Intravascular injections of a conditional replicative ... - CiteSeerX

Gene Therapy (2001) 8, 1627–1634  2001 Nature Publishing Group All rights reserved 0969-7128/01 $15.00 www.nature.com/gt

RESEARCH ARTICLE

Intravascular injections of a conditional replicative adenovirus (adl118) prevent metastatic disease in human breast carcinoma xenografts A Fabra2, C Parada1, A Vinyals2, P Martı´n Duque1, V Fernandez1, R Sanchez-Prieto1 and S Ramon y Cajal1 1

Department of Pathology, Clı´nica Puerta de Hierro, Madrid, Spain; and 2Institute of Recerca Oncologica, Barcelona, Spain

We describe a study showing that the adenovirus adl118, lacking both E1B proteins, very efficiently kills human malignant cells ‘in vitro’ and ‘in vivo’. Since many breast cancer patients do not have metastasis at the time of diagnosis, but finally develop it, we planned to study whether intravascular injections of adl118 could prevent metastatic development. We studied the effects of this mutant adenovirus in an orthotopic model of human breast carcinoma xenografts with the breast MB435-lung 2 cell line, which is highly metastatic in the lungs. In this study, all primary tumors were excised when they reached 50–100 mm3 volume in the animals. After surgery, 1010 p.f.u. of adl118 was intravenously injected into a random group of animals, either three times

during the first week only, or once every week. At death, almost all the control animals showed numerous lung metastases of large size, which were present in only 15–40% of the treated animals, depending on the size of the primary tumor at the time of excision. Overall survival was 50–70 days in control mice, and over 120 days in mice injected with adl118. Concomitant treatment with adl118 and cisplatin did not enhance the antitumor effects of adl118. With these results, we conclude that intravenous injection of conditional replicative adenovirus, after excision of the primary tumor, induces a clear decrease in the metastatic disease, and could be a new strategy in preventing tumor metastasis of breast carcinomas. Gene Therapy (2001) 8, 1627–1634.

Keywords: adenovirus; conditional replicative; breast cancer; metastasis

Introduction After nonmelanocytic skin cancer, breast cancer is the second most commonly diagnosed cancer in women. About 10% of women will suffer from breast cancer within their lifespan.1,2 One-third of surgical patients with lymph node-negative breast cancer will develop metastases, and even in patients with small tumors (T1N0) there is a 15–25% likelihood of a distant metastasis.1–3 These data imply that, in many cases, there are already disseminated cancer cells or micrometastases which are not detected at the time of surgery. These clinical-radiological data support the necessity of applying chemotherapy treatment after surgery in order to reduce the final progression of the tumor.4 One of the most promising approaches to oncological treatment is the utilization of selective replication viruses in tumor cells.5–9 This gene therapy approach is based on the fact that adenoviruses defective in replication in normal cells can be replicated in, and selectively kill, tumor cells.5,10 Of these defective adenoviruses, the following are quite promising: ONYX O15,8,10 which is defective in the region of the E1b gene that encodes protein 55K E1B, and which is already in clinical trial phase II–III; adeno-

Correspondence: S Ramo´n y Cajal, Departamento de Patologı´a, Clı´nica Puerta de Hierro, c/San Martı´n de Porres 4, 28035 Madrid, Spain Received 17 April 2001; accepted 24 August 2001

virus ⌬24,6,11 with a deletion in the E1a region that is implicated in binding with the protein of the retinoblastoma gene; and adenovirus adl118,9 which lacks both E1b proteins (19KE1B and 55KE1B). These adenoviruses can replicate in cells with functional alterations in the p53 protein or in the pRb pathway, which are altered in the majority of human tumors.12–14 Furthermore, with these viruses that are deficient in replication in normal cells, it is possible to obtain greater levels of viral particles ‘in vivo’ and, thus, have greater probability of infecting a much higher number of tumor cells. Previous studies have shown significant effects on head and neck tumors, as well as at other locations; a synergetic effect was produced with the conventional oncological treatments of chemotherapy and radiotherapy.8,9 We and others have studied the effects of adenovirus E1a on murine and human tumor cell lines, demonstrating that E1a very significantly increases sensitivity to DNA-damaging agents and decreases tumorigenicity of most cell lines in nude mice.15–19 Adenovirus adl118 carries a full E1a gene and does not express either of the E1b proteins that can block or suppress the pro-apoptotic E1a effect. Moreover, the absence of 55KE1b allows the adenovirus to replicate in cells with no functional p53, and the lack of 19KE1b can enhance spreading the adenovirus. Intratumor injections of this virus have been associated with significant reductions in tumor size and increases in the response to complementary treatments, such as cisplatin and radiotherapy.9

