Discussion

Would you simultaneously initiate vaccine and non-vaccine immunotherapy?

Ronald B. Natale, MD: Development of the vaccine therapy and of the non-vaccine therapy are currently processes of “one step at a time.” However, I believe that in the future there certainly will be a role for combining different approaches in combination immunotherapy trials. The first step is to develop each of the components of what will comprise those combinations downstream.

What is the role of chemotherapy with immunotherapy? Is there a concern that if you have a patient who has had a great deal of chemotherapy, they may fail to mount an immune response?

Natale: The way the current clinical trials are designed, I think it is rational to use a sequential approach. Early on in the development it is going to be important to look at the contribution of the immunomodulatory agent or the vaccine separately from chemotherapy. There may be a possibility at some time in the future to begin to combine them, although it probably will require increased understanding of the impact of chemotherapy on the immune system that we are trying to manipulate through the use of the immunomodulatory proteins or the vaccines.

What tissue is needed to test for MAGE expression on the tumor and is this test validated?

Charles A. Butts, MD: For the MAGRIT trial at least, there is a central laboratory that conducts the assessments, and I believe it is a reverse transcription polymerase chain reaction (RT-PCR) technique utilized to measure MAGE-A3, which is performed on the tumor cells themselves. In fact, there have been reports of circulating tumor cells, or bone marrow-detected tumor cells that have been MAGE-A3-positive, so it can be detected even at small sample size. But, for the study, it is the primary tumor or lymph nodes that are assessed using RT-PCR.

What is the histologic pathology of pseudoprogression and the mechanism?

Butts: I have seen a number of patients who actually have developed new lesions, some quite large, intraperitoneally or in the lung, who have gone on to have a biopsy. One patient in particular had resection of a mesenteric lesion; it turned out to be a large lymph node with many lymphocytes, and showed no evidence of cancer. Biopsies of lung lesions have shown the same — lymphocytic infiltration. Some lesions may be true progression, but we should not jump to that conclusion. If the patient is well and does not have new symptoms, it is reasonable to follow up to see if the lesions actually will reduce over time.

What is your opinion concerning prophylactic vaccination studies in patients identified to be at increased risk for developing disease (ie, patients with genetic, environmental, or habitual risk factors)?

Butts: I know there have been phase 1 trials examining patients with non-invasive disease, but I am not as familiar with those. Prophylactic vaccination is a novel concept and a long way from clinical use.

What will differentiate the patients who will respond to immune intervention from those that will not?

Butts: All of the phase 3 trials include immunological marker studies or biomarker studies as a substudy within them. So, in addition to the results of START and the MAGRIT trial, we hope to learn more of exactly that — which patients respond. Is it human leukocyte-antigen (HLA)-specific to certain vaccines? Certain HLA types that respond? Are there other immune markers that can predict response?

There is another MUC1 vaccine using a vaccinia virus for which preliminary trials suggest that patients who had high levels of activated natural killer cells in their blood may actually respond poorly to immune stimulus..

John Nemunaitis, MD: A lot of the assays used in the immunology field were developed years ago, and they are cellular-focused. We actually have found some evidence that looking at K cells is helpful in correlating with response. However, it is still a difficult surrogate parameter that we have to work with in immunology. We cannot yet identify the proper patients to enter into an immune trial, and this identification will vary from technology to technology as well.

Roy S. Herbst, MD, PhD: The problem is, where do you evaluate the markers? In the blood? Is that really what is important, or is it what is happening in the tumor micro-environment, which is more difficult to assess, that is really important?

Natale: I think that the difficulty we have had in identifying targets in squamous cell cancer gives us the opportunity to examine that important subset of patients with lung cancer. Interestingly, some of these antigens may be more strongly or more commonly expressed in patients with squamous cell cancer. I think immunological marker studies must be approached with an open mind and it should be taken into account that none of these approaches are mutually exclusive, but in the initial phases they must be developed individually.

Butts: We will learn somewhat retrospectively whether these agents may be more active in squamous vs. adenocarcinoma.

Did you use steroids as antiemetic therapy in randomized trials?

Butts: Steroids were allowed in all of these trials. In the original phase 2b trial in advanced-stage disease with Stimuvax, steroid premedication was allowed, and in the START trial steroid medication was allowed as an antiemetic. Steroids are also permitted for short-course treatment of patients who develop radiation pneumonitis; these are patients with stage III disease who have had chemotherapy and radiation. However, patients who require steroids for underlying disease or an immune problem were ineligible for the trials.

Have you see any evidence of a repeat of encephalitis or any autoimmune problem since the clinical hold on Stimuvax?

