However, the higher the preoperative CA125 level, the lower the predicted probability of achieving microscopic residual disease (P less than .01). For example, the probability was 33%, 27%, and 19% for women having a level of 500, 1,000, and 2,500 U/mL, respectively.
"It is important to remember that these curves only reflect data from women who achieved optimal cytoreduction and do not include those with suboptimal primary debulking. Therefore, the estimated predictions and probabilities are likely to decrease when applied to a preoperative population of unknown surgical outcome," he said. "But if one assumes a priori that you can achieve and will achieve optimal cytoreduction, then preoperative CA125 can estimate the likelihood of obtaining either microscopic or less than 1-cm residual."
In adjusted analyses, preoperative CA125 levels and extent of residual disease jointly predicted both progression-free survival (P less than .001) and overall survival (P = .04). For example, median overall survival ranged from 82 months in women having microscopic residual disease and a CA125 level of 35 U/mL to just 39 months in their counterparts with more residual disease and a CA125 level of 1,000 U/mL.
The change from preoperative to pretreatment CA125 levels predicted both progression-free survival (P less than .0001) and overall survival (P less than .0001). For example, median overall survival ranged from 60 months in women having a reduction in levels exceeding 80% to 45 months in their counterparts having stable or increasing levels.
In addition, among the group having a greater than 80% decline in CA125 level, survival was almost twice as long among those achieving microscopic residual disease, at 82 months, as among those with greater residual disease, at 48 months, "suggesting that residual disease rather than change in CA125 is more important to survival," Dr. Horowitz said.