Cancer Therapy: Preclinical

Intratumoral Coadministration of Hyaluronidase and Oncolytic Adenoviruses Enhances Potency in Metastatic Tumor Models Shanthi Ganesh, Melissa Gonzalez-Edick, Douglas Gibbons, Melinda Van Roey, and Karin Jooss

Abstract Purpose: Evaluate the codelivery of hyaluronidase enzyme with oncolytic adenoviruses to determine whether it improves the spread of the virus throughout tumors, thereby leading to a greater overall antitumor efficacy in tumor models. Experimental Design: The optimal dose of hyaluronidase that provided best transduction efficiency and spread of a green fluorescent protein (GFP)-expressing adenovirus within tumors was combined with oncolytic in tumor models to determine whether the combination treatment results in an improvement of antitumor efficacy. Results: In mice injected with the adenovirus Ad5/35GFP and an optimal dose of hyaluronidase (50 U), a significant increase in the number of GFP-expressing cells was observed when compared with animals injected with virus only (P < 0.0001).When the oncolytic adenoviruses Ad5OVor Ad5/35 OV (OV-5or OV5T35H) were codelivered with 50 U of hyaluronidase, a signifi- cant delay in tumor progression was observed, which translated into a significant increase in the mean survival time of tumor-bearing mice compared with either of the monotherapy-treated groups (P < 0.0001). Furthermore, the mice that received the combination of Ad5/35 OVand hyaluronidase showed the best antitumor efficacy. Importantly, the combination treatment did not increase the metastatic potential of the tumors. Lastly, the increase in virus potency observed in animals injected with both enzyme and virus correlated with enhanced virus spread throughout tumors. Conclusion: Antitumor activity and overall survival of mice bearing highly aggressive tumors are significantly improved by codelivery of oncolytic adenoviruses and hyaluronidase when compared with either of the monotherapy-treated groups, and it may prove to be a potent and novelapproachtotreatingpatientswithcancer.

The limited ability of Ad5-based oncolytic viruses to infect adenoviruses (7). Although the fiber-chimeric viruses enter tumor tumor cells efficiently and to spread throughout solid tumors cells more effectively than Ad5-based viruses, the inability of both may hamper their effectiveness in the clinic (1, 2). Due to of these viruses to spread efficiently throughout a tumor mass variable expression levels of the -adenovirus remains a major obstacle to the development of a successful receptor on tumor cells, the surface receptor used by Ad5 adenovirus-based therapy for solid tumors (9). Based on recent viruses to bind and enter cells, the initial transduction of tumor studies, connective tissue and extracellular matrix (ECM) com- cells by Ad5-based oncolytic adenoviruses is limited (3–7). By ponents apparently have a prominent role in inhibiting viral replacing the Ad5 fiber knob with the Ad35 knob, which spread after intratumoral administration of adenovirus (1, 9–11), recognizes the highly abundant CD46 receptor on tumor cells, and previously published data suggest that there is enhanced viral tumor cell transduction is significantly improved (7, 8). distribution within the tumor when ECM is degraded by The efficient viral entry of Ad5/35 fiber-chimeric viruses proteolytic such as trypsin or collagenase (12, 13). correlates well with their improved antitumor efficacy in multiple Hyaluronan is a high molecular weight nonsulfated glycos- different xenograft models compared with Ad5-based oncolytic aminoglycan that is an important constituent of the ECM surrounding cancer cells where it forms large networks with proteoglycans (14). By degrading the hyaluronan-rich matrixwith hyaluronidase, a naturally occurring enzyme that Authors’ Affiliation: Cell Genesys, Inc., South San Francisco, California depolymerizes the viscoelastic ECM components, the barrier for Received 10/25/07; revised 1/24/08; accepted 2/22/08. The costs of publication of this article were defrayed in part by the payment of page drug/virus dispersion throughout the interstitial spaces may be charges. This article must therefore be hereby marked advertisement in accordance improved (14, 15). The degradation of hyaluronan is likely to with 18 U.S.C. Section 1734 solely to indicate this fact. alter the configuration of the ECM and, thereby, to reduce Requests for reprints: Shanthi Ganesh, Cell Genesys, Inc., 500 Forbes interstitial fluid pressure, which is another obstacle to efficient Boulevard, South San Francisco, CA 94080. Phone: 415-260-3654; E-mail: [email protected]. spread of therapeutic agents during tumor treatment (16). F 2008 American Association for Cancer Research. Recombinant human hyaluronidase enzyme (rHuPH20) has doi:10.1158/1078-0432.CCR-07-4732 been developed and evaluated by Halozyme Therapeutics, Inc

www.aacrjournals.org 3933 Clin Cancer Res 2008;14(12) June 15, 2008 Downloaded from clincancerres.aacrjournals.org on September 27, 2021. © 2008 American Association for Cancer Research. Cancer Therapy: Preclinical

