Original Article

Gene Therapy (2008) 15, 911–920; doi:10.1038/gt.2008.21; published online 6 March 2008

Characterization of the adaptive and innate immune response to intravenous oncolytic reovirus (Dearing type 3) during a phase I clinical trial

C L White1, K R Twigger1, L Vidal1,2, J S De Bono2, M Coffey3, L Heinemann4, R Morgan5, A Merrick5, F Errington5, R G Vile6, A A Melcher5, H S Pandha4,7 and K J Harrington1,7

  1. 1The Institute of Cancer Research, Cancer Research UK Center for Cell and Molecular Biology, London, UK
  2. 2Drug Development Unit, The Royal Marsden Hospital NHS Foundation Trust, London and Sutton, UK
  3. 3Oncolytics Biotech Inc., Calgary, Canada
  4. 4Postgraduate Medical School, University of Surrey, Guildford, UK
  5. 5Leeds Institute of Molecular Medicine, St James‘s University Hospital, University of Leeds, Leeds, UK
  6. 6Molecular Medicine Program, Mayo Clinic, Rochester, MN, USA

Correspondence: Dr KJ Harrington, The Institute of Cancer Research, Cancer Research UK Center for Cell and Molecular Biology, Chester Beatty Laboratories, 237 Fulham Road, London SW3 6JB, UK. E-mail: kevinh@icr.ac.uk

7Joint senior authors.

Received 23 September 2007; Revised 21 December 2007; Accepted 23 December 2007; Published online 6 March 2008.



There is an emerging realization from animal models that the immune response may have both detrimental and beneficial therapeutic effects during cancer virotherapy. However, there is a dearth of clinical data on the immune response to viral agents in patients. During a recently completed phase I trial of intravenous reovirus type 3 Dearing (RT3D), heavily pretreated patients with advanced cancers received RT3D at doses escalating from 1 × 108 tissue culture infectious dose-50 (TCID50) on day 1 to 3 × 1010 TCID50 on 5 consecutive days of a 4 weekly cycle. A detailed analysis of the immune effects was conducted by collecting serial clinical samples for analysis of neutralizing anti-reoviral antibodies (NARA), peripheral blood mononuclear cells (PBMC) and cytokines. Significant increases in NARA were seen with peak endpoint titres >1/10000 in all but one patient. The median fold increase was 250, with a range of 9–6437. PBMC subset analysis showed marked heterogeneity. At baseline, CD3+CD4+ T cells were reduced in most patients, but after RT3D therapy their numbers increased in 47.6% of patients. In contrast, most patients had high baseline CD3+CD8+ T-cell levels, with 33% showing incremental increases after therapy. In some patients, there was increased cytotoxic T-cell activation post-therapy, as shown by increased CD8+perforin/granzyme+ T-cell numbers. Most patients had high numbers of circulating CD3−CD56+ NK cells before therapy and in 28.6% this increased with treatment. Regulatory (CD3+CD4+CD25+) T cells were largely unaffected by the therapy. Combined Th1 and Th2 cytokine expression increased in 38% of patients. These data confirm that even heavily pretreated patients are capable of mounting dynamic immune responses during treatment with RT3D, although these responses are not clearly related to the administered virus dose. These data will provide the basis for future studies aiming to modulate the immune response during virotherapy.


immune response, reovirus, neutralizing antibody, cytokine, T-cell subset, phase I trial



The development of oncolytic viruses for the treatment of cancer has been hindered by the paucity of immunocompetent experimental models. Thus, in many cases, it has not been possible to model comprehensively the interaction of the virus with the host immune system, because many of the viruses proposed for clinical use do not grow effectively in the rodent cell lines that can be grown as tumours in immune-competent animals. However, it is now clear that the three way interactions between the administered virus, the tumour and the host immune system are paramount in determining the therapeutic outcome of oncolytic virotherapy; moreover, it is not immediately apparent whether an intact immune system will augment, or diminish, therapy. Thus, antiviral immune responses may hinder both viral transduction and intratumoural spread, but may enhance antitumour clearance of virus-infected tumour cells. In addition, it has recently been shown that viral oncolysis stimulates effective antitumour immune responses and that therapy can, in some circumstances, be dependent upon intact T-cell responses in the treated host.1 Finally, it is also unclear how to best model the immune system of the cancer patients who will be treated with oncolytic viruses in early phase trials and who, in most cases, will have undergone multiple rounds of alternative therapies, many of which are profoundly immunosuppressive. Therefore, the opportunity to study patient immune parameters both before and after virus treatment represents a critical step in the future development of oncolytic viruses, especially in the context of systemic delivery to metastatic disease. In this respect, we have recently completed a phase I dose-escalation study of intravenously administered reovirus type 3 Dearing (RT3D) in patients with advanced cancers (Vidal et al., submitted).

