¹Department of Obstetrics and Gynecology, Division of Gynecologic Oncology, The University of Iowa Hospitals and Clinics, Iowa City, Iowa, USA

²Department of Obstetrics and Gynecology, University of Freiburg, Freiburg, Germany

³University of California, Los Angeles, California, USA

⁴Schering-Plough Research Institute, Kenilworth, New Jersey, USA

⁵University of Ulm, Ulm, Germany

Correspondence to: Dr Richard E Buller, Department of Obstetrics and Gynecology, Division of Gynecologic Oncology, 200 Hawkins Drive ¾ #4630 JCP, Iowa City, IA 52242-1009, USA. E-mail: richard-buller@uiowa.edu

^†Presented in part at the 7th International Conference on Gene Therapy of Cancer, San Diego, CA, November 19-21, 1998 and the 30th Annual Meeting of the Society of Gynecologic Oncologists, San Francisco, CA March 20-24, 1999.

Purpose: To determine the safety, gene transfer, host immune response, and pharmacokinetics of a replication-deficient adenovirus encoding human, recombinant, wild-type p53 (SCH 58500) delivered into the peritoneal cavity (i.p.) alone and sequentially in combination with platinum-based chemotherapy, of patients with recurrent ovarian, primary peritoneal, or fallopian tube cancer containing aberrant or mutant p53. Methods: SCH 58500 was administered i.p. to three groups of patients with heavily pretreated recurrent disease. Group 1 (n=17) received a single dose of SCH 58500 escalated from 7.5´10¹⁰ to 7.5´10¹² particles. Group 2 (n=9) received two or three doses of SCH 58500 given alone for one cycle, and then with chemotherapy for two cycles. The SCH 58500 dose was further escalated to 2.5´10¹³ particles/dose in group 2. A third group (n=15) received a 5-day regimen of SCH 58500 given at 7.5´10¹³ particles/dose per day i.p. alone for cycle 1 and then with intravenous carboplatin/paclitaxel chemotherapy for cycles 2 and 3. Results: No dose-limiting toxicity resulted from the delivery of 236/287 (82.2%) planned doses of SCH 58500. Fever, hypotension abdominal complaints, nausea, and vomiting were the most common adverse events. Vector-specific transgene expression in tumor was documented by RT-PCR in cells from both ascitic fluid and tissue biopsies. Despite marked increases in serum adenoviral antibody titers, transgene expression was measurable in 17 of 20 samples obtained after two or three cycles of SCH 58500. Vector was detectable in peritoneal fluid by 24 hours and persisted for as long as 7 days whereas none was detected in urine or stool. There was poor correlation between CT scans and CA125 responses. CA125 responses, defined as a greater than 50% decrement in serum CA125 from baseline, were documented in 8 of 16 women who completed three cycles of the multidose regimen. Conclusion: CT scans are not a valid measure of response to i.p. SCH 58500 due to extensive adenoviral-induced inflammatory changes. Intraperitoneal SCH 58500 is safe, well tolerated, and combined with platinum-based chemotherapy can be associated with a significant reduction of serum CA125 in heavily pretreated patients with recurrent ovarian, primary peritoneal, or fallopian tube cancer. Cancer Gene Therapy (2002) 9, 553-566 doi:10.1038/sj.cgt.7700472

p53; gene therapy; ovarian cancer; CA125

It is projected that 23,400 women will be diagnosed with ovarian cancer and 13,900 will die from this disease during the year 2001.¹ These statistics make ovarian cancer the fifth leading cause of death among women in the United States. Ovarian cancer offers several unique opportunities for novel therapeutic intervention. First, despite the tendency to present at advanced International Federation of Gynecology and Obstetrics (FIGO) stage reflected by the observation that nearly 73% of ovarian cancers are no longer confined to the ovary at diagnosis,² this cancer often remains confined within the abdominal cavity throughout its course.^3,4 Second, initial, complete clinical responses are the expected norm following surgical cytoreduction and adjuvant systemic chemotherapy. Unfortunately, recurrence, progression, and death from disease is the eventual outcome for more than 75% of women diagnosed with epithelial ovarian cancer. Finally, because both primary^5,6 and secondary^7,8 surgical cytoreduction are cornerstones of the therapeutic approach to this cancer, tissue samples are often available for molecular genetic studies. Such studies have resulted in a better understanding of some of the molecular changes associated with ovarian cancer and how they may influence prognosis or response to treatment.

Mutation of the p53 tumor suppressor gene is one of the most frequent molecular genetic changes in cancer.^9,10 Wild-type p53 functions include roles in DNA repair following G1 cell cycle arrest, and directing irreparably damaged cells toward apoptotic pathways, thus maintaining the integrity of the genome.¹¹ Both in vitro and in vivo evidence suggests that cells with altered p53 function may be less responsive to certain chemotherapeutics than those that are able to express wild-type p53.^12,13 p53 dysfunction frequently results from mutations that can generate both missense and nonsense inactivating mutations. Rare gain of function mutations has also been described.¹⁴ Nearly 70% of advanced stage ovarian cancers contain p53 mutations and many of these mutations render the cancers p53 null.^15,16,17 Overall, p53 null mutations can be associated with extremely poor prognosis reflected, at least in part, by early and distant metastasis.⁴ These observations suggest that p53 mutation is of fundamental importance in the progression of ovarian cancer.

Despite the association of distant metastasis with p53 null mutation, most ovarian cancers usually remain confined to the abdominal cavity throughout their course and provide a unique opportunity for regional delivery of therapeutic agents. Intraperitoneal delivery of chemotherapy can provide a pharmacokinetic advantage over intravenous dosing by maximizing delivery of drug directly to tumor and minimizing systemic side effects.¹⁸ A seminal study by the Gynecologic Oncology Group has demonstrated both response and survival advantage to women with minimal residual disease treated with intraperitoneal (i.p.) cisplatin after primary cytoreductive surgery for ovarian cancer.¹⁹ Thus, ovarian cancer is a unique model for gene replacement strategies.^20,21

Preclinical studies in several in vivo models have shown that delivery of wild-type p53 to tumor cells can be achieved.^{22,23,24,25,26,27,28,29,30,31} Extension of these studies, particularly in lung cancer, to phase I clinical trials has produced encouraging results.^{32,33,34,35,36} To date, gene transfer in these systems has been accomplished with cationic lipids and a variety of viral vectors including the retroviruses and adenoviruses.³⁷ The use of an adenoviral vector, which has been rendered replication deficient, offers several advantages for therapeutic gene replacement strategies in cancer.³⁷ First, in contrast to retroviruses, adenoviral vectors efficiently transduce both dividing and quiescent cells. Second, they can be produced in high titers with particle numbers approaching the number of target cancer cells. Third, a bystander effect has been observed to occur following dosing with adenoviral vectors. Fourth, adenoviral vectors do not integrate into the host genome minimizing concerns regarding insertional mutagenesis. Taken together, these observations encouraged us to undertake a phase I/II trial of human recombinant adenoviral p53 gene therapy with rAd/p53 (SCH 58500) in recurrent ovarian cancer. Preliminary results have been presented in part.³⁸ The objectives of the study were: (a) to determine safety and tolerability to SCH 58500 alone and in combination with chemotherapy, (b) to determine the ability to transfer wild-type p53 sequences into ovarian cancer cells in vivo, (c) to measure serum and ascites antibody responses to this form of therapy along with their influence on gene transfer, (d) to determine the pharmacokinetics of SCH 58500 in ascites and serum, and (e) to evaluate tumor response when multiple doses are delivered to patients over a 3-month period. Our findings indicate that gene transfer of SCH 58500 can be accomplished with minimal toxicity and that reduction in a surrogate marker, CA125, suggests the potential for clinical activity.

Methods

SCH 58500

SCH 58500 is a novel antineoplastic agent consisting of a recombinant adenoviral vector containing the cloned, human, wild-type p53 tumor suppressor gene cDNA, which is under the control of the human cytomegalovirus immediate early promoter/enhancer element. SCH 58500 is derived from a type 5 adenovirus, a common serotype belonging to subgroup C, which has been rendered replication-defective through deletion of the viral genes E1a, E1b, and protein IX.³⁹ Vector is produced using GMP standards and has been tested for the presence of viral, bacterial, and other contaminants.