Adenovirus prevents metastatic disease A Fabra et al

1628

Nevertheless, it was the fact that tumor cells infected with this adenovirus were never able to form tumors in nude mice that suggested to us that these cells would not be able to colonize and form metastases ‘in vivo’ if we inoculated the virus intravenously. For this reason, we proposed a study on a model of breast cancer implanted orthotopically in the mouse, and in which 100% of the tumors developed pulmonary metastases. In order to try to correlate the data with what normally occurs in human breast tumors, we extirpated the primary tumors once they reached an average size of 50–100 mm3, since we knew that most of the animals were going to develop massive lung metastases. In the days following surgery, adl118 was injected intravenously and the animals were left to survive for more than 100 days. Interestingly, in the four sets of independent experiments that were carried out, there was a very marked increase in the survival of the animals and a striking reduction in the number and size of pulmonary metastases. With these data, we propose that the post-operational administration of conditional replicative adenoviruses could be a new model for avoiding the development of metastatic disease in human tumors.

Results ‘In vitro’ sensitivity of the MB435-lung 2 human breast carcinoma cell line to infection with adl118 Since cell line MB435-lung 2 forms tumors in mouse breast tissue and is highly metastatic in the lungs, we chose this line to study its sensitivity to infection with adl118 ‘in vivo’. The selective replication and the oncolytic effects of this mutant adenovirus have been previously described.9 First, we analyzed the ‘in vitro’ sensitivity of MB435-lung 2 cells to infections with adl118 and wt adenovirus. As shown in Figure 1, marked lethality was observed with 20 and 100 p.f.u. at 3 days after infection. As previously reported,9 adl118 kills faster than wtAd (Figure 1). Likewise in replication studies, the adminis-

Figure 1 ‘In vitro’ infection of MB435-lung 2 cells with adenovirus adl118 and wt adenovirus. The breast carcinoma cell line MB435-lung2 was infected with 20 and 100 p.f.u. per cell. After 3 days, viable cells were stained with crystal violet. Note the marked decrease in cell viability in treated cells, compared with control cells. Adl118 kills more than wild type at 3 days with 20 p.f.u. Gene Therapy

tration of 1 p.f.u. per cell exhibited a clear lethality at 7– 10 days, and the supernatant of these cells also showed a clear parallel lethality on 293 cells, related to the number of days in cell culture with adl118. Infection with an adenovirus GFP showed green fluorescence in the cells, and no significant loss of cell viability, after 10 days (data not shown). Treatment with adl118 increases the survival of mice Because the MB435-lung 2 cells were not affected by infection with the adGFP virus, and were very sensitive to adl118, we did not inject adGFP ‘in vivo’. In the first in vivo experiment, we inoculated tumor cells in mouse breasts and extirpated the tumors when they reached an average volume of >50 mm3. To obtain that size, we had to wait for 4 weeks after injection of the cell line. From previous results with this model, we knew there would be a very high percentage of these animals that would die from lung metastases. Therefore, half of the mice were injected with three doses of adl118 at 1010 p.f.u. during the first week after extirpation of the primary tumor. They were either left to survive for more than 120 days, or they were killed when they presented a very poor general state. As shown in Figure 2a, 85% of the animals not treated with adll18 were killed before 80 days and showed massive metastases in the lungs. On the other hand, the groups of animals treated with adl118 survived more than 120 days, and only 30% showed micrometastases of less than 1 mm in diameter. This result was very significant (P ⬍ 0.01). Since in this experiment adl118 had been injected three times only during the first week, we decided to make a second experiment of similar characteristics, injecting adl118 once every week. As seen in Figure 2b, the results were worse, with 50% of the animals having lung metastases, 20% of which measured more than 1 mm in diameter. Furthermore, comparing only the percentages of metastases in the two groups of mice, the difference was significant (P ⬍ 0.01). Because 15% of the animals from the control group did not develop metastases, due to the size of their primary tumors, we planned a third experiment in which the primary tumors were left to grow until reaching a maximum size of >100 mm3 (about 5 weeks after injection of MB435lung 2). Under these conditions, 100% of the untreated mice developed massive lung metastases, whereas only about 20% of the group treated with a weekly injection of adl118 developed lung metastases (Figure 3). In this experiment, the difference in the two groups regarding matastases >1 mm was significant (P ⬍ 0.01). These results indicate that the effect of adl118 could be related to the size of the primary tumor and, therefore, to the presence of an already existent metastasis at the time of surgery. Since the presence of adenovirus particles was detected in the lung metastasis by in situ hybridization and by immunohistochemistry with antiE1a or hexon antibodies, we carried out a new experiment in which we injected, concomitantly, adl118 intravascularly and cisplatin intraperitoneally. The primary tumors were left to grow until reaching a size of >100 mm3, and then the animals were treated weekly with 1010 p.f.u. adl118 and 9 mg/kg CDDP. In one group of mice, only CDDP was injected. As seen in Figure 4, all the non-treated animals developed massive lung metastases, with a survival of 50–70 days. The group treated only with CDDP showed an average

