Among the most serious complications of immunosuppression therapy in organ transplantation is the high risk for neoplastic tumor recurrence and the development of de novo cancer. For example, recent studies in patients with bronchioloalveolar carcinoma of the lung or cholangiocarcinoma of the liver, receiving a potentially curative organ transplant, show a high tumor recurrence rate (1, 2). Furthermore, liver cancer is second only to age-related cardiovascular complications as the leading cause of late death in liver transplant recipients (3). Development of de novo malignancies represents an even greater threat to the long-term success of an organ transplant. In general, immunosuppressed transplant recipients have an increased risk of developing cancer, particularly skin cancer, posttransplant lymphoproliferative disease, Kaposi sarcoma, and gastrointestinal cancers (4). Furthermore, because the incidence of cancer in immunosuppressed transplant recipients becomes greater over time (4, 5), and new immunosuppressive strategies are expected to extend allograft survival, the incidence of cancer could increase even more. The inclusion of older recipients and donors in transplant programs will also likely increase this problem (6). Thus, cancer has become a major cause of morbidity and mortality in patients otherwise successfully treated by organ transplantation.
To date, little is known about the tumorigenic effects of immunosuppressive agents. However, the clinical observation of increased cancer incidence in transplant recipients has contributed to the view that immunosuppressive agents can promote tumor development, or at least are permissive of cancer progression. Indeed, clinical transplant studies suggest that CsA may be one drug linked to the increased cancer incidence (7). CsA use for other indications, such as psoriasis, is also associated with a high risk for (skin) cancer (8). This association is substantiated by experimental data showing that CsA enhances cancer cell growth characteristics (9), inhibits DNA repair mechanisms (7), and increases liver tumor recurrence in rats (10). Notwithstanding these observations, it has recently been discovered that not all immunosuppressive agents necessarily promote tumor growth and in fact can have antineoplastic activities. In particular, we reported that immunosuppressive doses of rapamycin have potent antiangiogenic properties that inhibit tumor growth (11), which has since been independently confirmed (12). Therefore, careful selection of immunosuppression in patients who have cancer or are at a high risk for developing cancer could improve their long-term prognosis. In the present study, we used mice-bearing tumors and an allogeneic heart transplant (HTx) to evaluate the potential effects of immunosuppressive doses of CsA and rapamycin.
MATERIALS AND METHODS
Animals and Tumor Cell Culture
Male BALB/c, BALB/c-severe combined immunodeficient (SCID), C3H, and C57BL/6J mice (Harlan Sprague-Dawley, Indianapolis, IN) were used. Surgical procedures were performed according to approved protocols.
CT26 cells used in our experiments were derived from a murine BALB/c colon adenocarcinoma (13); B16-F10 melanoma cells were derived from C57BL/6J mice. Tumor cells were maintained in cell culture, and for in vivo experiments, cells were always injected into syngenic mice.
Subcutaneous Tumor Models
To produce established tumors in mice, 1 × 106 tumor cells were injected subcutaneously in the dorsal region of syngenic BALB/c or C57BL/6J mice. In one group of experiments, T- and B-cell–deficient BALB/c-SCID mice were used. In all experiments, tumors were allowed to grow for 7 days without any treatment. At this time, rapamycin or CsA treatment was initiated for the remainder of the experimental period. Tumor volumes were estimated by measurements of the short (a) and long (b) axis of the mass, where V = π/6 ×0.5 ×a2x b. Animals were monitored on a daily basis for tumor size and their general condition. Mice were sacrificed when tumor complications occurred, which included signs of inactivity, cachexia, or decreased responsiveness.
Administration of Rapamycin and Cyclosporine A
Rapamycin (Wyeth Pharma GmbH, Münster, Germany) was administered intraperitoneally at a dose of 1.5 mg/kg/day beginning on day 7 relative to tumor implantation, which was day 0 of HTx; in the mice receiving only a HTx, drug treatment was also started on the day of transplantation (day 0). CsA (Sandimmune, Novartis, Basel, Switzerland) was administered intraperitoneally at doses of 10 or 40 mg/kg/day, according to the same schedule described for rapamycin. Control mice received 0.9% saline through the same route and according to the same schedule.
