• adenovirus;
  • gene therapy;
  • nonsmall cell lung cancer;
  • orthotopic model;
  • in vivo imaging


  1. Top of page
  2. Abstract
  6. Acknowledgements


Variable expression of the coxsackie and adenovirus receptor (CAR) has limited gene transfer efficacy to many types of tumors. Consequently, tropism-modified adenoviruses have been developed for enhanced infectivity. To the authors' knowledge, targeting approaches for nonsmall cell lung cancer (NSCLC) have not been comprehensively evaluated. The current hypothesis was that modified adenoviruses could be used for increasing gene transfer to and killing of NSCLC cells in vitro and in vivo.


Ten NSCLC cell lines were analyzed to represent the different NSCLC histologies. Because clinical tumors may differ from established cell lines, 6 clinical specimens fresh from patients were analyzed. For in vivo studies, a novel orthotopic murine model of advanced lung cancer was developed. Because tumor response is difficult to quantitate in orthotopic models, noninvasive imaging of green fluorescent protein (GFP) was utilized as a surrogate for tumor size measurements.


Adenoviruses whose capsids were modified with RGD-4C, the serotype 3 knob, or polylysine displayed increased gene transfer to NSCLC cell lines and clinical samples in comparison to serotype 5 viruses. Conditionally replicating oncolytic adenoviruses (CRAds) with the same modifications showed enhanced therapeutic efficiency in vitro and in vivo. The median survival of mice treated with Ad5.pK7-Δ24 or Ad5-Δ24RGD increased 37% (P<.01). GFP imaging allowed noninvasive individualized detection of response and recurrence.


Targeting of adenovirus to heterologous receptors can improve killing of NSCLC cells. Utilization of clinical samples and an orthotopic model of advanced lung cancer may provide clinically relevant translational data. Cancer 2006. © 2006 American Cancer Society.

Lung cancer is 1 of the most common cancers, and with 1.18 million deaths it is the leading cause of cancer-related mortality worldwide. Nonsmall cell lung cancer (NSCLC) accounts for approximately 80% of cases. For patients with recurrent or disseminated disease, there is a paucity of effective treatments. Chemotherapy, radiotherapy, small molecular inhibitors, and monoclonal antibodies have been rigorously studied and benefit some NSCLC patients, but advanced disease remains incurable and treatment usually influences survival only marginally. Unfortunately, most patients present with advanced disease, and therefore mortality rates are close to incidence rates and thus there is a need for new therapies.

Adenoviral gene therapy is a promising treatment alternative for many types of cancers refractory to current modalities.1 Importantly, clinical feasibility and utility to patients has been recently demonstrated in randomized clinical trials with glioma and head and neck cancer patients.2–4 The treatment of lung cancer with adenoviral transfer of p53 has been well tolerated and gene delivery has been demonstrated, but to date evidence of single-agent efficacy has been marginal.5–7 Available correlative data suggest that insufficient transduction of advanced tumor masses, due to poor intratumoral dissemination of the virus, frequently underlies low efficacy.8 Although transduced cells in and around the needle tract may be killed, the remaining cells quickly make up for the deficit. In combination with radiation or chemotherapy, the data appear more promising, but in contrast to head and neck cancer,4 randomized data are not yet available in lung cancer patients.9, 10

Adenoviruses based on serotype 5 (Ad5) are widely used because they can infect many cell types effectively, can be propagated to high titers, and do not integrate into the host genome. Adenovirus infection is initiated by binding of the knob domain of the fiber with the cellular primary receptor, followed by interaction of the RGD-4C (Arg-Lys-Asp) motif in the penton base with cellular αvβ integrins, which leads to the internalization of the virus. The primary receptor for adenovirus serotype 5 is the coxsackie/adenovirus receptor (CAR), whose normal function is not fully understood, but may relate to adhesion.11

Importantly, recent evidence suggests that CAR exhibits variable expression on cancer cells, which results in inefficient gene delivery with Ad5.12–14 Specifically, resistance may be due to low expression levels or aberrant localization at the cellular or tissue level.15 It has been reported that efficient gene delivery to target cells both in vitro and in vivo remains a major limitation for successful gene therapy, but to our knowledge, lung cancer substrates have not been previously studied critically with regard to their infectivity with Ad5.

To circumvent deficiency of CAR, viruses can be retargeted by improving their abilities to attach to heterologous receptors.12 One approach is to clone an integrin-binding RGD motif into the HI loop of the fiber knob or to construct a chimeric fiber with the knob domain of Ad3 in the Ad5 capsid,13 which allows internalization through the serotype 3 receptor. One can also add a polylysine tail to the C-terminal end of the fiber for binding to heparin sulfate proteoglycans (HSPGs).16

Replication-deficient adenoviruses have frequently shown high transduction and antitumor efficacy in animal models. However, advanced human solid tumor masses are larger and more complex than xenograft tumors, which restrict sufficient access of virions to tumor cells. For instance, necrotic, hyperbaric, hypoxic, and stromal areas can be problematic. To overcome this problem, conditionally replicating oncolytic adenoviruses (CRAds) have been developed.1 Their antitumor effect is based on oncolysis caused by replication per se, whereas normal tissues are spared due to low replication.