Tumor p53 mutation status

For screening, a representative primary or recurrent tumor sample from each patient who had signed informed consent was analyzed for p53 mutation by immunohistochemistry utilizing both Pab 1801 (diluted 1:40) and Pab 240 (diluted 1:20) antibodies (Pharmingen, San Diego, CA). A positive stain with either antibody was considered to reflect aberrant tumor p53 protein and confirmed eligibility. Although this finding does not always reflect a p53 mutation, most authors consider immunopositive tumor to contain dysfunctional p53.¹⁶ Sections with <10% of cells showing nuclear staining were considered negative. Such individuals were excluded from study entry unless a p53 DNA sequence abnormality could be documented.¹⁶

Antiadenovirus antibody assay

An ELISA was used to measure antiadenovirus antibodies specific for adenoviral coat proteins (antihexon antibodies) in serum and ascites. Samples were assayed in parallel with normal human serum and a ratio of sample titer versus normal human serum titer was calculated. If this ratio was less than 0.28 the sample was considered negative.

Patient eligibility and exclusion criteria

Only female patients at least 18 years of age previously treated with surgery and chemotherapy for ovarian, fallopian tube, or primary peritoneal carcinoma now presenting with pathologically confirmed recurrence of disease were eligible. An elevated CA125 was not required for entry. For those individuals without malignant ascites at recurrence, we required surgically documented i.p. disease accessible to laparoscopic or percutaneous biopsy. A tumor p53 mutation was required as described above. All treated individuals functioned with a Karnofsky performance status of at least 60% and a minimum life expectancy of 3 months. Standardized clinical laboratory tests were within normal limits. Previous whole abdominal radiotherapy was not allowed. Before the first treatment cycle a contrast study of the abdomen demonstrated free flow of instilled agent. Either a spiral CT with i.p. contrast, or i.p. Hypaque (Nycomed, Princeton, NJ) in 500 mL of normal saline followed in 30minutes with a conventional flat plate x-ray was used to determine adequate peritoneal distribution of the infuseate. Three eligible, consented patients did not receive treatment with SCH 58500 based on poor distribution of contrast. Initially, only patients serologically positive for antiadenovirus type 5 antibody at screening were treated.

Patients not eligible for the study included those pregnant or nursing, and those with presence of serious bacterial, viral, fungal, or parasitic infection. Patients with evidence of adenoviral infection, as determined by ELISA, at screening were excluded and the chronic use of immunosuppressant therapy or use of another investigational drug within 3 months of proposed treatment with SCH 58500 also resulted in exclusion. Known human immunodeficiency virus (HIV)-positive individuals were also excluded. Short-term bolus use of dexamethasone as an antiemetic, or as premedication for paclitaxel, was allowed.

Registration

An institutional human subjects review board approved informed consent was obtained before the performance of any test or evaluation not considered standard of care for patients with peritoneal carcinomatosis. The same consent detailed the treatment with SCH 58500 and alternatives. No patient received SCH 58500 without signing an informed consent.

SCH 58500 delivery

SCH 58500 was infused over 20 minutes into the peritoneal cavity via a Hickman (Bard Systems, Salt Lake City, UT), Tenckhoff (CR Bard, Murray Hill, NJ), or Porta Cath (SIMS Deltec, St. Paul, MN) catheter. In preliminary studies, all catheters were shown to be compatible with SCH 58500. The goals of the infusions were to use a constant volume for each dose, with the volume large enough to generate adequate i.p. distribution, while at the same time providing a tolerable total volume. To achieve these goals, some variability of infusion volumes was required. Patients with clinically significant, preexisting ascites underwent drainage of the ascites before dosing with SCH 58500. Patients in group 1then received SCH 58500 in 1000 mL of 0.9% NaCl. Group 2 and 3 patients received SCH 58500 in 250 mL, for 2 (Level 4), 3 (Level 5), and 5 (Level 6) days. By the end of five daily administrations (i.e., level 6), a total infusion volume of 1250 mL had been reached. Any additional ascites that accumulated during the course of administration of multiple doses was not removed except in one patient who had a large volume of ascites with her recurrent disease. Following this patient's first dose in cycle 1, the day 2 dose was delayed 24 hours to allow for ascites drainage. In the absence of ascites, each dose of SCH 58500 was infused in 500 mL of 0.9% NaCl so that by the end of five daily administrations (i.e., level 6), a total infusion volume of 2500 mL had been reached. Patients were then rotated every 15 minutes for 2 hours into Trendelenberg, right lateral, left lateral, and sitting positions.

Treatments

This sequential cohort, nonrandomized study, was conducted in three groups of patients. Table 1 outlines the treatment schema for i.p. SCH 58500. For group 1 patients, a single treatment dose of SCH 58500 was escalated from 7.5´10¹⁰ particles to 7.5´10¹² particles per dose in four steps. Three patients were to be treated with SCH 58500 at each dose level in this group. The decision to escalate or expand a dose level was based on review of safety data for the patients within the single-dose level cohort under study or after day 7 of the first cycle when multiple cycles were given. A single, potential dose-limiting toxicity (DLT; see Results) prompted us to expand level 2 from three to six patients. After initial safety data were obtained, two additional antiadenoviral antibody negative individuals were allowed to enter at level 1. Therefore, a total of 17 patients were treated in group 1. Patients treated in this group were allowed to enter the multiple-dose group (see below) if they continued to meet all eligibility criteria.

For group 2 patients (n=9), cytotoxic chemotherapy was added in cycles 2 and 3 to allow differentiation between SCH 58500 side effects when it was given alone in cycle 1 and those related to its combination with chemotherapy. The doseof SCH 58500 was further escalated to 2.5´10¹³ particles although single-day dosing was increased first to 2 and then to 3 days per treatment cycle. Six group 2 patients received single-agent i.p. cisplatin at 100 mg/m² on day 1 of cycles 2 and 3 for dose levels 4 and 5 only. A 30-minute infusion of cisplatin was delivered in 1 liter of 0.9% NaCl 1hour following the SCH 58500 infusion. The rest of the multiple-cycle patients were treated with intravenous chemotherapy. Paclitaxel at 175 mg/m² was infused over 3hours immediately before SCH 58500 on day 1 whereas carboplatin was infused over 30 minutes immediately after SCH 58500 on day 3 of cycles 2 and 3 at dose levels 5 and 6. The carboplatin dose was based on an area under the curve (AUC) of 6 mg/mL min with the GFR based on the Cockroff-Gault formula for creatinine clearance.⁴⁰ With multiple-dose, multiple-cycle regimens, paclitaxel was before the vector because of in vitro evidence that this agent enhances transfection efficiency of SCH 58500.²³

Once safety was confirmed by interpatient escalation, group 3 patients (n=15) were treated with intravenous carboplatin and paclitaxel in combination with SCH 58500 at 7.5´10¹³ particles, dose level 6. The number of doses of SCH 58500 was escalated from 3 to 5 per cycle. For patients in this group, either measurable or evaluable disease was required. Measurable disease was defined as a bidirectionally measurable lesion with clearly defined margins on physical exam or x-ray, computed tomography (CT), or magnetic resonance image analysis. Evaluable disease was defined as an elevated CA125 tumor antigen level greater than two times the institutional norm.

Tumor sampling: Twenty-four to 72 hours following single-dose SCH 58500, or 24 hours after the last dose of SCH 58500 in each multiple dose of the study agent, the peritoneal cavity was drained to obtain tumor cells. Patients who did not have ascites with recurrence of their cancer, or who had inadequate ascites following SCH 58500 dosing, were separately consented to laparoscopy for cycles 1 and 3 to obtain tumor and normal tissue samples for the various PCR studies. Pathologic, or cytologic, examination confirmed the presence of malignant cells in the samples of all patients.

Toxicity: The study design with escalating doses of SCH 58500 was aimed to determine dose-limiting toxicities utilizing standard WHO criteria. Any grade 4 (G1; WHO) toxicity, or a grade 3 (G3) toxicity lasting greater than 1week, was to be considered dose limiting (DLT), unless the event was obviously related to another procedure (e.g., anemia due to chronic test phlebotomy).

