Tipping 1.GIF
Celebration Thread! Biogen is going to the FDA with Aducan.
Re: Celebration Thread! Biogen is going to the FDA with Aducan.
You do not have the required permissions to view the files attached to this post.
Re: Celebration Thread! Biogen is going to the FDA with Aducan.
This figure is quite a mystery to me; I do not recall ever seeing something similar to it.
Figure 15 is another figure in Appendix 2 that is fudgy and not easy to read. Clearer figures would be very helpful.
The discussion about the lack of correlation between CDRsb and amyloid removal has been noted in a previous post.
Appendix 1 adds that at the group level this correlation is more apparent. This was also posted to the thread earlier. Page 220 Figure 12 actually shows a strong correlation at the group level. This figure does not add in the placebo 0,0 so the actual correlation is even larger than the figure makes it appear. Looking at Figure 12 this correlation is entirely lost and all we see is a mass of points. Without having a better view of the dataset it is difficult to comment further. There might be other features in the data that have been not disclosed that could help to explain the data cloud in Figure 12. For example, which dots represent mild and prodromal? Which epsilon 4+ or 4++? etc.
p.295 tries to move down into n= ~10 with tau etc. Probably best not to go there.
Err, Own Goal! "It can rule out a drug effect greater in magnitude than 0.262 with 95% confidence. " 301 high dose.
page 299. But Figure 13 on page 290 reported a 0.50 drug effect (1.58 n=77 - 1.08 n=77) for the 14 dose subgroup in 301.
Guess we can kiss that 95% CI good bye. 302 14 high dose in the same Figure 13 reports a 0.54 n=98 CDRsb benefit.
Page 300 small sample comparisons of sub-groups.
page 301 for the 301 trial consideration is given to the different age groups, this did not seem to be done for the 302 trial.
Shouldn't the analyses done be run on both datasets?
Figure 20 is verging on another own goal. Figure 13 suggested a 0.5 CDRsb benefit for 14 high doses in 301 with an average CDRsb change of 1.04. It wasn't sub-divided by epsilon 4, though this is a substantial benefit especially given that 301 as has been mentioned repeatedly in Appendix 2 reported a topline negative result. Also notice that even the high dose CDRsb at week 54 jumped around a fair amount; perhaps this level of variation is what we should expect.
Figure 21 is no less confusing. Do all of these figures show everything jumping around? What is the point then? The 302 placebo jump did seem odd, yet now with 301 what we see is that everything jumps around. Perhaps that is the behavior that we should expect. In figure 21, 301 APOE+ placebo jumps from Pre- to Post- and the in the high dose there is another dump and this time it dis downwards. The differences are not small; with high dose there is a drop at 78 weeks of ~0.50 CDRsb which is basically large; very large. That is exactly why you don't want to open the engine up; you start to see many things that are not easy to explain.
page 304 down to n=1 n=2? Isn't the idea that dose titration would be needed for those with ARIA because ARIA relates to more amyloid clearance and as seen here potentially better cognition a reasonable explanation? It is not even clear what dose titration even means for the placebo arm. Did some patients really need to be titrated down from their placebo? Sure, some on placebo experienced ARIA, though what happened then? They received a lower dose of placebo? Did placebos with ARIA also experience actual amyloid clearance or simply cognitive decline?
The issue about unblinding and placebo effect also need to be mentioned. I am unclear whether a placebo effect exists with AD. Is it not required for a placebo effect that the patient has sufficient mental ability? Does research demonstrate that placebo effects occur in AD?
Page 305, this is confusing some APOE 4+ who were in non-PV4 actually received 10 mg/kg dosing. Patients who migrated to PV-4 might have been selected for specific cognitive features. Very confusing. It shows how it is probably best not to go into some of these trial aspects too carefully.
Figure 23 is what we expect for the 301 high dose. As the trial continues and patients receive more high doses the CDRsb change improves for the high dose and the placebo regresses to its average which is largely stationary. The figure shows how 301 high dose was trending to significance at the last time point and this improvement appeared to be intensifying.
page 316, There is repeated mention made of how APOE- subgroup in 103 achieved a treatment benefit while in 302 it was the APOE + subgroup. Yet, the study populations were different in the two trials. 302 had earlier patients. Early APOE- patients might present particular problems in establishing true decline. Later in the disease process there van be a well-defined pattern of decline, yet at ~28 MMSE even with amyloid APOE- might pose particular problems in demonstrating consistent decline.