Heart Transplant Models
Two different allogeneic HTx models were used in this study. The first model involved the use of primary nonvascularized C3H (H-2k) ear–heart transplants (EHTxs) in BALB/c (H-2d) recipients, performed as previously described (14). Heart beat was verified daily by microscopic examination, and an electrocardiogram was used to confirm results on any allograft not showing a clear beat (15). In the second model, C3H abdominal-heterotopic cardiac transplants were performed in C57BL/6J (H-2b) mouse recipients, as previously described (16).
Aortic Ring Assay
Aortic ring assays were performed using a modification of the technique reported by Nicosia et al. (17). Briefly, thoracic aortae were harvested from male Wistar rats (Harlan Sprague-Dawley) and sectioned into 1-mm slices, which were placed on matrigel-coated, 24-well plates. DMEM medium (+0.3% fetal bovine serum) containing combinations of CsA (100 ng/mL), rapamycin (10 ng/mL), mouse anti-transforming growth factor (TGF)- β antibody (10 μg/mL; R&D Systems, Wiesbaden, Germany), or mouse IgG1 isotype control anti-body (10 μg/mL; R&D Systems) was added, and the plate was incubated for 4 days at 37°C, 5% CO2. Phase-contrast photomicrographs of the rings were taken to record vascular sprouting.
Allograft survival times and “start of beat” times in the EHTx model were statistically compared using the Mantel-Cox log-rank test. Tumor volumes are expressed as the mean ± standard error of mean.
Experimental mouse models were designed to simultaneously examine the effects of immunosuppression on allograft survival and growth of established tumors in a transplant recipient. Our intention was to mimic the clinically relevant situation in which a syngenic tumor has developed in a recipient with an allogeneic HTx. For this purpose, tumors syngenic to the chosen mouse recipient strain were allowed to become established before a subsequent allogeneic HTx was performed. The first model we developed involved the use of a colon adenocarcinoma (CT26 tumor cells) in syngeneic BALB/c mice and EHTx. Tumors were established subcutaneously for 7 days, followed by placement of a fully allogeneic C3H EHTx and initiation of CsA or rapamycin immunosuppression. Different combinations of immunosuppression treatment with and without tumor or transplantation were included as experimental groups. When only EHTx was performed, all allografts were rejected by 12 days without immunosuppressive treatment (Fig. 1A). CsA at a dose of 40 mg/kg/day prolonged allograft survival in most mice until at least day 28 after transplantation (P < 0.0001 vs. control). We also tested CsA at a dose of 10 mg/kg/day, but allograft survival was only prolonged for a few days, thus not providing enough protection against rejection for the desired tumor observation period (data not shown). A standard 1.5 mg/kg/day dose of rapamycin protected allografts to a similar degree as 40 mg/kg/day of CsA (P < 0.0001 vs. control) (Fig. 1A). A separate group of BALB/c mice received only subcutaneous CT26 tumors without an EHTx (Fig. 1B). Results from these mice showed that tumors developed at a similar rate for approximately 2 weeks before a distinction in growth rate first became apparent in the different treatment groups. Animals receiving CsA demonstrated a greater tumor size versus controls between days 14 and 26, and 50% of the mice were sacrificed because of tumor complications during this interval. Although the remaining CsA-treated mice were in good condition for a few more days, tumors continued to grow. In direct contrast, rapamycin treatment did not increase CT26 tumor growth, and in fact, tumors grew at a slightly slower rate until approximately day 26. At this point, because tumors had grown somewhat larger, and thus became more angiogenesis dependent, the neoplastic mass began to regress in size. Tumor volume effectively shrank throughout the remainder of the observation period. The definitive and most critical aspect of these experiments was to examine mice bearing both tumor and a heart allograft (Fig. 1C). Compared with saline-treated mice with tumor and an EHTx (no immunosuppression), CsA treatment once again promoted tumor growth. Moreover, for all CsA-treated mice, tumor expansion, resulting in animal death or sacrifice, accounted for the end of allograft survival (Fig. 1C–E). Notably, and in contrast with CsA, rapamycin treatment allowed for 100% animal survival, protected all allografts from immunologic destruction (P = 0.0004 vs. control), and inhibited growing tumors (Fig. 1C,F,G).