Adenovirus prevents metastatic disease A Fabra et al

1629

Figure 3 Lung metastases after excision of large primary breast carcinomas and injection of adenovirus adl118 once per week. In this experiment, orthotopic breast carcinomas were excised with a size of >100 mm3. All untreated mice showed massive lung metastases. In the group of mice treated with adl118 once per week, over 20% showed large lung metastases and about 45% micrometastases. Untreated mice were killed after 60 days, and treated mice after 120 days. Control mice (n = 8), treated with adl118 (n = 12).

Figure 2 Lung metastases after excision of small primary breast carcinomas and injection of adenovirus adl118. (a) After injections of adl118 during the first week only. Orthotopic breast tumors were excised when they reached a size of 50–100 mm3. 85% of untreated mice showed large lung metastases. In mice treated daily for 3 consecutive days after excision of the primary tumor, only about 30% showed micrometastases. Untreated mice were killed after 60–80 days, and treated mice after 120 days. Control mice (n = 10), treated with adl118 (n = 16). Micrometastases are nodules ⬍1 mm in diameter. (b) After injection of adenovirus adl118 once per week. In this experiment, orthotopic breast carcinomas were excised when they reached a size of 50–100 mm3, adl118 was injected once per week. Note that almost 20% of the mice showed large lung metastases and 30% micrometastases. Untreated mice were killed after 60–80 days, and treated mice after 120 days. Control mice (n = 10), treated with adl118 (n = 15).

survival of 100 days, and about 60% had macrometastases. The group treated with both adl118 and CDDP showed about 30% with macrometastases and, unexpectedly, another 50% of the animals showed micrometastases. The group of mice treated only with adl118 showed about 20% of massive lung metastases. The survival averages of mice treated with adl118 alone, with adl118 plus CDDP, or with CDDP alone are shown in Figure 5. Kaplan–Meier survival curve showed that all the untreated mice were dead before 70 days; that by combining CDDP and adl118, only 20% of the mice were still alive after 120 days; and that 60% of the mice treated only with adl118 were alive after 130 days (P ⬍ 0.001). Histological studies Histologically, lung metastases were moderately or poorly differentiated adenocarcinomas with necrosis and a high grade of atypia cytologica. The tumor cells dis-

Figure 4 Lung metastases after excision of large primary breast carcinomas and injection of adenovirus adl118 once per week and concomitant treatment with cisplatin. In this experiment, tumors were excised with a size >100 mm3. All untreated mice showed massive lung metastases. About 60% of the mice treated only with CDDP once per week showed large lung metastases and 30% showed micrometastases. In mice treated concomitantly with CDDP and adl118 once per week, about 30% of the mice showed large lung metastases and over 50% micrometastases. In mice treated with adl118, about 20% showed metastases and 45% micrometastases. Untreated mice were killed after 60–80 days and treated mice after 130 days. Control mice (n = 8), treated with adl118 (n = 12), treated with CDDP (n = 8) and treated with adl118 plus CDDP (n = 14).

played a high proliferation index, evaluated with the marker Ki67, and an intense positivity for the mp53 protein (Figure 6). No significant expressions of neu (cerbB2), or positivity for estrogen or progesterone receptors, were detected. In the study for the presence of viral products in the metastases, positivity was detected for protein E1a (by in situ hybridization or by immunohistochemistry) and for antibodies of adenoviral capsids. This positivity was localized at the interphase level between the areas of central necrosis and the tumor cells in the periphery. Expression of about 5–10% of the Gene Therapy