Nausea, vomiting, and anorexia were excluded as dose-limiting toxicities in patients receiving chemotherapy.

Patient monitoring: All single-dose patients were treated as inpatients. A qualitative ELISA kit (Cambridge Biotech, Worcester, MA) was used to confirm the absence of viral shedding in urine and stool samples before dosing, during treatment, and before hospital discharge. Vital signs were obtained before and periodically following the administration of SCH 58500. Physical exams, performance status, weight, and adverse event assessments were performed daily whereas the patients were hospitalized and at prescribed intervals following discharge: Day 7, 14, 21, and 2 months after dosing, then every 3 months until death. Laboratory data included serum and ascites sampling for pharmacokinetic studies, complete blood counts (CBC), fibrinogen, fibrin split products, PT, PTT, serum C₃, C₄, CH₅₀, electrolytes including magnesium, blood glucose, and CA125. Laboratory tests were performed at each visit, except CA125, which was monthly. Baseline abdominal and pelvic computed tomograms were obtained along with a chest x-ray before dosing. Follow-up scans were obtained at 28 days and as clinically indicated for patients who received multiple cycles of SCH 58500. Lesions were measured in two perpendicular directions. Standard response definitions were used, i.e., complete response (CR) required the disappearance of all gross evidence of disease for at least 4weeks; partial response (PR) required a reduction in lesion size in excess of 50% lasting at least 4 weeks; progressive disease was said to have occurred on the basis of a 25% increase in lesion size; all other measurable disease cases were considered to define stable disease (SD). As an additional measure of response, changes in serum CA125 were evaluated. The 50% and 75% CA125 responses as defined by Rustin et al^41,42 have been shown to correlate well with conventional CT response measures.

Documentation of gene transfer and viral persistence: Total RNA was extracted and cDNA prepared from ascites or tissue biopsies. The QIAamp 96 Spin Blood Kit (Qiagen, Valencia, CA) was used to extract viral DNA from serum. Polymerase chain reactions were carried out utilizing primers specific for both the adenovirus and the p53 gene as well as -actin or glyceraldehyde 3-phosphate dehydrogenase (G3PD) collectively referred to as housekeeping genes or HKG. The MIMICÔ (Clontech, Palo Alto, CA) reverse transcriptase technique allowed for semiquantitative comparisons of mRNA levels. Tripartite leader sequence-specific primers permitted the resolution of SCH 58500 sequence from host p53 sequence (See Results).

In situ PCR: Five-micrometer sections of formalin-fixed, paraffin-embedded tissue were placed on 1.2-mm silane-coated Perkin-Elmer (Foster City, CA) in situ PCR glass slides. Slides were baked 2-3 hours at 60°C to reduce RNA content. Slides were then treated sequentially with 0.02 N HCl, Proteinase K, and acetic acid. Thirty-five PCR cycles were carried out using dinitrophenyl-labeled primers (DNP)specific for SCH 58500. Following incubation with anti-DNP antibody conjugated to alkaline phosphatase, visualization was achieved by adding nitro-blue tetrazolium-5-bromo-4-chloro-3-indolyl phosphate as substrate and counterstaining with Nuclear Fast Red. Negative staining was pink, whereas positive staining was blue and nuclear.

Statistics: Differences in toxicities between SCH 58500 and SCH 58500 plus chemotherapy cycles were evaluated with Fisher's Exact test, 2-tailed. Mean CA125 changes were analyzed by ANOVA, or t tests as appropriate. A P value of <.05 was required for significance.

Results

Patient selection and characteristics

One hundred and fifty-five patients signed informed consent and entered into screening at three sites. Overexpression of p53 protein by immunohistochemistry was demonstrated for 79 of the 155 (51%) cancers tested. The Iowa site carried out p53 gene sequencing on 25 of 28 p53 immunonegative cancers screened at that institution. An additional 8 patients (7 with p53 null mutations) met the p53 eligibility criteria in this fashion. So that 57% (88/155) of patients screened were eligible for entry. Overall, 36 patients were dosed with SCH 58500. Five patients were treated on both the single-dose arm of the study and the later multiple-dose program. Therefore, 41% (36/88) of the eligible patients representing 23% (36/155) of the patients screened were actually treated with SCH 58500. The mean age of the individuals dosed was 60 years (range: 39-76). Demographic and disease-related parameters for this cohort of heavily pretreated individuals with recurrent peritoneal carcinomatosis are summarized in Table 2. Most individuals had recurrent ovarian cancer. The mean interval from primary diagnosis to dosing with SCH 58500 was 778 days (range: 115-2360 days). The mean platinum-free interval to dosing with SCH 58500 was 263days (range: 37-711 days). Nine of 36 patients had a platinum-free interval of more than one year. The mean number of prior chemotherapy regimens was 2.8 with 22% of individuals receiving four or more prior regimens and 33% receiving just one prior regimen. All individuals had previously received platinum-based chemotherapy, and all but two had prior treatment with a taxane. Three patients had recurrent disease evaluable only on the basis of laparoscopic biopsy. Fourteen patients could be considered to have small-volume disease, arbitrarily defined as less than or equal to 2 cm.

Toxicity

Two hundred and thirty-six different signs or symptoms were recorded as adverse events. These varied from single patient WHO grade 1 (G1) events such as increased earwax and nonspecific breast complaints to a G4 transient ischemic attack. Because adenoviral particles delivered in this study are more than a log higher than in any previously reported gene therapy trial, great attention was paid to complete reporting of all potential adverse events. From a practical standpoint, we have chosen to present all serious G3 or G4 events, but only the G1 and grade 2 (G2) events that occurred in three or more treated individuals. This of course underreports the total number of minor adverse events. Each treatment-related adverse event is recorded as the highest-grade toxicity experienced out of all treatment cycles received by that patient as explained in the legend to Table 3. The events are listed in this table on the basis of occurrence in either the single-dose or multiple-dose groups. To show that there was no cumulative toxicity, we have listed the five patients who were treated with multiple-dose SCH 58500 after completion of the single-dose portion of the study separately. The most common adverse events in the single-dose group included fever (47%), nausea (41%), edema (41%), abdominal complaints (41%), and anemia (29%). Seven patients experienced 5 different adverse events whereas only 1 patient had no adverse events at all. Eight G3 or G4 adverse events were reported in four patients. These included anemia (2:G3; 1:G4), abdominal complaints (1:G3), dehydration (1:G3), pain (1:G3), tachycardia (1:G3), and vomiting (1:G3). There was no unusual toxicity in the two serum antiadenoviral antibody negative patients dosed at level 1.

Fever was also the most common adverse event experienced by 100% of the multiple-dose patients. This sign developed within 2 to 4 hours of dosing. The highest reported temperature was 40.5°C. Four cycles were accompanied by G3 febrile responses (>40°C) among two different patients. After fever was noted in the initial dosing cohorts, patients were generally given prophylactic acetaminophen. The subsequent febrile responses were attenuated, but this may also have been due to the steroid premedication given before chemotherapy for cycles 2 and 3. Figure 1 demonstrates this observation graphically for a patient treated at dose level 5. In the multidose cohort, the next most frequent signs and symptoms related to SCH 58500 included hypotension (89%), a variety of abdominal complaints (79%), hypertension (68%), nausea (63%), tachycardia (58%), vomiting (58%), and fatigue (53%) ¾ often in the same patient and cycle as the hypertension was noted. The blood pressure changes prevalent in this group were not seen at all in the single-dose group, but they were generally considered mild because only one G3 toxicity occurred. All of these most common adverse events, except hypertension, also occurred in 100% of the single-dose patients who reenrolled in the multiple-dose regimen. However, there was no progression of toxicity grade in those re-treated relative to those initially treated at the same dose of SCH 58500.

