Re: CAL-101 for untreated symptomatic low-risk CLL

[Guest guest]

re: Actually, it's the 'symptomatic' low-risk untreated CLL patients that were

the focus of my comments. Symptomatic previously untreated CLL patients have

been the subjects of clinical studies of other CLL

drugs in which benefits have been assessed.

I see, Al. Sorry about that. Would you expect randomization to FCR / FC / a

kinase inhibitor would be an enrollment challenge in the symptomatic population?

Would it be challenging to compare results in these very different approaches?

What would be the primary endpoint for comparison?

Karl

[Guest guest]

At 07:47 AM 3/10/2011, karlamonyc wrote:

>a kinase inhibitor would be an enrollment challenge in the

>symptomatic population?

Karl, why is enrollment of symptomatic untreated patients expected to

be difficult?

It would seem if a patient is " symptomatic " , which presumably means

they have CLL-related pathologies (e.g. anemia, splenomegaly,

lymphadenopathy, autoimmune anemia, thrombocytopenia, etc.) that are

problematic enough to end the W & W phase and seek, with some urgency,

some type of treatment.

Are you thinking that such urgency might be more likely to attract

patients to proven (approved) therapies for their initial treatment?

Or are you thinking that a tyrosine kinase inhibitor would not have

sufficient efficacy for a prompt cessation of those problematic pathologies?

At 07:47 AM 3/10/2011, karlamonyc wrote:

>Would it be challenging to compare results in these very different

>approaches? What would be the primary endpoint for comparison?

The eligibility criteria for the ongoing " Study of CAL-101 and

Rituximab in Elderly Patients With Untreated CLL or SLL " were/are the

same types of symptomatic pathologies (listed above).

SEE: http://tinyurl.com/47fvxfc

As such, I'd expect the same (or similar) efficacy endpoints as in

that ongoing study of CAL-101+R in untreated CLL patients would be

acceptable for CAL-101 monotherapy of untreated CLL patients.

i.e.

Primary Outcome Measures: overall response rate

Secondary Outcome Measures: duration of response and

progression-free survival

At 07:47 AM 3/10/2011, karlamonyc wrote:

>Would you expect randomization to FCR / FC /

The above ongoing CAL-101+R study in symptomatic untreated patients

apparently does not involve a head-to-head comparison with an

approved therapy (e.g. FCR, FC), and, thus, one would not expect that

CAL-101 monotherapy in that same subgroup of CLL patients would

require that comparison.

However, I did comment that there may be marketing advantages to

having head-to-head comparison clinical results with FCR (the current

" gold-standard " ). Since the expected advantages of CAL-101 are

associated with much less toxicity (including less risk of

therapy-induced secondary leukemias, but that may be hard to show

statistically), clinical design should be such that those lower-tox

advantages (direct & indirect) are observable. How that is done

would be up to the clinical experts. An FCR comparison would not

seem necessary, but maybe nice, if feasible.

Al Janski

[Guest guest]

Hi Al,

Impossible to be certain if the design you propose presents

exceptional enrollment challenges without polling

oncologists and patients first, but I think you provided two

reasonable guesses.

My understanding being that the investigator/oncologist is

the key - if he or she buys in or not in a circumstance

where treatment is needed. ( " Buys in " meaning has genuine

uncertainty about which study is superior - and the study

doesn't pose an unnecessary risk to the patient)

.... However, given the favorable toxicity profile of the

kinase inhibitors the risk to the oncologist of being

second-guessed (blamed) for recommending the study seems

very low. If symptoms do not resolve, or progression

continues for patients receiving the kinase inhibitor you

can go off study - hopefully in a timely manner.

However, I expect that many eligible persons with a strong

preference for the low toxic continuous therapy will decline

participation, not wanting to risk randomization to a more

toxic protocol which they could receive in regular clinical

practice - unless there is no other way to try the kinase

inhibitor (KI) .. such as in a single arm phase II study.

So I expect that randomized studies comparing very different

protocols in respect to toxicity will have more significant

enrollment challenges compared to apples to apples

comparison (F-R vs B-R for example).

Regarding endpoints, I think that overall response is rarely

a persuasive estimate of clinical benefit. CR rates can be

persuasive in some settings, particularly if long-lasting,

but I expect it would be challenging for a kinase inhibitor

to compete with FCR or FR in the untreated population on

that measure.