A promising approach for development of CRAds is based on utilization of a 24-basepair (bp) deletion in the E1 gene that allows viruses to replicate preferentially in cells in which the Rb/p16 pathway is inactive, which could include most human cancer cells, including NSCLCs.17, 18 Recently, it is been demonstrated that an important determinant of the oncolytic potency of CRAds is their ability to infect target cells. Therefore, to realize the full potential of CRAds for treatment of advanced cancers, it is appealing to target the virus to receptors highly expressed in tumors.19

In the current study, we initially used replication-deficient adenoviruses to evaluate gene transfer mediated through non-CAR receptors. Next, we utilized the respective CRAds, all featuring an isogenic 24-bp deletion as a means for tumor selectivity. Because clinical tumors significantly differ from cell lines, we obtained and analyzed clinical specimens fresh from patients. In vivo studies were performed in a novel orthotopic murine model of advanced lung cancer. Because tumor response is not trivial to quantitate in orthotopic models, we utilized noninvasive green fluorescent protein (GFP) imaging as a surrogate for tumor size measurements.


  1. Top of page
  2. Abstract
  6. Acknowledgements

Cell Lines and Tissues

Human NSCLC lines and 293 human transformed embryonic kidney cells were obtained from the American TypeCulture Collection ([ATCC] Manassas, VA), except for 201T (provided by Jay D. Hunt at Louisiana State University, New Orleans, LA), Ma44-3 (provided by Kazuya Kondo at the Department of Oncological and Regenerative Surgery at the University of Tokushima, Japan), and LNM35/EGFP (provided by Takashi Takahashi of the Honda Research Institute, Japan). Enhanced GFP (EGFP) was inserted into this cell line with an adeno-associated virus (AAV-EGFP).20 All cell lines were cultured at 37°C in media recommended by suppliers in a humidified atmosphere of 5% carbon dioxide. Fresh NSCLC primary tissue samples were obtained after signed informed consent was obtained from patients diagnosed with advanced lung cancer at the Helsinki University Central Hospital. The study was approved by the hospital Ethics Committee.

Recombinant Adenoviruses

Ad5luc1, Ad5/3luc1, and Ad5lucRGD13 contain firefly luciferase transgene cassette in place of a deleted E1 region. Ad5(GL), Ad5.pK7(GL), and Ad5.pK7.RGD(GL)16 contain GFP expression cassette in addition to the luciferase cassette. Therefore, Ad5(GL) was used as the control for GL viruses, whereas Ad5Luc1 was used for the former group. Ad5LacZ and Ad5F/K2121 have a β-galactosidase expression cassette inserted in the place of deleted E1 region. Ad5-Δ24E3, Ad5/3-Δ24, Ad5-Δ24RGD22, and Ad5.pK7-Δ2423 have a 24-bp deletion in the CR2 region of the E1A. Replication-deficient viruses were propagated in 293 and CRAds in A549 cells. All viruses were purified in cesium chloride gradients. The particle concentration was measured at 260 nanometers (nm) and a standard plaque assay was performed to determine functional units.

Adenovirus-Mediated Gene Transfer Assays

Cells were infected in 2% groth medium (GM) in 24-well plates for 30 minutes at room temperature and luciferase assay was performed with the Luciferase Assay System (Promega, Madison, WI) after 24 hours, as reported.22 For immunohistochemistry, NCI-H460 cells were infected in chamber slides (Nunc, Naperville, IL). Next day, cells were fixed with 10% formalin and permeabilized with 0.1% Tween 20/phosphate-buffered saline (PBS). Blocking was done with 1% bovine serum albumin (BSA)/PBS for 30 minutes at room temperature. Antihexon was used as a primary antibody at a concentration of 1.5 μg/mL (LabFrontier, Seoul, Korea). Texas Red-labeled secondary antibody (AbCAM, Cambridge, UK) was used at a concentration of 2 mg/mL. Cells were stained with Hoechst dye (Oriola, Espoo, Finland), positive cells were counted, and fluorescence images obtained (Olympus Ix81, Biosystems, Munich, Germany). Luciferase analysis of tissue samples was performed as reported.14 Luciferase assay results were normalized to protein content by DC Protein Assay (Bio-Rad, Hercules, CA).