(17). Their studies have shown that rHuPH20 increases the Antitumor activity and survival of tumor-bearing animals in transduction efficiency of an adenovirus expressing either green both models are significantly improved by combination fluorescent protein (GFP) or the therapeutic gene PDGF-B in treatment with virus and enzyme, compared with treatment mouse skin (17). with either virus or enzyme alone. Moreover, these viruses in To test the effects of hyaluronidase on virus transduction and conjunction with enzyme do not increase the metastatic spread spread in solid tumors, we selected two distinct but highly of tumor cells in a highly metastatic model. Such combination metastatic tumor models: PC-3.luc (human prostate) and treatment may therefore offer a novel and improved oncolytic A375-mln1.luc (human ). Both of these models therapy for patients with cancer. were used previously to test the effect of relaxin expressed by oncolytic viruses on virus spread, and we were interested to evaluate hyaluronidase coinjected with oncolytic viruses in Materials and Methods these same models (11). Furthermore, PC-3 tumors have high levels of hyaluronan, and it was previously shown that Nonreplicating and replicating viruses. Ad5/35GFP is an E1-deleted, expression of hyaluronan biosynthetic and processing enzymes nonreplicating virus in which a GFP , under the control of promote growth and vascularization of PC-3 tumors in mice, a strong constitutive cytomegalovirus , has been incorporated thereby stimulating tumorigenicity and metastasis (18, 19). into the E1 region of an adenoviral vector that contains the fiber knob from Ad35 and the fiber shaft from Ad5 adenovirus (7). Oncolytic Tumor-produced hyaluronidase has also been shown to viruses OV-5 and OV-5T35H are conditionally replication-competent degrade hyaluronic acid into smaller oligomeric units and adenoviruses that were engineered to replicate selectively in tumor cells may possibly promote metastasis by inducing (7, 11). In these viruses, the tumor-specific E2F-1 promoter replaces the (18–24). Interestingly, hyaluronan and hyaluronidase have native E1a promoter to restrict viral replication to cells that have a been detected in the invasive areas of highly aggressive PC-3 defect in the pRb pathway (27, 28). In addition, the cDNA encoding tumors (18–24). A375-mln-1, a human melanoma model, is human granulocyte- colony-stimulating factor replaces the another highly metastatic tumor model, and overexpression of open reading frame encoding the viral 19-kDa glycoprotein in the E3 hyaluronidase in human melanoma cell lines has been shown region. With the aim of improving viral entry and spread throughout to induce vascularization of tumors in mice (20). On the other the tumor, the Ad5 fiber knob of OV-5 has been replaced with the Ad35 knob to generate OV-5T35H (7). hand, literature suggests that eradication of a hyaluronan-rich Tumor cell lines. The tumor cell lines used in the study were environment by i.v. administration of hyaluronidase to severe cultured in RPMI 1640 supplemented with 10% fetal bovine serum. combined immunodeficient mice bearing human tumor Human prostate carcinoma PC-3 cells were purchased from American xenografts could reduce the aggressiveness of tumors, thereby Type Culture Collection and have been modified to express potentially offering a new anticancer strategy (14, 25, 26). (PC-3.luc; ref. 30). A375-mln1.luc is a human cell line established by Based on the above findings and the important role of in vivo selection of lymph node metastases from A375-mice bearing s.c. hyaluronidase as a matrixdegradative protein, we hypothesized tumors (31). that rHuPH20 might facilitate spread of adenovirus within In vivo efficacy studies. Female NCR (nu/nu) mice (ages 4-6 wk; tumors due to the removal of the ECM that may act as a body weight, 18-20 g) were purchased from Simonsen Laboratories. Â 6 naturally occurring structural barrier to virus spread within Mice were injected s.c in the right flank with either 5 10 PC-3.luc cells or 2 Â 106 A375-mln1.luc cells (injection volume of 100 AL). tumors. When the PC-3.luc tumors reached the desired mean tumor volume We have coadministered rHuPH20 with a fiber-chimeric (125-150 mm3) as determined using the formula, Volume = W Â (L)2/ E1-deleted adenovirus expressing GFP (Ad5/35GFP) to animals 2, wherein W is width and L is length in millimeters, animals were bearing human prostate PC-3.luc tumors, monitored GFP randomly distributed into treatment groups (n = 3) and injected expression within the tumors, and compared it to tumors intratumorally with rHuPH20 enzyme (recombinant human hyaluron- injected with either virus or rHuPH20 alone. These studies idase from Halozyme Therapeutics, Inc.) in PBS at a dose of 800 U per show that rHuPH20, when coadministered with adenovirus, mouse per treatment, Ad5/35GFP virus at a dose of 1 Â 1010 viral 10 enhances virus spread throughout the tumor when compared particles (vp in PBS), or a combination of Ad5/35GFP virus (1 Â 10 A with virus treatments alone. vp) and rHuPH20 enzyme at a dose of 50, 200, or 800 U in 50 L PBS. To determine if the improved spread of these adenovirus Three days after virus injection, the tumors were harvested and immunohistochemical analysis was done to detect the GFP-positive vectors in the presence of enzyme translates into improved cells. Specifically, 5-Am serial sections were generated throughout the antitumor efficacy and survival, the PC-3.luc and A375- tumors and analyzed to select the section with peak GFP expression for mln1.luc xenograft models were used to test oncolytic each tumor. From the peak GFP expressing section, every third serial adenoviruses and rHuPH20 for their effects on survival. The section in both directions (total of 12 sections per tumor) was selected, adenoviruses used in these studies are the oncolytic virus OV-5 and a quadrant (consisting of 4 fields) in the center of these sections (containing the original Ad5 fiber knob) and OV-5T35H was analyzed at a Â10 magnification for the number of GFP-expressing (containing the Ad35 knob; ref. 7). In both viruses, the cells using an image analysis software program (Image Pro Plus; Media tumor-specific E2F-1 promoter replaces the native E1a promot- Cybernetics). The average number of GFP-expressing cells per field was er to restrict viral replication to cells that have a defect in the determined from 3 animals per group, analyzing a total of 144 fields pRb pathway (27, 28). In addition, these viruses carry the per group. For efficacy studies, mice bearing established PC-3.luc (100-150 cDNA encoding human granulocyte-macrophage colony- mm3 or f250 mm3) or A375-mln1.luc (100 mm3) tumors received stimulating factor in place of the open reading frame encoding four or sixintratumoral injections (50 AL) of PBS, OV-5 or OV-5T35H the non-essential viral 19-kDa glycoprotein in the E3 region virus (1 Â 1010 vp), rHuPH20 enzyme (50 U per mouse), or a (7, 29). OV-5 or OV-5T35H and rHuPH20 were also evaluated combination of virus and rHuPH20, once every other day. Tumor for their effects on metastatic spread of primary, metastatic progression was monitored with survival as the end point. Animals A375-mln1.luc tumors in this study. were euthanized when tumor volumes reached 2,000 mm3 or if the