Forty-seven G3 or G4 toxicities were reported in 13patients who received multiple-dose SCH 58500. Many of the new WHO G3 toxicities were probably related to chemotherapy because they usually appeared in cycles 2 and3. The patient with congestive heart failure also developed a G4 neutropenia with concomitant thrombocytopenia in cycle 3. A few new low-grade adverse events were reported when chemotherapy was combined with SCH 58500. These included lower extremity myalgias, myoclonus, ileus, gastritis with hematemesis, hyperactive bowel sounds, pulmonary hypertension, peripheral neuropathy, oliguria, mucositis, port site cellulitis, agitation, generalized weakness, and cachexia. Only three multiple-dose patients had £3 adverse events whereas 11 reported 10 adverse events. Overall, G3 toxicities accompanied approximately one-third of the treatment cycles. Antiemetics generally alleviated the gastrointestinal symptoms and were used prophylactically at the investigator's discretion. As the total amount of SCH 58500 delivered was increased, there was a trend toward more G3 adverse events: the number of adverse events went from 0.5 to 1.6 to 2.9 per patient as the treatment was advanced from single dose to level 4/5 and then to level 6. The addition of chemotherapy produced additional nausea and vomiting (P=.03, Fisher's exact test, 2-tailed). There was no trend for adverse events to worsen in a given individual as the number of doses delivered was increased. Likewise, there was no evidence of cumulative toxicity as patients progressed from the single-dose arm to treatment with multiple doses and multiple cycles.

One G4 toxicity occurred in a patient who became anemic in cycle 2. This complication along with the other G3 toxicities due to anemia occurred in individuals who were anemic at the start of the study and has been attributed to the volume of blood drawn for the multiple laboratory studies, anemia of chronic disease, and anemia secondary to chemotherapy treatments. There was no evidence of hemolysis in any patients. One individual with liver metastasis and progressive disease following single-dose SCH 58500 at level 2 developed a potential DLT reflected by an increase in alkaline phosphatase from 57 U/L at baseline to 742 U/L 28 days following dosing. This was accompanied by an increase in AST to 106 U/L and ALT to 111 U/L. She refused a follow-up CT scan and died 51 days after dosing. The family declined a request for an autopsy. Because of this adverse event, three additional patients were treated at this dose level before moving on to level 3. Therefore, although one cannot rule out SCH 58500 as a cause of this potential DLT, the investigator felt that the clinical course of this patient was quite consistent with progression of disease as the proximal cause of these events. Supporting this conclusion was the additional observation that no other patient, at any dose, developed evidence of G3 or G4 hepatic toxicity. Five individuals (14%) developed potentially worrisome small bowel obstructions between 2 and 8 months after initial dosing. These events occurred both on the single-dose (n=2) and multiple-dose (n=3) arms. Only one episode was attributed to SCH 58500, rather than to disease progression and/or underlying adhesions. All five resolved with conservative nonsurgical management. Two i.p. catheter-related infections also complicated treatment and led to patient removal before completion of the anticipated number of cycles. Both were associated with abdominal Hickman (Bard Systems, Salt Lake City, UT) catheters. One of these individuals developed vancomycin-resistant enterococcal peritonitis. She was found to be a nasal carrier of this organism. Another individual developed a sterile pelvic abscess. Both patients received only five doses of SCH 58500 alone before withdrawal from the multiple-dose arm. Overall, 82.2% of the planned doses of SCH 58500 were delivered. In addition to the catheter problems outlined above, failure to complete the planned number of cycles of chemotherapy plus SCH 58500 resulted from disease progression (two patients), side effects (one patient), and a withdrawn consent (two patients).

Pharmacokinetics: SCH 58500-specific PCR was carried out on serum samples of all patients during cycle 1. Samples were obtained pretreatment at 15, 30 minutes, 1, 2, 4, 6, 12, 24, 36, 48, and 72 hours; and days 7, 14, 21, and 28 following administration of SCH 58500. Detectable serum levels of SCH 58500 were found in seven patients. In four of these, the levels were detectable but not quantifiable. Only one patient had a quantifiable level after 24 hours. There was no vector shedding in either urine or stool of any patient as determined by ELISA assay. One patient underwent a therapeutic thoracentesis 72 hours after dosing with SCH 58500 on the single-dose arm. The pleural fluid was positive for vector by ELISA. Patient peritoneal fluid analysis consistently demonstrated the presence of viral DNA for 24 hours. For a subset of three patients, viral DNA was detected on day 6 for one patient and day 7 for two. ELISA positive peritoneal fluid was noted for periods in excess of 1 year following the last dose of SCH 58500. However, we were unable to culture live virus or demonstrate infectivity by the FACS assay⁴³ from the prolonged ELISA positive fluid.

Tumor sampling

Following cycle 1, 22 patients had ascites sampled, 5 had a laparoscopic biopsy, and 11 had both. Only ascitic fluid was sampled after cycle 2. Following cycle 3, 8 patients had ascitic fluid sampled, 5 underwent laparoscopic biopsy, and 1 had both procedures.

Determination of gene transfer

The unique tripartite leader sequence incorporated into the recombinant p53 gene sequence allowed us to differentiate mRNA expression due to transduced p53 gene from any host wild-type p53 mRNA co-isolated from contaminating normal cells. Figure 2 shows a gel containing both sample and MIMICÔ PCR reaction that demonstrates this principle. The equivalence of band intensities in lane 7 at a 1:4 dilution of template cDNA allows for the calculation of the number of molecules of p53 mRNA isolated from the sample normalized for the sample -actin message content. In this case 1.6 molecules of p53 transgene mRNA per 1000 molecules of -actin message were detected. Similar studies were carried out using mRNA isolated from cells separated from ascites or from biopsies obtained at the indicated times and cycles for all patients treated. Figure 3 summarizes these results. Transgene expression was seen at doses as low as 7.5´10¹⁰ particles and consistently at or above 7.5´10¹¹ particles per dose. Three samples were negative for -actin and were excluded from this analysis. In two cases samples thawed during shipment. In another patient, a tumor biopsy obtained at day 3 was negative; however, her ascites was positive at day 7. Overall, transgene expression at the RNA level occurred in 3 of 5, 4 of 4, 3 of 3, 8 of 11, 9 of 11, and 25 of 28 samples analyzed for SCH 58500 doses of 7.5´10¹⁰, 7.5´10¹¹, 2.5´10¹², 7.5´10¹², 2.5´10¹³, and 7.5´10¹³ particles per dose, respectively. The most significant observation from this analysis is that transgene expression was detectable in 17 of 20 (85%) samples following multiple dosing with SCH 58500.

Demonstration of vector-encoded DNA in tumor target cells

The RT-PCR transgene expression data presented above were generated from ascitic fluid cell pellets or tissue biopsies. Such samples may contain normal cells as well as tumor cells. Thus, whereas we have clearly demonstrated transgene expression in our biopsy and ascitic fluid samples, we have not demonstrated the presence of either agent or transgene product from within tumor cells. To achieve this goal, in situ PCR was carried out on sequential tissue samples from a single patient. The primers used were specific for SCH 58500. Figure 4A shows a sample obtained before dosing with SCH 58500. The pink stain indicates the absence of viral DNA. This contrasts with the blue nuclear stain of the sample shown in Figure 4B obtained after three cycles of SCH 58500. A negative control is shown in Figure 4C wherein Taq polymerase was omitted from the reaction. These results clearly demonstrate the presence of viral DNA within tumor cells. Finally, Figure 4D shows a hematoxylin and eosin stained section corresponding to the tissue sample in panels B and C. In this figure, apoptotic bodies and dying tumor cells are readily differentiated from healthy tumor cells deeper within the biopsy.

Antiadenoviral antibody response

Baseline serum antiadenoviral antibody titers ranged from 1:160 to 1:16,000 before the first dose of SCH 58500. A 2-fold rise in titer could be seen by day 3 following i.p. SCH 58500. Increases in titer on day 28 ranged from 2- to 1600-fold over screening values. For patients enrolled in the multiple-dose regimens, or those re-treated with SCH 58500, a transient decrease in antibody titer on the order of 2- to 4-fold was sometimes seen. Twelve- to fifty-fold increases over the baseline titer were observed for up to 11 months following a single dose of SCH 58500. With multiple dosing, continued increases in titer were measured to as high as 1:2,560,000. An immune response was documented in one of the two individuals treated at level 1 who entered with negative titers. There was no apparent correlation between change in antibody titer and alterations in CA125 levels (see below). Likewise, there was no correlation between the dose of SCH 58500 and the mean change in antibody titer or the mean change in antibody titer with transgene expression (data not shown).

