Wednesday, August 28, 2013

It is Just a Little Cancer. Don't Worry.

In a recent JAMA article there is a discussion of the alleged problem of over-diagnosis of cancer. The authors are trying to make the point that we now discover malignancies quite early and that for the most part they will not kill the patient so we should not be treating them.

They start by saying:

Screening for breast cancer and prostate cancer appears to detect more cancers that are potentially clinically insignificant.

 The operative word is potentially. The problem is that in PCa we still cannot differentiate between the 90% indolent and 10% deadly types of PCa. Ductal CIS of the breast has a similar situation. Thus one errs on the side of caution and advises the patient accordingly. They ultimately make the decision, often based upon their mindset rather than medical evidence.

They continue:

Optimal screening frequency depends on the cancer’s growth rate. If a cancer is fast growing, screening is rarely effective. If a cancer is slow growing but progressive, with a long latency and a precancerous lesion (eg, colonic polyps or cervical intraepithelial neoplasia), screening is ideal and less frequent screening (eg, 10 years for colonoscopy) may be effective. In the case of an indolent tumor, detection is potentially harmful because it can result in overtreatment. These observations provide an opportunity to refocus screening on reducing disease morbidity and mortality and lower the burden of cancer screening and treatments.

Again the issue of what is indolent.  That is the operative phrase and that unfortunately is the problem. Now here is the operative position:

National Cancer Institute convened a meeting to evaluate the problem of “overdiagnosis,” which occurs when tumors are detected that, if left unattended, would not become clinically apparent or cause death. Overdiagnosis, if not recognized, generally leads to overtreatment. 

Do we have the tools at this point to determine what is capable of being left unattended and what is not. If we diagnose a melanoma, just a little melanoma, just a few clusters about the rete, some small spreading down to the dermis, should we just wait? But even to get there we had to excise. What of those few hundreds of thousands of cells in the prostate that have a Gleason 7 grade at biopsy. Do we say, "see you in a couple of years".

They conclude:  

Physicians and patients should engage in open discussion about these complex issues. The media should better understand and communicate the message so that as a community the approach to screening can be improved.

Now the first part is often difficult.  You tell a patient they have cancer and you can predict the result. Get rid of it. Worse the family finds out and you have the whole clan on your back, and then their lawyers. So you suggest, cajole, guide, and hope for the best. The Press on the other hand is often placing gasoline on the flames. Then add the glory docs who take to TV and have to tell a tale a day and you have a massive stew just boiling.

Now let us review some of their data:
First, the incidence. The argument is that incidence has increased because we are diagnosing earlier. This is true in prostate, breast and melanoma. That means that in all cases we are finding early stage cancers that would not change the ultimate end state. Namely the patient would never die from this cancer.
Now mortality as above. In breast and prostate we see a drop of mortality. That is due to better treatments. But melanoma is up, and up a lot percentage wise. But I thought we were diagnosing earlier, but more people are dying. Why? No answer given.
Finally I did a survival analysis. They did not do this. Survival of prostate and breast is up. But is that because of better treatment or because the numerator is enlarged due to the over diagnosis? In which case we may actually have worse survival despite all the improved treatments. There is a problem of logic here. The argument is far from compelling.

Finally look at melanoma. Survival is up but I thought we had over diagnosed here as well. Makes no sense.

The story is more complex that the authors seem to state it. We are still left with telling what we see and doing the best we can.