To save time and space in my posts, as well as to help you understand what you read or hear. Since I am not a statistician by training, I apologize in advance if my statistical explanations are slightly off:

Types of trials:

Phase I – mostly focused on safety and toxicity, for brand new drugs. Can this be given to patients without killing them? What should the dose be? Then, secondarily, can we see any sign that it might be effective? May be preceded by Phase 0 or “first in humans”, using healthy volunteers with similar questions.

Phase II – Once you know the dose, now you have to see if it works in a specific disease. Or in combination with another drug(s) in that disease. So a Phase II trial gets a preliminary signal on efficacy or effectiveness. Sometimes these are randomized, comparing to another “standard” therapy. This starts to give a hint (although not definitive) if the treatment is better than other treatments.

Phase III – This is a large scale comparison of a new treatment (usually encouraged by Phase II results) to the older, established treatment. Phase III trials may also be referred to as “registration” trials, because they are designed to provide data allowing the FDA to “register” the drug for non-research use.

Phase IV – post-approval surveillance for unexpected or potential safety concerns.

Outcome measurements:

OS = overall survival, measured from study-specified start date (treatment initiation date, randomization date, or diagnosis date, etc) to date of death.

PFS = progression free survival, or time from study specified start date until evidence of disease progression (see RECIST).

DMFS = Distant metastasis free survival. Similar to PFS except excludes local recurrence.

CR = complete response, typically meaning all cancer has been eliminated.

PR = partial response, where a significant portion of the cancer has gone away or shrunk. Under RECIST criteria, this equates to a 30% reduction in the sum of the diameters of measurable target lesions, typically by CT or MRI scans without any new lesions appearing.

SD = stable disease, under RECIST criteria this means less than 20% disease progression but more than 30% disease reduction (ie, in between a PR and PD).

PD = progressive disease, by RECIST criteria, more than 20% increase in the sum of the diameters of measurable target lesions, or the appearance of any new lesions.

Endpoint = what is the study designed to measure and statistically compare? These can be either primary or secondary endpoints. The designation of endpoints is crucial to any study, and failure to meet the criteria for the primary endpoint of the study is generally the death-knell of that study.

p-value = the statistical likelihood that a comparison between two treatments is due to random chance alone or due to the difference in the treatments given. The number ranges from 0-1. A p-value of 1.0 means that the two treatments are statistically equivalent. The threshold for a statistically significant outcome is <0.05, technically meaning that there is a less than 5% chance that the measured difference in results is due to chance alone.

HR = hazard ratio. Another statistical measure to compare two treatments or interventions or study arms. A hazard ratio of 1.0 means that the two arms give equivalent outcomes. A hazard ratio above 1.0 means that the experimental study arm is more likely to cause the measured outcome than the control arm. A hazard ratio less than 1.0 means that the experimental study arm is less likely to cause the measured outcome than the control arm. Now for the hardest and most important part: the magnitude of the difference between the calculated HR and 1.0 tells you the percent change in the likelihood of the measured outcome between the two groups. Deep Breath. For example, an HR of 1.24 means that the experimental intervention results in an 24% increase in the occurrence of the measured outcome relative to the control arm. An HR = 0.68 means that the experimental intervention caused a 32% decrease in the measured outcome. As in, “treatment with DrugX compared to placebo resulted in an improvement in OS, with a ┬áHR = 0.68″, meaning that patients treated with DrugX were 32% less likely to die than those receiving placebo.

Confidence Intervals: or C.I., gives a range of results for HR or other statistically calculated results, meaning that there is a 95% (typical example) chance that the HR falls between the upper and lower limits. This is typically reported as “HR = 0.68, 95% C.I. 0.64-0.78″

Study assessments:

RECIST criteria = widely accepted and utilized published criteria to enable consistent quantitative evaluation of cancer patients undergoing therapy. Has the disease progressed? Has it shrunk? How much and how significant? The core component of RECIST tumor assessment is the sum of the maximal transverse diameters of all measurable cancer deposits or lesions. This is typically done on CT scans, although MRI scans (or calipers for skin nodules) are also used. PET scans alone are not usable, nor are bone scans.

Please comment if there are other acronyms that I am forgetting or seem mysterious, and I will check back and update.

Shopping Cart (0 Items)
Your cart is empty!

Subtotal: $0.00
Total: $0.00