Anti-CD40 monoclonal antibody therapy in combination with irradiation results in a CD8 T-cell–dependent immunity to B-cell lymphoma

Lymphomas generally exhibit exquisite radiosensitivity, and for many years irradiation has played an integral role in the treatment of a range of B-cell malignancies.^1,2  Localized irradiation remains the only potentially curative option for early-stage “low-grade” non-Hodgkin lymphomas (NHLs), and the use of total nodal irradiation in NHL has led to 10-year disease-free survival times, which are comparable to cytotoxic drugs.¹  Total body irradiation (TBI) has the additional theoretical advantage over chemotherapy of eradicating tumor in sanctuary sites.³ 

Integrating monoclonal antibody (mAb) with irradiation appears a promising strategy, given the radiosensitivity of lymphoma and the potency of a number of mAbs when used as single agents against B-cell antigens. In view of the lack of overlapping toxicity and the different mechanisms of action, mAbs are increasingly being successfully combined with chemotherapy and radiotherapy in the clinic.^4,5 

While the optimal dose and fractionation schedule of TBI and the relative merits of mAb-targeted radiation versus TBI remain the subject of ongoing investigation,⁶  it is becoming increasingly clear that mAb specificity is critical to improving therapeutic outcomes when mAb is combined with irradiation. Our group and other workers have previously shown that some signaling mAbs have additive antitumor activity when combined with irradiation.^7,8  A number of different antibody-mediated effector mechanisms including antibody-dependent cell cytotoxicity (ADCC), complement-dependent cytotoxicity (CDC), and cell surface signaling are likely to operate and may act in concert with the irradiation to enhance antitumor activity. Recent data suggest that some signaling mAbs that are able to regulate antigen presentation and cytotoxic T-cell (CD8⁺) responses may be effective as anticancer therapies.^9,10 

One such immunomodulatory mAb is anti-CD40, which has shown therapeutic activity in a range of preclinical models.^10,11  CD40 is a member of the family of tumor necrosis factor (TNF) receptor molecules and has a broad pattern of expression, including B cells and a variety of antigen-presenting cells (APCs).¹²  CD40 ligation triggers a range of cellular functions including the activation of APCs, “licensing” them to present antigen to cytotoxic T cells.¹³  These effects on the generation of CD8⁺ T cells have made anti-CD40 treatment particularly attractive for tumor immunotherapy in the appropriate immunization context. With large amounts of anti-CD40 mAb and optimal tumor burden, activation of tumor-specific T cells and durable responses in tumor-bearing hosts have been seen for a range of both CD40⁺ and CD40^– syngeneic B-cell lymphomas.^10,11  However, in each of these models it was clear that the immune system can be overwhelmed by a rapidly growing tumor in such a way that mice succumb to the malignancy before the anti-CD40 can promote an effective T-cell response. Combining anti-CD40 with an effective cancer treatment such as irradiation appears a promising strategy whereby the host immune system is simultaneously stimulated while debulking tumor and liberating antigens that might be recognized by responding T cells.

In this study we have investigated the efficacy and feasibility of combining TBI with anti-CD40 treatment. We have demonstrated for the first time that there is a dose-dependent additive effect between irradiation and anti-CD40 mAb, which affords long-term tumor protection via a CD8⁺ T-cell–dependent pathway. Moreover, we have shown that there is a clear radiation threshold for this therapeutic activity, and with increasing doses of irradiation T-cell activation is impaired and long-term tumor protection is lost. We believe these observations have important implications for a number of different immunotherapy strategies for cancer.