Reovirus is a non-enveloped, icosahedral, double-stranded RNA virus. It is a ubiquitous non-pathogenic agent, isolated from the respiratory and gastrointestinal tracts of humans2, 3 and it has not been associated with a specific clinical syndrome.4 Reoviruses are selectively able to kill cells with an activated Ras signalling pathway in vitro and in vivo,5, 6, 7, 8 which can occur through Ras mutation or aberrant expression of upstream mitogenic signals such as overexpressed or mutated receptor tyrosine kinases. Therefore, reovirus represents a potentially useful treatment for a wide variety of solid and haematological tumours including pancreatic, colorectal, thyroid and lung cancers and acute myelogenous leukaemia.9, 10, 11, 12, 13 RT3D is currently undergoing extensive evaluation in phase I and II clinical trials, either as a single agent or in combination with radiotherapy or cytotoxic chemotherapy.

Previous studies of viruses (either as vehicles for gene delivery or as oncolytic agents) have demonstrated that the occurrence of antiviral antibody responses may have an impact on their clinical efficacy.14 In contrast to other viral and gene therapy approaches (such as virally-directed enzyme prodrug therapy), RT3D does not carry an exogenous transgene and, therefore, antibody responses will only occur against the virus itself. In addition to adaptive humoral responses, viral administration may also affect circulating populations of T lymphocytes and NK cells and may have effects on circulating cytokine profiles.15 Such alterations to components of the adaptive and innate immune systems may, in turn, affect the antiviral and antitumour immune milieu.16 A knowledge of the effects of viral administration on immune parameters will be an important component of the future development of viral vectors as cancer therapeutics.

Most healthy adults have anti-reoviral antibodies, in keeping with a high incidence of subclinical infection in early life.3 When administering viruses as therapeutic agents, it is possible that the route of vector delivery may qualitatively and quantitatively influence the nature of the antiviral immune response. For example, intracranial and intralesional delivery (sites of relative immune privilege) may result in a different response to systemic intravenous administration.17, 18 However, for oncolytic agents to find a meaningful role in the treatment of advanced cancers, it is likely that they will need to be used in a systemic setting.

In the studies reported here, we exploited the important opportunity provided by our clinical trial of systemic delivery of RT3D to closely monitor the evolution of the adaptive and innate immune responses to the virus. Significantly, the cohorts of patients in this trial represent a relatively broad cross-section of patients likely to be treated in increasing numbers in the future in similar trials. Here, we show that antibody titres are indeed raised to high levels—even in patients who were pre-immune to the virus at the start of the treatment. Moreover, we observed important changes in patient immune cell populations associated with virus administration, which have not been previously predicted. We propose that studies such as those reported here are critical to gaining a complete understanding of how the administration of oncolytic viruses will affect patient's immune systems and how this information can, in turn, be used to modify and improve future protocols for both local and systemic delivery.



Titration of RT3D cytotoxicity against L929 cells

L929 cells were shown to be extremely sensitive to RT3D-induced cytotoxicity. For both the 1:2 (Supplementary Figure 1A) and the 1:10 (Supplementary Figure 1B) dilution series, cell survival was <20% at 48h across the range of dilutions between 1:1000–1:6400, equating to multiplicities of infection of 1400 and 219, respectively. Thereafter, subsequent analyses were conducted at a multiplicity of infection of 350 to ensure that levels of cytotoxicity in the positive control (that is, cells infected with RT3D without addition of neutralizing goat polyclonal anti-reovirus antibody) were consistently >80%.

Reoviral neutralization by goat polyclonal antibody

Experiments were performed in which 2-, 3-, 4- and 10-fold dilution series of the goat polyclonal antibody were incubated with a 1:4000 dilution of RT3D stock (representing a multiplicity of infection of 350 against 2.5 × 104 L929 cells per well). As before, cytotoxicity was determined by MTT assay at 48h after addition of virus. Representative data are shown for the 1:2 (Supplementary Figure 2A) and 1:10 (Supplementary Figure 2B) antibody dilution series. In all experiments, cells infected with RT3D at a multiplicity of infection of 350 without addition of neutralizing goat polyclonal anti-reovirus antibody showed at least 80% cell kill. Dilutions of the antibody up to 1:1000 were able to neutralize (at least partially) RT3D cytotoxicity such that survival was >20%. Beyond a polyclonal antibody dilution of 1:2048, no effective neutralization of reoviral cytotoxicity was seen. As a result of these studies, 1:3 dilution series of goat polyclonal anti-reovirus antibody and patient serum samples were used in all subsequent experiments to achieve a range of dilutions between 1:9 and 1:531441. This ensured that in all cases the cytotoxic effect of RT3D was demonstrated at a level of cell survival of 20% or less.

