68 – an important number for the Topotarget investor

2011-11-19 20:16, Edited at: 2012-02-01 18:54

On yesterdays telephone conference it was asked what the target number of events is for the ongoing CUP trial and the answer from the CEO was 68. A model based on expressing progression free survival (PFS) as a exponential decay function yielded values of PFS for the BelCaP combination to at least 13 months provided median PFS for CaP is 6.0 months, implying an increase of >100%. This contrasts strongly to the primary endpoint set to +60% and yields the intriguing question of whether this study may serve as a basis for registration. UPDATED 31 JAN 2012 END OF BLOG

Lets get right into the model. The study was started in february 2009 and since then 89 patients has been enrolled in either the Carboplatin+Paclitaxel (CaP) or the Belinostat+Carboplatin+Paclitaxel (BelCaP) arm. In the artificial world this means that 44.5 persons have been enrolled in the BelCaP arm. It has been reported that the study was fully enrolled by the end of december 2010 which means that the recruitment rate for the BelCaP arm has been 2.02 per month assuming a constant recruitment rate (44.5 divided by the time elapsed from study start to the time when the study was reported fully enrolled).

Now, in the Q3 2011 report, the following paragraph and especially the words highlighted in italics are of key importance:

“The PFS events are taking longer time to occur (i.e. the disease is progressing slower) than originally anticipated. Since the outcome of CLN-17 is event-driven, we now expect to report data in H1 2012. At this point in time, awaiting the target number of all events to occur, data collecting and cleaning are still on-going, and until this have been finalized, the code of randomization will be unavailable. Accordingly, we will not have a read-out based on the treatment allocation until the data have been analyzed.“

During the question session following the telephone conference it was asked what this number is and the answer was 68!

Now, this means that 68 disease progressions or deaths have not been recorded yet which implies that at least 21 (89-68) patients are still alive and show no progression in the disease.

A realistic model to predict PFS by the data available so far is to assume that PFS follows an exponetial decay function on the form PFS(%) = Aexp(-t/k) where A denotes the patients recruited in period A, t the time in months from treatment start and k the parameter associated with the treatment scheme (combination). Assuming a uniform recruitment rate approximately 2.02 persons were recruited each month in both arms.

--------------------------------------

CALCULATION EXAMPLE

The value of k for the CaP combination can be calculated to 5.1 if the median PFS is 3.5 months (k=-PFS/ln0.5).

The number of patients still not suffering from progression in the CaP arm can thus be calculated as the sum...

N(CaP)=2.02exp(-11/5.1)+2.02exp(-12/5.1)+...+2.02exp(-32/5.1)=1.3≈1

...where the first term denotes the number still not suffering from progression recruited in december 2010 and the last term the number still not suffering from progression recruited when the study was started. This means that roughly 1 person in the CaP arm has still not shown progression.

Given that at least 21 patients have not shown progression for the group as a whole it means that at least 20 persons are in the BelCaP arm. Using a similar expression but for an unknown k, this means that

N(BelCaP)=2.02exp(-11/k)+2.02exp(-12/k)+...+2.02exp(-32/k)=20

and it can be calculated that this gives k≈26 for the BelCaP combination.

This implies that PFS for both CaP and BelCaP can be expressed as

PFS(CaP)=exp(-t/5.1)

PFS(BelCaP)=exp(-t/26)

which are shown in the graph below. PFS (50%) is thus 3.5 months for CaP and at least 18 months for BelCaP implying an increase of 250% for this scenario.

----------------------------------------------

For larger values of PFS(CaP), estimated value of PFS(BelCaP) will be reduced significantly. Setting PFS(CaP) to 6 months which is close to the longest PFS of 6.4 months that has been recorded in a study, estimated value of PFS(BelCaP) is 13 months, implying an increase exceeding 100%. See values of PFS(CaP) and PFS(BelCaP) below

median PFS(CaP) median PFS(BelCaP)

6 13

8 11

10 9

12 8

However the median PFS for the CaP combination is important the results yields an intriguing question of whether 100% would be enough to serve as a basis for registration given that it is also stated in the Q3 report that:

“Finally, it is our current assessment that CLN-17 is unlikely to service as a basis for registration, unless extremely outstanding results are realized, due to the limited powering of the CLN-17 study.”

Now it is certainly true that 89 patients is a limited number but since the difference between the CaP and the BelCaP arm may exceed 100% its a possible scenario.

As final words; sales from Belinostat, if registered for CUP are expected to exceed 500 MUSD annually and it has also been shown that off-label sales of Belinostat for this indication will be significant prior registration if the BelCaP combination shows good results (see http://www.redeye.se/aktiebloggen/topotarget/topotarget-%E2%80%93-belinostat-will-be-used-label-cup-us).

The market cap of Topotarget is today below 50 MUSD.

-------------------------------------------------

UPDATE 31 jan 2012

Revised in that 33 persons were enrolled by jan 2010 and also with the possibility that the target number has not yet been reached.

PFS(CaP) PFS(BelCaP, met by 1 dec 2011) PFS(BelCaP, not yet met)

5 16(+220%) 18(+260%)

6 14.5(+140%) 16.5(+175%)

7 13(+86%) 15(+114%)

8 11.5(+44%) 14(+75%)

9 10.5(+17%) 12.5(+39%)

/