Abstract
Study of our first 201 cases of polycythemia vera and of cases reported in the literature shows a remarkable age distribution for this disease. The age at onset, age at diagnosis, age at first treatment, and age at death, as well as survival times, each fits a normal distribution.
In our cases, the median age at onset is 57 years with a standard deviation of 13 years, an entirely different distribution from that for the population of Oregon. This requires a logarithmic increase in relative age specific incidence from age 20 to 40, when there is little difference in the number of persons alive at each age, with a doubling of the proportion occurring in each 5-year interval every 7.5 years. After age 55, the proportion of new cases developing follows the slope of the number alive in the population. This means that the incidence remains almost constant after the peak age incidence is reached.
Unfortunately, no data exist to transform these figures to absolute values, but if enough of the population could be studied to give absolute values at any one point, all other points could be determined. The implication is that polycythemia vera is due to a single cause which is highly correlated with age.
The normal distribution of survival times means that polycythemia vera is not a malignant process since survival times in all malignancies studied fit a log normal distribution. The constant value of age at death means that age at first treatment is a most important prognostic factor.
The mean age at death of patients with polycythemia vera, treated with P32, is 69 years ± 1, and treated without radiation therapy is 65 years ± 1. Apparently, if a patient lives to be treated with P32 previous treatment by other modalities or no treatment prior to that, will not affect the years to be gained by P32 treatment. However, the total survival time is highly correlated with age and the sooner after onset that the patient can be treated the more likely it is he will benefit from P32 therapy.