BALB/c and CBA/H mice were supplied by Harlan UK (Blackthorn, United Kingdom). The A31¹⁴  and BCL₁¹⁵  B-lymphoma lines were maintained by in vivo passage in BALB/c and CBA/H mice, respectively. The πBCL₁ variant of the BCL₁ tumor can be maintained both in vivo and in liquid culture.¹⁶ 

All cell cultures were maintained in supplemented Dulbecco modified eagle medium (DMEM) (Life Technologies, Paisley, United Kingdom) containing glutamine (2 mM), pyruvate (1 mM), penicillin and streptomycin (100 IU/mL), amphotericin B (Fungizone; Bristol-Meyers Squibb, Chester, United Kingdom; 2 μg/mL), and 5% to 10% fetal calf serum (FCS; Myoclone; Life Technologies), except πBCL₁, which was maintained in supplemented RPMI (Life Technologies), plus 50 μmol/L 2-mercaptoethanol (2-ME) (BDH, Poole, United Kingdom) and 15% FCS.

Monoclonal antibodies used in this study were Mc106A5 (rat anti-BCL₁ Id)⁹ ; Mc39-12 (rat antimouse Fcμ)⁹ ; Mc39-16 (rat anti-A31 Id)⁹ ; 3/23 (rat antimouse CD40)¹⁷ ; 1D3 (rat antimouse CD19)¹⁸ ; TI2-3 (rat antimouse major histocompatibility complex [MHC] II; Tenovus Research Laboratory); HB79 (rat antimouse MHC I K^d and D^d, American Type Culture Collection, Manassas, VA); NIMR6 (rat antimouse CD22)¹⁹ ; YTA1.3.2 (rat antimouse CD4)²⁰ ; and YTS169 (rat antimouse CD8).²⁰  Hybridoma cells were maintained in stationary culture using 5% supplemented DMEM (as above). To purify the immunoglobulin G (IgG) mAb, supernatants were concentrated by membrane filtration (Amicon, Beverly, MA), precipitated with saturated ammonium sulfate, and dialyzed and fractionated on protein G (Pharmacia, Piscataway, NJ).

Groups of age-matched mice were injected intravenously with 10⁶ A31 (CBA/H) or BCL₁ (BALB/c) lymphoma cells on day 0. Total body irradiation (TBI) was conducted using a modified 225 kV x-ray unit (Gulmay, Chertsey, United Kingdom). TBI was given on day 10 at a dose rate of 1.15 Gy/min. In all cases animals were fed acidified water (pH 2.5; 1N HCl) supplemented with neomycin sulfate (2 g/L) (Sigma Aldrich, Poole, United Kingdom), starting 1 week prior to TBI and continuing for 4 weeks afterward. Animals were housed in a sterile maximizer unit (Techniplast, Kettering, United Kingdom). Mice were treated with intravenous mAb 4 hours after irradiation.

Both tumors develop primarily in the spleen, with a terminal leukemic overspill. Survival was monitored daily, and the results were analyzed using the χ² test of Peto.²¹  Animal immunotherapy was approved by the local ethical committee of Southampton University and was performed under a United Kingdom Home Office license.

Mouse splenocytes were analyzed by direct immunofluoresence staining using a FACSCalibur (Becton Dickinson, Mountain View, CA). Briefly, splenocytes at 10⁶/mL (100-μL sample) were incubated with fluorescein isothiocyanate (FITC)– or phycoerythrin (PE)–labeled mAb (50 μg/mL) for 20 minutes on ice, then washed with phosphate-buffered saline (PBS) containing 1% bovine serum albumin, and immediately analyzed.

To measure tumor growth kinetics following treatment and chart the progress of any immunologic response, groups of mice were inoculated with 10⁶ tumor cells intravenously on day 0 and treated on day 10 with TBI, mAb (1 mg intravenously), or a combination of the two. Mice were housed under sterile conditions (as before). Mice were killed at sequential time points after treatment, spleens removed, homogenized, washed once by centrifugation, and resuspended at 10⁶/mL in supplemented RPMI.

Tumor cells were detected by dual-fluorescence staining using FITC–Mc39-16 (rat anti-A31 idiotype)¹⁰  versus PE-ID3 (rat antimouse CD19). Cells were stained with mAb (50 μg/mL) on ice for 20 minutes, washed, and analyzed on a FACSCalibur (Becton Dickinson) as before.

