Identifying Driver Genes in Liver Cancer

Sean Cleary

Genomics is a baffling new world of biologic information that is slowly revolutionizing our once more comfortable world of surgical science. The first cancer genome sequence was reported in 2006 – less than 10 years ago. At Gallie Day and Grand Rounds, surgeons sit through dazzling arrays of data hoping for a few familiar clues to remind us how to link the leading edge of biology to the everyday world of the operating room and clinic.

Sean Cleary is a hepatobiliary surgeon who is helping sort through the genomic science of hepatocellular cancer. Sean has recently published a high impact paper identifying driver genes(1) – those few among the tens of thousands in the tumor genome that confer selective growth advantage on tumor cells. Sorting drivers from passenger genes that have no prognostic or therapeutic significance requires bioinformatics analysis of a large cohort of tumor samples. The UHN tissue bank, which now contains over 280 samples of liver cancers and nearby normal liver tissue, was the source of the study specimens.

“The incidence of liver cancer is growing fast as the fallout of the ‘60s, ‘70s, and ‘80s, when HIV, blood transfusion, and drug use caused an uptick in the incidence of hepatitis B and C. As a result, liver cancer is now becoming more and more prevalent in North America. Coupled with this is the increase in immigration to North America from Asia and Africa, especially in the Greater Toronto Area. We are the second largest volume surgical center for liver cancer in North America, behind only Mount Sinai in New York, because of immigration and because of our hepatologists, our transplant program, and our liver and pancreas oncology program. Provincial guidelines recommend that liver cancer surgery be referred to Toronto General, Sunnybrook and St. Joseph’s Hospital. The guidelines were developed by Cancer Care Ontario as another important contribution from our former Chairman and hepatobiliary surgeon Bernie Langer.

SPECIMEN ANALYSIS

“With my colleagues in North Carolina, we tried to broaden the genetic analysis, analyzing the whole exome (all of the genetic material which results in the production of a protein). Others have been working with multiple genes, or small segments of the exome. We developed the technology. We looked at our 87 initial specimens, the largest study so far. Studies from China and Europe have tended to be focused on one group of cancers, such as those caused by hepatitis B, C, or cirrhosis. Our sample includes all causes of liver cancer, a wider range in which we found ‘common drivers’. Although there is no one single gene, there are a dozen or so that are prevalent. In colon cancer the genomic analysis allows categorization in two groups. We have about six categories. Most colon cancers respond to certain regimens based on their genetic mutations. For example, no cetuximab unless a K ras mutation is found. This is a broad stroke toward the development of personalized treatment-response, versus no response, predicted on the basis of a particular mutation. Some mutations tell us how aggressive the cancer is, which improves our ability to prognosticate and will some day influence our choice of chemotherapy. We are hoping to achieve this improvement which we have definitely seen in colon cancer. For example, the P53 mutation leads to a particular protocol. A more familiar example is the presence of an estrogen receptor, determining the choice of treatment in breast cancer, or the B ras mutation, determining whether Brivenib is used or not in melanoma.

OPERATIVE TREATMENT

“Surgery is now safer. The mortality of hepatobilliary cancer surgery is less than 2%, having been as high as 10% in the past. We do many laparoscopic procedures, perhaps 40% of the operations for liver cancer in the last 3-5 years. That helps to reduce the morbidity and mortality of surgery, as there is less stress, faster recovery and shorter hospitalization. Surprisingly, the minimally invasive laparoscopic approach is also less expensive, even though disposable equipment is used(2). We dispose of our instruments rather than accept the expense of cleaning and sterilizing. There is less time in the postanesthesia recovery room, less time in the hospital and less time in the operating room. This was an eye-opener for our administration. The 5 year survival after surgery for liver cancer is now approximately 60%.

PERSONAL BACKGROUND

“Steve Gallinger has been my mentor and role model. He is a role model for all surgical scientists. He helped me to find a niche for my work. I began with collaboration with Derek Chiang in North Carolina. He was originally from Toronto and had studied at the Broad Institute at Harvard. He is a scientist who is very strong in information technology. I provide the clinical spin to our frequent communications, particularly during the development of this present study. I now have many partners here in Toronto. I was born in North York General, grew up all over the world, because of the career of my father who worked as an executive for Procter & Gamble. I lived in six countries (Canada, the United States, Belgium, England, Austria, and Switzerland) before I was 18. I came to Canada to attend Queen’s University and then medical school at Western University. Steve Gallinger and Bryce Taylor convinced me to take my surgical training in Toronto. Joining the Surgical Scientist Program was the best decision I have ever made. Steve Gallinger was my mentor and thesis advisor. I subsequently did a Master’s degree in epidemiology, which has been very helpful for the population studies that are required in the genetic epidemiology work that I am doing.

“My wife Janice and I have two boys, Owen and Christian, who are in French immersion in public schools and active hockey players. I play hockey two times a week, swim competitively, and ski. In my reading, I favour non-fiction. I will next go to Melbourne to talk about our work here to the GI and hepatobiliary surgeons.”

M.M.

REFERENCES:

1. Cleary SP, Ghanekar A, Chiang DY, et al. Identification of driver genes in hepatocellular carcinoma by exome sequencing. Hepatology 2013; 58:1693-1702.

2. Bhojani FD, Fox A, Pitzul K, Wei AC, Moulton CA, Okrainec A, Cleary SP. Clinical and economic comparison of laparoscopic to open liver resection using a 2-to-1 matched pair analysis; an institutional experience. Journal of the American College of Surgeons. 2012. 214(2):184-95.

TOWARD GENOMICALLY PERSONALIZED CARE The Sean Cleary paper is good news - the beginning of the journey toward the much lauded era of personalized care. Surgeons always personalize care as they tailor operations to the particular patient. The next generation of personalization is DNA and RNA sequencing, so we can sort tumors by their genomic and transcriptomic signatures, not just their stage. Steven Gallinger Sean’s work has been highly praised. He is networking with excellent scientific collaborators. The molecular classification schema that he is developing with his collaborators will have to be confirmed, but he is currently an important player in this field. The teams are complex - no one of them can accomplish the mission alone, but the teams can, as they work together. For example, Mike Taylor’s team and ours are very committed to this level of ‘big science’, using the Cloud to bring information together in a way that could never be accomplished with any number of individual computers. There is lots of room for discovery, and this paper on hepatocellular carcinoma shows the complexity of the problem. Most cancers show a low percentage of discreet groups and very few have a single mutation of the significance of the estrogen receptor in breast cancer or the BRCA1 mutation. The important work that Sean and his colleagues are doing is helping to weed out the noise and upregulate the important signals - the drivers as they are called in the title of his paper. The drivers may begin to influence the progression 18 years before a tumor is clinically evident. After initiation, second or third drivers join in. This is well understood in the colon cancer example - there are drivers that modulate the growth and progression of small adenomatous polyps, and transform them into invasive cancers. We now know about initiators of hepatoma like alcohol, hepatitis B and hepatitis C. One day we will be treating patients on the basis of prognosis, predicted by their genomic or expression signatures. Sean and his colleagues were able to pull this together as a well-managed team. It reminds me of the picture of the 20 people sitting in the control center for the Mars Rover. They all had important roles, and all were crying with relief and satisfaction when it landed. “We did it!” Steve Gallinger