Reoviral neutralization by clinical serum samples

Thirty-three patients (RL01–RL33) were tested for the presence of NARA. In 22 of the 31 patients (in three patients no pretreatment samples were available for assay), the pretreatment endpoint titre was >1/100 (that is, pretreatment serum needed to be diluted >100-fold to restore reovirus cytotoxicity to 80% or more). Pretreatment and peak endpoint NARA titres are summarized for the whole study population in Figure 1. Post-treatment endpoint titres were not assessable in patient RL29 because samples were not available. Peak endpoint titre increased to >1/10000 in all but one of the patients and reached >1/100000 in 23 of the 32 patients assessed. The median fold increase was 250, with a range of 9–6437.

Figure 1.
Figure 1 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact help@nature.com or the author

Pretreatment and peak endpoint neutralizing anti-reoviral antibody (NARA) titres for patients in the eight dose levels of the trial. Pretreatment endpoint titre was >1/100 (that is, pretreatment serum needed to be diluted >100-fold to restore reovirus cytotoxicity to 80% or more) in 22/31 patients. Peak endpoint titre increased to >1/10000 in 31/32 patients and to >1/100000 in 23/32 patients. Patient RL29 did not complete the first cycle of treatment, hence no post treatment titres are available. Asterisks indicate that patients RL05, RL06, RL09, RL10, RL11, RL13-RL16 and RL28 all reached endpoint titres of at least 1/531441.

Full figure and legend (109K)

Representative profiles for the evolution of the NARA response from patients in four different dose cohorts (cohorts 1, 3, 4 and 5) are presented in Figures 2a–d and Supplementary Figure 4. In addition, plots of the reciprocal of the endpoint titre over time are presented in Supplementary Figure 3. These data demonstrate a number of important features relating to the timing and magnitude of the immune response. In Figure 2a, the pretreatment serum was as effective as the goat polyclonal antibody in neutralizing RT3D cytotoxicity against L929 cells, suggesting the presence of preexisting immunity to RT3D in this patient. In contrast, Figure 2b clearly demonstrates that the patient was not pre-immune to RT3D or had very low levels of NARA, as the serum samples taken and the pretreatment at day 2 were less effective than the goat polyclonal antibody. Figures 2c and d demonstrate the opposite situation in which the patients had relatively higher (compared to patient from cohort 1 in Figure 4a) preexisting titres of NARA. Despite these differences in the pretreatment level of anti-reoviral antibody, all patients demonstrated an early right-shift in the curve such that the maximum NARA titre was reached by day 7 in 12 patients and at day 14 in 20 patients. Thereafter, the titre remained constant during subsequent treatment cycles for the majority of the patients. An apparent late ‘boost’ in the NARA titre was seen in three patients (RL09, RL18 and RL19) (Figure 2b, Supplementary Figure 4). The median final endpoint titre of NARA for the entire study population was 1:177147. There appeared to be a trend towards a higher median final endpoint titre for patients in cohorts 1–6 compared to cohorts 7 and 8: 1/177147 and 1/59049, respectively.

Figure 2.
Figure 2 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact help@nature.com or the author

Representative neutralizing anti-reoviral antibody (NARA) assay data from patients in four of the dose cohorts. (a) Patient RL01 from cohort 1. (b) Patient RL09 from cohort 3. (c) Patient RL11 from cohort 4. (d) Patient RL14 from cohort 5. An MTT assay was performed with absorbance read at 550nm and survival expressed as a percentage of a control population of untreated L929 cells. In each graph, the control curve represents data obtaining using the goat polyclonal antibody. The value of n=3 replicates for each serum sample and error bars represent the s.e.m.

Full figure and legend (260K)

Figure 4.
Figure 4 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact help@nature.com or the author

Serial peripheral blood mononuclear cell (PBMC) subset data from patients treated with RT3D. (a) CD3+CD4+ T cells. (b) CD3+CD8+ T cells. (c) CD8+granzyme B+Perforin+ cells. (d) CD3-CD56+ NK cells.

Full figure and legend (397K)