Adenovirus prevents metastatic disease A Fabra et al

1630

Discussion

Figure 5 Kaplan–Meier survival curves after excision of large human breast carcinomas in a nude mouse human-tumor xenograft model after treatment with adl118 or adl118 plus CDDP. The overall survival of animals treated only with adl118 was 60% after more than 130 days, but it was less than 70 days for all untreated mice. Concomitant treatment of CDDP and adl118 showed an intermediate survival. Control mice (n = 8), treated with adl118 (n = 12), treated with CDDP (n = 8) and treated with adl118 plus CDDP (n = 14).

tumor cells could be verified and, interestingly, the cells were arranged in clearly defined accumulations or groups (Figures 7 and 8). This positivity for viral sequences and proteins could be detected even 1–2 weeks after the last injection of adenovirus, and its distribution in groups revealed that, at this level, the virus could be replicating and infecting adjacent tumor cells. Side-effects No significant side-effects were detected ‘in vivo’, and no significant inflammatory changes could be detected in the liver or other organs of the mice in the necropsy analyses. Other analyses are in progress to evaluate systemic toxicity after intravascular injection of adl118, including liver enzymes and kidney function analyses. All the data required for the approval of this adenovirus for a clinical trial are being assembled.

In this paper, we show that the application of the defective adenovirus adl118 (with both E1b proteins absent), after excision of human breast tumors implanted orthotopically in mice, induced a clear reduction in the incidence of metastasis and significantly increased survival. In this model of metastatic breast cancer, the size of the primary tumor removed, correlated with both the presence of metastases and survival. Accordingly, the decrease in lung metastases after intravascular injections of adl118 was more significant when the primary tumors were less than 100 mm3. This can be interpreted as the presence, at the time of removing the primary tumor, of lung metastasis, where the adl118 virus probably had not spread effectively. Moreover, we deduce from these results that the antitumor action of adl118 must be preferentially on tumor cells circulating in the blood and/or on incipient micrometastases. Moreover, a greater incidence of metastasis was observed when adl118 was injected weekly during the 120 days of survival versus the three injections of adl118 that were made only during the first week after removing the primary tumor. Along this line, follow-up studies are underway to determine this selective effect on circulating tumor cells, based on a model of intravenous injection of tumor cells and the concomitant application of adenovirus adl118. The concomitant application of CDDP did not increase the adl118-mediated reduction in metastases. This discrete concomitant effect of CDDP and adl118 could be explained by the probable premature destruction of cells infected with adl118, blocking the replication of adl118 and its later propagation ‘in vivo’. Therefore, since we have detected the presence of the E1a protein in tumor cells 7–14 days after the last inoculation with adl118 and since previous studies9 we know that E1a expression induces chemosensitivity, we propose that the application of chemotherapy should be administered after several days of adenovirus injections, and not concomitantly. The presence of circulating tumor cells and of micrometastasis at the time of diagnosis of a very high number of human tumors seems to be unquestionable. In fact, the presence of tumor DNA in the plasma is already utilized as a dissemination and prognosis marker in some

Figure 6 Histological features and p53 expression of lung metastases. (a and b) Hematoxylin–eosin picture of a lung metastasis. The tumor shows features of a moderately differentiated adenocarcinoma infiltrating the lung (a, ×100; b, ×400). (c) With immunohistochemistry, most tumor cells showed high levels of p53 protein in the nuclei (×150). Gene Therapy

Adenovirus prevents metastatic disease A Fabra et al

1631

Figure 7 Immunohistochemistry of adenoviral proteins. (a) With an adenovirus capside antibody, groups of positive cells are clearly identified (×150). (b) With an anti-E1a protein antibody, similar groups of tumor cells are observed, expressing high levels of E1a protein in the nuclei (×150). (c) Most positive cells are located close to necrotic areas (×400).

Figure 8 In situ hybridization of E1a sequences. (a) Low-power magnification of a lung metastasis: several groups of positive cells are shown in the periphery of nodules and close to necrotic central areas (×50). (b) At higher magnification, intense E1a positivity is observed in the nuclei and cytoplasm of tumor cells adjacent to the necrotic focus (×150).

