Volume 18, Issue 10, Pages 1874-1884 (October 2010) Treatment of Cancer Patients With a Serotype 5/3 Chimeric Oncolytic Adenovirus Expressing GMCSF  Anniina Koski, Lotta Kangasniemi, Sophie Escutenaire, Sari Pesonen, Vincenzo Cerullo, Iulia Diaconu, Petri Nokisalmi, Mari Raki, Maria Rajecki, Kilian Guse, Tuuli Ranki, Minna Oksanen, Sirkka-Liisa Holm, Elina Haavisto, Aila Karioja-Kallio, Leena Laasonen, Kaarina Partanen, Matteo Ugolini, Andreas Helminen, Eerika Karli, Päivi Hannuksela, Saila Pesonen, Timo Joensuu, Anna Kanerva, Akseli Hemminki  Molecular Therapy  Volume 18, Issue 10, Pages 1874-1884 (October 2010) DOI: 10.1038/mt.2010.161 Copyright © 2010 The American Society of Gene & Cell Therapy Terms and Conditions

Figure 1 Ad5/3-D24-GMCSF induces cell killing and expression of functionally active GMCSF in vitro. Viability of (a) MDA-MB-436 and (b) A549 cells after infection with Ad5luc1, Ad5wt, and Ad5/3-D24-GMCSF. (c) GMCSF secretion by A549 cells after infection with 100 VP/cell Ad5/3-D24-GMCSF. (d) TF1 cells, whose viability is dependent on functional human GMCSF (hGMCSF) were cultured in the presence of 2 ng/ml hGMCSF or indicated amount of filtered supernatant from Ad5/3-D24-GMCSF-infected cells. Normal growth media was used as negative control. Viability of cells was determined after 5 days; viability of cells with commercial hGMSCF was set as 100%. hGMCSF, human granulocyte–macrophage colony–stimulating factor; VP, virus particles. Molecular Therapy 2010 18, 1874-1884DOI: (10.1038/mt.2010.161) Copyright © 2010 The American Society of Gene & Cell Therapy Terms and Conditions

Figure 2 Antitumor efficacy and tumor selectivity of Ad5/3-D24-GMCSF in immune-competent Syrian hamsters. Syrian hamsters were inoculated subcutaneously with HapT1 cells. (a,b) Tumors were injected with 1 × 108 VP/tumor on days indicated by arrows, 1/5 of dose was given intravenously on day 1. NaCl was used as mock treatment. *P ≤ 0.05 against mock. In b the effect of low-dose cyclophosphamide (CP) in conjunction with Ad5/3-D24-GMCSF was tested and found to increase efficacy. Figure indicates tumor volumes recorded 15 days after injection of virus. (c,d) HapT1 tumors were grown and injected once with 1 × 108 VP/tumor Ad5/3-D24-GMCSF. (c) Virus replication was studied with quantitative PCR. To evaluate tumor selectivity of the virus, livers of non-tumor-bearing hamsters were injected and no replication was seen. Viral E4 copy number was normalized to genomic DNA with GAPDH primers. (d) Human GMCSF concentration was measured in virus injected tumors, serum and livers of tumor-bearing hamsters and livers of non-tumor-bearing hamsters injected into the liver. Data are expressed as mean ± SEM, *P ≤ 0.05 and ***P ≤ 0.005 versus the 0.5-hour time point. GMCSF, granulocyte–macrophage colony–stimulating factor; VP, virus particles. Molecular Therapy 2010 18, 1874-1884DOI: (10.1038/mt.2010.161) Copyright © 2010 The American Society of Gene & Cell Therapy Terms and Conditions

Figure 3 Radiological responses to treatment with Ad5/3-D24-GMCSF and survival after treatment. (a) Pre- and (b) post-treatment computed tomography scans of K75 with complete resolution of pleural effusion, O129 with reduction of peritoneal tumors, and I98 illustrating reduction of a liver metastases (note: post-treatments scan for this patient is a magnetic resonance image). (c) Kaplan–Meier analysis of the survival of patients. Censored refers to patients who were still alive at the time of submission of the manuscript. Molecular Therapy 2010 18, 1874-1884DOI: (10.1038/mt.2010.161) Copyright © 2010 The American Society of Gene & Cell Therapy Terms and Conditions

Figure 4 Ad5/3-D24-GMCSF treatment influences adenovirus- and tumor-specific cytotoxic T-lymphocytes. Total peripheral blood mononuclear cells were isolated from treated patients before and a month after treatment and pulsed with (a) an adenovirus type 5 penton-derived peptide pool and (b) a survivin-derived peptide pool. Interferon-γ ELISPOT was performed. The value before treatment was assigned 100 and the value after treatment is expressed relative to that. Spot forming colonies (SFC) are expressed as a mean of triplicate experiments, and background (SFC without peptide) was subtracted. Patients were grouped depending on the change in SFC after treatment; increased (>50%), no change, and decreased (>50%). Molecular Therapy 2010 18, 1874-1884DOI: (10.1038/mt.2010.161) Copyright © 2010 The American Society of Gene & Cell Therapy Terms and Conditions

Figure 5 5/3 chimeric adenoviruses are able to transfect and kill cells in ex vivo ascites and pleural effusion samples. Before treatment cells from (a) pleural effusion of V136 and (b) ascites of K75 were infected with Ad5luc1 (Ad5 capsid) and Ad5/3luc1 (Ad5/3 capsid) at 5,000 VP/cell and analyzed for luciferase expression (expressed as relative light units, RLU). ***P < 0.005 against Ad5luc1. Cells from pretreatment samples of pleural effusion of (c) V136 and (d) M137 were infected with 100 VP/cell Ad5/3-D24-GMCSF or the nonreplicative control. Viability of cells was analyzed 6 days later. ***P < 0.005 against uninfected cells. (e) Copy numbers of Ad5/3-D24-GMCSF were analyzed with quantitative PCR from serum samples and a post-treatment ascites sample of O82. NA, not analyzed; VP, virus particles. Molecular Therapy 2010 18, 1874-1884DOI: (10.1038/mt.2010.161) Copyright © 2010 The American Society of Gene & Cell Therapy Terms and Conditions