A possible indication of an antiangiogenic effect of rapamycin came from the evaluation of EHTx in CT26 tumor-bearing mice. This indication is related to the presumed dependency of EHTx on the development of local neovascularization. In this respect we noted a difference between experimental groups when examining the time interval until each EHTx began beating visibly. Although CsA had no effect (vs. control) on the number of days needed for the EHTx to start beating, rapamycin caused a significant delay in beat initiation (Table 1). In fact, in one animal treated with rapamycin, a heart beat was not observed until day 27 after transplantation. It was surprising to see this long delay, but a beat suddenly became visible, and in this particular mouse, we sustained treatment to 50 days to find that the heart continued to beat strongly, with little change in tumor size.
Another interesting finding in this model was that EHTx survival seemed to be affected by the presence of tumor. More specifically, tumor-bearing mice not treated with any immunosuppression showed allograft survival to 17 days post-transplantation, and the hearts were still beating when the animals were sacrificed because of tumor complications (Fig. 1C). A similar effect was observed in rapamycin-treated mice: All tumor-bearing mice showed 100% allograft survival throughout the observation period, versus 57% allograft survival in mice without tumor (Fig. 1A,C).
A second model system was used whereby B16 melanoma cells were established in C57BL/6J mice that subsequently received a C3H heterotopic HTx. In addition to the advantage of using a primary vascularized allograft, B16 tumors were established subcutaneously, which is a “nearorthotopic” location. Results showed the same contrasting effects of CsA and rapamycin on tumor growth compared with our other tumor-transplant model. In mice receiving a C3H transplant without a tumor present, allograft survival was promoted by 40 mg/kg/day of CsA or 1.5 mg/kg/day of rapamycin (P = 0.002 vs. control) (Fig. 2A). Because we once again found that a 10 mg/kg/day dose of CsA increased allograft survival only slightly (to 12± 1.1 days), the higher drug dose was subsequently used. Mice with B16 tumors but no transplant showed a sharply contrasting, advanced and inhibited, growth pattern when treated with CsA and rapamycin, respectively (Fig. 2B). In mice with both a tumor and transplant, allograft survival was slightly extended by CsA (P = 0.01 vs. control), but allograft loss continued to be caused by progressive tumor growth; none of the allografts were lost as the result of rejection (Fig. 2C). Notably, animals treated with rapamycin showed excellent allograft survival (P = 0.001 vs. control), with simultaneous control of tumor growth throughout the observation period.
To test whether the rapamycin antitumor effect was dependent on a fully competent immune system, we subcutaneously implanted CT26 tumors in BALB/c-SCID mice, treated them daily with rapamycin, and measured tumor growth as usual. Tumor growth was inhibited in rapamycin-treated mice compared with saline-treated mice (Fig. 3A), similar to earlier results in immunocompetent mice (Fig. 1B). Therefore, rapamycin’s antitumor effect seems independent of a fully functional immune system.
Finally, we asked whether the stimulatory effects of CsA on tumor growth in transplant recipients could be reversed when combined with rapamycin treatment. Mice bearing a B16 melanoma received or did not receive a C3H heterotopic HTx (as before), and combination treatment was initiated with CsA and rapamycin. The results showed that tumors in recipients with or without an allogeneic HTx that were receiving combination treatment both demonstrated markedly reduced growth (Fig. 3B), similar to recipients receiving rapamycin treatment alone (Fig. 2B). Therefore, the antitumor action of rapamycin seems to dominate over CsA’s tumor growth-enhancing effects. Moreover, we tested the effects of CsA on in vitro angiogenesis and determined whether rapamycin could be influential. In an aortic ring-sprouting assay, CsA enhanced vascular sprouting, and the effect was reduced with the addition of anti-TGFβ antibody (Fig. 3C). Rapamycin blocked vascular sprouting and also abrogated aortic ring stimulation induced by CsA. Therefore, in these models, CsA’s immunosuppressive and proangiogenic effects do not hinder the antitumor-antiangiogenic activities of rapamycin.