In Vitro Cytotoxicity Assay

The 3‒(4.5‒dimethlythiazol‒2‒yl(‒5‒(3‒carboxymethoxyphenyl)‒2‒(4‒sulfophenyl‒2H‒tetrazolium, inner salt [“MTS”] cell killing assay was performed as reported22 and viability was measured using the CellTiter 96 AQueous One Solution Cell Proliferation Assay (MTS assay; Promega), when at least 1 tested virus showed a cytotoxic effect in the 2 weakest dilutions. Ad5.pK7-Δ24 became available during the in vivo part of the project and thus only LNM-35/EGFP cells were analyzed with MTS. The crystal violet cell-killing assay was performed as reported.19 Briefly, cells were stained with crystal violet when at least 1 of the tested viruses showed cytopathic effect in the 2 weakest dilutions.

Orthotopic Model of Locally Advanced Lung Cancer

Female NMRI nude mice ages 3 to 4 weeks were obtained from Taconic (Ejby, Denmark) and quarantined for 2 weeks. All animal experiments were approved by the Experimental Animal Committee of the University of Helsinki and the Provincial Government of Southern Finland. Mice were injected into the left lung with 2 × 106 LNM35/EGFP cells in a total volume of 200 μL using a syringe with a 27G needle. Before injection the mice were anesthetized with medetomidine (Domitor; Orion Pharma, Espoo, Finland) + ketamine (Ketalar;, Pfizer, New York, NY) at 1:2.

In Vivo Imaging

Orthotopic LNM-35/EGFP tumors were inoculated and their growth was monitored on days 3, 6, and 9 in the biodistribution assay and on Days 5, 12, and 19 in the therapeutic assay. Imaging was done by using the IVIS 100 Imaging System (Xenogen, Alameda, CA). Fluorescence emission was quantitated by determination of regions of interest, which were normalized to reference regions in the same image, but without tumor. Mice were anesthetized as above before viral injections and imaging.

Biodistribution Assay

Five days after injection of cells, 3 × 1010 virus particles (VP) of viruses was injected through the tail vein. After 48 hours, mice (n = 7 in each group) were injected with 0.1 mg/μL D-luciferin (Promega) into the tail vein and an image was taken 10 minutes later. Mice were killed immediately after imaging and organs were harvested for luciferase analysis as previously described14 and the data were normalized to protein concentration of the organs by DC Protein Assay (Bio-Rad).

Therapeutic Assay

Advanced disease was allowed to develop for 5 days and then mice were divided randomly into 5 groups (n = 12 mice per group). Viruses were injected into the tail vein (2 × 1010 VP/mouse). Mice were treated weekly 3 times (on Days 5, 12, and 19). Imaging of the mice for monitoring tumor growth was done at the same time points. The health of the mice was monitored daily and they were killed when there was evidence of pain or distress.

Statistical Analysis

Pairwise comparisons were performed with the Student t test and survival data were analyzed with a log-rank test using SPSS software (version 11.5; SPSS Inc., Chicago, IL).


  1. Top of page
  2. Abstract
  6. Acknowledgements

Modified Adenoviruses Show Enhanced Gene Transfer to NSCLC Cell Lines In Vitro

We hypothesized that by using retargeted adenoviruses we could enhance gene transfer to NSCLC cell lines. Ad5luc1, Ad5/3luc1, Ad5lucRGD, Ad5(GL), Ad5. pK7(GL), Ad5.pK7.RGD(GL), Ad5LacZ, and Ad5F/K21 contain transgene cassettes in place of a deleted E1 region, rendering them replication-deficient. Ad5-Δ24E3, Ad5/3-Δ24, Ad5-Δ24RGD, and Ad5.pK7-Δ24 have a 24-bp deletion in the CR2 region of the E1A. “5/3” refers to serotype chimeric viruses that utilize the adenovirus serotype 3 receptor for entry.22 RGD-modified viruses can utilize αvβ integrins for cellular entry, whereas polylysine (pK7) modification of the fiber knob c-terminus allows entry through heparin sulfate proteoglycans.

To test whether these modifications could increase gene transfer, we infected 10 NSCLC cell lines belonging to the 3 major histology groups of NSCLC (Fig. 1). Infection with Ad5lucRGD or AdpK7(GL) resulted in up to 270-fold (SW900) and 90-fold (Ma44-3, A549) higher luciferase expression compared with Ad5luc1, respectively. In nonmalignant 293 cells, the modified viruses were not much different from Ad5luc1 (within 2.5-fold). The highest gene transfer efficiency with Ad5lucRGD was seen in NCI-H460, NCI-H661, SW900, and 201T cells. Ad5.pK7(GL) was most effective in LNM35/EGFP, Ma44-3, NCI-H520, and NCI-H23 cell lines. To examine whether the increase in gene transfer was due to an increase in the proportion of cells infected, NCI-H460 cells were infected either with Ad5luc1 or Ad5lucRGD and stained with anti-hexon antibody. Ad5lucRGD resulted in more cells infected than Ad5luc1 (Fig. 1L and 1M).