Butts: I have not heard anything. I have seen all the assays. For the START trial, there was an interim analysis at 50% of the events that was reviewed by the Data and Safety Monitoring Board, who advised to continue the trial without any changes. A second planned interim analysis is expected to occur late in 2011.

With ipilimumab response rates relatively low, what are the response rates for PD-1?

Butts: The PD-1 trials are only phase 1. In the second phase 1 trial, there was 1 partial response and 5 stable diseases out of 11 patients.

Herbst: The Lung Group at Yale have treated a large number of patients on the expansion, and have 6 or 7 really good responses in squamous cell lung cancer. The group is small, but I think this may be an agent that goes forward in some trials.

Butts: In addition, I think the responses have been durable. The stable diseases included some lesions that had significant reduction; they just did not meet the criteria for partial response. That stable disease has been durable for many months in patients with lung cancer in a phase 1 trial in whom you would not expect to have durable stable disease for very long.

Natale: I think PD-1 is promising for the future.

Instead of autologous transduced tumor, why not make a mix of the shelf-transduced lung tumor lines, which would be easier to administer, quality-control, and prepare?

Nemunaitus: From a commercial standpoint, that would be the easier, more convenient way. The reason for the autologous tissue is that when you take the tissue out, you know that the antigens in that tissue are representative of the rest of the disease in that particular patient, which may not be the case with allogeneic cell lines.

If we make a separate vaccine for every patient, would a separate incubator be needed for every patient sample in a facility?

Nemunaitus: No. When the tissue comes in, it can be lysed and put into a single-cell suspension in an approximate 2-hour timeframe. Then it is placed in a flask where the gene is electroporated into the tumor tissue and incubated overnight. The next morning, samples are analyzed for the gene expression and any contaminants. Then, the samples are aliquoted and frozen. So, the process takes about 2 days, because time is needed for the gene transfer to work.

Is it ever going to be feasible to make a vaccine in this way?

Nemunaitus: We are working on constructing a “tissue disassociater,” where you core the blocks of large tumor tissue into equal-sized components, and from there the process I just described, which is currently performed by hand, is automated.

How much tissue is needed?

Nemunaitus: The amount of tissue has to be a little bit larger than a large marble, or just a little bit smaller than a golf ball.

Are there differences in sites of metastasis in terms of response?

Butts: Many of the trials looked at survival rather than response. There have been other immune agents that have shown response at multiple sites. In some of the trials, a good response in some areas was observed, while progression was detected in other areas. However, it is hard to say with certainty whether 1 site predominates in terms of response.

How are the TGFβ antisense transcripts able to block TGFβ in tumor cells remote from the injection site?

Nemunaitus: The vaccine is local, and the gene transfers into that local vaccination site. Once the immune system recognizes the tumor antigens that are placed locally and it is able to access the antigens without interference from the tumor inhibitors being produced by those vaccine cells, the T-cell system becomes activated. Those activated T cells are then able to penetrate the immunologic barriers of metastatic disease sites, which is demonstrated by some of the responses and some of the survival curves that we have been seeing in the patients.

Given that GVAX seems to have failed in prostate cancer, what is your rationale to believe that this approach may work in lung cancer?

Nemunaitus: There is not a clear answer on why the phase 3 trial was negative. The data that led to the phase 3 trial was a prostate-stabilization antigen (PSA) improvement in stabilization during a prolonged period of time. Although these data are encouraging, there was not a statistically significant survival advantage in the phase 2 data. There were not any partial or complete responses in these patients. Furthermore, allogeneic cell lines were used, so there is a possibility that the antigens in these cell lines were not as well-represented in the patient population.

What is more important: overall survival, hazard ratio, or median survival? How does durability of response fit in?

Natale: Overall survival is the primary endpoint that should be achieved. The question is how that should be measured. Median survival has been used, but that is just 1 point in the overall survival curve. The advantage of hazard ratio is that it measures the differences in survival of the various arms of the study. In terms of durability, although response rates remain an important parameter to examine in clinical trials, the more important parameters are the survival outcomes.

When you give ipilimumab, do the patients have excessive response to infections? If not, why not?

Butts: Infections really have not seemed to be a significant problem, possibly because ipilimumab specifically targets the T cells, so other innate immune cells, macrophages and leukocytes, are still functioning normally.

Nemunaitus: We have treated many patients and have not experienced any problems with infections. However, it is still important to educate the patient to be aware of this possibility, as it can occur at any time within a 12-week period. Patients will respond to steroids, but if the problem is not handled it could be fatal.

Are there any trials examining the combination of vaccines that mount immune response and VEGF or EGFR treatments?

Butts: Yes. There is a trial combining the Stimuvax with bevacizumab, a VEGF inhibitor.