Safety of Intravenous Administration of a Canarypox Virus Encoding The

Safety of intravenous administration of a canarypox virus encoding the human wild-type p53 gene in colorectal cancer patients Anand G Menon,1 Peter JK Kuppen,1 Sjoerd H van der Burg,2 Rienk Offringa,2 Marie Claude Bonnet,5 Bert IJ Harinck,1 Rob AEM Tollenaar,1 Anke Redeker,2 Hein Putter,4 Philippe Moingeon,5 Hans Morreau,3 Cornelis JM Melief,2 and Cornelis JH van de Velde1 1Department of Surgery, Leiden University Medical Center, Leiden, The Netherlands; 2Department of Immunohematology and Bloodbank, Leiden University Medical Center, Leiden, The Netherlands; 3Department of Pathology, Leiden University Medical Center, Leiden, The Netherlands; 4Department of Medical Statistics, Leiden University Medical Center, Leiden, The Netherlands; and 5Aventis Pasteur, Lyon, France.

Overexpression of p53 occurs in more than 50% of colorectal cancers. Therefore, p53 represents an attractive target antigen for immunotherapy. We assessed the safety of a canarypox virus encoding the human wild-type p53 gene given intravenously to end- stage colorectal cancer patients in a three-step dose escalation study aimed at inducing p53 immune responses. Patients with metastatic disease of p53-overexpressing colorectal cancers were vaccinated three times at 3-week intervals, each time with 106.5 7.0 7.5 CCID50 (CCID50 ¼ cell culture infectious dose 50%; group 1, n ¼ 5), 10 CCID50 (group 2, n ¼ 5) or 10 CCID50 (group 3, n ¼ 6). Vital signs and the occurrence of adverse events were monitored and blood was analyzed for biochemical and hematological parameters as well as signs of auto-immune safety. In all, 16 patients were enrolled and 15 patients completed three vaccinations. No anaphylactic reaction or unwanted auto-immune reactions were observed. A total of 16 serious adverse events (SAEs) occurred: 10 in group 1, three in group 2 and three in group 3. All SAEs were tumor-related complications. There was no difference in the frequency of adverse events between the three groups, except for fever. Fever was the only vaccination-related adverse event consistently observed and was most frequent and outspoken in the group 3 patients. The majority was a grade 1 or 2 fever (93%) and grade 3 fever (7%) was observed in three patients of group 3. Some patients showed humoral and cellular responses against p53, following vaccinations. After having completed his initial treatment cycle, one patient (group 2) received a second treatment cycle 7.5 of three doses of 10 CCID50 and subsequently showed stable disease. All other patients showed progressive disease. We conclude that ALVAC-p53 can be administered intravenously to colorectal cancer patients without serious toxicity or pathological autoimmunity and can induce immune responses against p53. Cancer Gene Therapy (2003) 10, 509–517. doi:10.1038/sj.cgt.7700600 Keywords: colorectal cancer; P53; immunotherapy; canarypox virus

umors are characterized by numerous changes in a antigens), viral antigens (e.g. HPV E6/E7), protein Tvariety of genes, resulting in altered products that can antigens (e.g. CEA, HER-2/neu, Ep-Cam) and mutated be unique targets for the immune system. Although the oncogene products (e.g. ras, p53, c-myc).1 majority are expressed on both tumor cells and normal The p53 gene is a tumor suppressor gene located on the cells, the expression of these tumor-associated antigens short arm of chromosome 17 that plays a pivotal role in (TAAs) has been found to differ significantly between ensuring genetic homeostasis.2 Mutations in the p53 gene normal cells and tumor cells, and thus may provide an are the most common genetic alteration in many tumors immunotherapeutic window. Potential targets include and result in the production of a defective protein with a carbohydrate antigens (e.g. gangliosides, blood groups significantly longer than normal half-life of p53.3 More than 50% of colorectal cancers demonstrate p53 overexpression, potentially allowing the tumor cells to be Received January 25, 2003. processed and the immunogenic epitopes to be presented Address correspondence and reprint requests to: Professor Dr CJH 4 van de Velde, Department of Surgery, K6-R, Leiden University to the immune system. Although p53 is an auto-antigen, it Medical Center, PO Box 9600, 2300 RC Leiden, The Netherlands. has been demonstrated that humoral and cellular E-mail: [email protected] immune responses against p53 can be detected in cancer Safety of ALVAC-p53 in colorectal cancer AG Menon et al 510 patients.5–10 Enhancing these responses could result in ALVAC-p53 used in the present study, with doses up to effective antitumor responses, but also induce a hazar- 10 times the proportional human dose, revealed no dous reaction to normal cells that have lowexpression toxicity or auto-immune phenomena upon clinical or on levels of this auto-antigen. histopathological evaluation.14 The safety of ALVAC Vaccination against p53 can be performed using vector-based vaccines has been well established in various genetically modified viral vectors allowing high expression clinical trials where more than 1500 individuals, including of the transfected gene within antigen-presenting cells, cancer patients, have received ALVAC encoding a variety thereby eliciting strong immune response against p53. The of genes such as the rabies glycoprotein gene,15,16 HIV canarypox virus (ALVAC) is a well-characterized viral envelope genes,17,18 CEA19 and even multiple genes such vector capable of infecting a variety of human cells. This as CEA and B7.1.20 In this study, we show that repeated vector, which cannot replicate in non-avian species, can i.v. administration of ALVAC-p53 in end-stage colorectal nevertheless express inserted genes for up to a fewdays. cancer patients is well tolerated and does not induce This results in the induction of strong humoral and pathological auto-immunity, while stimulating p53-speci- cellular responses against these gene products.11 Hurpin fic immune responses. et al.12 compared subcutaneous, intramuscular, intrader- mal and intravenous (i.v.) administrations of ALVAC- p53 in mice and showed that only the i.v. vaccination Methods route was capable of inducing CTL responses against p53. Subsequent studies showed that i.v. vaccination of mice Patients with ALVAC-p53 resulted in long-term protection against Patients older than 18 years with histologically proven challenges with p53 overexpressing tumors. Although, colorectal cancer and evidence by imaging techniques of this does not resemble the clinical situation in which irresectable disease were eligible for inclusion. Patients tumors are already present, Roth et al.13 also showed that with metastatic disease that was untreatable by conven- established p53-overexpressing tumors were successfully tional therapies and patients with metastatic disease that eradicated with i.v. ALVAC-p53 administrations. was potentially treatable, but who refused conventional The use of viral vectors in humans is associated with therapy were also eligible. The study protocol was safety concerns for the patients and the environment, but approved by the local and national medical ethics as the canarypox virus is unable to replicate in committees as well as by the biological safety committee mammalian cells, these concerns do not apply to and the Dutch Ministry of Health and Environment ALVAC-based vaccines. Studies performed on rhesus before the study started. The inclusion and exclusion macaques that received i.v. injections of the same lot of criteria are summarized in Table 1.

