Joshua J. Meeks, MD, PhD, discusses the current understanding of genomics in bladder cancer.
Joshua J. Meeks, MD, PhD
Bladder cancer researchers are still far away from fully understanding the genomics of the disease, according to Joshua J. Meeks, MD, PhD. While there is effort being put into understanding muscle-invasive bladder cancer (MIBC), the genomics of nonmuscle-invasive disease (NMIBC) are still obscure.
“We are getting a sense of the blueprint—we know who the major actors are, but we are not able to translate that into therapy,” said Meeks. “There is also the concept of heterogeneity, and I am not sure that we are able to predict response yet. We have trials asking that question, so we will get a lot of insight there.”
In an interview during the 2018 OncLive® State of the Science Summit™ on Renal Cell Carcinoma and Bladder Cancer, Meeks, an assistant professor of urology, Northwestern University Feinberg School of Medicine, discussed the current understanding of genomics in bladder cancer.Meeks: We try to think about the framework of bladder cancer by looking at MIBC and NMIBC. There has been a lot more effort put into the MIBC setting. Gene expression profiling, whole-exome sequencing, methylomics, and regulons of more than 400 patients have been nicely detailed by The Cancer Genome Atlas (TCGA). [This] really looks at what is involved in these tumors, and it gave us so much insight into these cancers.
What is on the forefront for NMIBC, and how is it different? Certainly, there are less mutations, but it looks like it is different. There are some different pathways that are activated in these tumors that are not present in MIBC. How many of these are targetable? We also have to deal with the heterogeneity. It is almost as if they are starting from different places and going in different directions. Dealing with that heterogeneity and how to approach these tumors will be the next iteration of how we take this descriptive work and put it into place.We think about bladder cancer as a smoking-related cancer. One of the things that I was shocked by, when I first read the TCGA when it came out in 2014, was that the mutations that were in the smokers were the same exact mutations that were in the nonsmokers. It didn't make any sense to me. Our group took the TCGA data, which is publicly available, and we broke everything down to smoking and nonsmoking to look at the differences. We thought that they would be different, but we were shocked to find that the overall number of mutations was not different. It is very different in lung cancer; there is a much higher rate of mutations in smokers than nonsmokers.
The one thing that we did find was that the signatures—the footprints of the cancer—were completely different in smokers and nonsmokers. Now, does that matter? It’s hard to say. We were also shocked when we looked at the gene expression profile. When we looked at the differential gene expression, there wasn't anything different. Looking at the smokers, we are not surprised that they have more mutations, but what we do not understand at all is where the cancer comes from in nonsmokers. That is the next step—finding out where these cancers come from, why do [the patients] have it, and if there are any ways to prevent or target it.T1 patients have a very challenging cancer. I tend to see 1 or 2 patients that have a T1 tumor per week. Patients are now getting the sense that these are not very simple cancers that you give a little Bacillus Calmette-Guérin (BCG) to and it will go away. These are very aggressive cancers, and the real concern here is that many of these are potentially MIBC. If you look at all-comers, one-third of them will not be alive at 3 years, and 15% will die from bladder cancer. Is there any way, from a molecular perspective, that we can look at these tumors and say, "Is this a bad actor?"
We compared the mutation profiles and could not see anything. I will say that we are limited by the number of patients. Tumor mutational burden (TMB) matters. Those who received BCG and responded had a higher TMB than nonresponders, which goes along with immunotherapy data showing that patients with higher TMB tend to respond. At progression, TMB was lost. We think it is an escape mechanism that the cancer is using.Unfortunately, p53 is not a biomarker. We find that it is present in over 70% of tumors that are in MIBC. Believe it or not, it was the same frequency in our T1 patients. [p53] is essential, and you are going to have loss of p53 and Rb1, that is reflected in our genetic models and mice.
The real interesting concept of subtypes, which are luminal papillary, true luminal, luminal-infiltrated, basal, and neuronal. As we know more about these tumors, subtyping is not only going to be prognostic but predictive of certain therapies—both chemotherapy and immunotherapy.There are multiple groups looking at DDR as a hotspot for chemotherapy response. If [the patient] had an alteration in a DDR pathway, they were more likely to respond to chemotherapy. It also seems to be a predictor of response to immunotherapy. There are very nice data suggesting that. We are not sure what the mechanism of that is. One hypothesis is that if the patient has a DDR mutation, they will have a high total TMB. Those 2 are probably linked. The more mutations, antigens, and neoantigens a patient has, the more likely that they are going to respond to immunotherapy. Thinking about who will get what drug, we certainly think that some of that decision will be made by DDR and TMB.There are some early readout data with pembrolizumab (Keytruda) and atezolizumab (Tecentriq) that showed a very nice response rate of 30% to 40%. If you look at enrichment of DDR as well as PD-L1 expression, you can drive that even higher. There are a number of trials out there, but none of the data are available in manuscript form. At meetings, it looks very promising. We may be able to give those to our patients before cystectomy.