2 Clarke Drive
Suite 100
Cranbury, NJ 08512
© 2024 MJH Life Sciences™ and OncLive - Clinical Oncology News, Cancer Expert Insights. All rights reserved.
At the 2011 symposium in San Francisco, California, Oncology & Biotech News spoke with the author of one of the featured presentations at the meeting
At the 2011 symposium in San Francisco, California, Oncology & Biotech News spoke with the author of one of the featured presentations at the meeting. Robert Rosenberg, MD, of the Technical University of Munich in Germany, presented results from a retrospective evaluation of the ColoPrint test in patients with stage II colon cancer. Derived from a genome-wide search, the 18-gene microarray test is designed to distinguish highrisk from low-risk patients, leading to more informed clinical decision making about patients who will and will not benefit from adjuvant chemotherapy. The end result should be more efficient use of therapy in patients who need it and avoidance of toxicity and cost in patients who are unlikely to benefit. As Rosenberg explained in the following discussion, the results he presented suggest that the test might be superior to currently available clinical information used to identify patients who have stage II colon cancer and are most likely to benefit from adjuvant chemotherapy.
OBTN: Could you briefly discuss the underlying technology of this test?
Rosenberg: I was not directly involved in the development of the test. I was involved in the evaluation of it, so my understanding of the ColoPrint test is from that perspective. The test was developed from the whole genome, not a candidate-gene approach, where you would look for the most important genes that might be involved in development of a kind of cancer. The candidate-gene approach was the approach used to develop the OncoType test. For the ColoPrint test [researchers] used patients from the Netherlands to identify a prognostic gene set. From the whole genome, they identified 18 genes that correlated with colon cancer in the patients.
Then in the second step, they transferred these 18 genes in a test set. They combined these 18 genes with housekeeping genes, so it's a microarray platform. When the RNA is extracted from the tissue, depending on the gene expression profile, they get values which were identified as high risk or low risk.
The test has been evaluated in 2 validation studies, and additional studies are ongoing. Additionally, a multicenter prospective evaluation is being conducted, involving centers in the United States, Europe, and Asia. That study will involve about 600 patients with colon cancer.
The cutoff points for this test appear to be really robust, so they can clearly say whether the patient is high risk or low risk. Other tests have intermediate areas where it is not clear whether the test correlates with a high-risk or low-risk patient.
Is the ColoPrint test designed for a specific group or subset of patients with colon cancer? What types of patients would be good candidates for the test?
It seems best suited for patients with stage II colon cancer. Those are patients who have medium- or large-size tumors--which can be T3 or T4 tumors--with no lymph node involvement or distant metastases. These are patients who generally have a good prognosis, with a 5-year survival in the range of 85% to 90%. We know that even in this stage there are 25% to 30% of patients who do not have such a good prognosis. The ColoPrint test uses different parameters where we can identify the 25% to 30% of patients who clearly have the worst prognosis. We haven't [yet established] hard parameters to identify these patients. We have the American Society of Clinical Oncology (ASCO) recommendations, but even [the ASCO parameters] are not very [specific]. So, it seems that this test will help us identify some of these patients for whom we do not have good parameters to determine their prognosis.
From the perspective of a practicing oncologist, could you take us through the steps or the process involved in obtaining results with the ColoPrint test?
We are surgeons, and when we perform the operation we take a specimen and give it immediately to the pathologist, who comes to the operating room. We have that advantage that might not be available everywhere--the pathologist in the operating room. The pathologist determines the tumor size, which is important to be sure that we achieved the correct resection and that we are fine with respect to the resection. Then we remove a 0.5-cm piece of the specimen and immediately insert the tissue into a stabilization solution provided by the [ColoPrint] manufacturer. The tissue is then sent to the manufacturer.
At the moment I cannot say how long the wait will be to get the test result for future patients. This was a retrospective analysis that we presented at this meeting. We are currently participating in a prospective multicenter clinical trial, and we don't receive the results immediately. I would guess that the results would be ready within 1, 2, 3 days, but I don't know for sure.
How might the results of the test influence your decision making about the management of a specific patient? If the results showed the patient to be low risk, high risk, or in the middle, how would that affect your approach to treatment?
Usually in stage II patients we don't give the patients adjuvant chemotherapy because we believe they have a good prognosis. But as I said before, there are about 25% to 30% of stage II patients [who] don't have such a good prognosis. If it seems quite sure--and so far the ColoPrint seems to work quite well in this respect--that we can identify the high-risk patients, then we would offer [those] patients adjuvant chemotherapy. On the other side, if we think they are low-risk patients, we can tell the patients they don't need adjuvant chemotherapy because they have a good prognosis.
Is there any role for serial or follow-up testing with the ColoPrint? Would there be any potential advantages to performing additional tests at various intervals after surgery?
I don't think so because the point of the test is to determine whether the patient is high risk or low risk. What we don't know is whether the high-risk patients will benefit from adjuvant chemotherapy. We believe they will but that has never been proven, even for the ASCO recommendations. We guess that the high-risk patients should be treated with adjuvant chemotherapy, but no one really knows if we achieve anything from that. The test can tell you whether a patient is high risk or low risk but nothing more. Maybe there will be other tests developed to see whether the patient responds to the chemotherapy, but we don't know that [yet].