Table 1 Inclusion and exclusion criteria for participation in the ALVAC-p53 study

Inclusion criteria Histologically proven colorectal cancer Evidence of irresectable or metastatic disease in patients untreatable by conventional therapies, or in patients who are treatable by conventional therapies, but who refuse these Immunohistochemically proven expression of p53 and MHC class I in at least 30% of tumor cells Serum CEA X10 mg/l Age over 18 years Negative pregnancy test (for women of child-bearing age) Use effective contraception during the trial and up to 3 months after the last injection (for men and women) WHO performance status of 0–1 Renal function and bilirubin levels within 1.5 Â the normal upper and lower levels, ASAT and ALAT within 3 Â the normal upper value and alkaline phospahatase within 5x the normal upper value Signed informed consent

Exclusion criteria Auto-immune disease Symptomatic viral or other infections HIV seropositivity or refusal to hear the result of the HIV test Organ grafts Life expectancy of less than 3 months History of allergy: severe asthma or atopic eczema or known allergy to egg products or neomycin History of severe neurological, cardiovascular, renal, hepatic, endocrine, respiratory or bone marrow dysfunction Known family history of Li-Fraumeni syndrome Chemotherapy (with the exception of nitrosurea), or radiation therapy within 4 weeks preceding inclusion Immunotherapy, chemotherapy using nitrosurea, or hormonal therapy (other than contraception) within 6 weeks preceding inclusion History of treatment with extracted growth hormone