To characterize the immunologic response to treatment, homogenates were labeled with either FITC-KT3 (rat antimouse CD3)²²  versus PE-YTS169 (rat antimouse CD8), or FITC-YTA1.3.2 (rat antimouse CD4)²⁰  to determine the number of splenic CD8 or CD4 T cells, respectively.

Measurement of in vitro apoptosis was conducted according to the method of Nicoletti et al.²³  Essentially, culture plates (24-well multiplates; Life Technologies) were coated with mAb (5 μg/mL) for 2 hours at 37°C prior to addition of cells (πBCL₁ lymphoma; 2 × 10⁵/mL). Cells were then immediately treated with 5 Gy external beam irradiation (dose rate 1.15 Gy/min as before) or left untreated. Controls were maintained under similar conditions to test samples while irradiation was conducted, reducing the risk of experimental differences due to variations in environmental parameters. Samples (approximately 5 × 10⁵ cells) were harvested 72 hours after treatment, washed once in PBS, and resuspended in hypotonic fluorochrome solution (50 μg/mL propidium iodide [PI], 0.1% [wt/vol] sodium citrate, 0.1% [vol/vol] Triton X-100) and then stored overnight at 4°C to allow staining of DNA. PI-stained nuclei were analyzed using a FACSCalibur as previously described.⁷ 

Clonogenic survival was determined using a serial dilution assay according to the method of Olive et al.²⁴  Briefly, πBCL₁ lymphoma cells were seeded into individual wells of a 96-well plate over a range of cell densities (density: 10 wells per cell density; 8 cell densities per dose) and treated with 0 to 5 Gy of irradiation as before. Cells were plated in 70% fresh medium and 30% conditioned medium. Control (anti-A31 Id) or anti-CD40 mAb was added at a final concentration of 5 μg/mL. After 14 days the fraction of wells containing more than 25 surviving cells was determined using a light microscope. The cell number that resulted in 30% empty wells was calculated and used to generate a clonogenic survival curve. Experiments were repeated at least 3 times.

To determine T-cell specificity, BALB/c mice were inoculated with tumor (10⁶ cells, intravenously) and treated with 5 Gy TBI and anti-CD40 (100 μg) on day 15, as before. Splenocytes were harvested from mice 5 days after treatment and levels of cytotoxic activity against target lymphoma cells were assessed using a standard chromium ⁵¹Cr-release assay.⁹  Briefly, ⁵¹Cr-labeled targets (A31 or πBCL₁) were prepared from spleens of terminal animals (A31) or cultured cells (πBCL₁) and resuspended at 10⁵/mL in supplemented RPMI. The effectors were splenocytes taken from animals 5 days after treatment with TBI, anti-CD40, or a combination of the two. For the assay, 50 μL aliquots of target cells (5 × 10³) were seeded in wells of a 96-well U-bottomed plate (Life Technologies) and 100 μL aliquots of effector cells were then added at various CD8⁺ T cells to target (E/T) ratios. Plates were centrifuged at 200g for 5 minutes at room temperature, incubated for 4 hours at 37°C, and recentrifuged at 500g for 5 minutes before 100 μL supernatant was harvested from each well to estimate the released ⁵¹Cr. Determinants were performed in triplicate and maximal release of ⁵¹Cr obtained using target cells to which 150 μLof 1% Nonidet P-40 had been added. The percentage of specific ⁵¹Cr release was calculated using the formula: ([sample release – background release]/[maximum release – background release]) × 100. Controls included effector and target cells incubated with a chemically engineered bispecific F(ab′)₂ antibody with specificities for murine CD3 and MHC II, capable of retargeting any T cell against any B-cell target. Details of this reagent and its activity have been described previously.²⁵  Finally, to show that killing activity was dependent on CD8⁺ T cells, blocking monoclonal antibody against MHC I (HB79, anti-K^d and D^d, American Type Culture Collection) was included at a final concentration of 25 μg/mL.