Would the ColoPrint have any value in assessing normal tissue adjacent to the tumor site to see whether the cancer has recurred or whether it has been eradicated by surgery, with or without chemotherapy? For example, if the gene expression or signature changed, would that provide useful information?
Yes, that's for sure, but that is not the purpose of this test. We performed our own tests to address that issue. We have examined tumor tissue--different tumor stages--and we have examined normal colon tissue and have compared the gene profiles of the 2. We see different expression profiles in all tumor stages, and the biggest difference is between normal tissue and cancer tissue. There are a large number of genes that play a role in cancer.
Could you briefly discuss the results that you reported at the Gastrointestinal Cancers Symposium?
This was the second clinical validation study of the ColoPrint test. We retrospectively evaluated the test, using specimens from 135 patients with stage II colon cancer treated from 1987 to 2003. The ColoPrint test identified 74% of the patients as being at low risk for recurrence, and the remaining 26% were high risk on the basis of the test results. Long-term follow-up results showed that the patients who were low risk by the ColoPrint had a 5-year distant metastasisfree survival of 94.9% compared with 80.5% for patients whose test results indicated they were at high risk. This means that 20% of patients whose tests indicated they were at high risk developed distant metastasis compared with 5% of the low-risk group.
In a univariate analysis, the ColoPrint test results were the only significant predictor of distant metastasis and were associated with a hazard ratio of greater than 4 for high risk versus low risk, and this value was highly significant. An interesting finding in this clinical series was that all available clinical parameters that usually indicate high-risk patients--such as age, tumor localization, tumor grade, number of lymph nodes assessed, and gender--were not significant predictors of the 5-year outcome. They showed some ability to predict high risk, but they were not significant. Only the ColoPrint results remained statistically significant in the multivariate analysis, and the hazard ratio once again exceeded 4.
That is a very promising finding. This may be due to the fact that we had quite a low rate of patients with stage II disease who developed distant metastases. Only 9% of patients developed distant metastases in our series. Normally you would expect a little higher rate of distant metastasis in this stage group. Another possible explanation is that we didn't have many patients with T4 disease; we had a lot of T3 tumors. We also didn't have too many patients with fewer than 12 lymph nodes detected. These are usually factors where you define the patients as high-risk. The established ASCO recommendations didn't work in our group.
The ASCO and ColoPrint determinations were discordant in 50% of the patients. The ASCO criteria were more likely to identify as high risk those patients that the ColoPrint identified as low risk, whereas the patients who were high risk by the gene test were almost equally likely to be identified as high or low risk by the ASCO criteria.
It will be interesting to find out whether the ASCO clinical factors can add to the value of the ColoPrint test. This issue is being evaluated in the ongoing prospective international clinical trial.
Has the test been used prospectively to make treatment decisions?
In the ongoing multicenter trial, that is being evaluated. We are participating in that trial and have enrolled a number of patients. It is my understanding that about half of the 600 patients have been enrolled, so we do not have results yet. However, an analysis will be performed to examine the correlation between ColoPrint and clinical data for the patients. Our expectation is that there will be no correlation between ColoPrint and the clinical variables because that should be the potential advantage of the ColoPrint, which was shown in the validation trials. The second aim will be to show a survival advantage. That data probably will not be available for another 3 or 4 years.
How does the ColoPrint test differ from currently available tests that are based on analysis of gene expression or gene profiles?
I have not worked with the OncoType test, but I am familiar with it and have heard presentations about it. The OncoType gene set is not a validated gene set; it's an established gene set. It doesn't work on fresh-frozen tissue; it works on paraffin-embedded tissue. I think most people would believe this is an advantage. In truth, it is not an advantage for us because we are used to working with fresh-frozen tissue. Usually you get better-quality RNA when you use fresh-frozen tissue. Maybe for some pathologists it is an advantage to use paraffin-embedded tissue.
Other differences are that the ColoPrint was developed from the whole genome and the OncoType was developed from candidate genes. ColoPrint has only 2 distinctions: high risk and low risk. OncoType has 3 distinctions: high risk, intermediate risk, and low risk. Intermediate risk is significant compared with the other groups. OncoType works by RT-PCR technology, whereas ColoPrint uses microarray technology, so there are normalization genes that are used to validate the results of the 18 genes.
Are there other aspects of the ColoPrint test, or colorectal cancer risk assessment in general, that you would care to mention?
As most oncology specialists know, the seventh edition of the TNM classification system was published last year. For colon cancer, several new subgroups and substages were introduced, with the aim to predict prognosis more precisely than before. At my hospital we have performed some analyses, and we believe that these subclassifications, based only on TNM, are very complicated. I believe the combination of TNM with genes may be the future to further improve the prediction of patient prognosis, which is our aim. We want to offer treatment only to patients who [will benefit from it]. We want to tell patients that they don't need chemotherapy if they have a good prognosis. I think the combination of clinical factors and genes will be very important in the future.
________________________________________________________
Charles Bankhead is a freelance medical writer in Houston, Texas. He has written extensively about oncology and regularly covers major national and international oncology conferences.
Published in Oncology & Biotech News. February 2011.