types of tumors.20,21 The control of metastatic dissemination of malignant tumors is one of the essential points in oncological treatment.4 In breast cancer cases, it is known that about 15–25% develop a distant metastasis that could not be observed initially.3 With the majority of disseminated human epithelial solid tumors, we have not made notable advances in recent years with conventional chemo- and radiotherapy protocols. The administration of conventional chemotherapy can obtain results in some cases, but the problems inherent in this treatment (ie side-effects and especially chemoresistence still continue).22 Diverse studies and clinical oncological protocols are being developed in order to try to reduce the incidence of metastatic disease in tumors of the breast and other locations where there is no evidence of tumor dissemination at time of extirpation.3 With the results obtained in this study, it is proposed that the intravascular administration of a conditional replicative adenovirus after surgical extirpation of the primary tumor can be an alternative treatment for reducing metastatic disease progression and increasing survival. In four independent experiments, clear increases were observed in the survival of mice with human breast tumors implanted orthotopically. Adenovirus adl118 has neither of the E1b proteins that counteract the proapoptotic effect of E1a, ie neither the 19KE1b protein, which inhibits apoptosis mediated by E1a, nor 55KE1b, which inactivates p53. In this way, adl118 can replicate in tumor cells and, furthermore, does not have the E1a effect ‘counteracted’. Moreover, a replicant adenovirus with a deletion of the 19KE1B kDa viral gene induces more apoptosis and spreads faster than the wild-type adenovirus.23 This faster replication, observed also with the mutant adenovirus adl118, would enhance the oncolytic effects of these viruses in a short period of time, which could be important for decreasing the immune response after treatment. Both effects (selective replication and chemosensitivity), we believe, are essential for obtaining significant clinical results in patients with disseminated tumors. Recently, the adenovirus ⌬24, an E1A mutant (adl922947, with a mutation in CR-2, the pRB-family-binding region of E1A), has shown a potent and selective systemic Gene Therapy

Adenovirus prevents metastatic disease A Fabra et al

1632

antitumoral efficiency.6,11 In fact, in a metastatic breast xenograft model, a marked decrease in lung metastasis was observed when adenovirus ⌬24 was injected intravenously.11 Nevertheless, there are some important differences between the metastatic model used with that mutant E1A adenovirus and the model used in our study. In that model, the primary breast carcinomas were kept growing during treatment and the mice were killed after 56–90 days. In our model, we propose the application of adl118, or another conditional replicative adenovirus, after excision of primary tumors, and leaving the mice to live for more than 120 days. This model is much closer to what usually happens in patients with breast carcinomas. Other approaches with viral vectors deleted in the E1 region, which carry inserted suppressor genes such as p53, p16, etc, have been described as exerting local effects on tumors.24–26 The fact that these adenoviruses are defective in replication limits the clinical application in treating disseminated tumors. Similar limitations can be found in the use of other vectors, such as retroviruses, liposomes, etc. Recently, the intratumoral injection of adp53 in prostate tumors implanted orthotopically has also demonstrated a reduction in tumor growth and metastasis progression in a significant percentage of animals.25 This reduction in the incidence of metastatic disease can be interpreted as secondary to the reduction in size of the primary tumor, and not as an elimination of circulating tumor cells. Such an approach, although revealing the efficiency of adenovirus p53 in prostate tumors, has, unfortunately, limitations due to the fact that adp53 is not replicated ‘in vivo’; its efficacy is limited to the percentage of cells infected initially.5 However, in the application of an adenovirus through systemic administration, a series of important points have to be considered, such as its possible toxicity and its elimination by the immunological response with neutralizing antibodies and macrophages. As reviewed by Alemany et al5 an adenovirus is not a blood-borne virus and its clearance from the blood is rapid through liver Kupffer cells.27 Another important point in the delivery of an adenovirus is the porosity of the endothelial barrier in most tissues; thus, only the spleen, liver and tumor vessels would allow the extravasation of an adenovirus.5 The main toxicity associated with adenovirus systemic delivery described so far is hepatocellular toxicity.5,7 The adenovirus can infect liver cells by the interaction of its capsid fiber with the coxsackie-adenovirus receptor (CAR); thus, CAR-binding ablation could be a way to decrease liver toxicity.28 Nevertheless, this toxicity seems to be dependent on the total number of inoculated viral particles. The main limitation, at least theoretically, in the systemic application of an adenovirus is the immune response which could neutralize its dissemination. Clinical trials with the adenovirus ONYX015, injected intratumorally, have not detected any clear correlation between the presence of neutralizing antibodies and clinical response.8 These studies have the important objection of using intratumoral injections of adenoviruses and, therefore, other factors such as antibody penetration in the tissue must be considered. Clinical trials addressing the effects of adenoviruses injected intravascularly are underway.8 Strategies to abrogate the development of neutralizing antibodies are also being carried out. In fact, interleukin-12 can induce a lymphocyte T-helper 1

Gene Therapy

response and decrease antitumoral cytotoxic T lymphocytes, which destroy infected cells and suppress the development of neutralizing antibodies.5,29 In conclusion, we have shown that intravascular injection of a conditional replicative adenovirus (lacking both E1b proteins), after excision of human breast carcinoma xenografts in mice, induces a clear decrease in number and size of lung metastases and increases the survival of the animal. We propose that the use of conditional replicative adenoviruses could be a new strategy for preventing metastasis in human breast cancer.