thumbnail image

Figure 1. Gene transfer to nonsmall cell lung cancer (NSCLC) cell lines with capsid modified luciferase expressing adenoviruses. (A-C) Large-cell carcinoma. (D-F) Squamous cell carcinoma. (G-J) Adenocarcinoma. Nonmalignant 293 cells (K) were included as a control. Luciferase activity (relative light units [RLUs]) was measured 24 hours after infection. Capsid-modified viruses are compared with the serotype 5 viruses (Ad5luc1, Ad5(GL), and Ad5LacZ), which were given a value of 1. To determine whether an increase in gene transfer was due to an increased number of cells infected, immunohistochemistry was performed with an antihexon antibody (L and M). The percentage of infected A549 cells was counted and results are reported relative to Ad5luc1, which was given a value of 1 (L). (M) Hoechst-stained and antihexon-stained cells. VP indicates virus particles. Error bars indicate standard error of the mean (±SE); *P < .05; **P < .01; ***P < .001 versus the Ad5 control.

Download figure to PowerPoint

Modified Adenoviruses Demonstrate Enhanced Gene Transfer to NSCLC Clinical Samples

The geno- and phenotype of cell lines may be quite different from clinical tumors. Therefore, gene transfer experiments were performed in primary NSCLC samples fresh from patients (Fig. 2). To avoid genotypic and phenotypic changes associated with acclimatization to in vitro conditions, analyses were performed immediately after surgery, without passaging. Four adenocarcinoma samples (Fig. 2A-C) were T2N0M0 and 1 sample (Fig. 2D) was T3N0M0 and 2 squamous cell carcinoma samples (Fig. 2E: T2N0M0; Fig. 2F: T4N0M0) were analyzed. The results were in keeping with the data from cell lines; Ad5.pK7(GL) showed the best gene transfer in comparison with the Ad5 control and in 5 of 6 cases also, versus the other modified viruses (all P<.05). To summarize the in vitro gene transfer data, Ad5lucRGD and Ad5.pK7(GL) enhanced gene delivery over the Ad5 controls in both NSCLC cell lines and clinical samples. There does not appear to be an obvious correlation between virus modification and histological subtype.

thumbnail image

Figure 2. Ad5.pK7(GL) showed increased infectivity of nonsmall cell lung cancer (NSCLC) primary tissue samples fresh from patients. (A-D) Adenocarcinoma. (E-F) Squamous cell carcinoma. Samples were incubated with viruses and luciferase activity was measured after 24 hours. Luciferase activity is expressed as relative light units (RLUs) normalized for total protein concentration. Modified viruses are compared with isogenic serotype 5 viruses Ad5luc1 or Ad5(GL), which were given a value of 1. Error bars indicate standard error (SE); *P < .05; **P < .01; ***P < .001 versus the Ad5 control.

Download figure to PowerPoint

In Vitro Cytotoxicity Assay

To test whether the improved gene transfer noted with the replication-deficient viruses could enhance the cytotoxic effect of the respective CRAds, we performed MTS assays with same panel of NSCLC cell lines (Fig. 3A). On NCI-H460, SW900, NCI-H520, Calu-3, and A549, Ad5/3-Δ24 displayed significantly increased cell killing compared with Ad5-Δ24E3, an isogenic control with the Ad5 capsid (all P<.05). However, Ad5-Δ24E3 was nearly as cytotoxic as Ad5/3-Δ24 in the NCI-H661, LNM35/EGFP, NCI-H23, and 201T cell lines. Oncolysis with Ad5/3-Δ24 and Ad5-Δ24E3 was significantly improved (P<.001) over the RGD motif containing virus Ad5-Δ24RGD in every cell line except SW900 and NCI-H520. Ad5luc1 was included as an E1-deleted control, and it did not cause oncolysis compared with uninfected cells (P > .05). To confirm the MTS results with a different approach, a crystal violet cell viability assay was performed with A549 cells. The data were in keeping with the MTS results, with Ad5/3-Δ24 delivering the best cell-killing efficacy in this cell line (Fig. 3B).

thumbnail image

Figure 3. Capsid-modified conditionally replicating oncolytic adenoviruses show effective killing of nonsmall cell lung cancer (NSCLC) cell lines. Cell viability was measured with MTS assay. The OD490 values of uninfected cells were set as 100%. (A) The data are expressed as means of quadruplicates ± standard error (SE). (B) To verify MTS results, A549 cells were infected with Ad5luc1, Ad5-Δ24E3, Ad5/3-Δ24, Ad5-Δ24RGD, or Ad300wt and after allowing for virus replication, live cells were stained with crystal violet. MTS indicates 3‒(4.5‒dimethlythiazol‒2‒yl(‒5‒(3‒carboxymethoxyphenyl)‒2‒(4‒sulfophenyl‒2H‒tetrazolium, inner salt.

