reduction of pathologies as endpoint?

[Guest guest]

Re: FDA

At 02:30 AM 1/11/2011, Terry Hamblin wrote:

>A trial of the enzyme inhibitors versus observation would not be

>ethical in this group.

At 08:36 AM 1/11/2011, " karlamonyc " wrote:

>Agree that randomizing to observation has ethical concerns - I

>suppose the control could also be investigator choice.

>I know that the previously untreated is an uncommon indication for

>testing new single agents - but .........

>However, those issues may be less for the kinase inhibitors, when

>such agents are not yet part of effective standard protocols? A

>reason for a pilot studies in that population? ... I couldn't say.

Might an observational group be ethical in the testing of these

kinase inhibitors in untreated patients who are borderline in their

need for treatment? For example, patients in the borderline group

who decide to be treated could receive a kinase inhibitor therapy and

patients who decide not to be treated would be observed. These

agents have been reported (in in oral sessions at scientific

meetings) to have low levels of toxicities (in refractory patients),

which would seem to aid the ethics of such a consideration.

Such a scenario might be designed to fit the priorities of patients

who are borderline because they have pathologies from CLL that are

causing them to think that the time may be coming soon to be

treated. Specifically, could the design be an objective of

reducing/eliminating those pathologies?, rather than a remission

(that knocks out all of the proliferation centers for CLL disease).

Comparisons of the magnitude of pathologies between the treated and

observational groups would be an endpoint. [e.g. pathologies such as

anemia, splenomegaly, lymphadenopathy, autoimmune anemia,

thrombocytopenia, etc., selected based on how a given agent is

expected to act biochemically]

A difficulty of that scenario would be patients in the observational

group who progress, after the study is started, to a point at which

they desire treatment. At that point, those patients would have to

be excluded, and as the study progresses, more and more such patients

in the observational group would need to be excluded when they decide

to be treated, which would seem to make the observational group's

results (i.e. level of pathologies) look better than they really

are. But maybe reduction in pathologies could still be demonstrated

even with such an advantage for the observational group.

Convincing the FDA that reduction of pathologies is a sufficient

endpoint would be difficult, if not near impossible. However, there

seems to a basis for argument, in that it is now generally agreed

that initiation of treatment should be based on the magnitude of such

CLL pathologies, and, thus, reducing/eliminating those pathologies

seems a valid clinical endpoint for discussion with the FDA.

The clinical observations thus far (in refractory patients) that

these kinase inhibitors are providing benefits (presumably reducing

pathologies) without response (i.e. continued lymphocytosis) seem to

fit this type of design.

If lower (less toxic) doses of kinase inhibitors [than what are

needed for refractory patients] are effective in reducing pathologies

of untreated patients, then there is less probability of

treatment-induced development of chemo-resistant side-populations of

CLL cells, which is thought to be a pathway in becoming refractory to

therapies.

Thus, in addition to living with reduced pathologies, maybe there is

some basis for longer overall survival. Survival is what the FDA

seems to be expecting for new therapies, regardless of mechanism of

action. However, demonstration of a benefit of overall survival from

a strategy of milder initial treatments would seem a complicated

endpoint for which to design, given that patients in an untreated

borderline group would probably end up taking many different pathways

for follow-up treatment. A lessor endpoint than survival may be a

better place to start for untreated borderline patients.

Al Janski