Materials and methods Viruses Wild-type Ad5 (wt300) and a recombinant replicationdeficient adenovirus, adl118 (provided by Dr Ginsberg, NIH, Rockville, MD, USA), were employed. The adenovirus adl118 contains a 543-base pair deletion in the E1B region and does not synthesize the 19K or 55K-E1B proteins.30 It also has a small deletion in the E3 region. Adenovirus E1-carrying the green fluorescence protein (GFP), was provided by Dr Alemany (University of Alabama). The adenoviruses were grown in the human embryonic kidney cell line 293. Viruses were purified, titrated by plaque assay and stored in 10% glycerol PBS at –70°C, as previously described.9,31 The 293 cells were grown in DMEM, supplemented with 0.01 m l-glutamine, antibiotics and 10% fetal calf serum. Viability assay Cells were seeded at 104 cells per well in 24-well plates and allowed to grow for 24 h. Then they were infected with adl118, with 10, 20 or 100 plaque-forming units (p.f.u.) per cell. After 3–7 days of incubation, the cell monolayers were fixed with 1% glutaraldehyde and stained with 0.1% crystal violet. Experiments were performed at least three times. Replication assays To analyze the replication efficiency of adl118 in the MB435-lung 2 cell line, virus supernatants were tested at serial dilutions to infect the 293 cells, as previously described.9 They were infected at an MOI of 1–10/cell and lysates were obtained 1, 2, 3 and 5 days after infection by three cycles of freezing and thawing. We then infected the 293 cells at 80% confluency with serial dilutions of the lysates. Forty-eight hours after infection, the wells were fixed in 1% glutaraldehyde and stained by the crystal violet method. Immunohistochemistry for adenovirus protein expression Immunohistochemical staining was performed on deparaffined sections using the avidin–biotin complex immunoperoxidase technique. We placed 4–5 ␮m sections on to DAKO Chem Mate 75-␮m slides (DAKO, Carpinteria, CA, USA). The slides were melted, dewaxed and incubated in DAKO sodium citrate buffer at a dilution of 1:10 in a pressure cooker. The pressure cooker was heated until the maximum pressure was reached. Immunohistochemical staining was performed on a Horizon DAKO (MESIP program) automated immunohistochemical stainer, according to the manufacturer’s instructions. The following antibodies were used: antibody M73

Adenovirus prevents metastatic disease A Fabra et al

(Oncogene Science, Uniondale, NY, USA); anti-adenovirus hexon (MAB805, Chemicon International, Temecula, CA, USA); anti-p53 (DAKO); anti-c-erbB2 (DAKO); anti-Ki-67 (DAKO); and anti-estrogen and progesteron receptors (DAKO). The cells were incubated for 1 h with the primary antibody, followed by a biotinylated goat anti-mouse secondary antibody and streptavidin-horseradish peroxidase conjugate. Diaminobenzidine (DAB) was used as the chromogen, and slides were counterstained with hematoxylin. In situ hybridization Deparaffined sections of lung metastases were rehydrated and treated with proteinase K. The DAKO GenPoint catalyzed signal-amplification system for in situ hybridization (KO620) was used according to the manufacturer’s instructions, with the following modifications. We made a PCR probe of a whole E1a DNA, employing dUTP-Dig. The primers E1aF: AGGCCACTCTTGAG TGCCAGGGAGTAGAG and E1aR: TTAACCACACACGCAATCACAGGTTTACAC amplified a fragment of 1082 bp (between nucletides 494 and 1576 of adenovirus Ad5); 50 ␮g of probe were added. Because we obtained a digoxigenin-marked probe, we used an antibody antidigoxigenin (Jackson Immunoresearch, West Grove, PA, USA) conjugated with peroxidase at 1:400 000. The stringent washes were in 0.5 × SSC, 1% SDS at 65°C for 10 min. Sections were counterstained with hematoxylin. Cell culture conditions Tumor-cell variants of the estrogen-independent MDAMB-435 human breast cancer cell line, which have different metastatic abilities for growing in the mammary fatpad of nude mice, have been previously described.32 The breast cell line MB435-lung 2 is spontaneously and highly metastatic in the lungs from primary tumors developed in mammary tissue;33 thus it was used for this study. Briefly, MB435-lung 2 cells were grown in Ham F-12 medium (GIBCO, Grand Island, NY, USA) supplemented with 10% fetal bovine serum, sodium pyruvate, nonessential aminoacids, l-glutamine, and a two-fold vitamin solution (GIBCO). The cultures were maintained on plastic and incubated in 5% CO2–95% air at 37°C in a humidified incubator. The cell line was examined for, and found to be free of, Mycoplasma (assayed by Gene Probe Mycoplasma TC, purchased from Gene-Probe, San Diego, CA, USA) before animal inoculation. Animals and spontaneous metastasis assays Female athymic BALB/c nude mice were obtained from the Animal Production Area of IFFA-CREDO (Lyon, France). Mice were used when 8 weeks old and were maintained in a laminar-flow cabinet under specific pathogen-free conditions, in accordance with institutional guidelines. Tumor cells were harvested from subconfluent cultures (50–70% confluence) by a short treatment with 0.25% trypsin and 0.02% EDTA solution. The cells were washed in supplemented medium and then resuspended in Hank’s balanced salt solution (HBSS) for injection. Only single cell suspensions of greater than 90% viability (determined by Trypan blue dye exclusion) were used for the studies in vivo. Suspensions of MB435-lung 2 cells were prepared at a concentration of 107 viable cells/ml.