Clin Cancer Res 2008;14(12) June 15, 2008 3934 www.aacrjournals.org Downloaded from clincancerres.aacrjournals.org on September 27, 2021. © 2008 American Association for Cancer Research. Hyaluronidase Enzyme Enhances Virus Potency tumors became necrotic. Animals from each group were selected at virus at a dose of 1 Â 1010 vp or rHuPH20 enzyme at a dose various time points after the initial treatment, bled, euthanized, and of 800 U per mouse, or a combination of Ad5/35GFP virus tumors collected. Tumors were cut into halves, with one half used to (1 Â 1010 vp) and rHuPH20 enzyme at doses ranging from determine viral spread by immunohistochemistry and the other half 50 to 800 U. Three days postinjection, the tumors (n = 3) were used to detect virus replication. For all in vivo efficacy studies, each harvested and the number of GFP-positive cells evaluated in group consisted of at least 10 mice. Tumor volume between different groups was compared by linear regression analysis. Survival times of tissue sections by immunohistochemistry as described in mice from different groups was compared by Kaplan-Meier analysis. material and methods. An average of 79.2 GFP-positive cells For histologic examination, organs were collected in 4% parafor- per field was found in tumors injected with virus plus maldehyde in PBS and left in the solution for 3 h for complete fixing. hyaluronidase compared with only 15.5 GFP-positive cells per The samples were then transferred to 30% sucrose solution for field in tumors injected with virus alone (Fig. 1). The two higher cryoprotection and left overnight at 4C. Samples were rinsed in PBS, doses of the enzyme (200 or 800 U) coinjected with virus blotted dry, and blocks of respective tissues were embedded in resulted in averages of 35.2 and 20.2 GFP-positive cells per field, ornithine carbamyl transferase (STAT Lab Medical Products) before respectively. These results confirmed that virus coinjected with freezing in a methyl butane/dry ice bath. Five-microliter serial sections 50 U of hyaluronidase leads to a more widespread distribution were collected and directly analyzed for GFP expression using of virus within tumors than virus injected alone or virus fluorescence microscopy. GFP-positive cells in the tumors was quantified by determining the average number of GFP-positive cells injected with higher doses of the enzyme, suggesting that per Â10 field of view using an image analysis software program (Image although spread of virus seemed to be enhanced with the two Pro Plus; Media Cybernetics). To detect the hexon-positive cells, 5-Am higher doses of hyaluronidase, it was optimal with the 50 U paraffin sections of tumor tissue were stained for the adenoviral hexon dose of the enzyme and, thus, this dose was considered the protein using Zymed’s Histomouse kit following the manufacturer’s optimal dose of the enzyme to increase virus spread throughout instructions (11). The remaining half of the tumor was used to isolate the tumor and selected for further studies. virus, and the titer was determined by infecting AE1-2a cells with the Intratumoral hyaluronidase coinjection with fiber-chimeric recovered virus (7, 11). To monitor metastasis of tumor cells, 6 lymph enhances the antitumor efficacy of the virus nodes (LN), including the axillary and inguinal nodes, and the lungs in the PC-3.luc model. To determine if the increased virus were collected at the selected time points, microinjected with luciferin transduction observed with the combination of E1-deleted substrate (15 mg/mL), and imaged after 2 min using the Xenogen imaging system (Xenogen Corp) to detect luciferase-positive metastases adenovirus and hyaluronidase in PC-3.luc tumors translated (11, 31). into enhanced efficacy, the fiber-chimeric oncolytic adenovirus was evaluated either as monotherapy or in combination with Results hyaluronidase in the PC-3.luc xenograft model with survival as the end point. Intratumoral coadministration of hyaluronidase enzyme Nude mice bearing 100 to 150 mm3 PC-3.luc tumors received with Ad5/35GFPfiber-chimeric adenovirus enhances virus a total of four injections (50 AL intratumorally), given once transduction/spread within established (PC- every other day, of PBS, OV-5T35H (1 Â 1010 vp per injection), 3.luc) tumors. To determine whether degradation of the rHuPH20 enzyme (50 U per injection), or the combination of ECM component hyaluronan by hyaluronidase enhances virus these same doses of virus and enzyme. Tumor progression was transduction of solid tumors, Ad5/35 E1-deleted adenovirus- monitored over time with survival as the end point. None of the expressing GFP was tested in combination with rHuPH20 in the animals exhibited any symptoms of toxicity or body weight loss PC-3.luc xenograft model. In this model, female nude mice as a result of any of the treatments (data not shown). were injected s.c. with 5 Â 106 PC-3.luc cells. Once the tumors In the groups treated with either PBS or enzyme alone, tumor reached a volume of 125 to 150 mm3, tumor-bearing mice volumes increased by 9.3- and 11-fold, respectively, over a (n = 3) were injected intratumorally with either Ad5/35GFP period of 33 days. In the group treated with OV-5T35H alone,

Fig. 1. Coadministration of recombinant hyaluronidase increases the spread of Ad5/ 35GFP virus in the PC-3.luc model. Nude mice (n = 3 per group) bearing s.c PC-3.luc tumors of f15 0mm 3 were injected intratumorally with either hyaluronidase (rHuPH20; 800 U per tumor) orAd5/ 35virus (1 Â1010 vp) or a combination of Ad5/35GFP virus (1 Â1010 vp) and rHuPH20 at doses of 50, 200, or 800 U. Three days postinjection, tumors were collected and GFP expression evaluated in tissue sections by immunohistochemical analysis. (magnification, Â5).The number of GFP-positive cells was quantified by counting all positive cells in a central quadrant (4 fields) of 12 tissue sections localized at the area of peak GFP expression and averaging themfrom144 evaluated fields using an image analysis software program(Image Pro Plus; Media Cybernetics).

