Need To Monitor Multiple Outcomes

It is commonly accepted that adaptive monitoring (interim analyses) of clinical trials can lower the risk that future patients will receive a therapy already shown to be ineffective/toxic in earlier patients. Adaptive monitoring of most LI trials is limited to interim analyses of response rate, with response defined so as to include MR, which typically is observed much more commonly than is CR.28-36 The presence of MR indicates that the LI has activity and thus might be worthy of further investigation. However, as emphasized above,the relationship between MR and survival or QOL generally is unknown. Thus, making response the sole focus of interim analyses overlooks the reality that patients are likely to be concerned with response only to the extent that it is known to lead to longer survival and/or a better QOL. This is particularly true in secondary AML/MDS given the short life expectancy of patients with this condition. Indeed, because formal "stopping rules" in trials of LIs are based on response and not on survival, a scenario in which all patients on a trial "respond" but nonetheless die sooner than might be expected if standard HI could, in principle, lead to erroneous continuation of the trial. Because death rates in LI trials are typically monitored informally and on an ad hoc basis, such trials often have very undesirable "operating characteristics" (OCs). OCs include quantities such as the probability of early termination if a treatment arm is truly associated with a higher mortality rate, or the probability of selecting a given treatment if it is truly superior to others, as well as expected sample size.50

Survival time usually cannot be fully assessed until several months after entry onto such a trial, while response (MR or CR) can. Because, however, appreciable numbers of secondary AML/MDS patients die within a few months of presentation (e.g., 20-30% at 2 months if given HI), some information about survival is available early on. It follows that both response and survival should be formally monitored several months after beginning a given therapy. Accrual into a treatment arm would stop if either the death rate was too high or the response rate too low. It is possible that a high early death rate might transform into a low later death rate. Similarly, a low response rate might have no effect on survival. However, it is unlikely that patients would accept such possibilities, which should be assessed retrospectively.

Another example is of the use of investigational therapy during remission induction in younger secondary AML/MDS patients in whom SCH produces CR rates in excess of 60% (Table 7.3). The primary goal of using investigational therapy during induction is to improve RFS. Obviously then, this outcome must be monitored. However, because the investigational therapy might reduce a relatively high existing CR rate, it is also important to monitor CR rate. Trade-offs between

CR and RFS can be specified. For example, a high probability that there will be a 5% absolute decrease in CR rate might not cause early stopping, provided there is an equally high probability of a 20% improvement in RFS. These considerations underscore the desirability of including mechanisms for multiple-outcome monitoring in trials in secondary AML/MDS.5152

How To Add Ten Years To Your Life

How To Add Ten Years To Your Life

When over eighty years of age, the poet Bryant said that he had added more than ten years to his life by taking a simple exercise while dressing in the morning. Those who knew Bryant and the facts of his life never doubted the truth of this statement.

Get My Free Ebook

Post a comment