Single inoculations at 1 × 106 cells/50 ␮l in HBSS were used for orthotopic implantation into the mammary fat pad (mfp) of anesthetized females, as described. Growth of the tumors was monitored twice a week, and tumor volumes were calculated from caliper measurements of two orthogonal diameters (x and y), using the formula: volume = (1/2 xy2). Primary tumors were excised from the anesthetized mice when they reached 50–100 mm3 volume, and skin incisions were closed with wound clips. Treatment started after surgery. Briefly, 1010 p.f.u. of adenovirus dl118 was intravenously injected, either three times during the first week only, or once every week. In another treatment, CCDP (9 mg/kg) was administered i.p. once a week concomitantly with the injection of adl118. Mice were killed and examined at different times. All visceral organs were examined for the presence of tumor growth by gross and histological criteria. Lungs were removed, fixed and screened for metastases. Tumors were considered as metastases when they had a size greater than 1 mm diameter, and as micrometastases when they were less than 1 mm diameter.

1633

Statistical analysis The comparison of two proportions in independent samples was estimated using the chi-square test, the continuity correction chi-square and the Fisher’s exact test. The cumulative probability of survival was determined by the Kaplan–Meier method and statistical significant differences with the log-rank test. All P values were twosided and values of 0.05 or less were considered to indicate statistical significance. The SPSS v.90 software was used for all statistical analyses.

Acknowledgements We thank Dr R Alemany for providing the adGFP and for his critical reading of the manuscript. We thank S Jones for preparing the manuscript, Isabel Millan for the statistical anaylsis and Javier Hernandez for the in vivo studies. This work was supported by grants from the ‘Fondo Espan˜ ol de Investigaciones Sanitarias’, FIS 99/0504, and FIS 00/ 903 (AF); the Comunidad de Madrid, CAM 98/08.1/2, Aventis; and the Areces Foundation.

References 1 Hellman S. Natural history of small breast cancers. J Clin Oncol 1994; 12: 2229–2234. 2 Alberg AJ, Singh S, May JW, Helzlsouer KJ. Epidemiology, prevention, and early detection of breast cancer. Curr Opin Oncol 2000; 6: 515–520. 3 Fisher B et al. Prognosis and treatment of patients with breast tumors of one centimeter or less and negative axillary lymph nodes. J Natl Cancer Inst 2001; 2: 112–120. 4 Fisher B et al. Tamoxifen and chemotherapy for lymph nodenegative breast cancer. J Natl Cancer Inst 1997; 89: 491–496. 5 Alemany R, Balague C, Curiel D. Replicative adenoviruses for cancer therapy. Nat Biotechnol 2000; 7: 723–727. 6 Fueyo J et al. A mutant oncolytic adenovirus targeting the Rb pathway produces anti-glioma effect in vivo. Oncogene 2000; 19: 2–12. 7 Heise C, Kim DH. Replication-selective adenoviruses as oncolytic agents. J Clin Invest 2000; 7: 847–851. 8 Khuri FR et al. A controlled trial of intratumoral ONYX-015, a selectively replicating adenovirus, in combination with cisplatin Gene Therapy