Mice were inoculated with tumor and treated with 5 Gy TBI and anti-CD40 as before. Just prior to the peak of the CD8 response (typically day 4 after treatment), spleens were removed and the remaining tumor was depleted using FITC-labeled anti-CD19 mAb, followed by anti-FITC MACS beads and LS columns (Miltenyi Biotec, Bergisch Gladbach, Germany) as described in the manufacturer's instructions. Splenocytes from immunized mice were then mixed with 10⁴ fresh lymphoma cells at a ratio of 100:1 CD8⁺ T cells to tumor cells and given by intraperitoneal injection to naive mice. Control animals received tumor alone or tumor plus normal splenocytes. Survival was recorded daily.

Therapy was assessed in 2 B-cell lymphoma models, A31 and BCL₁, which both express similar levels of the CD40 antigen (approximately 4 × 10⁴ molecules per cell).⁹  Previous work in these B-cell lymphoma models has shown that anti-CD40 mAb was highly effective at promoting T-cell responses but that once the tumors had reached a critical size the immune system became overwhelmed before immunity could be established. Flow cytometry estimates suggested that treatment with anti-CD40 mAb could promote immunity in mice while the splenic lymphoma load remained below 10%. Building on these observations, we first tested the relative efficacy of a combination of anti-CD40 and TBI in comparison with anti-CD40 or irradiation alone. Mice received 10⁶ tumor cells, intravenously on day 0, and were treated on day 10 with either TBI (1 to 8 Gy), anti-CD40 mAb (1 mg intravenously), or a combination of the two, whereby the mAb was given 4 hours after irradiation. Figure 1A illustrates that treatment of the A31 lymphoma with TBI alone results in only a modest increase in survival of just a few days over that seen in control animals (PBS treated). Likewise, treatment with anti-CD40 mAb alone provided minimal improvement in survival (Figure 1B). However, when anti-CD40 and TBI were used in combination, there was a clear radiation dose–dependent increase in survival with 2 Gy providing around 6 days of advantage over controls, 4 Gy around 35 days, and 5 Gy consistently affording long-term protection (more than 100 days) to more than 80% of animals (Figure 1B). A similar radiation dose response was also observed in combination with another anti-CD40 mAb, 1C10,²⁶  (data not shown). Interestingly, as the dose of radiation was increased further to 8 Gy there was a significant decrease in therapeutic efficacy compared with the 5 Gy plus anti-CD40–treated cohort (P < .02), with no long-term survivors (Figure 1B). Similar results were obtained in a second lymphoma model, BCL₁ (Figure 1C), where long-term survival was obtained using as little as 100 μg anti-CD40 plus TBI. This tumor model is more sensitive to anti-CD40 treatment than the A31 model and, although some tumor protection is observed, no long-term survival is seen when treating with 100 μg mAb alone. Long-term survival was, however, seen with anti-CD40 mAb when the dose was increased 10-fold higher (1 mg) (data not shown).

We further assessed the effect of mAb dose on tumor protection by using the A31 tumor model, which was considerably more resistant to the effects of anti-CD40 mAb alone. Mice were inoculated with tumor as before and treated on day 10 with 5 Gy TBI plus a range of anti-CD40 doses. Figure 2A shows that the A31 lymphoma is still highly sensitive to combination treatment with doses of mAb as low as 0.25 mg, with more than 60% of animals becoming long-term protected at all doses. Clearing of tumor was not seen with mAb directed to other B-cell antigens, such as CD19, MHC II, and CD22 (Figure 2B). Thus, long-term eradication of tumor appears to be dependent on the inherent activity of the anti-CD40 mAb.

