balancing success and failure
posted on
Nov 19, 2019 06:52PM
I have taken the event data from slides 22 and 32 to examine the incidence of these events in the 4 subgroups, each treated with apabetalone and placebo. These numbers are derived by dividing the number of events in each category by the number of patients in each group. This is a measure of event risk in each group.
Event Incidence in Patient Groups (Events/Patients)
|
CKD |
n=263 |
nCKD |
n=2151 |
|
L LDL |
n=1191 |
H LDL |
n=1224 |
|
|
|
|
|
|
|
|
|
|
|
apabet |
control |
apabet |
control |
|
apabet |
control |
apabet |
control |
narrow MACE |
0.118 |
0.215 |
0.101 |
0.11 |
|
0.081 |
0.131 |
0.125 |
0.117 |
nMACE (un exc) |
0.100 |
0.216 |
0.093 |
0.102 |
|
0.074 |
0.122 |
0.112 |
0.111 |
broad MACE |
0.118 |
0.222 |
0.119 |
0.125 |
|
0.091 |
0.144 |
0.146 |
0.132 |
CV death, nf MI |
0.109 |
0.203 |
0.091 |
0.103 |
|
0.072 |
0.121 |
0.112 |
0.111 |
CHD death MI |
0.109 |
0.190 |
0.089 |
0.102 |
|
0.072 |
0.119 |
0.108 |
0.107 |
nonfatal MI |
0.082 |
0.124 |
0.062 |
0.071 |
|
0.052 |
0.079 |
0.074 |
0.078 |
CV death |
0.055 |
0.105 |
0.035 |
0.037 |
|
0.020 |
0.052 |
0.053 |
0.040 |
Stroke |
0.018 |
0.039 |
0.014 |
0.010 |
|
0.008 |
0.015 |
0.019 |
0.013 |
all-cause death |
0.091 |
0.150 |
0.046 |
0.044 |
|
0.035 |
0.060 |
0.065 |
0.054 |
Hosp. for CHF |
0.027 |
0.092 |
0.024 |
0.032 |
|
0.017 |
0.040 |
0.031 |
0.040 |
Hosp. CVD Evts |
0.018 |
0.222 |
0.055 |
0.055 |
|
0.047 |
0.065 |
0.057 |
0.056 |
H. CHF or death |
0.082 |
0.163 |
0.055 |
0.066 |
|
0.035 |
0.079 |
0.078 |
0.078 |
Some take-home messages from the table by patient group:
CKD group: A very high risk group that formed 10.5% of the trial population but that suffered between 15 and 24% of most adverse events (in the combined apabetalone and placebo groups). Apabetalone provided a large benefit in this group by approximately halving the incidence rates for most events relative to the control group. In essence, being a CKD patient approximately doubled your MACE risk in this trial and apabetalone almost halved it again, bringing it back closer in line with the nonCKD patients. When BKC pointed out that the CKD group accounted for almost all of the success in BoM, I didn't realize right away that two separate factors contributed to this: the small CKD group suffered a disproportionate number of total events and apabetalone had the greatest effect in this group. An excellent combination for an excellent result!
nonCKD group: A lower risk group that formed almost 90% of the trial population but accounted for only about 75-80% of MACE incidents. Apabetalone tended to show some benefit (primarily in CV deaths & nonfatal MI) but the benefit was much less than in the CKD group. Only modestly efficacious here.
Low LDL group: A low risk group that formed half of the of the trial population. Apabetalone provided a substantial additional benefit to this group in all event categories; the apabetalone patients of this group had the lowest incidence rate for every category of MACE and hospitalization listed. Good benefit here.
High LDL group: A high risk group constituting half of the trial population in which apabetalone provided no apparent benefit for any adverse event. This is the group (half of trial patients!) that caused BoM to miss topline. Why did these relatively high risk patients not respond when the even higher risk CKD patients did respond? Do they need earlier intervention with apabetalone or a higher dose of apabetalone? Did drug compliance obscure results here? Perhaps these patients simply don't need apabetalone.
So, let’s see some more analysis before we criticize the academics and FDA for any caution they may show or BP for not lavishing untold billions upon us. There is much more to learn from this trial, especially when the focus isn’t all on the positives (as good as they may be).
Jupe