Cancer Gene Therapy Safety of ALVAC-p53 in colorectal cancer AG Menon et al 511 Treatment and follow-up schedule toxicity tests, pyrogen assay identification, viral potency test and a particle count test. Patients were enrolled into three sequential groups of five individuals each and received three i.v. injections of ALVAC-p53 at 3-week intervals. At each of the three Statistical analyses injections, patients in group 1 received a tenth (106.5 The w2 test, Kruskal–Wallis test (CEA) and the ANOVA CCID50) of the total dose of ALVAC-p53, patients in (age and diagnosis metastases) tests were performed to 7.0 group 2 received a third (10 CCID50) of the total dose analyze the differences between clinical and tumor and patients in group 3 received the total dose (107.5 characteristics between groups 1 to 3. A Po0.05 indicates CCID50). Patients were hospitalized for 24 hours after a statistically significant difference. vaccination and vital signs (heart rate, blood pressure, rectal temperature and respiration rate) were measured just before and after vaccination, as well as 30 minutes, Results 1 hour, 6 hours and 24 hours after vaccination. Between hospital visits, patients documented symptoms and body Patients temperature in a safety diary. A clinical history was taken A total of 16 patients were enrolled in this dose-escalation and a physical examination was performed on each visit. study between April 1999 and December 2000 at the Adverse events (clinical and biochemical) were scored Department of Surgery of the Leiden University Medical according to the World Health Organization (WHO) Center. Characteristics of the study population are shown recommendation for grading acute and subacute toxic in Table 2. There were 12 men and four women with a effects.21 Blood was obtained for biochemical (renal and mean age of 60 years (range 42–71 years). Most patients liver function, C-Reactive Protein), hematological (he- (15/16) presented with metastatic disease, with the moglobin level, hematocrit, leucocytes and differentia- majority of patients suffering from liver (14/16 patients) tion, platelets), auto-immunity (anti-DNA antibodies and and/or lung metastases (8/16 patients). Although, not immune-complexes (ICX)) and CEA level assays, before statistically significant, the patients in the lowest dose each vaccination and during follow-up visits. Anti-DNA group (group 1) were younger on average and seemed to antibodies,22 ICX levels23 and immune responses against have a higher tumor burden, as was reflected in the higher ALVAC and p53 were measured as previously de- average CEA level at the start of the study and the higher scribed.24 IgG antibody titers against ALVAC are frequency of treatment with second line chemotherapy presented as mean OD measured at 415 nm (dilution and isolated liver perfusion. Only two out of five patients 1:800) minus the background. IgG antibody titers against in group 1 were able to complete the planned 14 week p53 were considered positive when 41.1 U/ml. Patient follow-up after the last vaccination, whereas four out of visits were scheduled as follows: pre-inclusion visit (PV, a five patients in group 2 and five out of six patients in maximum of 2 weeks before V1); first vaccination (V1, group 3 completed the planned 14 week follow-up after week 0); second vaccination (V2, week 3); third vaccina- the last vaccination. There were no statistically significant tion (V3, week 6); V4 week 7; V5 week 8; V6 week 14 and differences in patient- and tumor characteristics between V7 at week 20. All patients underwent a CT scan of the the three groups (P-values not shown). chest and abdomen a maximum of 2 weeks before the first vaccination (PV) and on the last follow-up visit, 14 weeks Alterations in the treatment and follow-up scheme after the last vaccination (V7). Patient 02 (group 1) developed a urosepsis 16 days after the first vaccination owing to progression of a local Recombinant ALVAC-p53 vaccine recurrence of a rectal carcinoma into the bladder wall. A The ALVAC-p53 (vCP207; Aventis Pasteur, Lyon, urosepsis had previously occurred, 2 months prior to France) recombinant, expressing the human wild-type inclusion. Treatment and follow-up were postponed 2 p53 gene, was constructed using the canarypox virus- weeks and radiation therapy was started to control the based vector ALVAC. The vCP207 recombinant was pain and rectal blood loss caused by the local recurrence. generated by co-transfection of ALVAC-infected primary Patient 09 (group 2) requested chemotherapy after the chick-embryo fibroblasts with an insertion plasmid and third vaccination. The second line chemotherapy was noninfectious purified ALVAC genomic DNA, leading to started between V5 and V6 and resulted in WHO grade 4 the integration of the foreign gene expression cassette into diarrhea approximately 10 days later (45 days after the the viral genome via homologous recombination. The last ALVAC-p53 vaccination) which required hospitaliza- vaccine was purified twice through a sucrose gradient12 tion. During hospitalization, the patient developed an and titrated based on the evaluation of the infectious titer ileus and became critically ill. His condition improved of the preparation on a quail cell line, QT35 (continuous without surgical intervention. Patient 13 (group 3) was cell line derived from a chemically induced tumor in discontinued from the study after the second vaccination quail). The titer is expressed in CCID50/ml (number of because of disease progression (progression of recurrence cell culture infectious dose 50% per ml). Other quality of rectal carcinoma into the bladder wall) and was assays performed include pH measurement, appearance, replaced to be able to analyze five patients having osmolality measurement, extractable volume, endotoxin received all three vaccinations with the full dose. Patient content, bacterial and fungal sterility tests, abnormal 17 (group 3) requested chemotherapy after completing the

Cancer Gene Therapy Safety of ALVAC-p53 in colorectal cancer AG Menon et al 512 Table 2 Patient and tumor characteristics of the 16 patients who received intravenous vaccinations with different doses of ALVAC-p53

All patients Group 1 Group 2 Group 3 (n=16) (n=5) (n=5) (n=6)

Gender Male 12 3 5 4 Female 4 2 0 2

Age (years) Mean 60 53 62 65 Range 42 – 71 42 – 66 46 – 69 57 – 71

Metastases Liver 14 5 4 5 Lungs 8 3 3 2 Other 9 4 2 3

Diagnosis (months) Mean 25 35 14 25 Range 6 – 66 17 – 66 7 – 21 6 – 56 Therapy First line chemotherapy 15 5 5 5 Second line chemotherapy 7 4 2 1 ILP 8 4 1 3 Rtx 4 0 1 3 Other 5 2 0 3

WHO performance score Normal 5 2 1 2 Restricted 11 3 4 4 CEA level (mg/l) Mean 562 1275 348 145

Abbreviations: ILP=isolated liver perfusion; RTx=radiation therapy; WHO=World Health Organization.