Download figure to PowerPoint

Biodistribution of Capsid-Modified Viruses in an Orthotopic Model of Advanced Lung Cancer

In this model, the tumor first spreads in the injected lung, then progresses to regional lymph nodes in the mediastinum and dissemination is subsequently observed in both lungs, bronchi, and later into the peritoneum (Fig. 4). After tumor inoculation and engraftment, viruses were injected intravenously. Although the majority of luciferase expression was detected in the liver, tumor transduction could also be demonstrated at the time of dissection. Despite some trends, there were no significant differences noted between viruses in any of the analyzed organs.

thumbnail image

Figure 4. An orthotopic model of disseminated lung cancer was developed and cells were labeled with green fluorescent protein (GFP) to facilitate noninvasive monitoring. LNM35/EGFP (2 × 106) cells were injected directly into the left lung at a final volume of 200 μL. Development of the tumor was followed by fluorescent imaging. (A-C) Pictures were taken 3 days, 6 days, and 9 days after injection. (Panel C, Insert:) A dissected primary tumor. For analysis of biodistribution, orthotopic advanced lung cancer was inoculated as above, followed by 3 × 1010 virus particles (VP) of viruses injected intravenously. (D) 48 hours later, D-luciferin was injected and whole body luciferase imaging was performed. Then tumors (E) and other organs were collected and (F) luciferase was extracted. Results are presented as relative light units (RLUs) normalized to the protein concentration and serotype 5 control viruses Ad5luc1 and Ad5(GL) were set as 1. Error bars indicate the standard error (SE).

Download figure to PowerPoint

Therapeutic Effect in an Orthotopic Murine Model of Advanced Lung Cancer

The median survival times were 19 days, 22 days, 23 days, 26 days, 26 days, and 26 days for no virus, Ad5luc1, Ad5/3-Δ24, Ad5-Δ24E3, Ad5.pK7-Δ24, and Ad5-Δ24RGD, respectively (Fig. 5A). In pairwise comparisons, survival with Ad5-Δ24E3, Ad5.pK7-Δ24, and Ad5-Δ24RGD was significantly enhanced compared with the untreated group (P<.01), whereas Ad5/3-Δ24 and Ad5luc1 did not demonstrate any significant enhancement. The CRAds were not significantly different from each other. Treatment with Ad5.pK7-Δ24 and Ad5-Δ24RGD increased the median survival by 7 days (37%). One mouse (8.3%) treated with Ad5Δ24RGD survived until the end of the experiment (93 days).

thumbnail image

Figure 5. (A) Therapeutic effect in an orthotopic murine model of advanced lung cancer. LNM35/EGFP cells were injected into the lungs of nude mice and the tumor was allowed to disseminate for 5 days before injection of 2 × 1010 virus particles (VP) of viruses, which was repeated twice at 1-week intervals. In pairwise comparisons, Ad5-Δ24E3, Ad5.pK7-Δ24, and Ad5-Δ24RGD demonstrated enhanced therapeutic efficacy when compared with untreated mice (P<.01). Ad5/3-Δ24 and Ad5luc1 were not different from the untreated mice. (B) Noninvasive imaging of tumor load was performed weekly. Ad5-Δ24E3-, Ad5.pK7-Δ24-, and Ad5-Δ24RGD-treated groups displayed rather stable green fluorescent protein (GFP) emission up to Day 12. Suggesting cell killing by the viruses between Days 12 and 19, GFP emission reduced by an average of 35% in the treated mice. In contrast, untreated mice displayed an increase in fluorescence by Day 19. Tumor load, as estimated by GFP emission, was approximately 8-fold smaller in Ad5-Δ24E3-, Ad5.pK7-Δ24-, and Ad5-Δ24RGD-treated mice compared with the untreated group (P<.01). * indicates P < .05; **P < .01; ***P < .001. Error bars indicate standard error (SE).

Download figure to PowerPoint

Noninvasive Imaging of Therapeutic Effect inan Orthotopic Model of Advanced Lung Cancer

The therapeutic effect provided by the viruses was evaluated with noninvasive in vivo imaging on Days 5, 12, and 19 (Fig. 5B). When fluorescence readings were averaged for all CRAd-treated mice, light emission increased until Day 12, and then decreased with an average of 34% toward Day 19. On a group-by-group basis, Ad5-Δ24E3-, Ad5.pK7-Δ24-, and Ad5-Δ24RGD-treated groups demonstated decreases of 43%, 21%, and 10% between Days 12 and 19. Concomitantly, GFP emission increased by 75% in untreated mice. On Day 19, tumor load was significantly smaller in Ad5-Δ24E3-, Ad5.Δ24.pK7-, and Ad5-Δ24RGD-treated mice versus untreated mice, as estimated by GFP emission (P < .01). Despite a more homogeneous genetic background than noted in patients, and identical tumor initiation in all mice, there was individual variation between treatment results. In Figure 6 we grouped Ad5-Δ24RGD-treated mice according to response. Some mice had progressive disease, whereas the disease stabilized in others and some had dramatic responses. We believe that this variation resembles the variation often observed when patients are treated with an effective therapeutic, which might support the clinical relevance of the model.

thumbnail image

Figure 6. Noninvasive imaging reveals individual variation in mice treated with conditionally replicating oncolytic adenoviruses (CRAds). Examples of treatment results in mice treated with Ad5-Δ24RGD. The far left panel indicates the day of imaging and each mouse is identified with a number. Mice 1, 7, and 12 had progressive disease; Mice 5, 9, and 11 had stable disease; and the mice in the right panel achieved tumor response (Mice 6, 8, and 10).