We next investigated the mechanisms behind the tumor protection and whether the combination of irradiation and the anti-CD40 mAb were acting directly on the tumor cells, resulting in an additive or synergistic increase in tumor cell kill. In vitro apoptosis and clonogenic survival assays were therefore performed. The πBCL₁ lymphoma cells, a variant of the BCL₁ tumor, were irradiated and then cultured in the presence of mAb. The levels of cell death were determined by flow cytometry using PI staining to measure DNA fragmentation. Figure 3A shows the DNA profile of cells treated with control mAb, anti-CD40, or anti-Fcμ for 72 hours. The percentage of DNA in the sub-G₁/G₀ phase of the cell cycle is shown gated and represents fragmented DNA present in apoptotic cells.²³  Anti-CD40 had no effect on the DNA profile compared with controls, while as expected anti-Fcμ induced cell death (12%). When cells were treated with 5 Gy external beam irradiation, there was a marked increase in DNA fragmentation (49%). In the presence of anti-Fcμ this increased still further (63%), indicating an additive effect between the mAb and irradiation as we have previously described.⁷  However, in stark contrast there was no increase in apoptosis over controls when irradiation was combined with anti-CD40 mAb. Likewise, there was no significant difference in clonogenic survival between control and anti-CD40–treated cells (Figure 3B).

Following on from these in vitro observations, we decided to address the issue of a potential additive effect between the radiation and the mAb through direct cytotoxicity in an in vivo setting. If treatment with anti-CD40 had the effect of sensitizing lymphoma cells to irradiation, we might expect such an interaction to be independent of tumor load and to be potentially more effective with smaller tumor burdens. Mice were therefore inoculated with between 10⁶ and 10³ tumor cells intravenously on day 0 and were then treated as before on day 10 with anti-CD40 mAb in combination with 5 Gy TBI. The results shown in Figure 4 reveal a critical threshold for the number of tumor cells required for the combination therapy to deliver long-term tumor protection. With the lowest tumor burden of 10³ (Figure 4D), lasting tumor immunity was not seen and only when the inoculum was increased to above 10⁴ tumor cells was long-term tumor protection seen (Figure 4A-C). These results suggest that there is not a direct process of sensitization occurring in vivo, but instead indicate that a critical tumor load is required in order to achieve long-term protection, pointing instead to there being an immunologic component to the therapy.

To try to identify the mechanism of tumor clearance in vivo, tracking experiments were performed to monitor the kinetics of tumor growth and assess the nature of the possible immune response. Mice were inoculated with 10⁶ tumor cells intravenously on day 0 and treated as before on day 10 with anti-CD40 mAb in combination with TBI. Animals were subsequently killed at sequential time points after the treatments and the number of splenic tumor cells was determined using flow cytometry. Figure 5 shows that the kinetics of tumor growth in control and anti-CD40–treated animals is similar and that they become terminally ill due to massive splenic tumor just 8 days after treatment. TBI results in a radiation dose–dependent decrease in tumor cell number, and 8 Gy clears a whole log more tumor than 5 Gy alone. However, tumor regrowth occurs rapidly after TBI when used alone, and similar amounts of tumor to those found before treatment are reached 10 to 12 days after treatment for both 5 and 8 Gy. In stark contrast, with the addition of anti-CD40 mAb the combination treatment results in an initial and then much more pronounced and sustained second decrease in the number of tumor cells. By day 12 with the combination therapy, less than 1% of the initial tumor burden present prior to treatment remains. Interestingly, there appears to be little difference in the profile of tumor eradication between mice receiving anti-CD40 and 5 or 8 Gy of TBI, although there are major differences in overall survival.

