Reactivity of a monoclonal antibody with human ovarian carcinoma

"This was really the beginning of my huge interest in ovarian cancer, and also the introduction to my great mentor, Bob Bast , who is professor of oncology at MD Anderson Cancer Center in the US. This paper was introduced to me by my late father-in-law, who knew Professor Bast from conferences they went to together. He showed me this paper as an example of scientific research and it was fascinating. It was the first time that CA 125, which is the marker for ovarian cancer, was discovered. This was the paper that described that discovery. Currently, every woman who has an ovarian cyst, or has ovarian cancer, gets a blood test measuring the CA 125, so it’s a paper of huge significance in the field of ovarian cancer. What fascinated me about it was the clarity of thought, and the process by which Bob Bast and his colleagues on that paper found the OC125, which we now refer to as CA 125. What they were doing was raising antibodies to different proteins collected from ovarian cancer cells, and then trying, by using very sound scientific methodology, to go through those antibodies until they found one that reacted with most ovarian cells, but reacted with very little or none of the normal tissue. That was how the CA 125 was discovered. 125 is, as Bob tells me, the 125th attempt to find that antibody, so they named it OC125. What happens is that ovarian cancer cells produce a lot of a certain type of protein. It’s a glycosylated protein, so a protein that has glucose and other sugar components attached to it. It’s located on the outside of ovarian cancer cells, so it’s shed into the blood stream. What we do is we get a blood sample, and now have an antibody—which is the CA 125—which will react with that antigen if it’s there and gives a measurement, a level. So if a patient has cancer, these measurements will go very high. Normally, it’s less than 35, but it can go to levels of more than 1,000, say. As the patient gets treated with chemotherapy, the level of CA 125 goes down to normal again. We then use it for follow-up. So if it starts going up again we can detect recurrence, if the tumour recurs. For screening, there is a huge UK-based study going on . The final results from that—to assess whether it’s going to be useful as a screening tool or not, in combination with ultrasound—are awaited, and we should be getting those in the next year or so. My colleagues Ian Jacobs and Usha Menon started this study many years ago. Yes, that’s very wise. Yes, I agree. Not really. There are some new potential markers, including ones that we found in our own lab, like Sox2. But nothing has really reached the level of testing that CA 125 has had. It’s been around since 1981, so it has been exhaustively tested. Nothing else has been tested to the same rigour for us to know whether it’s going to be useful or not. No, not really. I mean, this study for evaluating CA 125, the conduct of the study and then awaiting for the results, has taken—if I’m not mistaken—10 years, or even more. For just that piece of information, 120,000 women were recruited. It’s a huge study over a large number of years. Yes, of course. Big data can take a lot of forms. It could be medical records. It could be social media. It could be DNA sequencing data. Circulating DNA, RNA, or all of the above, integrated into one big data set from a large number of patients can be very, very useful. We have seen, for example, how useful they are in epidemiological studies. In the ‘ Million Women Study ’ with medical record data and data about patient history—whether they have taken contraceptive pills or not, whether they have had tubal ligation or not—we’re already seeing the huge power of this sort of data collection. For example, the link between taking the oral contraceptive pill and the reduction of the risk of ovarian cancer by 50% or so . These studies look at a huge number of women, their social life, their medical treatments. I think we are now in a phase where we are expanding the sources of these medical and non-medical records. This is the direction. Plus, if you take into account the DNA sequencing data, then clearly there is an enormous amount of information that can be mined."