www.aacrjournals.org 3935 Clin Cancer Res 2008;14(12) June 15, 2008 Downloaded from clincancerres.aacrjournals.org on September 27, 2021. © 2008 American Association for Cancer Research. Cancer Therapy: Preclinical the tumor volume increased by only 3.3-fold (Fig. 2A), whereas rHuPH20 injected animals with a MST of 65.5 days (OV-5T35H in the group treated with the combination of virus and enzyme, versus rHuPH20 or PBS; P < 0.01). Animals treated with both the tumor volume increased only 1.3-fold during the same time OV-5T35H virus and rHuPH20 showed a further improvement frame. These data suggest that the fiber-chimeric oncolytic virus in survival with a MST not yet reached by day 115 (OV- in combination with rHuPH20 was more efficacious than the 5T35H+rHuPH20 versus OV-5T35H; P < 0.05; Fig. 2B). virus alone. Complete tumor regression was noted in 2 of 10 To determine if this potent combination therapy of virus mice treated with either OV-5T35H alone or with OV-5T35H and enzyme could also control larger PC-3.luc tumors, animals and enzyme. with established PC-3.luc tumors at a size of f250 mm3 were Animals treated with PBS alone had a mean survival time treated with a total of 6 injections (50 AL intratumorally), (MST) of 38.5 days with no animals alive by day 62, and given once every other day, of PBS, OV-5T35H virus (1 Â 1010 animals treated with rHuPH20 alone had a MST of 35 days vp per injection), rHuPH20 enzyme (50 U per injection), or a with no animals alive by day 69. In contrast, animals treated combination of OV-5T35H and rHuPH20, and animals were with OV-5T35H alone survived significantly longer than PBS or monitored over time to determine survival. Animals treated with the combination of OV-5T35H plus hyaluronidase showed a statistically significant survival advantage (P < 0.01; MST not reached) compared with animals treated with PBS (MST, 40 days), ruHuPH20 (MST, 38.5 days), or virus alone (MST 50, days), thereby emphasizing the potency of this combination therapy, even in animals with a very large tumor burden (data not shown). Intratumoral hyaluronidase coinjection with oncolytic adenovi- rus enhances the antitumor efficacy of the virus in A375-mln1.luc tumor-bearing mice. The combination of oncolytic virus and rHuPH20 enzyme was further tested in the highly metastatic A375-mln1.luc xenograft model to evaluate the potency of the virus and enzyme combination in another tumor model. Both the parental Ad5-based OV-5 virus and the fiber-chimeric OV- 5T35H virus were tested in the A375-mln1.luc model because the Ad5-based virus transduces this tumor less efficiently than the chimeric virus (11), and thus, this model can test the effect of rHuPH20 treatment on tumor transduction by two oncolytic viruses with very different potencies. Nude mice were injected s.c with 2 Â 106 A375-mln1.luc cells. When the tumors reached a volume of 100 mm3, tumor- bearing mice received a total of four injections (50 AL intratumorally), given once every other day, of PBS, OV-5 or OV-5T35H (1 Â 1010 vp per injection), rHuPH20 (50 U per injection), or a combination of either virus and enzyme. None of the animals exhibited any symptoms of toxicity or body weight loss as a result of these treatments (data not shown). In the groups treated with PBS or enzyme alone, the tumor volume increased 6-fold over a period of 22 days (Fig. 3A). In contrast, in the groups treated with either oncolytic virus alone or in combination with enzyme, there was a significant delay in tumor progression. Specifically, in the OV-5–treated mice, the tumor volume increased 5-fold, whereas in the OV-5T35H– treated animals, tumor volume increased only 3.5-fold during the same time period (OV-5T35H versus OV-5; P < 0.05). When OV-5 virus was combined with rHuPH20, the overall rate of tumor progression decreased compared with OV-5 treatment alone, with a 3.5-fold increase in tumor volume, which is similar to the increase in tumor volume observed with OV- 5T35H monotherapy. When mice were treated with OV-5T35H Fig. 2. Coadministration of recombinant hyaluronidase increases the antitumor virus and enzyme, tumor progression was further delayed efficacy of oncolytic fiber chimeric viruses in the PC-3.luc model. Nude mice (n = 13) bearing s.c PC-3.luc tumors (100-150 mm3) received a total of 4 compared with the OV-5T35H–treated group (P < 0.0001) as intratumoral injections (50 AL) of PBS, rHuPH20 (50 U per mouse), OV-5T35H well as to the OV-5 plus enzyme-treated group (P < 0.0001), virus (1 Â1010 vp per injection), or coinjection of OV-5T35H plus rHuPH20 once every other day.Tumor progression (A) and survival (B) were monitored over time. exhibiting an increase in tumor volume of 1.2-fold by day 22 A statistically significant delay in tumor progression was observed between the (Fig. 3A). Complete tumor regression was noted in 2, 4, and vector-treated groups compared with the PBS or rHuPH20-injected groups 5 of 15 mice treated with either OV-5T35H, OV-5 with enzyme, (OV-5T35H and OV-5T35H+rHuPH20 versus PBS or rHuPH20; P < 0.0001) and between the virus plus enzyme ^ treated group compared with the virus or OV-5T35H with enzyme, respectively, by day 65. As only ^ treated group (OV-5T35H+rHuPH20 versus OV-5T35H, P < 0.0001). described above, tumor progression rates correlated with

Clin Cancer Res 2008;14(12) June 15, 2008 3936 www.aacrjournals.org Downloaded from clincancerres.aacrjournals.org on September 27, 2021. © 2008 American Association for Cancer Research. Hyaluronidase Enzyme Enhances Virus Potency