Measures of response

Table 4 compares conventional CT response determinations to the change in CA125 measured from study entry to study exit. Three tumors, all in group 1, were CA125 negative (<35 U/dL) at the time of enrollment. Although all three had biopsy-proven recurrence of disease, none had CT-measurable disease either. Six other individuals with elevated CA125 levels did not have CT-measurable disease at study entry. Two individuals without CT-measurable disease were treated in both group 1 and group 2. There were no CR or PRs documented by CT scan. On the contrary, the best CT responses were four cases of SD, three from group 1 and one from group 2. The most striking feature of the follow up CT scans was the frequency that disease progression was called on the basis of the development of new lesions ¾ documented in 18 treatment regimens. For nine of these cases, apparent disease progression was accompanied by at least a 26% decrease in CA125 from baseline. In several of these cases, apparent CT progression was found at laparoscopy to represent a pocket of inflammatory cells. Five of nine patients treated in groups 2 and 3 with purported CT disease progression demonstrated at least a 50% CA125 response. In contrast, for six of nine group 1 patients, the development of new CT lesions was accompanied by at least a 25% increase in CA125 disease. Together these observations are consistent with the hypothesis that the new CT lesions often occurred due to SCH 58500-induced inflammatory changes rather than disease progression. This conclusion prompted us to carry out a more detailed analysis of the response to SCH 58500 on the basis of the associated CA125 change from baseline.

Serum CA125 levels were measured immediately before dosing with SCH 58500 and following each treatment cycle. Two of the three patients with baseline CA125s <35 U/dL more than doubled their CA125 during study. CA125 could thus be considered a valid response parameter for all but a single patient. In addition, it is clear that the inflammatory response initiated by SCH 58500 did not uniformly give rise to an increase in CA125 by itself. The percent change in CA125 was then calculated for each individual for each treatment cycle, and overall at 28 days after study completion. Comparison of CA125 levels following treatment with SCH 58500 alone at £2.5´10¹¹ particles/dose to treatment at 2.5´10¹² particles/dose demonstrated a mean increase in serum CA125 of 94% versus a mean decrease of8.5% at the higher dose (P=.07, 2-tailed, unequal variances). Thus, SCH 58500 alone at higher doses provides a favorable change in CA125 not seen at lower doses of vector. Figure 5 summarizes the CA125 response data. Dosing with SCH 58500 alone resulted in a mean decrease in CA125 of 33.6% for the 16 of 41 women whose CA125 levels declined during the 28 days following dosing. These declines ranged from 4% to 77%. One additional patient's CA125 was unchanged at 46 U/mL. In contrast, with the addition of chemotherapy for cycle 2, 15 of 18 women demonstrated a mean decrement of 47.7% in their pre-cycle 2 CA125 levels. This result indicates enhanced CA125 response over treatment with SCH 58500 alone. For cycle 3, 11 of 16 of treated women showed a mean decline in CA125 levels of 36.6% compared to cycle 2 day 28 CA125. Thus, a continued response was seen in excess of that seen with SCH 58500 alone. Overall, 2 of 14 women demonstrated a 50% or greater decline in CA125 following a single dose of SCH 58500. For the 16 women who completed all three multiple-dose cycles, 8 registered a CA125 decline 54% from study entry. Among all responders, the average decline in CA125 was 60.1% (P=.06 vs SCH 58500 alone). Favorable changes in CA125 levels were independent of the time interval from initial diagnosis to SCH 58500 dosing. Likewise, the number of prior treatment cycles and regimens did not preclude a CA125 response and there was no apparent relationship between transgene expression and CA125 response.

Discussion

Because p53 tumor suppressor gene dysfunction is seen in 50-60% of all human malignancies,^9,10 this gene has become a leading candidate for clinical studies involving gene transfer technology for the treatment of cancer. Preclinical studies utilizing a variety of cell lines have shown efficient transduction, cell cycle arrest, apoptosis, and enhanced cell death following treatment with adenoviral constructs containing wild-type p53 gene sequence alone and in combination with cytotoxic chemotherapy.^{22,23,24,25,26,27,28,29,30,31} Although some evidence suggests that this effect may not be solely dependent on the presence of mutant p53, others have found greater efficacy when the endogenous p53 is mutant.^31,44 Results from in vitro xenograft models of several malignancies, including ovarian cancer, suggest promise for the strategy of p53 gene replacement as a novel cancer therapeutic approach.^25,30,31,45 There is no apparent effect of wild-type p53 overexpression on normal tissue such as fibroblasts.⁴⁶ Of particular relevance is the preclinical observation that the effects of p53 gene replacement are synergistic with both cisplatin and paclitaxel, the two mainstays of ovarian cancer chemotherapy.^22,23,30 Most clinical data to date with p53 gene replacement are limited to intratumoral injections,^{32,33,34,35,36} in contrast to the body cavity exposure of the large surface area of the peritoneal cavity exposed to SCH 58500 in the present study. Finally, although carcinogenesis clearly involves multiple gene defects, data support a therapeutic approach that corrects only a single, critical gene defect.^47,48

Intraperitoneal therapy of ovarian cancer was initially reported in 1955 by Weisberger et al.⁴⁹ In the past 5-10years encouraging results from the i.p. delivery of a variety of chemotherapeutics and biologics have been reported both for primary therapy and for small-volume recurrent or persistent disease.^{19,50,51,52,53,54,55} These studies have suggested the importance of treating small-volume disease and have established safety and symptom data to which one can then compare results of i.p. gene therapy. Indeed, encouraged by these data, phase I trials of the herpes simplex thymidine kinase/ganciclovir system,^56,57,58 and adenoviral E1a gene therapy,⁵⁹ have been initiated for recurrent ovarian cancer. Early results of phase I/II retroviral and adenoviral BRCA1 i.p. gene replacement have also been published.^60,61

Several potential limiting factors associated with i.p. drug delivery of gene therapy per se have been identified. For example, the uniformity of drug distribution is always of concern. For the present study, all patients were required to have widespread i.p. distribution verified by a pretreatment radiologic study before initial dosing. The fraction of the cancer cells that needs to be transduced in order for a clinical effect to be measured is unknown. It is clear that not all tumor target cells will be transduced, especially with a single administration of vector because the depth of penetration into tumor appears limited.⁶² Furthermore, there is concern that the accumulation of adhesions and the host immune response may prevent effective gene transfer with repetitive dosing of a viral vector. In the present study, multiple laparoscopies on the same patient provided the opportunity to demonstrate that individual inflammatory response was highly variable and that peritoneal distribution can clearly change over time.

The present study was designed to determine the safety of the SCH 58500 adenoviral vector delivered into the peritoneal cavity of women with refractory ovarian cancer. No maximum tolerated dose (MTD) was established as the protocol-defined DLT was not met. The doses delivered ranged from 7.5´10¹⁰ to 7.5´10¹³ particles per i.p. infusion. The highest dose tested was limited by practical considerations including the i.p. delivery volume for multiple-day dosing regimens. Tolerance to SCH 58500 was excellent with manageable toxicity. Aside from fever, the toxicity profile, even with multiple cycles was similar to that reported for i.p. chemotherapy in general.^19,50,51 Overall, 82.2% of the planned doses were delivered and this included 219 of 270 (81%) doses on the multiple-dose/multiple-cycle regimens. By way of comparison, 84% of the planned i.p. chemotherapy was delivered in a similarly sized study by Morgan et al⁵¹ whereas 76.8% of the planned i.p. cisplatin doses were delivered in the large cooperative group study reported by Alberts et al.¹⁹ Progression of disease was the most common reason for incomplete dosing rather than side effects in the present study.