vaccination schedule without significant problems. This probably (5%) or definitely (5%) related to ALVAC-p53. was started after V6 (8 weeks after the last vaccination). Fever was the most common ALVAC-p53-related event Patient 08 (group 2) requested three additional vaccina- (see Table 3 and Fig 1). Grade 2 fever (38–401C) occurred tions with the full dose as compassionate use after having significantly more often in group 3 patients and grade 3 received three vaccinations with 1/3 dose. This was fever (rectal temperature 4401C) was observed in group 3 performed outside the study protocol after receiving only. There were no signs of concomitant hemodynamic informed consent and approval from the appropriate or respiratory instability and the temperature normalized committees. within 24 hours after vaccination, either spontaneously or after treatment with acetominophen (paracetamol). There were no clinical signs of treatment-related toxicity to the Clinical safety lungs. None of the patients experienced an anaphylactic reaction. A total of 382 clinical events (CE; 122 in group 1, 124 in group 2 and 136 in group 3) were observed during Biochemical safety evaluation the study (V1–V7). This consisted of 16 serious adverse Hurpin et al.12 showed that after i.v. administration of events (SAEs; 10 in group 1, three in group 2 and three in ALVAC-p53 to mice, the virus was predominantly group 3) and 366 adverse events (AEs: 112 in group 1, 121 located in the lungs, liver and spleen. Therefore, toxic in group 2 and 133 in group 3). All SAEs were related to cytopathological effects caused by the virus were to be disease progression-related complications. The most expected in these organs. Abnormalities in liver biochem- frequent CE were: fever, alterations in liver biochemistry istry (ASAT, ALAT, alkaline phosphatase and bilirubin) and tumor-related complications ((abdominal) pain, were frequent, but it should be noted that 14/16 patients anemia and fatigue). The majority of clinical events had advanced liver metastases at the start of the study. (79%) was not related to administration of ALVAC-p53, Abnormal transaminase levels (ASAT440 U/l; ASA- but was a result of disease progression or intercurrent T445 U/l) and alkaline phosphatase levels (4120 U/l) diseases. The other events were either possibly (11%), were already present before vaccination in 9/16 patients

Cancer Gene Therapy Safety of ALVAC-p53 in colorectal cancer AG Menon et al 513 Table 3 Number of clinical events and number of subjects with clinical events per dose group

Group 1 (1/10 dose) Group 2 (1/3 dose) Group 3 (total dose)

Number of events grade 1–4 Number of events grade 1–4 Number of events grade 1–4 (number of patients) (number of patients) (number of patients)

123412341234

Fever 3 (2) 8 (4) — — — 9 (4) — — 2 (2) 29 (6) 4 (3) — Abdominal pain 1 (1) 1 (1) — — 3 (2) 9 (4) 3 (1) 1 (1) 2 (1) 3 (3) — — Pain 3 (2) 1 (1) 1(1) — 2 (1) 3 (2) — — 2 (1) 3 (3) — — Myalgia 2 (2) 1 (1) 1 (1) — 2 (1) — — — — — — — Fatigue 1 (1) 3 (3) — — 1 (1) 2 (2) 1 (1) — 4 (2) 3 (3) — — Cardiovascular 2 (2) 2 (2) — — 3 (3) — — — 2 (2) 4 (3) — — Cough 2 (1) 1 (1) 1 (1) — 1 (1) 1 (1) — — — 1 (1) — — Dyspnea — — 2 (2) — 2 (2) 1 (1) — — — 1 (1) — — Nausea 2 (2) 2 (1) — — — 2 (2) — — 3 (3) — — — Vomiting 1 (1) 4 (3) — — — 2 (2) — — — 1 (1) — — Anorexia — 3 (3) — — 2 (2) — — — 1 (1) 1 (1) — — Diarrhea — 1 (1) — — 3 (2) 1 (1) — 1 (1) — — — — Infection 1 (1) 2 (1) 1 (1) — 2 (2) — — — 3 (3) 3 (2) 1 (1) — Ascites — 2 (2) 1 (1) — 1 (1) — — — — — — — Edema — 2 (2) — — 1 (1) 1 (1) — — — 1 (1) — — Disease progression — — 1 (1) 3 (3) — — — — 1 (1) — — 1 (1) ASAT 7 (3) 6 (3) 2 (2) — 5 (3) 2 (2) 1 (1) 2 (2) 8 (4) 4 (2) — 1 (1) ALAT 3 (2) 2 (2) — — 4 (3) 2 (2) 1 (1) — 4 (2) 1 (1) 1 (1) — Alkaline phosphatase 5 (3) 6 (5) 3 (3) — 3 (2) 3 (3) 3 (2) — 4 (4) 5 (4) 1 (1) 1 (1) Bilirubin 4 (3) 1 (1) 1 (1) — 2 (1) — 1 (1) 2 (1) — — 1 (1) — Anemia 6 (3) 1 (1) 1 (1) — 7 (4) 1 (1) — — 7 (4) 2 (2) — — Leucopenia — — — — 4 (2) 1 (1) 1 (1) 1 (1) 1 (1) — — — Platelets — — — — — — — — 3 (2) 1 (1) — — Neurological — 2 (2) — — — — — — 2 (1) 1 (1) — —

A dash indicates that no adverse events occurred for that particular adverse event.