47.5 days, respectively. Lastly, animals treated with OV-5T35H and enzyme had not reached the MST by day 65, and 67% of the animals remained alive at that time point. In summary, the tumor growth rate and survival data confirmed that the combination of oncolytic adenovirus and rHuPH20 has greater antitumor efficacy compared with virus monotherapy in animals bearing highly metastatic A375- mln1.luc tumors. Furthermore, the enzyme combined with the fiber-chimeric oncolytic virus OV-5T35H provided the most efficacious treatment with a statistically significant survival advantage over animals treated with any of the other therapies. Improved viral spread and replication in the tumor underlie the enhanced potency of oncolytic adenovirus when coinjected with rHuPH20. To understand the mechanism underlying the increased therapeutic efficacy and survival observed with the combination treatment of oncolytic adenovirus and rHuPH20, the number of adenoviral hexon-positive cells and infectious virus particles within the tumors was determined in the above study. Tumors were harvested on day 10 after initiation of treatment, and the number of hexon-positive cells was roughly estimated by evaluating immunohistochemical stainings (Fig. 4A). Staining and evaluation of hexon-positive cells in tumor sections (magnification, Â10) showed that tumors treated with the combination of virus and enzyme had more hexon-positive cells compared with tumors treated with the parental or chimeric oncolytic virus alone. Hexon-positive cells also seemed to be lower in number in tumors treated with OV- 5, either as monotherapy or in combination with enzyme, compared with tumors treated with OV-5T35H monotherapy or OV-5T35H in combination with rHuPH20. No hexon- positive cells were found in control tumors injected with either PBS or enzyme alone (data not shown). To evaluate whether an increase in hexon-positive cells correlates with the presence of higher amounts of infectious virus within tumors, infectious virus particles in tumors were measured by titering tumor homogenates on AE1-2a cells (7). No significant difference in the amount of infectious virus was found in tumors from animals treated with either OV-5T35H or the combination of OV-5 and rHuPH20. However, there was 10-fold less infectious virus present in tumors treated with OV- 5 compared with tumors treated with either OV-5T35H alone or the combination of OV-5 and enzyme. Only a small increase in the level of infectious virus was detected in tumors treated with the combination of OV-5T35H and enzyme compared Fig. 3. Coadministration of hyaluronidase increases the antitumor efficacy of with tumors treated with OV-5T35H alone (Fig. 4B). Ad5 (OV-5) and fiber-chimeric oncolytic viruses (OV-5T35H) in a spontaneous metastatic melanoma (A375-mln1.luc) model. Nude mice (n = 16) bearing s.c Thus, the degree of virus spread and number of infectious A375-mln1.luc tumors (100 mm3) received 4 injections of either PBS, rHuPH20 particles in the A375-mln1.luc tumors following the different enzyme (50 U per mouse per injection), virus (OV-5 or OV-5T35H; 1 Â1010 vp), or in vivo coinjection of OV-5 or OV-5T35H plus rHuPH20 once every other day.Tumor treatments correlated well with antitumor efficacy , progression (A) and survival (B) were monitored over time. A statistically suggesting that extensive virus spread and a high level of infectious significant delay in tumor progression was observed between the vector-treated virus particles in tumors resulted in reduced tumor progression groups compared with the PBS or rHuPH20-treated groups (P < 0.0001) and between the virus plus enzyme ^ treated group compared with the virus and increased overall survival of tumor-bearing animals. only ^ treated group (P < 0.05), as well as between the chimeric vector plus Fiber-chimeric oncolytic virus coinjected with hyaluronidase enzyme ^ treated group compared with the Ad5 parental orAd5 parental plus controls metastases in A375-mln1.luc tumor bearing mice. One enzyme ^ treated groups (OV-5T35H+rHuPH20 versus OV-5 or OV-5+rHuPH20, P < 0.05). potential concern with hyaluronidase injections into tumors was whether cotreatment with the enzyme might increase the metastatic spread of the primary tumor. To address this improved survival. Animals treated with PBS or enzyme had concern, hyaluronidase only or in combination with the MSTs of 24 and 27 days (Fig. 3B), whereas animals treated with chimeric virus was injected into A375-mln1.luc tumors and OV-5 monotherapy had a MST of 40 days. In contrast, animals metastatic spread of tumor cells to LN and lungs was monitored. treated with either OV-5T35H monotherapy or the combina- Specifically, established A375-mln1.luc tumors of f100 mm3 tion of OV-5 and enzyme survived longer, with MSTs of 45 and received a total of 4 injections (50 AL intratumorally), once every

www.aacrjournals.org 3937 Clin Cancer Res 2008;14(12) June 15, 2008 Downloaded from clincancerres.aacrjournals.org on September 27, 2021. © 2008 American Association for Cancer Research. Cancer Therapy: Preclinical other day, of either PBS, rHuPH20 (50 or 250 U per injection), not increase the metastatic potential of the highly metastatic OV-5T35H virus (1 Â 1010 vp per injection), or OV-5T35H A375-mln1.luc tumor cells. (1 Â 1010 vp per injection) with rHuPH20 (50 U per injection). Mice were euthanized on days 8 and 13, and the presence of Discussion luciferase-positive tumor cells in LN and lungs was determined by measuring the level of luciferase activity in pooled organs after The relatively limited ability of Ad5-based oncolytic viruses direct injection with luciferin (11). to transduce a broad range of tumor cells and to spread Total photon counts observed in LN and lungs at both time efficiently throughout solid tumors may hamper their effec- points in the groups treated with oncolytic virus were greatly tiveness in the clinic. A fiber knob replacement strategy, reduced when compared with the organs from tumor-bearing generating Ad5/35 oncolytic viruses with chimeric fiber animals treated with either PBS or hyaluronidase alone proteins, has previously been shown to improve the transduc- (Fig. 5A-C), indicating that the metastatic spread of tumor cells tion efficiency of such viruses significantly (7, 11). However, was reduced in virus-treated animals. At day 8, there was a there is still a need for approaches that improve spread of these f10-fold reduction in photon counts in LN and lungs in the viruses throughout tumors. groups treated with either OV-5T35H or OV-5T35H plus In a previous study, engineering the matrix-degrading protein rHuPH20 compared with the groups treated with either PBS relaxin into an Ad5/35 fiber-chimeric virus improved both or enzyme alone. However, there was no difference in photon transduction efficiency and virus spread within tumors (11). To counts between the groups treated with OV-5T35H or the increase ECM degradation during virus transduction without combination of OV-5T35H with enzyme at this time point. modifications to the virus genome, a different approach to At day 13, there was no difference in photon counts in LN improve viral spread was used in this study. Ad5-based or Ad5/ and lungs of animals injected with PBS alone or treated with 35 fiber chimeric viruses were coinjected intratumorally with either 50 or 250 U of enzyme. Interestingly, there was a 50-fold the rHuPH20, which has been shown to degrade hyaluronan reduction in photon counts in LN and lungs of mice treated transiently in vivo (13, 16). with the chimeric virus plus rHuPH20 compared with those Although hyaluronidase and relaxin (11, 32) both degrade treated with the chimeric virus alone. These data suggested that components within the ECM thereby increasing virus spread, the enzyme either alone or in combination with OV-5T35H did hyaluronidase coinjection has the advantage of not requiring

