Second cancers in CLL

I have been thinking about the question of second malignancies in patients with CLL. This will not be the definitive article but I want to set out the hazards in attempting such an analysis.

It has long been assumed that second malignancies were common in CLL. When I was first appointed in 1974, one of my colleagues, David Beresford pointed out that he had written an article in 1951 which detailed 100 patients with CLL who had also had a second malignancy.

Here are some of the reasons that a patient with CLL might also have another cancer.

1] It might be a coincidence of old age. One in three of us is going to develop cancer, usually in old age. If we live long enough with one cancer we still have a one in three chance of getting a different tumor. People with CLL often live long enough.

2] CLL impairs immunity. If cancers are caused by a virus then impaired immunity to that virus might make a cancer more likely.

3] There might be immune surveillance against mutations that lead to cancers. This immune surveillance might be impaired in CLL along with other immune functions.

4] Both radiotherapy and chemotherapy may be oncogenic and could conceivably induce second cancers.

5] Chemotherapy is immunosuppresive and this might reduce immune surveillance.

6] There might be an unknown oncogenic or genetic factor that induces both CLL and other cancers.

7] There might be something about CLL that causes it to transform into another type of cancer.

8] CLL is hard to diagnose accurately. Perhaps patients had lymphoma all along, but a spillover into the blood was misdiagnosed as CLL.

Epidemiological studies might be expected to solve some of these puzzles but there are hazards in interpreting genetic studies:

1] CLL is usually undiagnosed until the individual has a blood test. Cancer patients have frequent blood tests but random controls do not. Therefore one would expect that cancer patients would have their CLL diagnosed while controls would not.

2] CLL patients see doctors regularly so if a cancer is present it is likely to be found, whereas controls don't see doctors regularly and therefore they may die of something else with their cancer undiagnosed.

Therefore, as I look at the published studies over the next few weeks I must be careful that I am not being hoodwinked.