Vector-specific gene transfer and mRNA expression of SCH 58500 was seen at doses as low as 7.5´10¹⁰ particles/single dose and was frequently detected in patients that received 7.5´10¹¹ particles/dose. It seemed desirable to increase the dose level and number of doses to a maximum based on the theoretical tumor burden within the peritoneal cavity and the need to maximize exposure of tumor cells to SCH 58500. Preclinical modeling indicated that multiple fractionated doses of SCH 58500 had greater efficacy than a single bolus injection.²²

Early concerns that the presence of serum neutralizing antibodies to the adenovirus might limit its effectiveness, particularly with repetitive exposure are not borne out by our results.⁶³ Preclinical work with immunized rodents treated with intratumoral injection of an adenoviral vector expressing IL-12 demonstrated minimal reduction in transfer efficiency.²⁵ Despite the generation of increased antiadenoviral antibody titers to SCH 58500 in all treated patients, we were also able to demonstrate transgene expression after multiple cycles of dosing. There was no obvious enhanced transgene expression in the two individuals who were treated at level 1 because of no demonstrable adenoviral immunity. Not all patients underwent sampling with each cycle of treatment, due to the invasive nature of laparoscopy. Nonetheless, our data clearly show the presence of transgene expression in RNA isolated from both ascitic fluid and tumor biopsies. The alternative explanation of persistent, stable expression of SCH 58500 over time is inconsistent with in vitro and in vivo preclinical observations.

For a single case, in situ PCR data confirmed gene transfer in tumor cells obtained at laparoscopic biopsy. It is not possible to determine the percent of tumor cells transduced because of variability in the size of the biopsies obtained and the variation in the depth of SCH 58500 penetration. For example, in the case of a 3-mm biopsy with 1 mm of penetration and 100% transduction to the level of penetration, one might infer 33% transduction efficiency. However, because the size of the lesion is unknown, the true transduction efficiency cannot be calculated. Similarly, a smaller (2 mm) biopsy from the same site would provide a different estimate of transduction efficiency. This important parameter cannot be estimated nearly as well in human clinical trials as it can be in cell culture, or in orthotopic animal models with smaller and more uniform lesions. Further in situ PCR studies are ongoing and will be the subject of a separate report (S Wen et al, in preparation).

Several investigators have postulated that adenoviral transfection efficiency is determined by the presence of coxsackie viral receptor (CAR) on the surface of epithelial cells.^64,65 We did not have sufficient samples to test this hypothesis as an explanation for the failure to achieve transfection in all samples collected or the differential expression of transgene, which varied between barely detectable to 89,000 copies/per copy of -actin. Variations in CAR receptor levels, however, may explain differences in transfection efficiency between ovarian cancer cell lines transduced with SCH 58500 in vitro.⁶⁶

The inclusion of a subset of patients who received multiple courses of SCH 58500, both alone and in combination with chemotherapy, provided the opportunity not only to compare cumulative toxicity but also to gain preliminary data relevant to clinical response. As we have demonstrated, the combination of SCH 58500 with conventional chemotherapy for ovarian cancer added little to the toxicity of SCH 58500 alone. The frequent appearance of new CT-measurable lesions during the course of treatment with SCH 58500 accompanied by concomitant dramatic decreases in CA125 suggests that for gene replacement studies utilizing adenoviral vectors, CT scans are not a valid means to assess response. Also supporting this conclusion is the observation of mixed clinical responses observed in the same individual with objective responses of some lesions accompanied by the simultaneous development of new lesions in the same individual. Fortunately, other studies have demonstrated that CA125 responses to ovarian cancer treatment correlate very well with CT responses when CT is a valid measure of response.^41,42 Because CA125 responses also correlate well with overall survival,^67,68,69,70 they should not be dismissed out of hand. Indeed, because inflammatory changes in the peritoneal cavity may effect modest elevations of CA125 independent of ovarian cancer,^71,72,73,74 the interpretation of the overall responses in this study solely on the basis of CA125 response in the face of extensive inflammation, may actually serve to underestimate true response rates. The number of CA125 responders and the degree of response observed in groups 2 and 3 is remarkable based on the heavily pretreated nature of these patients. These data suggest that SCH 58500 has no negative impact on clinical outcome expected from standard chemotherapy treatments. Finally, it should also be noted that our multiple-dose cohort contained bulky tumor deposits, not the most optimal group to study i.p. regimens of any type.⁷⁵ We conclude that SCH 58500 is safe, well tolerated, and in combination with platinum-based chemotherapy provides response data to justify its further clinical testing for efficacy in the newly initiated phase III trial for front-line treatment of minimal residual ovarian cancer after primary surgical cytoreduction.

Acknowledgements

The following individuals contributed significantly to the development, monitoring, and/or execution of this trial: from Iowa ¾ Barrie Anderson MD, Joel Sorosky, MD, Anil Sood MD, Teresa Benda, RN, Karen Powliss, RN, Linda Sanders, BS, Melanie Hatterman, BS; from UCLA ¾ Natalie Uhorne, Lenore Gordon, Lisa Yanemoto, Malgarzata Beryt; from SPRI ¾ Michelle Kerin, Mary Ann Fritz, PhD, L Nielsen PhD, Shu FenWen, PhD; from Ulm ¾ Dres T Hawighorst, S Regele, K Maidel, Ms T Kohler, Dres E Stickeler, T Einzmann, and Ms L Walz.

1 Greenlee RT, Hill-Harman MB, Murray T, Thun M. Cancer statistics, 2001. Ca-Cancer J Clin 2001; 51: 15-36. MEDLINE

2 International Federation of Gynecology and Obstetrics Annual Report on the Results of Treatment in Gynecological Cancer Stockholm: FIGO, 1994, 24.

3 Geisler J, Geisler H. Brain metastases in epithelial ovarian carcinoma. Gynecol Oncol 1995; 57: 246-249. Article MEDLINE

4 Sood AK, Sorosky J, Dolan M et al. Distant metastases in ovarian cancer: why do they occur? Clin Cancer Res 1999; 5: 2485-2490. MEDLINE

5 Griffiths CT. Surgical resection of tumor bulk in the primary treatment of ovarian carcinoma. Natl Cancer Inst Monogr 1975; 42: 101-104. MEDLINE

6 Hoskins WJ, McGuire WP, Brady MF et al. The effect of diameter of largest residual disease on survival after primary cytoreductive surgery in patients with suboptimal residual epithelial ovarian carcinoma. Am J of Obstet Gynecol 1994; 170: 974-979.

7 Podratz KC, Schray MF, Wieand HS et al. Evaluation of treatment and survival after positive second-look laparotomy. Gynecol Oncol 1988; 31: 9-24. MEDLINE

8 Hoskins WJ, Rubin SC, Dulaney E et al. Influence of secondary cytoreduction at the time of second-look laparotomy on the survival of patients with epithelial ovarian carcinoma. Gynecol Oncol 1989; 34: 365-372. MEDLINE

9 Levine AJ, Momand J, Finlay CA. The p53 tumour suppressor gene. Nature 1991; 351: 453-456. MEDLINE

10 Harris CC. Structure and function of the p53 tumor suppressor gene: clues for rational cancer therapeutic strategies. J Natl Cancer Inst 1996; 88: 1442-1455. MEDLINE

11 Lee JM, Bernstein A. Apoptosis, cancer and the p53 tumour suppressor gene. Cancer Metastasis Rev 1995; 14: 149-161. MEDLINE

12 Hawkins DS, Demers GW, Galloway DA. Inactivation of p53 enhances sensitivity to multiple chemotherapeutic agents. Cancer Res 1996; 56: 892-898. MEDLINE

13 Kirsch DG, Kastan MB. Tumor-suppressor p53: implications for tumor development and prognosis. J Clin Oncol 1998; 16: 3158-3168, (Review). MEDLINE

14 Blandino G, Levine AJ, Oren M. Mutant p53 gain of function: differential effects of different p53 mutants on resistance of cultured cells to chemotherapy. Oncogene 1999; 18: 477-485. MEDLINE

15 Wen W, Reles A, Runnebaum IB et al. P53 mutations and expression in ovarian cancers: correlation with overall survival. Int J Gynecol Pathol 1999; 18: 29-41. MEDLINE

16 Skilling JS, Sood AK, Niemann T et al. An abundance of p53 null mutations in ovarian carcinoma. Oncogene 1996; 13: 117-123. MEDLINE

17 Shahin M, Sood AK, Hughes JH et al. The prognostic significance of p53 tumor suppressor gene alterations in ovarian cancer. Cancer 2000; 89: 2006-2017. Article MEDLINE

18 Dedrick RL. Theoretical and experimental bases of intraperitoneal chemotherapy. Semin Oncol 1986; 12: 1-6.

19 Alberts DS, Liu PY, Hannigan EV et al. Intraperitoneal cisplatin plus intravenous cyclophosphamide versus intravenous cisplatin plus intravenous cyclophosphamide for stage III ovarian cancer. N Engl J Med 1996; 335: 1950-1955. MEDLINE