Fig. 4. Coadministration of rHuPH20 enzyme increases virus spread in A375-mln1.luc tumors. Nude mice (n = 16) bearing s.c A375-mln1.luc tumors at a size of (100 mm 3) received four intratumoral injections of either PBS, rHuPH20 enzyme (50 U per injection), virus (OV-5 or OV-5T35H; 1 Â1010 vp), or coinjection of OV-5 or OV-5T35H plus rHuPH20 once every other day. Ten days after the initial viral injection, tumors were harvested (n =3)and(A) virus distribution was evaluated by determining the number of hexon-positive cells within tumors by immunohistochemistry on paraffin-embedded tumor tissue (Â10). B, the amount of infectious virus present in tumors was determined by titering tumor lysates on AE1-2A cells.

Clin Cancer Res 2008;14(12) June 15, 2008 3938 www.aacrjournals.org Downloaded from clincancerres.aacrjournals.org on September 27, 2021. © 2008 American Association for Cancer Research. Hyaluronidase Enzyme Enhances Virus Potency

Fig. 5. Evaluation of the effect of rHuPH20 enzyme on metastasis of A375-mln1.luc tumor cells. Nude mice (n = 10) bearing s.c A375-mln1.luc tumors (100 mm3) received a total of four intratumoral injections (50 AL) of PBS, rHuPH20 enzyme (50 or 250 U per injection), OV-5T35H virus (1 Â1010 vp per injection), or coinjection of OV-5T35H plus rHuPH20 once every other day.The level of metastasis was evaluated on (A)day8and(B) day 13 postinjection in isolated LN and lungs by injecting them with luciferin to determine photon counts using the IVISR Xenogen system. The combined total average photon counts from LN and lungs is shown (n =5).C, actual IVIS images of LN and lung on Day 13.

www.aacrjournals.org 3939 Clin Cancer Res 2008;14(12) June 15, 2008 Downloaded from clincancerres.aacrjournals.org on September 27, 2021. © 2008 American Association for Cancer Research. Cancer Therapy: Preclinical manipulation of the adenovirus genome and can therefore be context of the relatively poorly transducing OV-5 virus, added to different viruses. In addition, by knowing the half-life coinjection of rHuPH20 significantly improved virus spread of hyaluronidase, one has control over the amount of enzyme that also was reflected in a significant survival advantage. Over- delivered to the tumor. In contrast, viruses expressing relaxin all, more infectious virus was observed in tumors treated with will produce relaxin as long as they replicate within the tumor OV-5T35H compared with OV-5, indicating a greater level of cells, which might differ between tumors. intratumoral virus replication and spread of the chimeric virus. The current data show that coinjection of hyaluronidase at a In summary, the increase in virus potency observed in animals concentration of 50 U with an E1-deleted Ad5/35 fiber treated with both enzyme and virus correlates with an improved chimeric virus significantly improves viral transduction of PC- spread of virus within the tumors, as shown by the presence of 3.luc tumors. Fifty enzyme units was selected from pilot increased numbers of hexon-positive cells, which in turn correlates experiments showing that tumors injected with less (10 U; with an increase in the amount of infectious virus present within data not shown) or more enzyme (200 or 800 U) showed fewer these tumors. Because the majority of human cancers express GFP-positive cells than tumors that were injected with 50 U of the CD46 receptor of the Ad5/35 chimera more consistently and enzyme. Ten enzyme units might be too little to break down at higher levels than the coxsackievirus-adenovirus receptor of the ECM within tumors well enough to allow effective virus the Ad5 virus, these findings suggest that the combination of spread; however, the reason for why 200 and 800 U of enzyme hyaluronidase and chimeric virus may be a more optimal therapy. did less well then 50 U is currently not known. One potential It is important to consider the possibility that the enzyme, explanation could be that higher concentrations of enzyme, which degrades the ECM, could increase metastatic spread of which is premixed with the virus before injection into the tumor cells not killed by the oncolytic virus. To address this tumor, are slightly toxic to the virus; however, in vitro studies concern, highly metastatic A375-mln1.luc tumors were injected did not support this hypothesis. Due to lack of access to reagents with the chimeric virus with or without hyaluronidase, or with that allow quantification of hyaluronic acid content within enzyme alone, and the animals were screened for the presence of tumors, it was impossible to look at effect of various tumor cells in LN and lungs at two time points. Metastasis to LN hyaluronidase concentrations on overall hyaluronic acid con- and lung at day 13 was greatly reduced in animals treated with tent within the tumor models used in this study. Nevertheless, chimeric virus and hyaluronidase compared with animals based on the optimal virus spread observed with 50 U of treated with virus or enzyme alone. These results are most likely enzyme, this concentration was selected for further studies. due to the reduced growth of primary tumors treated with the The observed increase in virus transduction as a result of combination compared with tumors treated with either virus or hyaluronidase treatment correlates with improved antitumor enzyme alone, because metastatic spread is believed to be activity when oncolytic adenoviruses are coinjected with related to the size of the primary tumor in this model. hyaluronidase in murine tumor models. Combination treat- Hyaluronidase by itself, even at a 5-fold higher dose, does not ment with both chimeric Ad5/35 virus and enzyme improves increase the metastatic spread of the A375-mln1.luc tumors. the oncolytic potency of the virus and significantly increases This observation is in agreement with a report from Bookbinder survival of tumor-bearing mice. Importantly, this combination et al. (17), demonstrating that hyaluronidase enhances the also effectively controls the growth of tumors that are twice as infusion rates and penetration of molecules up to 200 nm in large (250 mm3) as those typically used in these models and diameter, a size range that would allow adenoviruses but not significantly prolongs the survival of the treated animals tumor cells to penetrate throughout tissues. Another explana- emphasizing the overall potency of this combination. tion could be that the better spread of virus within tumors To determine if hyaluronidase coinjection has a different effect treated with the enzyme resulted in more effective killing of the on the potency of viruses that transduce tumors with different tumor cells by the increased amounts of virus, and that the virus- efficiencies, the parental OV-5 and fiber-chimeric oncolytic virus mediated killing of tumor cells preceded the escape of the cells OV-5T35H were tested with and without enzyme in the highly from the primary tumor mass into the vascular system. metastatic A375-mln1.luc melanoma tumor model. In this In conclusion, the data presented here show that combina- model, the combination of hyaluronidase and either virus tion treatment of solid tumors with an oncolytic virus in which resulted in improved antitumor efficacy and prolonged survival the Ad5 fiber knob has been replaced with the Ad35 knob with of tumor-bearing mice compared with monotherapy with either human hyaluronidase significantly improves transduction virus. It is notable that coinjecting hyaluronidase with the efficiency and spread of virus through the tumor, resulting in chimeric virus increases survival significantly, and the MST is not prolonged survival of the tumor-bearing animals without attained by day 65, with 67% of animals still alive at this time increasing the metastatic potential. Thus, this combination of point. Interestingly, the increase in viral titer observed in the oncolytic adenovirus and hyaluronidase may offer a novel and tumors treated with the combination of OV-5T35H and enzyme promising therapy for the treatment of patients with cancer. was only f2- to 3-fold compared with tumors treated with OV- 5T35H alone, whereas there was a f10-fold increase in virus Disclosure of Potential Conflicts of Interest titers in tumors treated with OV-5 with enzyme compared with OV-5 alone. Apparently, the enzyme enhanced the spread of the No potential conflicts of interest were disclosed. potent chimeric virus to a lesser extent than the spread of the Ad5-based parental virus, perhaps due to the higher initial Acknowledgments infectivity of the chimeric virus in the A375.mln-1.luc tumor WethankB.Batiste,S.Ulufatu,J.HoandT.LangerfromtheanimalfacilityandM. model. Nevertheless, the addition of rHuPH20 treatment to the Gebreyesus, P. Janaswamy from the cell culture group of Cell Genesys for their chimeric virus resulted in a pronounced survival advantage excellent work; Halozyme therapeutics, Inc. (San Diego, CA) for kindly providing the compared with treatment with the chimeric virus alone. In the hyaluronidase enzyme and Dr. Peter Working for critically reviewing the manuscript.