Despite similarities in tumor eradication, there is a profound difference in the number of splenic cytotoxic CD8⁺ T cells. Figure 6 shows the number of CD8⁺ cells present in the spleens of animals 5 days after treatment. Mice treated with 5 Gy TBI plus anti-CD40 show a dramatic increase in the number of CD8⁺ T cells with a rapid expansion that peaks from day 5 to 7, which is highly significant compared with mice given TBI alone (P < .01). Importantly, this increase coincides with the observed decrease in tumor burden and does not occur in nontumor-bearing mice receiving similar treatment, indicating that the response is dependent on tumor. Moreover, at combinations of 8 Gy irradiation this T-cell response does not occur. Neither does it occur at lower doses such as 2 Gy (data not shown). Importantly, absolute lymphocyte numbers were greatly increased in tumor-bearing animals, a phenomenon that has been observed before,¹⁰  presumably as a result of the considerably larger spleen size. Furthermore, mice treated with combinations of irradiation and mAb of other specificities, such as anti–MHC II, develop tumor at a similar rate to controls and show no expansion of CD8⁺ T cells (not shown), which again points directly to the response being specific to anti-CD40 mAb.

Despite the rapid expansion of CD8⁺ cells following treatment with 5 Gy TBI plus anti-CD40, the total number of CD8⁺ cells was still less than in control or anti-CD40 alone–treated groups. However, only the combined irradiation and anti-CD40 treatment groups provided long-term tumor clearance. This apparent paradox can perhaps be explained by looking at splenic CD8⁺ T-cell activity. Splenocytes from the combination treatment group show a highly significant (P < .01) increase in cytotoxic activity against syngeneic lymphoma cells in vitro (Figure 7A), compared with splenocytes from anti-CD40–treated, control, or irradiated mice, which show little or no activity. Cytotoxicity was specific in that splenocytes killed only syngeneic lymphoma cells and was strictly CD8⁺ T-cell dependent, because killing was blocked by mAb to MHC I (Figure 7B). To ensure that the A31 lymphoma was susceptible to killing, splenocytes were incubated with a bispecific monoclonal antibody (anti–MHC II × anti-CD3)²⁵  that retargeted effector T cells through CD3 against B-cell targets through MHC II (Figure 7B, bispecific antibody [BsAb]). Together, these data indicate that the CD8⁺ population in the TBI plus anti-CD40–treated mice is expanding as part of a specific response against the lymphoma.

To confirm the importance of CD8⁺ T cells to in vivo protection, therapy was conducted in the presence of T-cell–depleting mAb. Mice received anti-CD8 mAb (YTS169, 0.5 mg/d) or anti-CD4 mAb (YTA1.3.2, 1 mg/d) 1 day prior to treatment and subsequently every 3 days to a total of 4 inoculations. Figure 8A shows that the therapeutic efficacy of TBI and anti-CD40 is lost in mice depleted of CD8⁺ but not CD4⁺ T cells. These data strongly suggest that the combination therapy is able to eradicate lymphoma through the generation of a CD8⁺ T-cell response. Furthermore, mice previously inoculated with lymphoma that had become long-term disease-free survivors with combination TBI and anti-CD40 were reinoculated with fresh tumor (10⁶ cells intravenously). Whereas control animals developed tumor as usual, 80% of previously treated mice were able to reject tumor and remained disease free for over another 100 days (Figure 8B). The rechallenge experiments suggest that the response generated by anti-CD40 and irradiation is able to provide antitumor immunity for the long-term survivors.

Finally, to confirm that the combination therapy results in protective immunity, we performed adoptive transfer studies. Splenocytes were therefore harvested from treated mice just prior to the peak of the primary response and were transferred in to naive recipients together with fresh lymphoma cells at an effector-target ratio of 100:1. While cohorts receiving tumor plus splenocytes from naive, untreated controls had a median survival time of around 35 days (Figure 9), similar to that of animals given PBS, those given tumor plus splenocytes from treated animals showed a significant increase in protection (P < .01), with a median survival of around 50 days. Thus, it appears that adoptive transfer of T cells from mice given combination therapy is able to delay the rate of lymphoma growth in naive recipients.