20 Gomez-Navarro J, Siegal GP, Alvarez RD et al. Gene therapy: ovarian carcinoma as the paradigm. Am J Clin Pathol 1998; 109: 444-467. MEDLINE

21 Baselga J. New horizons: gene therapy for cancer. Anti-Cancer Drugs 1999; 10: 39-42. MEDLINE

22 Nielsen LL, Gurnani M, Syed J et al. Recombinant E1-deleted adenovirus-mediated gene therapy for cancer: efficacy studies with p53 tumor suppressor gene and liver histology in mouse tumor xenograft models. Hum Gene Ther 1998; 9: 681-694. MEDLINE

23 Nielsen LL, Lipari P, Dell J et al. Adenovirus-mediated p53 gene therapy and paclitaxel have synergistic efficacy in models of human head and neck, ovarian, prostate, and breast cancer. Clin Cancer Res 1998; 4: 835-846. MEDLINE

24 Nielsen LL, Dell J, Maxwell E et al. Efficacy of p53 adenovirus-mediated gene therapy against human breast cancer xenografts. Cancer Gene Ther 1997; 4: 129-138. MEDLINE

25 Bramson JL, Hitt M, Gauldie J et al. Pre-existing immunity to adenovirus does not prevent tumor regression following intratumoral administration of a vector expressing IL-12 but inhibits virus dissemination. Gene Ther 1997; 4: 1069-1076. MEDLINE

26 von Gruenigen VE, Santoso JT, Coleman RL et al. In vivo studies of adenovirus-based p53 gene therapy for ovarian cancer. Gynecol Oncol 1998; 69: 197-204. MEDLINE

27 Kim J, Hwang ES, Kim JS et al. Intraperitoneal gene therapy with adenoviral-mediated p53 tumor suppressor gene for ovarian cancer model in nude mouse. Cancer Gene Ther 1999; 6: 172-178. MEDLINE

28 Anderson SC, Johnson DE, Harris MP et al. P53 gene therapy in a rat model for hepatocellular carcinoma: intra-arterial delivery of a recombinant adenovirus. Clin Cancer Res 1998; 4: 1649-1659. MEDLINE

29 Gurnani M, Lipari P, Dell J et al. Adenovirus-mediated p53 gene therapy has greater efficacy when combined with chemotherapy against human head and neck, ovarian, prostate, and breast cancer. Cancer Chemother Pharmacol 1999; 44: 143-151. Article MEDLINE

30 Fujiwara T, Grimm EA, Mukhopadhyay T et al. Induction of chemosensitivity in human lung cancer cells in vivo by adenovirus-mediated transfer of the wild-type p53 gene. Cancer Res 1994; 54: 2287-2291. MEDLINE

31 Zhang W-W, Fang X, Mazur W et al. High-efficiency gene transfer and high-level expression of wild-type p53 in human lung cancer cells mediated by recombinant adenovirus. Cancer Gene Ther 1994; 1: 5-13. MEDLINE

32 Roth JA, Nguyen D, Lawrence DD et al. Retrovirus-mediated wild-type gene transfer to tumors of patients with lung cancer. Nat Med 1996; 2: 985-991. MEDLINE

33 Nemunaitis J, Swisher SG, Timmons T et al. Adenovirus-mediated p53 gene transfer in sequence with cisplatin to tumors of patients with non-small-cell lung cancer. J Clin Oncol 2000; 18: 609-622. MEDLINE

34 Clayman GL, El-Naggar AK, Lippman SM et al. Adenovirus-mediated p53 gene transfer in patients with advanced recurrent head and neck squamous cell carcinoma. J Clin Oncol 1998; 16: 2221-2232. MEDLINE

35 Swisher SG, Roth JA, Nemunaitis J et al. Adenovirus-mediated p53 gene transfer in advanced non-small-cell lung cancer. J Natl Cancer Inst 1999; 91: 763-771. Article MEDLINE

36 Cusack JC, Spitz FR, Nguyen D et al. High levels of gene transduction in human lung tumors following intralesional injection of recombinant adenovirus. Cancer Gene Ther 1996; 3: 245-249. MEDLINE

37 Nielsen LL, Maneval DC. P53 tumor suppressor gene therapy for cancer. Cancer Gene Ther 1998; 5: 52-63. MEDLINE

38 Buller RE, Pegram M, Runnebaum I et al. A phase I/II trial of recombinant adenoviral human p53 (SCH 58500) intraperitoneal (IP) gene therapy in recurrent ovarian cancer. Gynecol Oncol 1999; 72: 452-453.

39 Wills KN, Maneval DC, Menzel P et al. Development and characterization of recombinant adenoviruses encoding human p53 for gene therapy of cancer. Hum Gen Ther 1994; 5: 1079-1088.

40 Cockcroft DW, Gault MH. Prediction of creatinine clearance from serum creatinine. Nephron 1976; 16: 31-41. MEDLINE

41 Rustin GJ, Nelstrop AE, McClean P et al. Defining response of ovarian carcinoma to initial chemotherapy according to serum CA 125. J Clin Oncol 1996; 14: 1545-1551. MEDLINE

42 Rustin GJ, Nelstrop AE, Bentzen SM et al. Use of tumour markers in monitoring the course of ovarian cancer. Ann Oncol 1999; 10: S21-S27.

43 Musco ML, Cui S, Small D et al. A comparison of flow cytometry and laser scanning cytometry for the intracellular evaluation of adenoviral infectivity and p53 protein expression in gene therapy. Cytometry 1998; 33: 290-296. Article MEDLINE

44 Wolf JK, Mills G, Bazzet L et al. Adenovirus-mediated p53 growth inhibition of ovarian cancer cells is independent of endogenous p53 status. Gynecol Oncol 1999; 75: 261-266. Article MEDLINE

45 Mujoo K, Maneval DC, Anderson SC et al. Adenoviral-mediated p53 tumor suppressor gene therapy of human ovarian carcinoma. Oncogene 1996; 12: 1617-1623. MEDLINE

46 Liu T-J, el-Naggar AK, McDonnell TJ et al. Apoptosis induction mediated by wild-type p53 adenoviral gene transfer in squamous cell carcinoma of the head and neck. Cancer Res 1995; 55: 3117-3122. MEDLINE

47 Chen P-L, Chen Y, Bookstein R et al. Genetic mechanisms of tumor suppression by the human p53 gene. Science 1990; 250: 1576-1580. MEDLINE

48 Fujiwara T, Grimm EA, Roth JA. Gene therapeutics and gene therapy for cancer. Curr Opin Oncol 1994; 6: 96-105. MEDLINE

49 Weisberger AS, Levine B, Storaasli JP. Use of nitrogen mustard in treatment of serous effusions of neoplastic origin. J Am Med Assoc 1955; 159: 1704-1707.

50 Berek JS, Markman M, Stonebraker B et al. Intraperitoneal interferon- in residual ovarian carcinoma: a phase II gynecologic oncology group study. Gynecol Oncol 1999; 75: 10-14. Article MEDLINE

51 Morgan RJ Jr, Braly P, Cecchi G et al. Phase II trial of intraperitoneal cisplatin with intravenous doxorubicin and cyclophosphamide in previously untreated patients with advanced ovarian cancer ¾ long-term follow-up. Gynecol Oncol 1999; 75: 419-426. Article MEDLINE

52 Barakat RR, Almadrones L, Venkatraman ES et al. A phase II trial of intraperitoneal cisplatin and etoposide as consolidation therapy in patients with stage II-IV epithelial ovarian cancer following negative surgical assessment. Gynecol Oncol 1998; 69: 17-22. Article MEDLINE

53 Muggia FM, Liu PY, Alberts DS et al. Intraperitoneal mitoxantrone or floxuridine: effects on time-to-failure and survival in patients with minimal residual ovarian cancer after second-look laparotomy ¾ a randomized phase II study by the Southwest Oncology Group. Gynecol Oncol 1996; 61: 395-402. Article MEDLINE

54 Markman M, Reichman B, Hakes T et al. Impact on survival of surgically defined favorable responses to salvage intraperitoneal chemotherapy in small-volume residual ovarian cancer. J Clin Oncol 1992; 10: 1479-1484. MEDLINE