Clin Cancer Res 2008;14(12) June 15, 2008 3940 www.aacrjournals.org Downloaded from clincancerres.aacrjournals.org on September 27, 2021. © 2008 American Association for Cancer Research. Hyaluronidase Enzyme Enhances Virus Potency

References 1. Parato KA, Senger D, Forsyth PA, Bell JC. Recent longs survival of tumor-bearing mice. Cancer Res subcutaneous growth and vascularization in immuno- progress in the battle between oncolytic viruses and 2007;67:4399^407. compromised mice. Am J Pathol 2002;161:849 ^ 57. tumors. Nature 2005;5:965 ^ 76. 12. Kuriyama N, Kuriyama H, Julin CM, Lamborn K, 23. Lokeshwar VB, Cerwinka WH, Isoyama T, Lokesh- 2. Post DE, Khuri FR, Jonathan K Simons W, Van Meir Israel MA. Pretreatment with protease is a useful war BL. HYAL1hyalurinidase in prostate cancer: a tu- EG. Replicative oncolytic adenoviruses in multimodal experimental strategy for enhancing adenovirus- mor promotor and suppressor. Cancer Res 2005;65: cancer regimens. Hum GeneTher 2003;14:933 ^ 46. mediated cancer . Hum GeneTher 2000; 7782^9. 3. Okegawa T, Li Y, Pong RC, Bergelson JM, Zhou J, 11: 2 2 19 ^ 3 0 . 24. Posey JT, Soloway MS, Ekici S, et al. Evaluation of HsiehJT.The dual impact of coxsackie and adenovirus 13. Kuriyama N, Kuriyama H, Julin CM, Lamborn K, the prognostic potential of hyaluronic acid and hyal- receptor expression on human prostate cancer gene Israel MA. Pretreatment withprotease is ausefulexper- uronidase (HYAL1) for prostate cancer. Cancer Res therapy. Cancer Res 2000;60:5031 ^ 6. imental strategy for enhancing adenovirus-mediated 2003;63:2638^44. 4. LiY,Pong RC, BergelsonJM, et al. Loss of adenoviral cancergene therapy. Cancer Res 2001;61:1805 ^9. 25. Bertrand P, Girard N, Delpech B, Duval C, Anjou D, receptor expression in human bladder cancer cells: a 14. Shuster S, Frost GI, Csoka AB, Formby B, Stern R. Dauce JP. Hyaluronan and hyaluronectin in the extra- potential impact on the efficacy of gene therapy. Hyaluronidase reduces human xeno- cellular matrix of human breast carcinomas: compari- Cancer Res 1999;59:325^ 30. grafts in SCID mice. Int J Cancer 2002;102:192 ^ 7. son between invasive and noninvasive areas. Int J 5. Bergelson JM, CunninghamJA, Droguett, et al. Iso- 15. Victor R, Chauzy C, Girard N, et al. Human breast- Cancer 1992;52:1 ^ 6. lation of a common receptor for coxsackie B viruses cancer metastasis formation in a nude-mouse model: 26. Gotte M, Yip GW. heparanase, hyaluronan, and and adenoviruses 2 and 5. Science1997;275:1320^ 3. studies of hyaluronidase, hyaluronan and hyaluronan- CD44 in cancers: a breast carcinoma perspective. 6. Miller CR, BuchsbaumDJ, Reynolds PN, et al. Differ- binding sites in metastatic cells. Int J Cancer 1999;82: Cancer Res 2006;66:10233^ 7. ential susceptibility of primary and established human 77 ^ 83. 27. Hallenbeck PL, Chang YN, Hay C, et al. A novel tu- cells to adenovirus infection: targeting via the 16. Brekken C, Bruland OS, Davies CD. Interstitial fluid mor-specific replication-restricted adenoviral vector epidermal growth factor receptor achieves fiber re- pressure in human osteosarcoma xenografts: signifi- for gene therapy of . Hum ceptor-independent gene transfer. Cancer Res 1998; cance of implantation site and the response to intratu- GeneTher 1999;10:1721 ^ 33. 8:5738^48. moral injection of hyaluronidase. Anticancer Res 28. Jakubczak JL, Ryan P, Gorziglia M, et al. An onco- 7. Reddy PS, Ganesh S,Yu DC. Enhanced gene transfer 2000;20:3503^12. lytic adenovirus selective for retinoblastoma tumor and oncolysis of and melanoma 17. Bookbinder LH, HoferA, Haller MF, et al. A recom- suppressor protein pathway-defective tumors: depen- cells by fiber chimeric oncolytic adenoviruses. Clin binant human enzyme for enhanced interstitial trans- dence on E1A,the E2F-1promoter, and viral replication Cancer Res 2006;12:2869 ^ 78. port of therapeutics. J Controlled Release 2006;114: for selectivity and efficacy. Cancer Res 2003;63: 8. Nilsson M, Ljungberg J, RichterJ, et al. Development 230 ^ 41. 