.... So I expect the design would require Progression Free

Survival (PFS) as the primary measure of efficacy, which

like for any study could take many years to show a

difference ... after you achieve enrollment. Also, for PFS

the issue of how you define a response might go away as even

stable disease would not be counted as a progression event.

But, let's say that PFS is better for FC by 12 months, how

do you account for the difference in toxicity? For FC, you

may have a significant hit on immunity, and perhaps no impact

for those who had the KI? ...

Finally, how does exposure to either protocol influence the

response to next therapy? The appeal of the KIs being the

expected impact is nil ... but can you know this for sure

without studying it? So a conditional approval might be

appropriate requiring a confirmatory study to make sure that

exposure to the kinase inhibitors (with continuous dosing

over many years) doesn't have a dark side.

So for others unfamiliar with the long path to marketing

approval, you can see why it takes so long to evaluate any

new drug: The discovery and development phase (yrs), animal

studies (yrs), phase I dose finding studies (yrs), phase II

to refine the dose and see if it deserves full study (yrs),

time to figure out the best study design to prove efficacy

(mos.), enrollment time (yrs), time needed to measure

duration of responses (yrs), FDA review (6-9 mo), the

manufacturing piece ... adding up often to 10 years.

Karl

> Karl, why is enrollment of symptomatic untreated patients

expected to be difficult?

(snipped by moderator)

> Al Janski

[Guest guest]

Karl, thanks for your insights.

I suspect you may misunderstand what I've suggested for a Phase 2

study of CAL-101 monotherapy, which may have resulted in your

perception of complications (which may not exist) for such a Phase 2 study.

However, because those complications (from randomization of some

patients to other more toxic therapy, e.g. FCR) may exist for the

Phase 3 study of CAL-101, it is less likely that Calistoga could be

interested in affording a Phase 2 monotherapy study, regardless of

whether monotherapy may provide longer PFS for low-risk patients.

See detailed comments below.

At 09:48 AM 3/11/2011, karlamonyc wrote:

>Impossible to be certain if the design you propose presents

>exceptional enrollment challenges without polling oncologists and

>patients first....

The CAL-101 monotherapy study I am suggesting (as a straw-person) is

exactly the same study that is ongoing now for the assessment of

CAL-101+Rituxan, a Phase 2 study.

SEE: http://tinyurl.com/47fvxfc

The only difference is the exclusion of Rituxan therapy.

Generally, since Rituxan is the only difference between the trial I'm

proposing and that ongoing Phase 2 trial that was deemed worthy of

being conducted, how is it that the inclusion of Rituxan apparently

solved each of the issues you raise as possible barriers to

initiating the same (hypothetical) trial without the Rituxan treatment?

More specifically, most of the issues you raise relate to

randomization of study patients between being treated with CAL-101

monotherapy and a more toxic treatment (e.g. FCR, FR,

etc.). However, the CAL-101+Rituxan Phase 2 study does not include

randomization of some patients to more toxic treatments. All the

patients in that ongoing study are being treated with

CAL-101+Rituxan. The protocol indicates " Study Design: Allocation:

Non-Randomized " . As such, why would a CAL-101 monotherapy Phase 2

study require randomization of patients between CAL-101 monotherapy

and a more toxic treatment?

If the FDA will require CAL-101 (and other new CLL drugs) to be

compared head-to-head with approved (more toxic) CLL therapy (e.g.

FCR), I expect those comparisons may not be expected until Phase 3

studies. As such, the ongoing Phase 2 study of CAL-101+Rituxan, as

well as an identical (hypothetical) Phase 2 study of CAL-101

monotherapy, would not be expected to have some patients randomized

to approved (more toxic) CLL therapies.

It is at that Phase 3 stage that the randomization issues you raise

would then need to be taken into account. Because of costs, such a

complicated Phase 3 study probably will not be done with more than

one CAL-101 therapy regime.

As such, Calistoga may not be able to afford the lower tox (vs.

CAL-101+Rituxan) Phase 2 CAL-101 monotherapy trial of untreated

patients, since it is likely only the " low-risk " patients who will

benefit more (in PFS) from that lower-tox monotherapy.

In retrospect, CAL-101+Rituxan may be a pretty good bet as the

selected therapy for a single Phase 3 study of untreated symptomatic

patients, in that the balance of efficacy/toxicity of CAL-101+Rituxan

(vs. an approved CLL drug) may provide the most 'comparative' benefit

for the broader spectrum of untreated patients.

