HEMONC TODAY's one-on-one interview with Soonmyung Paik, MD

Issue: March 25, 2008

Editor's note: At the 25th Annual Miami Breast Cancer Conference, HemOnc Today sat down with Soonmyung Paik, MD, director, Division of Pathology, National Surgical Adjuvant Breast and Bowel Project in Pittsburgh, to discuss the B-31 trial and trastuzumab use in breast cancer patients. The following is an excerpt from the interview.

Essentially, we have data from our clinical trial, B-31, that we presented at last year’s ASCO meeting that shows that trastuzumab seems to work even in the so-called HER2-negative population. Trastuzumab was developed as a targeted therapy for HER2 and in the metastatic disease setting, it did not work in HER2-negative disease patients. Therefore, the question now is, why did we find a benefit in the HER2-negative subset in our trial? When we go back and look at the design of our trial, all the patients who entered the trial were supposed to be HER2-positive, but we had HER2-negative patients and that is the main criticism of our data.

What we did was allow any HER2 testing performed in any lab. Earlier on in our trial, we found that there are serious problems in quality assurance of HER2 testing in community based clinical labs. We found about a 20% rate of false positives, so those cases that were originally diagnosed as HER2-positive in community pathology labs turned out to be HER2-negative. Therefore, you could argue, how do you know the central lab is correct and the community lab is wrong? There are better ways to test and one way is to compare the rate of agreement among central labs regarding HER2 status that are high to the rate of disagreements that are low.

We performed ribonucleic acid expression analysis on the subset of the B-31 samples using about three or four different methodologies and all of those assays showed that the message RNA level for the HER2 is very low and similar to the level of any HER2-negative tumors from studies other than B-31. The hallmark of the amplification phenotype for the HER2 is that since it is a genomic lesion in the chromosome 17, HER2 genes and other genes around it behave in a similar way. Therefore, not only should HER2 be low, but also the genes around it should be low as well and that is exactly what we found.

We mapped all the expression data of the position of those genes and the only difference between what we call HER2-negative and HER2-positive tumors is that the genes in the HER2 locals are down. That biologically proves that these are really HER2-negative tumors and shows that they are somehow called HER2-positive by some maps, which could imply some kind of selection bias although they are really part negative. Our data shows that there is no selection bias, that they are really negative patients and the data still shows the benefits. There must be some other compounding variables that are predictable of trastuzumab response.

We performed a microanalysis looking at around 44,000 genes and in those samples we found that there are many genes that predict response to trastuzumab, but HER2 is not one of them. Many of the genes we found were predictive genes and their expression levels did not correlate at all with the expression level of HER2. Therefore, that directly proves that HER2 is not the predictor.

This raises a serious question: Is the design of trastuzumab completely wrong or is there a fundamental balance for difference in the behavior of metastatic disease and adjuvant disease? One possibility is that what the adjuvant trastuzumab does is prevent the progression of the subset of breast cancer cells in HER2-negative tumors that are HER-positive. Some studies have shown that in so-called HER2-negative tumors there may be a subset of cells that may actually be HER2-amplified or HER2-positive and they actually are driving the progression of the tumor or metastasis. So it is possible that in a micrometastatic site we can detect if they are actually HER2-positive although their index tumor is HER2-negative and maybe trastuzumab in the adjuvant setting is preventing the growth of those cells, which would make sense—that the HER2 response is not determined by the original HER2 status.

We have many things to work out and we actually have a clinical trial concept that is being reviewed by the National Cancer Institute. So we are hoping we can launch a clinical trial testing the trastuzumab in HER2-negative patients with all those [other] studies to address the various questions. – Interview by Paul Burress