and 11/16 patients, respectively. There were no significant Immune responses and auto-immune safety differences in transaminase levels, alkaline phophatase The patients in this study were significantly immune- and bilirubin levels between the three groups. The compromised compared to healthy controls as they did electrolytes and renal function also showed no significant not respond as well to control antigens such as M. changes associated with vaccination. C-reactive protein 24 tuberculosis, tetanus toxoid and C. albicans. Antibody levels also did not showchanges consistent withan and cellular responses against ALVAC were detected in induction of a generalized immune response after all but one patient. Antibody responses (IgG) against p53 vaccination. were present before the vaccination in seven patients. In addition IgG (n ¼ 3) responses against p53 were induced Hematological evaluation in all three dose groups (Table 4). Cellular responses All five patients in group 1, 3/5 patients in group 2 and 4/ against p53 were only induced in patients who had 6 patients in group 3 were anemic at baseline, as seen in received the full dose.24 Although anti-p53 immune many end-stage cancer patients. Hemoglobin levels responses were induced, no clinical signs of unwanted declined during follow-up in nine patients, mainly in vaccine-induced auto-immune responses were detected. groups 1 (4/5 patients) and 2 (3/5 patients) and were There was no serological evidence for an induction of a related to disease progression. No differences in WHO pathological auto-immune response, as illustrated by the anemia toxicity grades were apparent between the three results of the anti-DNA antibody- and immune complex groups (see Table 3). Leucopenia occurred mainly in assays. Anti-DNA antibodies were not present in any of group 2 and was attributable to second-line chemother- the patients before and after vaccination. Immune apy treatment of patient 09. The altered platelet levels in complexes were not present or induced in 10 out of 16 two patients of group 3 were already present at baseline patients. Two patients (02 and 07) had positive immune and did not worsen after vaccination. complexes before the first vaccination, and another four

Cancer Gene Therapy Safety of ALVAC-p53 in colorectal cancer AG Menon et al 514 a th rd 41 1/10 dose ( n=5) 1/3 dose ( n=5) total dose (n=6) compassionate use (n=1)

35 0 6 12 18 24 06121824 0 61218 24 0 6 12 18 24 b 41 C ° 40

TEMPERATURE 36

35 0 61218 24 0 6 12 18 24 0 6 12 18 24 0 6 12 18 24 c 41

35 0 6 12 18 24 0 61218 24 0 61218 24 0 6121824

HOURS AFTER VACCINATION Figure 1 Patients received three i.v. vaccinations with ALVAC-p53. Patients received vaccinations at t ¼ 0 hours. The temperature was measured 5, 30 minutes, 1, 6, 10 and 24 hours after the vaccination. The mean rectal temperature and 95% confidence intervals (bars) are shown per dose group and vaccination (vaccinations 1–3 in a–c, respectively). All patients treated with the full dose developed fever 4 381C after the second and third vaccinations. One patient (08) received three vaccinations with 1/3 dose and later received compassionate treatment with the full dose. This patient is shown separately from the three dose groups and illustrates that the typical vaccine-induced fever associated with the full dose did not occur. This is probably caused by regular use of nonsteroidal anti-inflammatory drugs to alleviate tumor-induced abdominal pain.

patients had borderline levels of immune complexes after vaccination. Following the vaccination cycle with the full the vaccination scheme was completed. The formation of dose, anti-ALVAC antibodies were further boosted. The these immune complexes, however, was not related to mean OD at 415 nm increased from 0.55 before the first ALVAC-p53. These data indicate that although responses vaccination with the full dose to a maximum of 2.4 one against p53 were induced, this was not accompanied by week after the third vaccination. An IgG response against the induction of a pathological auto-immune response. p53 was present in this patient before participation and Of note, patient 08 received as part of compassionate p53-specific antibody levels increased during both the first treatment another vaccination treatment with the full and second vaccination cycle with 1/3 and the full dose, dose of ALVAC-p53. Anti-ALVAC antibodies were respectively. The IgG level increased from 1.1 U/ml before induced after the first vaccination cycle and antibody receiving 1/3 dose to 3.9 U/ml after completing the levels were further enhanced after the second vaccination vaccination cycle with the full dose. A cellular response treatment. The mean OD (415 nm, dilution 1:800) was not detected after vaccinating with 1/3 dose and could increased from 0.04 before the first vaccination with 1/3 not be tested after the full dose, because no material was dose to a maximum of 0.92 two weeks after the third available for these analyses.