Download figure to PowerPoint


  1. Top of page
  2. Abstract
  6. Acknowledgements

The inherent potential of adenoviral gene therapy for treatment of advanced cancer has recently been demonstrated in landmark randomized clinical trials.2–4 With regard to lung cancer, randomized trials have not yet been completed. Phase I and II trials suggest excellent safety, and although there are some examples of tumor responses, the existing correlative data parallels other tumor types and suggests that transduction of solid tumor masses has been mostly limited to the needle tract.5–8 Therefore, more effective treatment approaches may be required for clinical utility. One such approach is the utilization of replication-competent viruses such as CRAds that provide dramatic local amplification of effect. In essence, CRAds deliver “readministration in situ,” and in theory cell killing can proceed as long as there are tumor cells. Moreover, dissemination can occur at distant sites for eradication of metastases.24

In the current study, we used capsid-modified replication-deficient adenoviruses and the respective CRAds. All CRAds had an isogenic 24-bp deletion in the E1A region for tumor-selective replication and featured an intact E3 region for effective oncolysis. Modified adenoviruses displayed increased gene delivery to NSCLC cell lines (Fig. 1) and also to fresh clinical samples (Fig. 2). Enhanced cell killing also was observed in vitro in a cytotoxicity assay (Fig. 3). There was a tendency for RGD-4C- and pK7-modified viruses to provide the best gene transfer rates, whereas 5/3 chimerism was compatible with rapid viral packaging, and therefore highly effective oncolysis. This is in accord with previous data that suggest that, although 5/3 chimeric CRAds are highly potent oncolytic agents, packaging of RGD-4C- or pK7-containing virions may occasionally be less rapid, perhaps due to the “sticky” nature of these moieties.

In vivo efficacy after systemic administration is affected not only by the speed of oncolysis, but also biodistribution, intratumoral retention and bioavailability, access from the bloodstream, and other factors. For example, preliminary data suggest that adenovirus binding to blood factors may have a role in hepatic sequestration, which can be reduced by using fiber-mutated viruses.25 Therefore, in vivo data should be obtained in a model that resembles the clinical situation as closely as possible. In particular, there is an increasing amount of data emphasizing the crucial role of the anatomic location of the tumor. The “seed and soil” hypothesis implies that the surroundings of a tumor play an important role in the behavior of the tumor, including response to treatment.26 Therefore, we developed an orthotopic mouse model of advanced lung cancer (Fig. 4). In this model, a large tumor develops rapidly in the injected lung of the mouse and dissemination is seen to both lungs, the mediastinum, bronchi, and later to the peritoneum.

Underlining the disseminated nature of the disease, surgery cannot cure advanced lung cancer. Thus, treatment approaches should utilize systemic delivery. In the current study, we investigated the biodistribution of capsid-modified viruses in mice with advanced lung cancer. Differences between the groups were small for most tissues. Nevertheless, it is noteworthy that in addition to tumor cells, Ad5lucRGD and Ad5.pK7(GL) appeared to also effectively transduce the lung parenchyma. At this point, we do not have definite evidence regarding whether this is due to transduction of infiltrating tumor cells or transduction of normal parenchyma, but we suspect the latter. This finding might have relevance for the treatment of other lung diseases. Interestingly, there appeared to be a difference between the left and right lung with Ad5.pK7(GL). This suggests that the tumor may cause metabolic differences that are reflected in the biodistribution. None of the analyzed tissues demonstrated much transduction by Ad5/3luc1, which suggests that this virus is mostly cleared by macrophage lineage cells such as Kupffer cells of the liver. This is in accord with previous data, although it is clear that the route of administration affects biodistribution.13 Importantly, most viruses used allowed transduction of the tumor after intravenous delivery, which is the first requirement for possible systemic therapeutic efficacy (Fig. 4E).