PBMC subsets

Peripheral blood mononuclear cells (PBMC) subsets were analysed pre- and post-viral therapy in 21 patients from cohorts 3–8. Absolute numbers of cells were expressed as a proportion of total PBMC. Initially, we confirmed that our methodology yielded results that were in line with other published series by measuring all subsets in four healthy volunteers who had no active medical problems and took no medications (Figure 3). Following exposure to RT3D and PBMC subset analysis showed quite marked heterogeneity (Figures 4a–l), but certain trends were observed and in some patients a correlation between specific subsets and therapeutic cycles was apparent. There was evidence of reduced CD3+CD4+ T-cells pretreatment (compared to the normal controls) in virtually all patients reflecting lymphopenia, perhaps associated with high disease burdens and previous extensive treatment. However, as a result of reoviral therapy, CD3+CD4+ T-cell numbers increased in ten (47.6%) patients (RL08, RL09, RL13, RL17, RL19, RL23, RL24, RL26 and RL27). Two patients (RL09 and RL19) apparently showed cyclical increases in CD3+CD4+ T cells over more than one cycle of treatment (Figures 4a–c). In contrast, the majority of patients had relatively high CD3+CD8+ T-cell levels (Figures 4d–f) relative to controls with seven (33.3%) patients (RL09, RL15, RL21, RL24, RL25, RL26 and RL27) showing incremental increase after therapy. In a proportion of patients, there was an indication of increased cytotoxic T-cell activation post-therapy as evidenced by an increase in CD8+perforin/granzyme+ T-cell numbers. Such changes were seen in five patients (RL09, RL13, RL17, RL19 and RL31) (Figures 4g–i). Similarly, compared with controls most patients had high numbers of CD3−CD56+ NK cells in the circulation before therapy and in six (28.6%) patients (RL09, RL13, RL22, RL25, RL26 and RL27) this increased with treatment, with apparent correlation with cycles of reovirus therapy in two patients (RL09 and RL22) (Figures 4j–l). Although regulatory (CD3+CD4+CD25+) T cells were seen to increase in response to therapy in two patients (RL09 and RL19), in all other patients the levels of CD3+CD4+CD25+ T cells were unaffected by the therapy (data not shown).

Figure 3.
Figure 3 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact help@nature.com or the author

Peripheral blood mononuclear cell (PBMC) subset data from four individual normal volunteers (V1–V4). (a) Data for CD3+CD4+, CD3+CD8+ and CD3−CD56+ cells presented as percentage of total PBMC. (b) Data for CD4+CD45RA+, CD4+CD45RO+ and CD3+CD4+CD25+ expressed as percentage of CD4+ cells. (c) Data for CD8+CD45RA+, CD8+CD45RO+ and CD8+granzyme B+Perforin+ expressed as percentage of CD8+ cells.

Full figure and legend (65K)

In view of previous clinical trial data demonstrating that a higher CD4/CD8 ratio may be a good prognostic feature,19 we analysed this ratio over cycles 1 and 2. These data are presented in Figure 5 and demonstrate that in most patients the CD4/CD8 ratio remained relatively constant. It was not possible to correlate changes in CD4/CD8 ratios with clinical response because there were no objective responses in this population.

Figure 5.
Figure 5 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact help@nature.com or the author

CD4+/CD8+ ratios for patients treated with RT3D. In spite of the trend to increased CD4+ and CD8+ levels in 47.6 and 33% of patients, respectively, the ratio remained unchanged in the majority of patients.

Full figure and legend (61K)

Cytokine analysis

Serum cytokine samples were analysed during pre- and post-viral therapy in 21 patients from cohorts 3–8. There was no consistent pattern of cytokine response to reovirus across the cohorts, indicating the absence of a dose-dependent effect. Overall, combined Th1 and Th2 cytokine expression increased in eight (38.1%) patients. A cyclical increase in IL-5 expression was seen in four patients (RL09, RL12, RL15 and RL21). A similar pattern was seen for IL-8 and IL-6 for two patients (RL22 and RL29); IL-2 for three patients (RL09, RL19 and RL22) and IL-12p40 for two patients (RL09 and RL12). Patient RL09 had the most consistent cyclical cytokine responses that are represented in Supplementary Figure 5.

Correlation of immune profile with disease response or treatment-related toxicity

A summary for each patient of the initial diagnosis, previous treatments and numbers of cycles of RT3D is provided in Table 1. It was not possible to draw any conclusions about the relationship between adaptive and innate immune reactions and tumour response because none of the patients demonstrated an objective partial or complete response. Eight patients developed grade 3 lymphopenia (five patients—RL01, RL22, RL27, RL29 and RL31) or neutropenia (three patients—RL20, RL26 and RL32). Interestingly, seven of these patients were treated in the two highest dose cohorts (1 × 1010 and 3 × 1010 for 5 consecutive days) and four of them failed to generate endpoint NARA titres of >1/100000. Only two patients (RL03 and RL05) developed grade 3 infective episodes and both generated robust NARA responses (endpoint titres >1/100000). There was no clearly defined relationship between PBMC or cytokine profiles and toxicity.



The paucity of truly meaningful animal model systems in which it is possible to measure host immune responses to either locally, or systemically, administered oncolytic viruses makes the use of patient trials critically important. Here, we have studied the immune responses to intravenously delivered RT3D in a group of patients with established cancers, most of whom had been heavily pretreated.

Most notably, we observed profound effects on the levels of circulating anti-reoviral antibodies. Despite the fact that this was a heterogeneous group of patients who had received extensive prior treatment (surgery, radiotherapy and chemotherapy) for a variety of underlying malignant diagnoses, they were all capable of developing robust NARA responses such that the median endpoint antibody titre was 1/177147 and the median increase above baseline was 250-fold. These data suggest that the notion that trial patients with extensive pretreatment will be severely immune compromised may not be as true as is commonly perceived—at least with regards to their ability to mount effective antibody responses.