Cancer Gene Therapy Safety of ALVAC-p53 in colorectal cancer AG Menon et al 515 Tumor response treatment cycle with 1/3 dose. After the second treatment cycle with the full dose, the follow-up CT scan, performed CEA levels increased in 15/16 patients during the study. 7 weeks after the last vaccination, revealed stable disease Patient 10 showed a relatively stable CEA level during the (see Fig 2). study. It was 196 mg/l before the first vaccination and 171 mg/l 14 weeks after the last vaccination. Disease progression, however, was outspoken clinically and radiographically. Metastases were evaluated by CT scan 14 weeks after Discussion the last vaccination, except for patients 02, 09 and 17 for In this study we administered a recombinant canarypox whom CT scans were performed 2, 4 and 8 weeks after the virus encoding the human wild-type p53 gene (ALVAC- last vaccination, respectively. All patients showed pro- p53) intravenously to 16 end-stage colorectal cancer gressive disease, including patient 08 after his first patients to assess safety, immunological reactions and tumor response. Immune responses were also evaluated and both p53-specific humoral and cellular responses Table 4 Vaccination-induced humoral and cellular immune re- were observed.24 The patient who showed stable disease sponses against ALVAC and p53 after additional compassionate use-vaccinations with the full dose, illustrates that immune reactions may take time Group 1 Group 2 Group 3 to develop and may need boosting to achieve clinical effects. Importantly, the antitumor effect occurred in the Anti-ALVAC absence of damage to healthy cells and the vaccine-related Antibodies (IgG) 4 5 5 Cellular 5 4 5 toxicity was minimal, even after repetitive administra- Anti-p53 tions. Fever was the only adverse event that was Antibodies (IgG) 1 1 1 consistently related to ALVAC-p53 treatment and was Cellular 0 0 2 most outspoken in patients who received the highest dose of ALVAC-p53. However, there was evidence in one

Figure 2 Patient 08 (group 2) received three vaccinations with 1/3 dose between November 20, 1999 and January 31, 2000. A baseline CT was made 1 week prior to the first vaccination with 1/3 dose (a) and shows two liver metastases (white arrows). At 20 weeks (May 9, 2000) after the first vaccination, a follow-up CT scan (b) was made and showed disease progression of the liver metastases. The lung metastases had also grown (not shown). At 6 months (November 6, 2000) after this scan, compassionate treatment with the full dose was started. His baseline (October 30, 2000, c) and follow-up (January 29, 2001, d) CT scans after compassionate treatment showed stable disease.

Cancer Gene Therapy Safety of ALVAC-p53 in colorectal cancer AG Menon et al 516 patient, that vaccine-induced fever could be prevented The safety profile described here and the ability to with the use of nonsteroidal anti-inflammatory drugs. induce anti-p53 responses have prompted us to further Other adverse events consisted mainly of tumor- optimize the ALVAC-p53 vaccine, for instance by using related complications, such as abdominal pain, fatigue, multiple tumor antigens (e.g. CEA, Ep-Cam), adding co- elevated liver enzymes and anemia. These symptoms stimulatory signals, and most importantly, by applying occur frequently in end-stage colorectal cancer pa- this strategy to patients who might truly benefit from this tients.25–27 immune-mediated anti-cancer approach, that is, patients Mutations in the p53 gene occur in approximately 50% with minimal residual disease. of malignancies and usually result in overexpression in the malignant cells only.3 The combination of increased steady-state levels and predominant expression by malignant cells provided an opportunity for a p53-specific Acknowledgments immunotherapeutic intervention. Experimental animal models have shown that the induction of anti-p53 We are very grateful to Mrs NG Ensink, Mr RLP van responses can be used to eradicate p53-overexpressing Vlierberghe and Dr AR Wafelman for their help in tumors without toxicity or autoimmunity.28,29 The safety preparing the vaccine for each vaccination. We thank Mrs of the ALVAC canarypox virus as a vector as well as its MG Kallenberg-Lantrua and Mrs AMEG Voet-van den ability to induce an immune response against the product Brink for assistance during the vaccinations, outpatient