55 Markman M, Brady MF, Spirtos NM et al. Phase II trial of intraperitoneal paclitaxel in carcinoma of the ovary, tube, and peritoneum: a Gynecologic Oncology Group Study. J Clin Oncol 1998; 16: 2620-2624. MEDLINE

56 Link CJ Jr, Moorman D, Seregina T et al. A phase I trial of in vivo gene therapy with the herpes simplex thymidine kinase/ganciclovir system for the treatment of refractory or recurrent ovarian cancer. Hum Gene Ther 1996; 7: 1161-1179. MEDLINE

57 Alvarez RD, Curiel DT. A phase I study of recombinant adenovirus vector-mediated intraperitoneal delivery of herpes simplex virus thymidine kinase (HSV-TK) gene and intravenous ganciclovir for previously treated ovarian and extraovarian cancer patients. Hum Gene Ther 1997; 8: 597-613. MEDLINE

58 Hasenburg A, Tong XW, Rojas-Martinez A et al. Thymidine kinase (TK) gene therapy of solid tumors: valacyclovir facilitates outpatient treatment. Anticancer Res 1999; 19: 2163-2165. MEDLINE

59 Hortobagyi GN, Hung MC, Lopez-Berestein G. A phase I multicenter study of E1A gene therapy for patients with metastatic breast cancer and epithelial ovarian cancer that overexpresses HER-2/neu or epithelial ovarian cancer. Hum Gene Ther 1998; 9: 1775-1798. MEDLINE

60 Tait DL, Obermiller PS, Redlin-Frazier S et al. A phase I trial of retroviral BRCA1sv gene therapy in ovarian cancer. Clin Cancer Res 1997; 3: 1959-1968. MEDLINE

61 Tait DL, Obermiller PS, Hatmaker AR et al. Ovarian cancer BRCA1 gene therapy: phase I and II trial differences in immune response and vector stability. Clin Cancer Res 1999; 5: 1708-1714. MEDLINE

62 Grace MJ, Xie L, Musco ML et al. The use of laser scanning cytometry to assess depth of penetration of adenovirus p53 gene therapy in human xenograft biopsies. Am J Pathol 1999; 155: 1869-1878. MEDLINE

63 Hitt M, Addison CL, Graham FL. Human adenovirus vectors for gene transfer into mammalian cells. Adv Pharmacol 1997; 40: 137-206. MEDLINE

64 Bergelson JM, Cunningham JA, Droguett G et al. Isolation of a common receptor for Coxsackie B viruses and adenoviruses 2 and 5. Science 1997; 275: 1320-1323. Article MEDLINE

65 Yingming L, Pong R-C, Bergelson JM et al. Loss of adenoviral receptor expression in human bladder cancer cells: a potential impact on the efficacy of gene therapy. Cancer Res 1999; 59: 325-330. MEDLINE

66 Pegram M, Tseng Y, Baldwin RL et al. Expression of coxsackie and adenovirus receptor (CAR) correlates with efficiency of adenovirus transduction of human ovarian carcinoma cells. Gynecol Oncol 1999; 72: 452.

67 Buller RE, Berman ML, Bloss JD et al. CA 125 regression: a model for epithelial ovarian cancer response. Am J Obstet Gynecol 1991; 165: 360-367. MEDLINE

68 Buller RE, Vasilev S, DiSaia PJ. CA125 kinetics: a cost effective clinical tool to evaluate clinical trial outcomes in the 1990s. Am J Obstet Gynecol 1996; 174: 1241-1254. MEDLINE

69 van der Burg MEL, Lammes FB, van Putten WLJ et al. Ovarian cancer: the prognostic value of the serum half-life of CA125 during induction chemotherapy. Gynecol Oncol 1988; 30: 307-312. MEDLINE

70 Rustin GJS, Gennings JN, Nelstrop AE et al. Use of CA-125 to predict survival of patients with ovarian carcinoma. J Clin Oncol 1989; 7: 1667-1671. MEDLINE

71 Yedema CA, Kenemans P, Thomas CM et al. CA125 serum levels in the early post-operative period do not reflect tumour reduction obtained by cytoreductive surgery. Eur J Cancer 1993; 29A: 966-971. MEDLINE

72 Van Der Zee AG, Duk JM, Aalders JG et al. The effect of abdominal surgery on the serum concentration of the tumour-associated antigen CA 125. Br J Obstet Gynaecol 1990; 97: 934-938. MEDLINE

73 Talbot RW, Jacobsen DJ, Nagorney DM et al. Temporary elevation of CA 125 after abdominal surgical treatment for benign disease and cancer. Surg Gynecol Obstet 1989; 168: 407-412. MEDLINE

74 Zeimet AG, Offner FA, Marth C et al. Modulation of CA-125 release by inflammatory cytokines in human peritoneal mesothelial and ovarian cancer cells. Anticancer Res 1997; 17: 3129-3131. MEDLINE

75 Howell SB, Zimm S, Markman M et al. Long-term survival of advanced refractory ovarian carcinoma patients with small-volume disease treated with intraperitoneal chemotherapy. J Clin Oncol 1987; 5: 1607-1612. MEDLINE

Figure 1 Typical febrile response to SCH 58500 over time with multiple doses and cycles. indicates dose of i.p. SCH 58500 delivered. SCH 58500 toxicity by treatment cycle.

Figure 2 MIMICÔ PCR assessment of gene transfer. The numbers correspond to lane numbers in a 3% agarose gel. Total RNA was extracted from a tumor biopsy obtained laparoscopically from a patient 72 hours after administration of a single dose of 2.5´10¹² particles of SCH 58500. For this sample, the effect of serial dilution of -actin message template cDNA prepared from tissue RNA and added to a MIMICÔ PCR reaction is reflected by the decreasing intensity of the upper band in lanes 2-5 of the agarose gel. A precisely calculated amount (500 molecules) of -actin MIMICÔ has been spiked into the PCR reaction and results in the generation of the lower band in the same lanes. Lanes 11 (500 molecules of -actin MIMICÔ) and 12 (100,000 molecules of -actin MIMICÔ) have been used for quantitative calculations. Similarly 500 molecules of the p53 MIMICÔ have been spiked into the PCR reactions run in lanes 6-8 from serial dilutions of the template cDNA. In these lanes the MIMICÔ product is the lower band and corresponds to the single band in lane 1 (500 molecules of p53 MIMICÔ without template cDNA). The upper band in lanes 6-8 represents p53 product containing the tripartite leader. In the absence of transfection, as in lanes 9 and 10, no upper band is seen because the wild-type p53 sequence does not contain sequence that will bind the leader sequence specific primers.

Figure 3 p53 gene transfer following multidose i.p. delivery of SCH 58500. MIMICÔ PCR reactions were carried out as described in the legend to Figure 1. Measurable levels of mRNA are plotted in the graphs according to cycle and dose of SCH 58500. Seven additional samples were RT-PCR positive, but at expression below levels that could be quantitated. Only 9 of 62 samples expressing -actin were negative for p53 transgene expression.

Figure 4 Analysis of tumor biopsies after i.p. SCH 58500. In situ PCR measurement of viral DNA. Panel A: Biopsy from a patient before SCH 58500 (´400). Panel B: Biopsy from a patient after three cycles of SCH 58500 (´200). Panel C: 5-m section from the same sample as 3B but a negative control based on omission of Taq polymerase from the PCR reaction (´40). Panel D: An H and E section from the same sample (´200).

Figure 5 CA125 responses following treatment with SCH 58500. In cycle 1 (C1: ) all patients received SCH 58500 alone. Cycles 2 and 3 (C2, C3: ) includes all patients who received chemotherapy in addition to SCH 58500. The mean decline in serum CA125 was calculated for responders only. Overall () is the average percent decline at the end of the study for individuals who responded relative to their screening CA125 level. In each case, CA125 is measured over a 28-day interval or at the beginning of the subsequent treatment cycle. The error bars are 95% confidence limits of the mean.

Table 1 SCH 58500 treatment regimens

Table 2 Study cohort demographics

Table 3 Treatment-related adverse events

Table 4 Relationship of CT-based response to CA125 response