AJR 2001; 177:989-992
Stanley RJ.
Over the past few years I have become increasingly concerned about the use of diagnostic imaging for the purpose of screening adults. In my pursuit of furthur information on this topic, I have leaned heavily on various studies. William Black and Gilbert Welch, from the Dartmouth-Mary Hitchcock Medical Center [1-4]. I have also found consdiderable useful information in the published works of Alan Morrison and Philip Cole, which will be referenced later in my talk.
Let me initiate this discussion by asking you a question: How many islands are off the coast of Maine? The answer is that it depends on how closely you look and also on what constitutes an island. A large-scale map of the coast of Maine shows nothing that resembles an island. However, when we look at a regional map of the Bangor area, we discover that there are multiple islands, some of which are named and some of which might be too small to deserve a name. If one can see a giant rock only during low tide and not during high tide, does that rock qualify as an island?
Likewise, we may ask the question: How many renal cell cancers exist in an adult screening population? The answer is that it depends on how closely you look and on what constitutes a renal cell carcinoma. The prevalence of these tumors in patients who had no symptoms suggestive of renal cell carcinoma before death depends on how closely their kidneys are examined after death.
The prevalence of clinically unsuspected renal tumors at autopsy, when the threshold size for the macro inspection is 2 cm, is 1-2% [5]. However, when kidneys are examined at autopsy by serial 2- to 3-mm sectioning, the prevalence of tumors exceeds 22% [6]. The current age-specific prevalence rate of kidney cancer is 0.03-0.06% in the general population and 0.3% in men more than 70 year sold [7]. Fewer than 0.5% of all deaths in the United States in 1986 were attributed to kidney cancer (8897 cancers of the kidney and renal pelvis divided by 2,105,361 total deaths) [8]. With a cute rate of approximately 50%, only 1% of the adult population develops potentially lethal renal cancer. However, if the detection threshold is lowered to the serial sectioning level (2-3 mm) and histology remains the gold standard, 22% of the screened population could be said to have pathologically proven renal cell carcinoma.