12

three different estimates of risk for custom MACE. The message is quite clear that there

13

is no evidence of increased risk based on these estimates. The only difference we see is

14

that the ratio results are borderline significant, that’s these results, while the risk

15

difference results are not. They are not statistically significant. This difference may be

16

due to the fact that the three trials had no custom MACE events.

5

CPK increases had a subsequent SMQ event and found only one placebo patient with

6

such an event. The reanalysis of SMQ without these events results in an odds ratio much

7

closer to the one observed for “custom MACE” and for the company’s Primary MACE

8

endpoint and that's this estimate depicted here. So the estimate is about 0.5 with an upper

9

bound of about one, and you can see the difference between when you include the CPK

10

increases.

18

plot of the custom MACE results using the short-term data. I wanted to show you this

19

because we initially thought that our focus would be primarily on the short-term data and

20

not the short-term plus long-term data. For this plot, a half is added to each cell of the

21

corresponding two by two table of studies with no events in at least one arm. This

22

addition called a continuity correction enables one to compute an odds ratio for these

23

studies and also allows one to compute an overall estimate including all eight studies.

24

For seven of the eight studies, three of which had no events, a continuity correction is

11

So now let's look more closely at the custom MACE results. Here are the

17

So we will look next to the Forest plot. I am showing you first a Forest

Scribes, LLC Toll Free 1-800-675-8846 www.scribesllc.com

1

the SMQ endpoint. She mentioned how the majority of the events were recorded as

2

increased CPK; 50 of the 58 SMQ events for Saxagliptin and 14 of the 25 for placebo.

3

We thought it was important to do an analysis without these events.

4

So using the raw adverse event data, I looked to see if patients with the

83