These results are significant in various ways. First, it is likely that this humoral immune response will represent a very significant barrier to reovirus administered by the intravenous route, reaching and colonizing disseminated tumour deposits in patients once it is fully boosted. Therefore, we propose that the most effective systemic delivery of RT3D, and other oncolytic agents, will be achieved through rapid, repeated doses of virus at high titres within the first week of treatment, before the NARA response has been boosted. Second, these results provide a clear rationale for specific interventions to attenuate the development of the NARA response. It has been demonstrated in an animal model that combining RT3D with cyclophosphamide results in significant blunting of the NARA response and recovery of reoviral titres between 107–108 plaque forming units per mg from tumour tissue (Qiao et al., submitted). Importantly, this animal model has also allowed us to explore the importance of the NARA response to the safety profile of RT3D. Cyclophosphamide dose schedules that blunt the NARA response are effective and safe but those that completely ablate detectable NARA, are associated with viral replication in normal organs and severe toxicities. These effects can be recapitulated by repeated RT3D injections into B-cell knockout mice (Qiao et al., submitted). Therefore, the development of a NARA response should be seen as a double-edged sword. On one hand, it has the potential to prevent RT3D from exerting a potent antitumour effect and, on the other, it can protect patients from unwanted toxic effects of systemically administered virus.

A strategy of attenuating the NARA response would appear to be worthy of examination in carefully designed phase I clinical trials. The ability to characterize the effects of cyclophosphamide administration on the evolution of the NARA response will be an essential component of the safety assessment of such a regimen. By developing a rapid 48-h assay that gives a readout of the ability of the NARA response to antagonize viral cytotoxicity, we are in a position to develop clinical protocols in which viral dosing decisions can be made on the basis of almost instant feedback from clinical trial samples. Phase I studies of combinations of RT3D with conventional cytotoxic drugs that are in common use in cancer patients (carboplatin plus paclitaxel, docetaxel and gemcitabine) are already underway and characterization of the effect of these regimens on NARA responses is being performed in our laboratory.

The different levels of NARA responses in the various cohorts also require explanation. In patients in the first six cohorts, only 1 of 19 (5.3%) failed to achieve an endpoint titre of >1/100000. In contrast, 8 of 14 (57%) patients in the two highest dose cohorts (7 and 8) failed to generate this level of NARA response. It is noteworthy that only one patient in cohorts 1–6 developed grade 3 lymphopenia or neutropenia, compared to seven patients in cohorts 7 and 8. Significantly, four of these seven patients failed to generate endpoint NARA titres >1/100000. Therefore, it is possible that increased leukopenia associated with high-dose systemic virotherapy may limit the extent of the neutralizing antibody response. Another factor that may have contributed to the reduced levels of NARA endpoint titre in the two highest dose cohorts was prior exposure to lumbosacral or pelvic radiotherapy. This treatment is known to be associated with myelosuppression and may have blunted the patient's immune responsiveness. Only 4 of 19 (21%) patients in cohorts 1–6 received radiotherapy to the lumbosacral spine or pelvis, compared to 7 of 14 (50%) patients in cohorts 7 and 8. It is unlikely that exposure to cytotoxic chemotherapy played a role in the differences in endpoint titres seen between cohorts 1–6 and 7 and 8 since the mean number of chemotherapy regimens received by the patients in each of these groups were 1.58 and 1.57, respectively.

In addition to these data on changes in levels of NARA, we also observed some important changes in PBMC subsets. Despite heavy prior treatment, the baseline levels of CD3+CD8+ T cells were relatively normal (although those of the CD3+CD4+ T cells were reduced). However, encouragingly, a third of patients demonstrated increases in CD3+CD4+ and CD3+CD8+ T cells during treatment with RT3D. In some cases these changes were marked (for example, RL09) and showed a very clear cyclical pattern relative to the administration of the study agent. Similarly, the level of CD3−CD56+ NK cells was increased during treatment with RT3D in a number of patients. We believe that these results are significant because they indicate that these patients, despite prior myelotoxic treatment, can still respond immunologically in a fairly robust manner. Without further investigation of the specificity of these cell subsets, it is not possible to be definitive about the significance of the increases. However, it is tempting to speculate that they would be beneficial to the therapeutic effects elicited by the RT3D agent. This could be either through the clearance of virally infected tumour cells themselves (antiviral T-cell responses) or through the generation of genuine antitumour immune specificities (antitumour T cells). Vile and colleagues have recently reported that the efficacy of oncolytic virotherapy with vesicular stomatitis virus is critically dependent upon intact NK and CD8+ T-cell compartments.1 Therefore, the presence of increased CD8+ and NK cell populations following reovirus therapy is encouraging. We were unable to detect any significant alteration of Treg responses in the patients treated with reovirus. Once again, in animal models using vesicular stomatitis virus it has been reported that Treg depletion actually decreased vesicular stomatitis virus-mediated virotherapy, co-incident with increased antiviral immune responses and viral clearance from the tumour.1 Conversely, it has been shown that Treg depletion in the absence of oncolytic virotherapy augments antitumour immune responses. Therefore, we will develop future protocols in which Treg numbers, and activity, are altered in patients receiving RT3D, to test their role in mediating antiviral and antitumour immune responses.