The model developed here proved to be highly aggressive, which resembles the clinical behavior of many advanced large-cell NSCLCs. Nevertheless, an increase in survival was noted with Ad5.pK7-Δ24, Ad5-Δ24RGD, and Ad5-Δ24E3. The relative differences between biodistribution data and efficacy data may be due to many factors. For instance, expression of the respective receptors may change as the tumor progresses, which would affect the survival experiment, whereas the biodistribution experiment reflects a single timepoint. Moreover, it is possible that when the viruses reach the virus from the intravenous direction, they come across CAR-expressing cell types such as endothelial and stromal cells, but once they get inside the tumor the milieu might be somewhat different. It might be possible that HSPGs and integrins are expressed more inside the tumor, which would contribute to higher oncolytic efficacy than what would be expected based on biodistribution of RGD- and pK7-modified viruses.

Survival is an important endpoint, but it is not very sensitive to differences between different treatments.27 Moreover, survival does not provide information regarding the dynamics of tumor response and recurrence, and therefore longitudinal monitoring systems may have advantages over survival alone.28 We developed a model that allows longitudinal monitoring of tumor burden over time by analyzing GFP expression (Fig. 6). This approach might be further improved with another fluorophore such as red fluorescent protein, or a luminescent reaction such as luciferin conversion, both of which penetrate tissues more effectively than GFP.

Analogous to the situation in patients, some mice responded well, whereas others benefited little (Fig. 6). Impressively, some mice demonstrated a dramatic response, but developed disease recurrence soon after, despite weekly injections of virus. Further work is needed to characterize the biologic reasons behind this phenomenon, and this model allows access to the necessary substrates. If CRAds eventually become a treatment option for lung cancer patients, it will be important to understand the mechanisms of resistance for development of interventions and next-generation agents. We believe that the aggressive nature of the model and individual variation between mice resembles the clinical situation noted in patients.

Three of the analyzed CRAds were promising in vivo. This finding is useful because treatment with adenovirus results in a neutralizing antibody response. It has been shown previously that relatively modest changes in the capsid of the virus allows escape from neutralization, suggesting conformation sensitivity of the antibodies.14, 27, 29 Therefore, one could envision utilizing isogenic viruses for subsequent rounds of treatment to retain effective transduction.29

In conclusion, three adenoviral capsid modifications were found to increase gene transfer to NSCLC cells in sensitive in vitro assays utilizing both cell lines and unpassaged clinical samples. Effective killing of NSCLC cells was observed with the respective CRAds and this translated into a therapeutic benefit in a highly aggressive orthotopic model of large-cell NSCLC. Nevertheless, only 1 mouse survived long-term, and therefore further improvements are required. Alternatives include combination with potentially synergistic treatment approaches such as chemotherapy or radiation therapy, or arming of CRAds with immunostimulatory or cytotoxic transgenes. Most importantly, these data set the stage for clinical testing of these CRAds in patients with NSCLC refractory to current modalities. Recent clinical breakthroughs in glioma and head and neck cancer also provide hope for similar results in other treatment-resistant cancers.2–4


  1. Top of page
  2. Abstract
  6. Acknowledgements

We thank Drs. Takashi Takahashi, Jay D. Hunt, and Kazuya Kondo for cell lines.