of the incorporated gene has been well established in visits and data entry. Mrs E Kampert’s contributions to various malignancies30,31 and infectious diseases.15,32–34 the establishment of the safety guidelines for this study Other viral vectors may be less efficient owing to pre- and the help of Mrs WM Meershoek-Klein Kranenbarg existing anti-viral antibodies,35 the immunogenicity of the and Mr J Junggeburt with setting-up the databases and viral vector or toxicity.36,37 For instance, a study in which performing the data analyses are also greatly acknowl- adenovirus encoding wild-type p53 was injected intratu- edged. Mrs C Di Vita-Feignier and Mrs N van Lommel morally in patients with advanced non-small cell lung were of great assistance during the checks of the case cancer did not result in the induction of an anti-p53 report forms. Most importantly, we thank the patients response.35 and their families for their participation in this study. Despite the significant immune suppression in the end- stage cancer patients participating in our trial, immune responses against p53 were induced in the group that received the highest ALVAC-p53 dose.24 This positive References immune response was not accompanied by significant clinical responses, but this was not surprising in view of 1. Henderson RA, Finn OJ. Human tumor antigens are ready the fact that these patients had large tumor burdens that to fly. Adv Immunol. 1996;62:217–256. were diagnosed 25 months on average before treatment 2. Lane DP. Cancer. p53, guardian of the genome. Nature. 1992;358:15–16. with ALVAC-p53. As we evaluated the tumor response a 3. Hollstein M, Sidransky D, Vogelstein B, Harris CC. p53 long time (14 weeks) after completing vaccinations, it is mutations in human cancers. Science. 1991;253:49–53. possible that ALVAC-p53 did induce a clinical response, 4. Tollenaar RA, van Krieken JH, van Slooten HJ, et al. but that this response was no longer detectable 14 weeks Immunohistochemical detection of p53 and Bcl-2 in color- later. Other investigators report anti-tumor responses in ectal carcinoma: no evidence for prognostic significance. Br end-stage patients, but assess these earlier after stopping J Cancer 1998;77:1842–1847. the experimental treatment, for example, from 0 weeks,25 5. Soussi T. p53 Antibodies in the sera of patients with various 2 weeks, 4 weeks,26 4–6 weeks,38 or 8 weeks39 after types of cancer: a review. Cancer Res. 2000;60:1777–1788. stopping treatment. However, we observed disease 6. Shiota G, Ishida M, Noguchi N, et al. Circulating p53 progression at occasional earlier radiographic evalua- antibody in patients with colorectal cancer: relation to clinicopathologic features and survival. Dig Dis Sci. tions, such as in patients 02 (1/10 dose), 09 (1/3 dose) and 2000;45:122–128. 17 (full dose) who underwent CT scans on week 2, 4 weeks 7. Tilkin AF, Lubin R, Soussi T, et al. Primary proliferative T and 8 weeks after the last vaccination, respectively. Most cell response to wild-type p53 protein in patients with breast interestingly, we documented stable disease 7 weeks after cancer. Eur J Immunol. 1995;25:1765–1769. the completion of a second cycle of treatment with the full 8. Houbiers JG, Nijman HW, van der Burg SH, et al. In vitro dose in a patient who had had a first cycle of treatment induction of human cytotoxic T lymphocyte responses with 1/3 dose previously. The observed increase in IgG against peptides of mutant and wild-type p53. Eur J antibody titer against p53 after the vaccination cycles may Immunol. 1993;23:2072–2077. indeed reflect the slowinduction of an immune response 9. van der Burg SH, de Cock K, Menon AG, et al. Long induced by the vaccinations. Alternatively, it may reflect lasting p53-specific T cell memory responses in the absence 40,41 of anti-p53 antibodies in patients with resected primary an increase in tumor burden. Although this was colorectal cancer. Eur J Immunol. 2001;31:146–155. observed in only one subject, this indicates that it might 10. Ropke M, Hald J, Guldberg P, et al. Spontaneous human be necessary to repeat vaccinations more often, as squamous cell carcinomas are killed by a human cytotoxic T 42 suggested by others, while using NSAIDs to prevent lymphocyte clone recognizing a wild-type p53-derived the postvaccination fever. peptide. Proc Natl Acad Sci USA. 1996;93:14704–14707.