The problem of cancer in transplant recipients has been documented for more than 30 years, but new approaches to deal with neoplasms in immunosuppressed patients have not been realized. A special dilemma is created in this clinical situation, because removal of potential cancer-promoting immunosuppressive drugs risks rejection of the allograft. Therefore, once cancer does occur in a transplant recipient, important decisions must follow. This decision process raises questions including the following: Should immunosuppression be stopped or reduced (risk allograft rejection)? Should the patient continue receiving a normal-dose immunosuppressive therapy, and should standard therapy for the specific cancer be initiated (attack the cancer in the face of full immunosuppression)? Should the immunosuppressive drug regimen be changed, rather than lowered, to minimize cancer progression or even simultaneously attack the cancer? Here we show that two premier immunosuppressive agents in organ transplantation today, used at doses needed to protect allografts from rejection in mice, can indeed positively or negatively influence tumor progression.
Our consideration of CsA in this investigation was based on previous evidence, largely correlative data (18), that it may promote cancer progression. In fact, immunosuppressive doses of CsA did enhance the growth of a colon adenocarcinoma and melanoma in our mouse models. The increase in tumor growth was consistent between the two tumor types, and mouse strains, and did not appear to be dependent on the presence of a primary nonvascularized or vascularized allograft. Furthermore, our data clearly demonstrate that the side effects of tumor progression, leading to death, are largely responsible for allograft loss in a tumor-transplant situation, delineating the fundamental problem existing in the clinic. However, most important, a similar enhancement of established tumor growth did not occur with normal rapamycin-based immunosuppression. Furthermore, tumors in rapamycin-treated mice actually decreased in size by a deterioration process that started at 3 to 4 weeks, and allografts remained functional. The anti-cancer effect of rapamycin was not surprising, because we have shown that this substance, in a nontransplant-related situation, has a potent antiangiogenic effect that inhibits tumor development (11). A critical point is that the drug doses used for immunosuppression were effective in controlling tumor progression while simultaneously protecting allografts from immunologic rejection. A corollary to this issue is that we also tested higher doses of rapamycin (4.5 mg/kg/day) in our tumor-transplant system, but we did not observe a greater reduction in tumor growth (data not shown). It is also interesting to note that when we used a suboptimal immunosuppressive dose of CsA (10 mg/kg/day) in mice with B16 tumors, the tumor growth rate increased to a similar degree as with the higher 40 mg/kg/day dose (data not shown). Together these results indicate that although CsA protects allografts in a tumor-transplant situation, cancer progression is enhanced by continued treatment. In direct contrast, rapamycin in the same situation shows promise to simultaneously protect allografts and control cancer advancement.
Although this study does not concentrate on the mechanism of the rapamycin effect on tumors, inhibition of angio-genesis remains a primary consideration. Indeed, we show that aortic ring sprouting is blocked by rapamycin and, in direct contrast, that CsA promotes vascular sprouting by a mechanism that is at least partially related to TGFβ. Notably, the in vitro proangiogenic effect of CsA is negated by rapamycin. The explanation for this effect remains unclear, but one possibility is that TGFβ, with previously reported proangiogenic effects (19), may be inhibited at a production or signaling level by rapamycin. However, in endothelial cells, we have not observed lower TGFβ mRNA levels with rapamycin treatment (11), although this does not rule out an effect on TGFβ activity. It is notable also in the present study that anti-TGFβ antibody did not completely reverse in vitro angiogenesis stimulated by CsA, suggesting that other factors are involved and that more mechanistic studies are needed. Our in vivo data also indicate that rapamycin is antiangiogenic. More specifically, we noted a significant delay in beat initiation in mice that underwent EHTx and were treated with rapamycin. These data do not specifically show that vascularization of ear-hearts was inhibited by rapamycin, but ear-hearts do require neovascularization, which is a process that generally takes approximately 4 to 6 days (20), coinciding with the time it takes for normal EHTxs to begin beating. In the same respect, inhibition of neovascularization could also affect wound healing (21), which is a side effect of rapamycin treatment (22). Therefore, in addition to taking advantage of rapamycin’s antiangiogenic effect against cancer, recognition of this activity could be useful in designing strategies to avoid its negative side effects.