Finally, our analysis of cytokine responses in these patients was difficult to interpret. In this respect, the heterogeneity of the patient population may have been a disadvantage. It seems probable that specific cytokine responses, which correlate with treatment success, or failure, will be highly dependent upon multiple factors including tumour type, HLA type, tumour location and cytokine response. Future studies focusing on specific patient groups, which can be related to immunological endpoints as well as metrics of treatment success, will be required before reproducible cytokine changes can be associated with either positive or negative biological responses. Nonetheless, the observation that over one-third of patients showed increased levels of Th1/Th2 cytokines is itself encouraging and consistent with our results from the studies of PBMC subsets showing increased immune stimulation in these patients by the virus treatment. These data also contrast with those from a study of adeno-associated virus in Canavan disease, in which selective induction of the Th2 cytokine IL-10 was seen.15

In summary, we have analysed the immune responses to intravenous administration of an oncolytic agent in a highly valuable, and relevant, population of patients. Our data indicate that future protocols should be biased towards rapid, repeated administrations of virus before the full force of the NARA response is developed. They also support the development of interventions aimed at blunting this NARA response, although our pre-clinical data also suggest that maintaining a baseline NARA level is necessary to restrict systemic spread/toxicity of the virus. We have also shown that even extensively treated patients show clear signs of T-cell-mediated immune stimulation following virus delivery. Although we have not yet shown the specificity of such responses, these results suggest that intravenous administration of RT3D is associated with immune stimulation that could significantly enhance the efficacy of oncolytic virotherapy. These data will drive the development of new trials that take into account the complex interactions between virus, tumour and patient immunity to maximize the chances of therapeutic success.


Materials and methods

Cell lines

L929 (mouse fibroblast) were cultured in Dulbecco's modified Eagle's medium (5% fetal calf serum (v/v), 1% glutamine (v/v), 0.5% penicillin/streptomycin) at 37°C and 5% CO2. Plating and reovirus infection were carried out in Dulbecco's modified Eagle's medium containing 2% (v/v) fetal calf serum, 1% (v/v) glutamine and 0.5% (v/v) penicillin/streptomycin.

Reovirus stocks

RT3D stocks at 3.45 × 1010 tissue culture infectious dose (TCID50) per millilitre were obtained from Oncolytics Biotech Inc. (Calgary, AB, Canada) and stored in the dark at neat and 1:50 concentrations either in PBS (phosphate buffered saline) or Dulbecco's modified Eagle's medium containing 2% (v/v) fetal calf serum, 1% (v/v) glutamine and 0.5% (v/v) penicillin/streptomycin at −80°C.

MTT assay

Cell viability was quantified using an MTT (3-(4,5-dimethyl-2-thiazolyl)-2,5-diphenyltetrazolium bromide) assay, 20μl of MTT (Sigma-Aldrich, Poole, UK) at 5mgml−1 in PBS was added to treated cells in a 96-well plate. After 4h incubation at 37°C, crystals were solubilized in dimethyl sulphoxide and absorbance was measured at 550nm on a SPECTRAmax 384 plate reader (Molecular Devices, Downington, CA, USA).

Trial subjects and dose escalation schema

Thirty-three patients (23 men, 10 women, median age 59 (range 32–80) years) were enrolled into a phase I dose escalation study of intravenously administered single-agent RT3D and treated over eight dose levels with a total of 76 cycles (median 2, range 1–6). The majority of the patients had treatment-refractory advanced-stage cancer. Details of their diagnoses, prior therapies and the number of cycles of RT3D received by individual patients are presented in Table 1. None of the patients received additional concomitant anticancer therapy (chemotherapy, radiotherapy or biological therapy) while they were treated with RT3D. Patients provided serial blood samples for analysis of NARA titres (33 patients), PBMC subset (21 patients) and plasma cytokine analysis (21 patients). RT3D was administered as a 1h intravenous infusion once every 4 weeks for the first cohort, over 3 consecutive days every 4 weeks for the second cohort and over 5 consecutive days every 4 weeks for the third and subsequent cohorts. The dose escalation scheme is summarized in Table 2, which also identifies the number of patients in each cohort who participated in the immune profiling studies. The clinical trial protocol was approved by the local Research Ethics Committee and all patients gave written informed consent to participation and provision of study samples.

Schedule of clinical samples

Patients provided blood samples while pretreatment and at multiple time points during the first week and then weekly thereafter. Precise timings of samples for NARA, PBMC subset and cytokine analysis varied between individual patients, according to their clinical course. The level of NARA in clinical samples was estimated in 33 patients (3 patients per cohort, except cohorts 3 and 8–4 patients and cohort 7–6 patients). For the individual patient cohorts, a variable number of samples were available for analysis, depending on the length of time the patients remained on the clinical trial protocol. Data were available for PBMC subset and cytokine analysis from 21 patients.