Cancer Gene Therapy Safety of ALVAC-p53 in colorectal cancer AG Menon et al 517 11. Paoletti E. Applications of pox virus vectors to vaccination: human carcinoembryonic antigen and the B7.1 co-stimula- an update. Proc Natl Acad Sci. USA. 1996;93:11349–11353. tory molecule. Cancer Immunol Immunother. 2000;49:504– 12. Hurpin C, Rotarioa C, Bisceglia H, et al. The mode of 514. presentation and route of administration are critical for the 27. Scheithauer W, Kornek GV, Ulrich-Pur H, et al. Oxaliplatin induction of immune responses to p53 and antitumor plus raltitrexed in patients with advanced colorectal immunity. Vaccine. 1998;16:208–215. carcinoma: results of a Phase I-II trial. Cancer. 13. Roth J, Dittmer D, Rea D, et al. p53 as a target for cancer 2001;91:1264–1271. vaccines: recombinant canarypox virus vectors expressing 28. Vierboom MP, Nijman HW, Offringa R, et al. Tumor p53 protect mice against lethal tumor cell challenge. Proc eradication by wild-type p53-specific cytotoxic T lympho- Natl Acad Sci USA. 1996;93:4781–4786. cytes. J Exp Med. 1997;186:695–704. 14. Rosenwirth B, Kuhn EM, Heeney JL, et al. Safety and 29. Mayordomo JI, Loftus DJ, Sakamoto H, et al. Therapy of immunogenicity of ALVAC wild-type human p53 (vCP207) murine tumors with p53 wild-type and mutant sequence by the intravenous route in rhesus macaques. Vaccine. peptide- based vaccines. J Exp Med. 1996;183:1357–1365. 2001;19:1661–1670. 30. Paoletti E, Tartaglia J, Cox WI. Immunotherapeutic 15. Cadoz M, Strady A, Meignier B, et al. Immunisation with strategies for cancer using poxvirus vectors. Ann NY Acad canarypox virus expressing rabies glycoprotein. Lancet. Sci. 1993;690:292–300. 1992;339:1429–1432. 31. Tartaglia J, Bonnet M, Berinstein N, et al. Therapeutic 16. Fries LF, Tartaglia J, Taylor J, et al. Human safety and vaccines against melanoma and colorectal cancer. Vaccine immunogenicity of a canarypox–rabies glycoprotein recom- 2001;19:2571–2575. binant vaccine: an alternative poxvirus vector system. 32. Egan MA, Pavlat WA, Tartaglia J, et al. Induction of Vaccine. 1996;14:428–434. human immunodeficiency virus type 1 (HIV-1)-specific 17. Ferrari G, Berend C, Ottinger J, et al. Replication-defective cytolytic T lymphocyte responses in seronegative adults by canarypox (ALVAC) vectors effectively activate anti-human a nonreplicating, host-range-restricted canarypox vector immunodeficiency virus-1 cytotoxic T lymphocytes present (ALVAC) carrying the HIV-1MN env gene. J Infect Dis. in infected patients: implications for antigen-specific im- 1995;171:1623–1627. munotherapy. Blood. 1997;90:2406–2416. 33. Belshe RB, Stevens C, Gorse GJ, et al. Safety and 18. Clements-Mann ML, Weinhold K, Matthews TJ, et al. immunogenicity of a canarypox-vectored human immuno- Immune responses to human immunodeficiency virus (HIV) deficiency virus Type 1 vaccine with or without gp120: a type 1 induced by canarypox expressing HIV-1MN gp120, phase 2 study in higher- and lower-risk volunteers. J Infect HIV-1SF2 recombinant gp120, or both vaccines in serone- Dis. 2001;183:1343–1352. gative adults. NIAID AIDS Vaccine Evaluation Group. J 34. Berencsi K, Gyulai Z, Gonczol E, et al. A canarypox vector- Infect Dis. 1998;177:1230–1246. expressing cytomegalovirus (CMV) phosphoprotein 65 19. Marshall JL, Hawkins MJ, Tsang KY, et al. Phase I study in induces long-lasting cytotoxic T cell responses in human cancer patients of a replication-defective avipox recombi- CMV- seronegative subjects. J Infect Dis. 2001;183:1171– nant vaccine that expresses human carcinoembryonic anti- 1179. gen. J Clin Oncol. 1999;17:332–337. 35. Yen N, Ioannides CG, Xu K, et al. Cellular and humoral 20. von Mehren M, Arlen P, Tsang KY, et al. Pilot study of a immune responses to adenovirus and p53 protein antigens in dual gene recombinant avipox vaccine containing both patients following intratumoral injection of an adenovirus carcinoembryonic antigen (CEA) and B7.1 transgenes in vector expressing wild-type. P53 (Ad-p53). Cancer Gene patients with recurrent CEA-expressing adenocarcinomas. Ther. 2000;7:530–536. Clin Cancer Res. 2000;6:2219–2228. 36. van der Eb MM, Cramer SJ, Vergouwe Y, et al. Severe 21. World Health Organization. WHO Handbook for Reporting hepatic dysfunction after adenovirus-mediated transfer of Results of Cancer Treatment. Geneva:WHO offset publica- the herpes simplex virus thymidine kinase gene and tion, 1979. ganciclovir administration. Gene Therapy 1998;5:451–458. 22. Aarden LA, de Groot ER, Feltkamp TE. Immunology of 37. Curiel D, Reynolds P. Gene therapy death highlights the DNA. III. Crithidia luciliae, a simple substrate for the remaining challenges. Helix 2000;9:21–25. determination of anti-dsDNA with the immunofluorescence 38. Ward RL, Packham D, Smythe AM, et al. Phase I clinical technique. Ann NY Acad Sci. 1975;254:505–515. trial of the chimeric monoclonal antibody (c30.6) in patients 23. Hay FC, Nineham LJ, Roitt IM. Routine assay for the with metastatic colorectal cancer. Clin Cancer Res. detection of immune complexes of known immunoglobulin 2000;6:4674–4683. class using solid phase C1q. Clin Exp Immunol. 1976;24:396– 39. Wittig B, Marten A, Dorbic T, et al. Therapeutic 400. vaccination against metastatic carcinoma by expressionmo- 24. van der Burg SH, Menon AG, Redeker A, et al. Induction dulated and immunomodified autologous tumor cells: of p53-specific immune responses in colorectal cancer a first clinical phase I/II trial. Hum Gene Ther. patients receiving a recombinant ALVAC-p53 candidate 2001;12:267–278. vaccine. Clin Cancer Res. 2002;8:1019–1027. 40. Houbiers JG, van der Burg SH, van de Watering LM, et al. 25. von Mehren M, Arlen P, Gulley J, et al. The influence of Antibodies against p53 are associated with poor prognosis granulocyte macrophage colony-stimulating factor and of colorectal cancer. Br J Cancer 1995;72:637–641. prior chemotherapy on the immunological response to a 41. Soussi T. The humoral response to the tumor-suppressor vaccine (ALVAC- CEA B7.1) in patients with metastatic gene-product p53 in human cancer: implications for carcinoma. Clin Cancer Res. 2001;7:1181–1191. diagnosis and therapy. Immunol Today. 1996;17:354–356. 26. Horig H, Lee DS, Conkright W, et al. Phase I clinical trial of 42. Zinkernagel RM. Immunity against solid tumors? Int J a recombinant canarypoxvirus (ALVAC) vaccine expressing Cancer. 2001;93:1–5.

Cancer Gene Therapy