Integrating mAb and irradiation is emerging as an important new approach for lymphoma therapy.^7,8  We have demonstrated for the first time in 2 different tumor model systems that irradiation and anti-CD40 can work in concert to eradicate lymphoma and produce long-term survival when anti-CD40 treatment or irradiation alone is ineffective in extending survival. The combination of anti-CD40 and irradiation is able to generate a T-cell immunity that provides long-term protection consistently in more than 80% of treated animals. The tumor protection seen is critically dependent on the radiation dose. The CD8⁺ T-cell response is essential to provide long-term tumor protection, and as the radiation dose increases, the therapeutic efficacy of the combination of anti-CD40 and irradiation decreases secondary to the immunosuppressive effects of irradiation on the CD8⁺ T-cell response.

These data suggest that TBI, which is primarily thought to be immunosuppressant, when used at a modest level in combination with anti-CD40 mAb can eradicate tumor but still allow stimulation of host immunity. TBI is known to lead to chronic lymphocytopenia, and the immunosuppressive effects are used as part of the conditioning regimens prior to bone marrow transplantation.²⁷  We have observed similar lymphocytopenia in our model system, with a dose-dependent reduction of lymphocyte numbers. CD8⁺ T cells appear to be among the most sensitive lymphocyte populations, and after 5 Gy TBI the CD8⁺ population is severely reduced (up to 75% compared with controls) for more than 50 days after irradiation (data not shown). Despite this depletion, it is still possible to generate a potent CD8⁺ T-cell response following the same dose of irradiation in the presence of anti-CD40 mAb and tumor.

In this study we have shown that radiation-induced tumor apoptosis alone is inadequate in generating long-term tumor clearance in animals and that an anti-CD40–driven CD8⁺ T-cell activation is required to achieve long-term survival. Treatment with TBI alone, even at a higher single dose of 8 Gy, was unable to enhance survival by more than a few days, and even when supralethal doses of irradiation are followed by bone marrow transplantation all animals developed tumor after immune reconstitution (data not shown). In addition, the lack of T-cell expansion we observed in vivo after higher doses of TBI plus anti-CD40 appears to account for the decrease in therapeutic efficacy.

Anti-CD40 has been shown to protect mice against tumor in a range of lymphoma models.^10,11  However, tumor protection was found to be dependent not only on large tumor burdens but also on large doses of anti-CD40 mAb, indicating the necessity of sufficient tumor antigen for the successful eradication of tumor. Delivering large doses of anti-CD40 mAb to patients may potentially induce autoimmune disorders.^28,29  We have shown that it is possible to use much smaller doses of anti-CD40 mAb with irradiation to produce long-term tumor protection. Lower doses of mAb may be sufficient to activate CD8⁺ cells against tumor antigen liberated by irradiation-induced apoptosis but may be insufficient to induce autoimmunity triggered, for example, by breaking tolerance to self-antigen.

We have made a number of important observations that all indicate that the tumor eradication seen here is not due to the direct additive cytotoxic effects of irradiation and anti-CD40 on tumor cells but is instead due to the generation of an active immune response. This contrasts to previously published work from our laboratory and others that has demonstrated a significant increase in the tumor cell kill with mAb such as anti-idiotype (anti-Id)⁷  and anti-CD20³⁰  when combined with irradiation. This additive effect is believed to be related to the combined cytotoxic effect of signaling mAb and irradiation. Although others have reported that anti-CD40 signaling can induce apoptosis of both normal³¹  and malignant B cells,³²  we did not see such an effect in our lymphoma models even when using secondary cross-linking mAb (data not shown). While there is a clear radiation dose–dependent increase in apoptosis of the πBCL₁ lymphoma line in vitro that is enhanced in the presence of mAb to the B-cell receptor such as anti-μ and anti-Id,⁷  there is no additive effect between radiation and anti-CD40 mAb. Neither does culturing lymphoma in the presence of anti-CD40 mAb affect clonogenic survival. This suggests that ligation of CD40 is not sensitizing the tumor to radiation-induced cell death (or vice versa). Similar results were obtained using a second in vitro cell line BCL₁-3B3 (data not shown).