How much of a role, if any, does the immune system play in the antitumor effect of rapamycin? Data from this study showing that rapamycin has an antitumor effect in SCID mice indicate that the immune system may not have a dominant role. Although this is contrary to the common presumption that inhibition of the immune system would likely promote tumor growth, the balance of effects seem to be in favor of an antitumor action. In addition to angiogenesis, rapamycin can also directly inhibit tumor cell proliferation (23), including the CT26 cells used in the present study (24). Furthermore, as we have recently reviewed (25), mTOR (rapamycin sensitive) signaling pathways are critical for the growth and survival of different types of cancer cells, and rapamycin has also been shown to inhibit tumor metastasis in mice (26). Therefore, because rapamycin has multiple antitumor activities that could potentially dominate over any proneoplastic effects of immunosuppression, we cannot rule out a functional role of the immune system in our tumor models.
An incidental finding in our study was that the presence of tumor, at least in some cases, seemed to have a positive effect on allograft survival. For instance, in the EHTx model, both control and rapamycin-treated mice showed improved transplant survival in the presence of tumor (Fig. 1A,C). On a cautionary note, however, a similar allograft-promoting effect with the B16 tumor was not observed in our second model. These variable results could be the result of immunologic dissimilarities between tumor cell types, because some cancer cells may down-regulate the immune response. For instance, presentation of certain tumor-associated antigens may promote an antigen-presentation environment favoring tolerance development toward allograft antigens (27), or “spreading” of T-cell unresponsiveness. Moreover, cytokines produced by some tumor cells, such as interleukin-10, could have an immunosuppressive effect that promotes immune escape (28). In addition to affecting T-cell proliferation, interleukin-10 is known to inhibit T-helper type 1 cell expansion and dendritic cell maturation (29). In this same respect, vascular endothelial growth factor, produced in high amounts by many tumor cells, including the CT26 tumor cells used in our study (11), has been suggested to inhibit immunity to cancer cells by causing a maturational defect of dendritic cells (30). The role of immunity in the situation in which tumors exist in transplant recipients creates a complex situation that will require more in-depth investigations.
Our study suggests that specific immunosuppressive agents may promote or inhibit cancer progression in mice while exerting their prescribed protective effects on organ allografts. In this respect, although CsA protected allografts from rejection, it facilitated tumor growth at immunosuppressive concentrations. In contrast, rapamycin demonstrated an ability to attack growing tumors while simultaneously protecting allografts from immunologic destruction. This dual role should be considered and may prove to be of critical importance in clinical situations in which transplant recipients are at high risk for tumor recurrence or have posttransplant de novo cancer.
1. Garver RI, Zorn GL, Wu X, et al. Recurrence of bronchioloalveolar carcinoma in transplanted lungs. N Engl J Med 1999; 340: 1071.
2. Meyer CG, Penn I, James L. Liver transplantation for cholangiocarcinoma: results in 207 patients. Transplantation 2000; 69: 1633.
3. Fung JJ, Jain A, Kwak EJ, et al. De novo malignancies after liver transplantation: a major cause of late death. Liver Transpl 2001; 7: S109–S118.
4. Penn I. Occurrence of cancers in immunosuppressed organ transplant recipients. Clin Transplant 1998; 147.
5. Sheiner PA, Magliocca JF, Bodian CA, et al. Long-term medical complications in patients surviving > or = 5 years after liver transplant. Transplantation 2000; 69: 781.
6. Sheil AG. Donor-derived malignancy in organ transplant recipients. Transplant Proc 2001; 33: 1827.
7. Herman M, Weinstein T, Korzets A, et al. Effect of cyclosporin A on DNA repair and cancer incidence in kidney transplant recipients. J Lab Clin Med 2001; 137: 14.
8. Marcil I, Stern RS. Squamous-cell cancer of the skin in patients given PUVA and ciclosporin: nested cohort crossover study. Lancet 2001; 358: 1042.
9. Hojo M, Morimoto T, Maluccio M, et al. Cyclosporine induces cancer progression by a cell-autonomous mechanism. Nature 1999; 397: 530.
10. Freise CE, Ferrell L, Liu T, et al. Effect of systemic cyclosporine on tumor recurrence after liver transplantation in a model of hepatocellular carcinoma. Transplantation 1999; 67: 510.
