The longevity, reckoned from the symptomatic onset of acute leukemia as well as from the initiation of treatment, of a group of 89 cases including children and adults, treated in this clinic since March, 1954, was found to exceed that of a similar group of 78 patients treated in preceding years in the same clinic. The median survival of all cases considered as a unit was increased from three to six months, as measured from the date treatment was begun. The change was found to be statistically significant as well as clinically important. The longer survival found in the 1954-57 series correlated with the greater use of three, rather than only two, chemotherapeutic agents.
On analysis, improved survival was statistically proven only for individuals with lymphoblastic leukemia; for these, the median survival was 10 months after the symptomatic onset of leukemia and 7½ months after the start of treatment. Within this group, there was an even greater improvement for those whose leukocyte values immediately prior to the initiation of therapy were less than 10,000 per cu.mm.; half of these patients lived more than 12 months after treatment was begun. The reason for the exceptional longevity of patients with the lower pre-treatment leukocyte values is unknown. It appears that one can attribute at least part of the improvement to therapy.
The type of statistical analysis used presents the minimum figures for improvement in median survival. When the ultimate longevity is known for the 17 per cent of the cases surviving at the closing date of the study an additional small increment may be found.
The relative inefficacy of chemotherapy in myeloblastic leukemia should be a signal for greater screening of new agents in this type of tumor.