NCRI 2013

Comparing toxicity profiles for 72 hour infusional and conventional bleomycin in testicular cancer

Dr Jonathan Shamash - Barts Cancer Institute, London, UK

We presented the results of a phase III study looking at the way we administer bleomycin in conventional chemotherapy for testicular cancer. Usually bleomycin has been given as a weekly short infusion as part of the BEP chemotherapy protocol but the main problem with it when it has been administered this way is that there is a risk of lung damage, sometimes this has proved to be fatal. It has been suggested by quite a few people and animal studies that if you alter the way in which you give the drug you can lessen that side effect, in particular by avoiding the peak levels of the drug associated with short infusions. So what we set out to do was to give the drug over a 72 hour continuous infusion and this was randomised against giving it in the conventional way. 210 patients were randomised, 105 in each arm of the study, and unfortunately it showed there was no difference in the development of lung toxicity as judged by changes on the CT scan, which most people consider to be the gold standard. However, it certainly gave us a lot of information about the way in which lung damage develops and how possibly we should be screening people who are at risk of lung damage. So what was quite clear was that patients who were older in this study, over the age of 30, and that is the median age of patients going in to this treatment, were much more likely to get lung damage but surprisingly patients who smoked, we’ve always associated that as a major risk factor, did not seem to be at high risk.

We also looked at pulmonary function testing, these are the breathing tests that patients have before they start the treatment. Again, many have assumed that if those tests show poor lung function we shouldn’t go ahead and give bleomycin. What we found was that wasn’t the case; patients with poor lung function were no more likely to develop damage to the lungs than those with good lung function. But what was clear was that the damage that appeared started to appear at around three weeks into the treatment, as judged on the CT scan. So indirectly what we concluded is that, unless there’s a very good reason, all patients should be offered bleomycin in their initial therapy but that we should consider stopping early if lung damage started to appear on scans in the first three weeks of treatment. In addition, when we looked at questions addressing quality of life, it was clear that the development of a dry cough was certainly associated with the development of scarring of the lungs on CT. So, again, development of dry cough in particular was a warning that perhaps we should stop the treatment and we think that the interpretation of that would therefore be that those patients who develop that symptom should also have early CT scanning.

What is it about this treatment that causes lung damage?

No-one is quite sure exactly how the drug does it but it seems to act indirectly and cause oxidative metabolites to occur which cause inflammation within the lung. Some of this risk may be inherited and, in addition, clearly patients with pre-existing poorly functioning lungs, a lot of scarring, or probably lots of, in some of the cases of testicular cancer, patients with extensive lung disease do seem to be at much higher risk of developing these problems. One of the big difficulties has been that if you try to treat it once it’s fully established it can prove very difficult and treatments are not uniformly successful and unfortunately it’s quite clear that a few patients will die from complications of the drug, even though the cancer has been cured and that has led to people really backing away from using it whenever there has been any fear. What you need to remember is that if you remove the drug from the combination, the cure rate will go down and so what we’re advocating is that we, if at all possible, try to include the drug at the beginning in that particular combination but withdraw it early if there’s any early signs of damage occurring.

What percentage of people can it kill?

Overall it would be less than 2% of patients and in a good risk population, as we studied, it would be less than 1%. We had no treatment related deaths in this study.