14 9 0 ^ 9. of an adenoviral vector systemwith adenovirus sero- 18. Simpson MA. Concurent expression of hyaluronan 29. RameshN,GeY,EnnistD,etal.CG0070,acondi- type 35 tropism; efficient transient gene transfer into biosynthetic and processing enzymes promotes tionally replicating GM-CSF armed oncolytic adenovi- primary malignant hematopoietic cells. J Gene Med growth and vascularization of prostate tumors in mice. rus for the treatment of bladder cancer. Clin Cancer 2004;6:631 ^ 41. AmJ Pathol 2006;169:247 ^ 57. Res 2006;12:305 ^ 13. 9. Harrison D, Sauthoff H, Heitner S, Jagirdar J, Rom 19. Lokeshwar VB, Rubinowicz D, Schroeder GL, et al. 30. Rubio N, Villacampa MM, Hilali El, Blanco J. Meta- WN, Hay JG. Wild-type adenovirus decreases tumor Stromal and epithelial expression of tumor markers static burden in nude mice organs measured using xenograft growth, but despite viral persistence com- hyaluronic acid and HYAL1hyaluronidase in prostate prostate tumor PC-3 cells expressing the luciferase plete tumor responses are rarely achieved-deletion of cancer. J Biol Chem2001;276:11922 ^32. gene as a quantifiable tumor cell marker. Prostate the viral E1^ 19kD gene increases the viral oncolytic 20. Liu D, Pearlman E, Diaconu E, et al. Expression of 2000;44:133^43. effect. HumGeneTher 2001;12:1323^ 32. hyaluronidase by tumor cells induces angiogenesis 31. Lin JL, Lalani AS, Harding TC, et al. Inhibition of 10. Sauthoff H, HuJ, Maca C, et al. Intratumoral spread in vivo. Proc Natl Acad Sci U S A 1996;93:7832^ 7. lymphogeneous metastasis using adeno-associated of wild-type adenovirus is limited after local injection 21. Kovar JL, Johnson MA, Volcheck WM, Chen J, virus-mediated gene transfer of a soluble VEGFR-3 of human xenograft tumors: virus persists and Simpson MA. Hyaluronidase expression induces decoy receptor. Cancer Res 2005;65:6901^ 9. spreads systemically at late time points. Hum Gene prostate tumor metastasis in an orthotopic mouse 32. KimJH, Lee YS, KimH, et al. Relaxin expression Ther 2003;14:425^ 33. model.AmJPathol2006;69:1415^26. fromtumor-targeting adenoviruses and its intratu- 11. Ganesh S, Gonzalez M, Idamakanti N, et al. Relaxin- 22. Simpson MA, Wilson CM, McCarthy JB. Inhibition moral spread, induction, and efficacy. Natl expressing, fiber chimeric oncolytic adenovirus pro- of prostate tumor cell hyaluronan synthesis impairs Cancer Inst 2006;98:1482^ 93.

www.aacrjournals.org 3941 Clin Cancer Res 2008;14(12) June 15, 2008 Downloaded from clincancerres.aacrjournals.org on September 27, 2021. © 2008 American Association for Cancer Research. Intratumoral Coadministration of Hyaluronidase Enzyme and Oncolytic Adenoviruses Enhances Virus Potency in Metastatic Tumor Models

Shanthi Ganesh, Melissa Gonzalez-Edick, Douglas Gibbons, et al.

Clin Cancer Res 2008;14:3933-3941.

Updated version Access the most recent version of this article at: http://clincancerres.aacrjournals.org/content/14/12/3933

Cited articles This article cites 32 articles, 14 of which you can access for free at: http://clincancerres.aacrjournals.org/content/14/12/3933.full#ref-list-1

Citing articles This article has been cited by 1 HighWire-hosted articles. Access the articles at: http://clincancerres.aacrjournals.org/content/14/12/3933.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://clincancerres.aacrjournals.org/content/14/12/3933. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.

Downloaded from clincancerres.aacrjournals.org on September 27, 2021. © 2008 American Association for Cancer Research.