  1. Top of page
  2. Abstract
  6. Acknowledgements
  • 1
    Kanerva A, Hemminki A. Adenoviruses for treatment of cancer. Ann Med. 2005; 37: 3343.
  • 2
    Immonen A, Vapalahti M, Tyynela K, et al. AdvHSV-tk gene therapy with intravenous ganciclovir improves survival in human malignant glioma: a randomised, controlled study. Mol Ther. 2004; 10: 967972.
  • 3
    Xia ZJ, Chang JH, Zhang L, et al. Phase III randomized clinical trial of intratumoral injection of E1B gene-deleted adenovirus (H101) combined with cisplatin-based chemotherapy in treating squamous cell cancer of head and neck or esophagus. Ai Zheng. 2004; 23: 16661670.
  • 4
    Peng Z. Current status of gendicine in China: recombinant human ad-p53 agent for treatment of cancers. Hum Gene Ther. 2005; 16: 10161027.
  • 5
    Tursz T, Cesne AL, Baldeyrou P, et al. Phase I study of a recombinant adenovirus-mediated gene transfer in lung cancer patients. J Natl Cancer Inst. 1996; 88: 18571863.
  • 6
    Schuler M, Rochlitz C, Horowitz JA, et al. A phase I study of adenovirus-mediated wild-type p53 gene transfer in patients with advanced non-small cell lung cancer. Hum Gene Ther. 1998; 9: 20752082.
  • 7
    Swisher SG, Roth JA, Nemunaitis J, et al. Adenovirus-mediated p53 gene transfer in advanced non-small-cell lung cancer. J Natl Cancer Inst. 1999; 91: 763771.
  • 8
    Puumalainen AM, Vapalahti M, Agrawal RS, et al. Beta-galactosidase gene transfer to human malignant glioma in vivo using replication-deficient retroviruses and adenoviruses. Hum Gene Ther. 1998; 9: 17691774.
  • 9
    Schuler M, Herrmann R, De Greve JL, et al. Adenovirus-mediated wild-type p53 gene transfer in patients receiving chemotherapy for advanced non-small-cell lung cancer: results of a multicenter phase II study. J Clin Oncol. 2001; 19: 17501758.
  • 10
    Swisher SG, Roth JA, Komaki R, et al. Induction of p53-regulated genes and tumor regression in lung cancer patients after intratumoral delivery of adenoviral p53 (INGN 201) and radiation therapy. Clin Cancer Res. 2003; 9: 93101.
  • 11
    Cohen CJ, Shieh JT, Pickles RJ, Okegawa T, Hsieh JT, Bergelson JM. The coxsackievirus and adenovirus receptor is a transmembrane component of the tight junction. Proc Natl Acad Sci U S A. 2001; 98: 1519115196.
  • 12
    Bauerschmitz GJ, Barker SD, Hemminki A. Adenoviral gene therapy for cancer: from vectors to targeted and replication competent agents [review]. Int J Oncol. 2002; 21: 11611174.
  • 13
    Kanerva A, Mikheeva GV, Krasnykh V, et al. Targeting adenovirus to the serotype 3 receptor increases gene transfer efficiency to ovarian cancer cells. Clin Cancer Res. 2002; 8: 275280.
  • 14
    Kanerva A, Wang M, Bauerschmitz GJ, et al. Gene transfer to ovarian cancer versus normal tissues with fiber-modified adenoviruses. Mol Ther. 2002; 5: 695704.
  • 15
    Anders M, Christian C, McMahon M, McCormick F, Korn WM. Inhibition of the Raf/MEK/ERK pathway up-regulates expression of the coxsackievirus and adenovirus receptor in cancer cells. Cancer Res. 2003; 63: 20882095.
  • 16
    Wu H, Seki T, Dmitriev I, et al. Double modification of adenovirus fiber with RGD and polylysine motifs improves coxsackievirus-adenovirus receptor-independent gene transfer efficiency. Hum Gene Ther. 2002; 13: 16471653.
  • 17
    Kaye FJ. RB and cyclin dependent kinase pathways: defining a distinction between RB and p16 loss in lung cancer. Oncogene. 2002; 21: 69086914.
  • 18
    Sherr CJ. Cancer cell cycles. Science. 1996; 274: 16721677.
  • 19
    Hemminki A, Dmitriev I, Liu B, Desmond RA, Alemany R, Curiel DT. Targeting oncolytic adenoviral agents to the epidermal growth factor pathway with a secretory fusion molecule. Cancer Res. 2001; 61: 63776381.
  • 20
    He Y, Rajantie I, Pajusola K, et al. Vascular endothelial cell growth factor receptor 3-mediated activation of lymphatic endothelium is crucial for tumor cell entry and spread via lymphatic vessels. Cancer Res. 2005; 65: 47394746.
  • 21
    Sarkar D, Su ZZ, Vozhilla N, et al. Targeted virus replication plus immunotherapy eradicates primary and distant pancreatic tumors in nude mice. Cancer Res. 2005; 65: 90569063.
  • 22
    Kanerva A, Zinn KR, Chaudhuri TR, et al. Enhanced therapeutic efficacy for ovarian cancer with a serotype 3 receptor-targeted oncolytic adenovirus. Mol Ther. 2003; 8: 449458.
  • 23
    Ranki T, Kanerva A, Sarkioja M, et al. A heparin sulphate targeted conditionally replicative adenovirus Ad5.pK7-Δ24, for the treatment of advanced breast cancer. Genetherapy. 2006. In press.
  • 24
    Choy G, Choyke P, Libutti SK. Current advances in molecular imaging: noninvasive in vivo bioluminescent and fluorescent optical imaging in cancer research. Mol Imaging. 2003; 2: 303312.
  • 25
    Shayakhmetov DM, Gaggar A, Ni S, Li ZY, Lieber A. Adenovirus binding to blood factors results in liver cell infection and hepatotoxicity. J Virol. 2005; 79: 74787491.
  • 26
    Mueller MM, Fusenig NE. Friends or foes — bipolar effects of the tumour stroma in cancer. Nat Rev Cancer. 2004; 4: 839849.
  • 27
    Hemminki A, Zinn KR, Liu B, et al. In vivo molecular chemotherapy and noninvasive imaging with an infectivity-enhanced adenovirus. J Natl Cancer Inst. 2002; 94: 741749.
  • 28
    Le LP, Le HN, Dmitriev IP, et al. Dynamic monitoring of oncolytic adenovirus in vivo by genetic capsid labeling. J Natl Cancer Inst. 2006; 98: 203214.
  • 29
    Mastrangeli A, Harvey BG, Yao J, et al. “Sero-switch” adenovirus-mediated in vivo gene transfer: circumvention of anti-adenovirus humoral immune defenses against repeat adenovirus vector administration by changing the adenovirus serotype. Hum Gene Ther. 1996; 7: 7987.