NARA assay

The methodology used for analysis of NARA was modified from Yang et al. (2004).20 Ten millilitre blood samples were obtained in the week before the start of treatment and weekly after each cycle of RT3D administration. Within a maximum of 4h of the sample being taken, clotted blood samples were centrifuged at 3000r.p.m. for 10min at room temperature and serum was stored at −80°C in 500μl aliquots until analysis. Samples from individual patient cohorts were batched and analysed simultaneously.

To determine a suitable virus dilution for subsequent assay, L929 cells were plated in 96-well plates at 2.5 × 104cells per well and incubated overnight at 37°C and 5% CO2. RT3D stock (3.5 × 1010ml−1) was added in two dilution series (2- and 10-fold) across the plate such that the final dilutions of the two series were 1:204800 and 1:10.12 After 2h, the RT3D inoculum was removed and replaced with growth medium. After a further 48h, cell survival was measured by MTT assay.

To establish a suitable dilution series for the estimation of neutralizing antibody levels in the clinical specimens, the above experiment was repeated with a constant titre of RT3D (known to cause 80% cell death) that was pre-incubated with a dilution series of goat polyclonal anti-reoviral antibody. Experiments were performed with 2-, 3-, 4- and 10-fold dilution series of the goat polyclonal antibody and cell survival was measured at 48h by MTT assay as before. For experiments with the clinical serum samples, a positive control of goat polyclonal antibody was always included and the patient sera were initially heat-inactivated at 56°C for 30min.

PBMC preparation and analysis

Human PBMC were prepared by centrifugation of fresh human peripheral blood over Ficoll-Hypaque (Sigma-Aldrich). Thereafter, the plasma was decanted and stored at −70°C for subsequent cytokine assay (see below). PBMC were divided into 1 × 105 aliquots and stored in freezing mixture (12% dimethyl sulphoxide, VWR International Ltd., Poole, UK) and stored at −70°C until analysis. For external staining, the cells were washed with FACS buffer (PBS plus 0.6% bovine serum albumin and 0.6% sodium azide) and then stained in combinations of monoclonal antibodies in 96-well U-bottomed microtitre plates. The following monoclonal antibodies were used for staining (plus the appropriate isotype controls): CD3 (Clone HIT3A-IgG2a); CD45RA (HI100-IgG2b) and granzyme B (Clone GB11-IgG1)—all conjugated to fluorescein isothiocyanate; CD25 (Clone M-A251-IgG1); CD45RO (Clone UCHL1-IgG2a); CD56 (Clone B159-IgG1); Perforin (Clone δG9-IgG2b)—all conjugated to phycoerythrin; CD4 (Clone RPA-T4-IgG1); CD8 (Clone HIT8a-IgG1)—both conjugated to phycoerythrin-cyanin 5. All monoclonal antibodies were supplied by BD Biosciences-Pharmingen (Oxford, UK). Cells and antibodies were incubated on ice for 30min and then washed twice in FACS buffer before being re-suspended in FACS FIX (440ml SDW+60ml 10 × PBS+15ml 38% formaldehyde solution). The cells were either stored in the dark at 4°C before analysis or analysed immediately on a Beckman-Coulter EPICS XL flow cytometer. For internal staining (that is, perforin and granzyme B), a BD Pharmingen Cytofix/Cytoperm kit was used in conjunction with CD8 external staining.

Cytokine analysis

IL-2, IL-4, IL-5, IL-6, IL-8, IL-10, IL-12, IL-13, IL-17, IFN-α, IFN-γ and TNF-α cytokines were detected by Luminex multiplex bead assay according to the manufacturer's instructions (Biosource, Nivelles, Belgium) and analysed using STarSystems software (Applied Cytometry Systems, Sheffield, UK). IFN-β was detected by ELISA kit according to the manufacturer's instructions (PBL Biomedical Laboratories, Piscataway, NJ, USA).