11. Guba M, von Breitenbuch P, Steinbauer M, et al. Rapamycin inhibits primary and metastatic tumor growth by antiangiogenesis: involvement of vascular endothelial growth factor. Nat Med 2002; 8: 128.
12. Luan FL, Ding R, Sharma VK, et al. Rapamycin is an effective inhibitor of human renal cancer metastasis. Kidney Int 2003; 63: 917.
13. Brattain MG, Strobel-Stevens J, Fine D, et al. Establishment of mouse colonic carcinoma cell lines with different metastatic properties. Cancer Res 1980; 40: 2142.
14. Babany G, Morris RE, Babany I, et al. Evaluation of the in vivo dose-response relationship of immunosuppressive drugs using a mouse heart transplant model: application to cyclosporine. J Pharmacol Exp Ther 1988; 244: 259.
15. Fey TA, Krause RA, Hsieh GC, et al. Improved methods for transplanting split-heart neonatal cardiac grafts into the ear pinna of mice and rats. J Pharmacol Toxicol Methods 1998; 39: 9.
16. Larsen CP, Morris PJ, Austyn JM. Migration of dendritic leukocytes from cardiac allografts into host spleens. A novel pathway for initiation of rejection. J Exp Med 1990; 171: 307.
17. Nicosia RF, McCormick JF, Bielunas J. The formation of endothelial webs and channels in plasma clot culture. Scan Electron Microsc 1984; Pt 2: 793.
18. Dantal J, Hourmant M, Cantarovich D, et al. Effect of long-term immunosuppression in kidney-graft recipients on cancer incidence: randomised comparison of two cyclosporin regimens. Lancet 1998; 351: 623.
19. Dickson MC, Martin JS, Cousins FM, et al. Defective haematopoiesis and vasculogenesis in transforming growth factor-β 1 knock out mice. Development 1995; 121: 1845.
20. Guba M, Cernaianu G, Koehl G, et al. A primary tumor promotes dormancy of solitary tumor cells before inhibiting angiogenesis. Cancer Res 2001; 61: 5575.
21. te Velde EA, Voest EE, van Gorp JM, et al. Adverse effects of the antiangiogenic agent angiostatin on the healing of experimental colonic anastomoses. Ann Surg Oncol 2002; 9: 303.
22. Kings-Biggs MB, Dunitz JM, Parks SJ, et al. Airway anastomotic dehiscence associated with use of sirolimus immediately after lung transplantation. Transplantation 2003; 75: 1437.
23. Grewe M, Gansauge F, Schmid RM, et al. Regulation of cell growth and cyclin D1 expression by the constitutively active FRAP-p70S6K
pathway in human pancreatic cancer cells. Cancer Res 1999; 59: 3581.
24. Seeliger H, Guba M, Koehl G, et al. Blockage of 2-deoxy-D-Ribose induced angiogenesis with Rapamycin counteracts a thymidine phosphorylase-based excape mechanism available for colon cancer under 5-fluorouracil therapy. Clin Cancer Res 2004; 10: 1843.
25. Guba M, Graeb C, Jauch KW, et al. Pro and anti-cancer effects of immunosuppressive agents used in organ transplantation. Transplantation 2004; in press.
26. Luan FL, Hojo M, Maluccio M, et al. Rapamycin blocks tumor progression: unlinking immunosuppression from antitumor efficacy. Transplantation 2002; 73: 1565.
27. Thomson AW, Lu L. Are dendritic cells the key to liver transplant tolerance? Immunol Today 1999; 20: 27.
28. Garcia-Hernandez ML, Hernandez-Pando R, Gariglio P, et al. Interleukin-10 promotes B16-melanoma growth by inhibition of macrophage functions and induction of tumour and vascular cell proliferation. Immunology 2002; 105: 231.
29. Lee WC, Qiani S, Wan Y, et al. Contrasting effects of myeloid dendritic cells transduced with an adenoviral vector encoding interleukin-10 on organ allograft and tumour rejection. Immunology 2000; 101: 233.
30. Gabrilovich DI, Chen HL, Girgis KR, et al. Production of vascular endothelial growth factor by human tumors inhibits the functional maturation of dendritic cells. Nat Med 1996; 2: 1096.