy: A comparison with sys- 5. Grand RJA, Grant ML, Gallimore PH: Enhanced expression of temic chemotherapy. Cancer Res 49:3380-3384, 1989 p53 in human cells infected with mutant adenoviruses. Virology 23. Heise C, Lemmon M, Kirn D: Efficacy with a replication- selective adenovirus plus cisplatin-based chemotherapy: Dependence 6. Lowe SW, Ruley HE, Jacks T, et al: p53-dependent apoptosis on sequencing but not p53 functional status or route of administration.
modulates the cytotoxicity of anticancer agents. Cell 74:957-967, 1993 7. Kaye SB: Ovarian cancer, from the laboratory to the clinic: 24. Rosenshein N, Blake D, McIntyre PA, et al: The effect of Challenges for the future. Ann Oncol 7:9-13, 1996 volume on the distribution of substances instilled into the peritoneal 8. Yew PR, Kao CC, Berk AJ: Dissection of functional domains in the adenovirus 2 early 1B 55K polypeptide by suppressor-linker 25. Gross ML, Somani P, Ribner BS, et al: Ceftizoxime elimination insertional mutagenesis. Virology 179:795-805, 1990 kinetics in continuous ambulatory peritoneal dialysis. Clin Pharmacol 9. Kao CC, Yew PR, Berk AJ: Domains required for in vitro association between the cellular p53 and the adenovirus 2 E1B 55K 26. Zinkernagel RM: Immunology taught by viruses. Science 271: 10. Bischoff JR, Kirn DH, Williams A, et al: A mutant adenovirus 27. Yang Y, Nunes FA, Berencsi K, et al: Cellular immunity to viral which selectively replicates in tumour cells with non-functional p53.
antigens limits E1B-deleted adenoviruses for gene therapy. Proc Natl 11. Ganly I, Kim YT, Hann B, et al: Replication and cytolysis of an 28. Khuri FR, Nemunaitis J, Ganly I, et al: A controlled trial of E1B-attenuated adenovirus in drug resistant ovarian tumour cells is intratumoral ONYX-015, a selectively-replicating adenovirus in com- associated with reduced apoptosis. Gene Therapy (in press) bination with cisplatin and 5-fluorouracil in patients with recurrent 12. Hall AR, Dix BR, O’Carroll SJ, et al: p53-dependent cell head and neck cancer. Nat Med 6:879-885, 2000 death/apoptosis is required for a productive adenovirus infection. Nat 29. Baxter LT, Zhu H, Mackensen DG, et al: Physiologically based pharmacokinetic model for specific and non-specific monoclonal anti- 13. Harada J, Berk A: p53-independent and -dependent require- bodies and fragments in normal tissues and human tumour xenografts ments for E1B-55kD in adenovirus type 5 replication. J Virol 73:5333-5344, 1999 in nude mice. Cancer Res 54:1517-1528, 1994 14. Goodrum FD, Ornelles DA: p53 status does not determine 30. Shisler J, Duerksen HP, Hermiston TM, et al: Induction of outcome of E1B 55kilodalton mutant adenovirus lytic infection. J Virol susceptibility to tumour necrosis factor by E1A is dependent on binding to either p300 or p105-Rb and induction of DNA synthesis. J Virol 15. Heise C, Ganley I, Kim YT, et al: Efficacy of a replication- sensitive adenovirus against ovarian carcinomatosis is dependent on 31. Lowe SW, Bodis S, McClatchey A, et al: p53 status and the tumour burden, viral replication and p53 status. Gene Ther 7:1925- efficacy of cancer therapy in vivo. Science 266:807-810, 1994 32. Sanchez-Prieto R, Quintanilla M, Cano A, et al: Carcinoma cells 16. Boyle JO, Koch W, Hruban RH, et al: The incidence of p53 become sensitive to DNA-damaging agents by the expression of the mutations increases with progression of head and neck cancer. Cancer adenovirus E1A gene. Oncogene 13:1083-1092, 1996 33. Harris CC: Structure and function of the p53 tumour suppressor 17. Ganly I, Eckhardt SG, Rodriguez GI, et al: A phase I study of gene: Clues for rational cancer therapeutic strategies. J Natl Cancer Inst ONYX-015, and E1B attenuated adenovirus, administered intratumor- ally to patients with recurrent head and neck cancer. Clin Cancer Res 34. Vasey PA, Jones NA, Jenkins C, et al: Cisplatin, camptothecin and Taxol sensitivities of cells with p53-associated multidrug resis- 18. Vasey PA: Rationale for and complications of intraperitoneal chemotherapy. CME J Gynaecol Oncol 3:83-89, 1999 35. Lavarino C, Pilotti S, Oggionni M, et al: p53 gene status and 19. Heise C, Sampson-Johannes A, Williams A, et al: ONYX-015, response to platinum/paclitaxel-based chemotherapy in advanced ovar- an E1B gene-attenuated adenovirus, causes tumour-specific cytolysis ian carcinoma. J Clin Oncol 18:3936-3945, 2000 Downloaded from www.jco.org on August 17, 2006 . For personal use only. No other uses without permission. Copyright 2002 by the American Society of Clinical Oncology. All rights reserved.

Source: http://www.sunwaybio.com.cn/download/gw200204.pdf