  1. Diaz RM, Galivi F, Kottke T, Wongthida P, Qiao J, Thompson J et al. Oncolytic immunovirotherapy for melanoma using vesicular stomatitis virus. Cancer Res 2007; 67: 2840–2848. | Article | PubMed | ChemPort |
  2. Rosen L, Evans HE, Spickard A. Reovirus infections in human volunteers. Am J Hyg 1963; 77: 29–37. | PubMed | ISI | ChemPort |
  3. Tai JH, Williams JV, Edwards KM, Wright PF, Crowe Jr JE, Dermody TS. Prevalence of reovirus-specific antibodies in young children in Nashville, Tennessee. J Infect Dis 2005; 191: 1221–1224. | Article | PubMed |
  4. Rosen L, Hovis JF, Mastrota FM, Bell JA, Huebner RJ. Observations on a newly recognized virus (Abney) of the reovirus family. Am J Hyg 1960; 71: 258–265. | PubMed | ChemPort |
  5. Hashiro G, Loh PC, Yau JT. The preferential cytotoxicity of reovirus for certain transformed cell lines. Arch Virol 1977; 54: 307–315. | Article | PubMed | ISI | ChemPort |
  6. Coffey MC, Strong JE, Forsyth PA, Lee PW. Reovirus therapy of tumours with activated Ras pathway. Science 1998; 282: 1332–1334. | Article | PubMed | ISI | ChemPort |
  7. Strong JE, Coffey MC, Tang D, Sabinin P, Lee PW. The molecular basis of viral oncolysis: usurpation of the Ras signaling pathway by reovirus. EMBO J 1998; 17: 3351–3362. | Article | PubMed | ISI | ChemPort |
  8. Hirasawa K, Nishikawa SG, Norman KL, Coffey MC, Thompson BG, Yoon CS et al. Systemic reovirus therapy of metastatic cancer in immune-competent mice. Cancer Res 2003; 63: 348–353. | PubMed | ISI | ChemPort |
  9. Grünewald K, Lyons J, Fröhlich A, Feichtinger H, Weger RA, Schwab G et al. High frequency of Ki-ras codon 12 mutations in pancreatic adenocarcinomas. Int J Cancer 1989; 43: 1037–1041. | Article | PubMed | ISI | ChemPort |
  10. Bos JL, Fearon ER, Hamilton SR, Verlaan-de Vries M, van Boom JH, van der Eb AJ et al. Prevalence of ras gene mutations in human colorectal cancers. Nature 1987; 327: 293–297. | Article | PubMed | ISI | ChemPort |
  11. Lemoine NR, Mayall ES, Wyllie FS, Farr CJ, Hughes D, Padua RA et al. Activated ras oncogenes in human thyroid cancers. Cancer Res 1988; 48: 4459–4463. | PubMed | ISI | ChemPort |
  12. Rodenhuis S, Slebos RJ, Boot AJ, Evers SG, Mooi WJ, Wagenaar SS et al. Incidence and possible clinical significance of K-ras oncogene activation in adenocarcinoma of the human lung. Cancer Res 1988; 48: 5738–5741. | PubMed | ISI | ChemPort |
  13. Needleman SW, Kraus MH, Srivastava SK, Levine PH, Aaronson SA. High frequency of N-ras activation in acute myelogenous leukemia. Blood 1986; 67: 753–757. | PubMed | ChemPort |
  14. Juillard V, Villefroy P, Godfrin D, Pavirani A, Venet A, Guillet J-G. Long-term humoral and cellular immunity induced by a single immunization with replication-defective adenovirus recombinant vector. Eur J Immunol 1995; 25: 3467–3473. | Article | PubMed | ISI | ChemPort |
  15. McPhee SW, Janson CG, Li C, Samulski RJ, Camp AS, Francis J et al. Immune responses to AAV in a phase I study for Canavan disease. J Gene Med 2006; 8: 577–588. | Article | PubMed | ISI | ChemPort |
  16. Sanchez-Perez L, Gough M, Qiao J, Thanarajasingam U, Kottke T, Ahmed A et al. Synergy of adoptive T-cell therapy and intratumoral suicide gene therapy is mediated by host NK cells. Gene Therapy 2007; 14: 998–1009. | Article | PubMed | ChemPort |
  17. Bramson JL, Hitt M, Gauldie J, Graham FL. Pre-existing immunity to adenovirus does not prevent tumor regression following intratumoral administration of a vector expressing IL-12 but inhibits virus dissemination. Gene Therapy 1997; 4: 1069–1076. | Article | PubMed | ISI | ChemPort |
  18. Fujiwara T, Tanaka N, Kanazawa S, Ohtani S, Saijo Y, Nukiwa T et al. Multicenter phase I study of repeated intratumoral delivery of adenoviral p53 in patients with advanced non-small-cell lung cancer. J Clin Oncol 2006; 24: 1689–1699. | Article | PubMed | ChemPort |
  19. Hernberg M, Muhonen T, Turunen JP, Hahka-Kemppinen M, Pyrhönen S. The CD4+/CD8+ ratio as a prognostic factor in patients with metastatic melanoma receiving chemoimmunotherapy. J Clin Oncol 1996; 14: 1690–1696. | PubMed | ChemPort |
  20. Yang WQ, Lun X, Palmer CA, Wilcox ME, Muzik H, Shi ZQ et al. Efficacy and safety evaluation of human reovirus type 3 in immunocompetent animals: racine and nonhuman primates. Clin Cancer Res 2004; 10: 8561–8576. | Article | PubMed | ISI |

Supplementary Information accompanies the paper on Gene Therapy website (http://www.nature.com/gt)