Adenovirus prevents metastatic disease A Fabra et al

1634 9

10

11

12 13 14 15 16 17

18

19

20

21

Gene Therapy

and 5-fluorouracil in patients with recurrent head and neck cancer. Nat Med 2000; 8: 879–885. Martin-Duque P et al. Adenovirus lacking 19-kDa and 55-kDa E1B genes exerts a marked cytotoxic effect in human malignant cells. Cancer Gene Ther 1999; 6: 554–563. Bischoff JR et al. An adenovirus mutant that replicates selectively in p53 deficient human tumor cells. Science 1996; 274: 373–376. Heise C et al. An adenovirus E1A mutant that demonstrates potent and selective systemic anti-tumoral efficacy. Nat Med 2000; 10: 1134–1139. Paulovich AG, Toczyski DP, Hartwell LH. When checkpoints fail. Cell 1997; 88: 315–321. Levine AJ. p53, the cellular gatekeeper for growth and division. Cell 1997; 88: 323–331. Hunter T. Oncoprotein networks. Cell 1997; 88: 333–346. Frisch SM. Antioncogenic effect of adenovirus E1A in human tumor cells. Proc Natl Acad Sci USA 1991; 88: 9077–9081. Mumryk JS. Tumor suppressive properties of the adenovirus 5 E1A oncogene. Oncogene 1996; 13: 1581–1589. Sa´ nchez-Prieto R et al. Carcinoma cell lines become sensitive to DNA-damaging agents by the expression of the adenovirus E1a gene. Oncogene 1996; 13: 1083–1092. Sa´ nchez-Prieto R et al. In vivo antitumor effect of retrovirusmediated transfer of the adenovirus E1a gene. Cancer Gene Ther 1998; 4: 215–224. Yujiao Chang J et al. The tumor suppression activity of E1A in HER-2/neu-overexpressing breast cancer. Oncogene 1997; 14: 561–568. Silva JM et al. Presence of tumor DNA in plasma of breast cancer patients: clinicopathological correlations. Cancer Res 1999; 13: 3251–3256. Chen X et al. Detecting tumor-related alterations in plasma or serum DNA of patients diagnosed with breast cancer. Clin Cancer Res 1999; 9: 2297–2303.

22 Olivotto Y et al. Adjuvant systemic therapy and survival after breast cancer. N Engl J Med 1994; 30: 805–810. 23 Sauthoff H, Heitner S, Rom WN, Hay JG. Deletion of the adenoviral E1b-19kD gene enhances tumor cell killing of a replicating adenoviral vector. Hum Gene Ther 2000; 11: 379–388. 24 Ghaneh P et al. Adenovirus-mediated transfer of p53 and p16INK4a results in pancreatic cancer regression in vitro and in vivo. Gene Therapy 2001; 8: 199–208. 25 Eastham JA, Grafton W, Martin CM, Williams BJ. Suppression of primary tumor growth and the progression to metastasis with p53 adenovirus in human prostate cancer. J Urol 2000; 3: 814– 819. 26 Clayman GL et al. Adenovirus-mediated p53 gene transfer in patients with advanced recurrent head and neck squamous cell carcinoma. J Clin Oncol 1998; 16: 2221–2232. 27 Alemany R, Suzuki K, Curiel DT. Blood clearance of adenovirus type 5 in mice. J Gen Virol 2000; 11: 2605–2609. 28 Roelvink PW et al. Identification of a conserved receptor-binding site on the fiber proteins of CAR-recognizing adenoviridae. Science 1999; 286: 1568–1571. 29 Yang Y, Trinchieri G, Wilson JM. Recombinant IL-12 prevents formation of blocking IgA antibodies to recombinant adenovirus and allows repeated gene therapy to mouse lung. Nat Med 1995; 1: 890–893. 30 Babiss LE, Fisher PB, Ginsberg HS. Effect on transformation of mutations in the early region 1b-encoded 21 and 55-kilodalton proteins of adenovirus 5. J Virol 1984; 52: 389–395. 31 Horwitz MS. Adenoviruses. In: Fields BN, Knipe DN, Howley PM (eds). Fields Virology, third edn. Lippincott-Raven: Philadelphia, 1996, pp 2149–2171. 32 Zhang RD, Fidler IJ, Price JE. Relative malignant potential of human breast carcinoma cell lines established from pleural effusions and a brain metastasis. Invas Metast 1991; 4: 204–215. 33 Price JE et al. Breast cancer cell line isolated from metastases in different organs of nude mice. Int J Oncol 1994; 5: 459–467.