In addition, as the number of tumor cells inoculated was increased, a threshold response was seen and, above 10⁴ tumor cells, long-term disease-free survival was achieved in most animals after irradiation with anti-CD40 mAb. This argues against irradiation simply acting as a means of tumor debulking, allowing the stimulation of APCs after anti-CD40 mAb to clear residual disease. We believe instead that irradiation-induced apoptosis liberates a critical tumor antigen load that works in concert with anti-CD40–activated APCs to provide long-term tumor protection.

Stimulation of APCs via CD40 is critical for their activation and also to the effector function of CD8⁺ T cells, bypassing cognate T-cell help.^13,33,34  Irradiation is well known to induce large amounts of apoptosis in lymphoma cells,³⁵  and this type of cell death is conducive to antigen release. Moreover, evidence is emerging that suggests exogenous antigen can be taken up from dying cells by APCs and presented to CD8⁺ T cells through the process of “cross-priming.”³⁶  Cross-priming appears to be most efficient when antigen is acquired from apoptotic cells by professional APCs, such as dendritic cells.³⁷  Thus, the balance of irradiation-induced tumor antigen release versus immunosuppression may determine the degree of tumor protection. With lower doses of irradiation, the level of tumor cell death or antigen release may be insufficient to supply enough antigen to prime the immune response, despite activation of APCs by anti-CD40 mAb. Above a certain threshold, increasing doses of irradiation may generate larger amounts of apoptosis but impair the host immune response. This effect at higher doses of irradiation could either be secondary to a direct impairment of T-cell survival and function or by switching DCs from being immunogenic to tolerogenic.³⁸  At the optimal radiation dose it is clear that a protective T-cell response is being generated. Splenocytes harvested from combination therapy–treated animals show tumor-specific activity in vitro that can be blocked by mAb to MHC I, indicating that activity is CD8⁺ T-cell mediated. Splenocytes from control animals or those given either irradiation or anti-CD40 mAb alone showed little or no activity. The fact that a substantial expansion of tumor-reactive CD8⁺ T cells is observed only in lymphoma-bearing animals given combination therapy further indicates that the tumor itself is important in T-cell proliferation. It was further confirmed that these T cells are responsible for protection by in vivo depletion of CD8 T cells with mAb. Here, the response was completely abrogated, consistent with observations in the immunotherapy model.¹⁰  It is possible that other cytotoxic cellular populations such as natural killer (NK) cells may also be depleted by such an approach. However, it has previously been shown that NK cells do not play an essential part in anti-CD40 immunotherapy,¹¹  and while it seems unlikely that they have a role here, this could be addressed by performing combination therapy in NK-depleted mice, for example, using anti-ASGM1 antiserum.

Further evidence supporting CD8⁺ T cells to be the main immune effector mechanism is provided by adoptive transfer experiments. Here, T cells taken just prior to the peak of the primary therapeutic response are able to delay tumor growth and enhance survival of naive recipients. While no long-term protection was observed, this is most likely due to the fact that the T cells are taken directly ex vivo and are not restimulated in vitro. Successful eradication of tumor has been shown to be highly dependent on the dose and timing of transfer, and it is possible that coadministration of interleukin-2 (IL-2) or other adjuvant may improve efficacy by altering the kinetics of the response.^39,40 

In conclusion, we believe that long-term tumor protection is dependent on the balance between radiation-induced tumor apoptosis and anti-CD40 activation of APCs to cross-present tumor antigen to CD8⁺ T cells. Our data suggest that combining anti-CD40 mAb and radiation therapy may be a promising approach in the clinic for treatment of B-cell lymphomas. Further work is required to increase our understanding of the other stimuli that may act in concert with the APC stimulatory pathway in the host immune response to tumor rejection and the importance of recognition and trafficking of tumor antigen in this response.

We thank all the members of the Tenovus Research Laboratory who provided expert technical support. In particular, we thank Sam Martin, Maureen Power, and Dr Ruth French for invaluable technical assistance and thank colleagues in the Cancer Sciences division for helpful comments about the manuscript.