If CAL-101 achieves approval as such a combination, low-risk

untreated symptomatic patients may have to wait until after CAL-101

is approved as such a combination and only be able to receive CAL-101

monotherapy as an off-label use post-approval, without clinical proof

of longer PFS from monotherapy.

Al Janski

=====

At 09:48 AM 3/11/2011, karlamonyc wrote:

>Hi Al,

>Impossible to be certain if the design you propose presents

>exceptional enrollment challenges without polling

>oncologists and patients first, but I think you provided two

>reasonable guesses.

>

>My understanding being that the investigator/oncologist is

>the key - if he or she buys in or not in a circumstance

>where treatment is needed. ( " Buys in " meaning has genuine

>uncertainty about which study is superior - and the study

>doesn't pose an unnecessary risk to the patient)

>

>... However, given the favorable toxicity profile of the

>kinase inhibitors the risk to the oncologist of being

>second-guessed (blamed) for recommending the study seems

>very low. If symptoms do not resolve, or progression

>continues for patients receiving the kinase inhibitor you

>can go off study - hopefully in a timely manner.

>

>However, I expect that many eligible persons with a strong

>preference for the low toxic continuous therapy will decline

>participation, not wanting to risk randomization to a more

>toxic protocol which they could receive in regular clinical

>practice - unless there is no other way to try the kinase

>inhibitor (KI) .. such as in a single arm phase II study.

>

>So I expect that randomized studies comparing very different

>protocols in respect to toxicity will have more significant

>enrollment challenges compared to apples to apples

>comparison (F-R vs B-R for example).

>

>Regarding endpoints, I think that overall response is rarely

>a persuasive estimate of clinical benefit. CR rates can be

>persuasive in some settings, particularly if long-lasting,

>but I expect it would be challenging for a kinase inhibitor

>to compete with FCR or FR in the untreated population on

>that measure.

>

>... So I expect the design would require Progression Free

>Survival (PFS) as the primary measure of efficacy, which

>like for any study could take many years to show a

>difference ... after you achieve enrollment. Also, for PFS

>the issue of how you define a response might go away as even

>stable disease would not be counted as a progression event.

>

>But, let's say that PFS is better for FC by 12 months, how

>do you account for the difference in toxicity? For FC, you

>may have a significant hit on immunity, and perhaps no impact

>for those who had the KI? ...

>

>Finally, how does exposure to either protocol influence the

>response to next therapy? The appeal of the KIs being the

>expected impact is nil ... but can you know this for sure

>without studying it? So a conditional approval might be

>appropriate requiring a confirmatory study to make sure that

>exposure to the kinase inhibitors (with continuous dosing

>over many years) doesn't have a dark side.

>

>So for others unfamiliar with the long path to marketing

>approval, you can see why it takes so long to evaluate any

>new drug: The discovery and development phase (yrs), animal

>studies (yrs), phase I dose finding studies (yrs), phase II

>to refine the dose and see if it deserves full study (yrs),

>time to figure out the best study design to prove efficacy

>(mos.), enrollment time (yrs), time needed to measure

>duration of responses (yrs), FDA review (6-9 mo), the

>manufacturing piece ... adding up often to 10 years.

>Karl

[Guest guest]

Hi Al,

At least I'm consistently getting it wrong. My assumption was that your interest

was in a study design that could demonstrate safety and efficacy of the kinase

inhibitors sufficient for marketing approval, which generally requires a

randomized study,

.... unless the population is refractory to standard therapy, in which case even

responses rates in single arm studies could be the basis for accelerated

approval as that would address an unmet need as in the approval of ofatumumab

for refractory cll.

(See http://www.fda.gov/ohrms/dockets/ac/09/briefing/2009-4444b1-02-GSK.pdf )

or http://tinyurl.com/o4eccq

I suppose the phase II you've cited (Rituxan + CAL101 in untreated elderly

patients) was exploratory ... to see if the outcomes show signs that it might

compare well with standard options for this population.

So depending the result in that single arm study, the sponsor might feel

confident to prospectively evaluate R+CAL vs. R+Chlorambucil, or vs.

Chlorambucil monotherapy in the untreated symptomatic population as a study to

submit for marketing approval.

.... As you know, the random selection to the competing arms is needed to solve

the problem of the confounding influence of patient selection bias in single-arm

studies, making positive outcomes and adverse events difficult to interpret.

Anyhow, I agree that enrollment is much less challenging in phase II studies ...

and these can be our best opportunities to try protocols that appear favorable

to the standard of the care.

Hope this helps.

Karl

(quoted message* snipped by moderator)

* /message/14690