What is of interest is that even in the 103 the APOE+ showed CDRsb improvement of ~0.9. This is not completely out of line with what was seen in the 301 and 302 14 high doses. Perhaps this is a guideline for how epsilon 4s respond to aducan.
page 320, "A joint rank analysis of CDRsb change and survival gave an estimated week 78 difference in rank sums of 64.05 +/- 28.63, p=0.0256." Does this mean that patients with worse cognitive change had worse survival? If so, this highlights the importance that aducan be approved. Even early patients experience survival differences based upon cognitive changes? And earlier we saw that the 14 high dose patients in both 301 and 302 had better cognitive outcomes?
Page 322, looks at the different CDRsb changes by age. Once again it needs to be remembered that those groups that have not had much cognitive decline can not show much cognitive improvement. Perhaps this could have been better illustrated in some of the figures. For some patients, they might have only been able to show a 1 CDRsb improvement before hitting 0. Those who are younger might be especially sensitive to such improvements as it might allow them to return to work etc.. With those who are older, larger changes might relate more to regaining functional independence. It might have been better to simplify the trial design. Demonstrating larger cognitive gains in older patients would probably have been easier. In over 75 the CDRsb gain they saw was ~0.45, though it was 0.783 in 302.
Page 332, shows a 1.01 difference in 302 for the mild high group versus placebo. It reversed in 301 to -0.2774.
Figure 15 is another figure in Appendix 2 that is fudgy and not easy to read. Clearer figures would be very helpful.
The discussion about the lack of correlation between CDRsb and amyloid removal has been noted in a previous post.
Appendix 1 adds that at the group level this correlation is more apparent. This was also posted to the thread earlier. Page 220 Figure 12 actually shows a strong correlation at the group level. This figure does not add in the placebo 0,0 so the actual correlation is even larger than the figure makes it appear. Looking at Figure 12 this correlation is entirely lost and all we see is a mass of points. Without having a better view of the dataset it is difficult to comment further. There might be other features in the data that have been not disclosed that could help to explain the data cloud in Figure 12. For example, which dots represent mild and prodromal? Which epsilon 4+ or 4++? etc.
p.295 tries to move down into n= ~10 with tau etc. Probably best not to go there.
Err, Own Goal! "It can rule out a drug effect greater in magnitude than 0.262 with 95% confidence. " 301 high dose.
page 299. But Figure 13 on page 290 reported a 0.50 drug effect (1.58 n=77 - 1.08 n=77) for the 14 dose subgroup in 301.
Guess we can kiss that 95% CI good bye. 302 14 high dose in the same Figure 13 reports a 0.54 n=98 CDRsb benefit.
Page 300 small sample comparisons of sub-groups.
page 301 for the 301 trial consideration is given to the different age groups, this did not seem to be done for the 302 trial.
Shouldn't the analyses done be run on both datasets?
Figure 20 is verging on another own goal. Figure 13 suggested a 0.5 CDRsb benefit for 14 high doses in 301 with an average CDRsb change of 1.04. It wasn't sub-divided by epsilon 4, though this is a substantial benefit especially given that 301 as has been mentioned repeatedly in Appendix 2 reported a topline negative result. Also notice that even the high dose CDRsb at week 54 jumped around a fair amount; perhaps this level of variation is what we should expect.
Figure 21 is no less confusing. Do all of these figures show everything jumping around? What is the point then? The 302 placebo jump did seem odd, yet now with 301 what we see is that everything jumps around. Perhaps that is the behavior that we should expect. In figure 21, 301 APOE+ placebo jumps from Pre- to Post- and the in the high dose there is another dump and this time it dis downwards. The differences are not small; with high dose there is a drop at 78 weeks of ~0.50 CDRsb which is basically large; very large. That is exactly why you don't want to open the engine up; you start to see many things that are not easy to explain.
page 304 down to n=1 n=2? Isn't the idea that dose titration would be needed for those with ARIA because ARIA relates to more amyloid clearance and as seen here potentially better cognition a reasonable explanation? It is not even clear what dose titration even means for the placebo arm. Did some patients really need to be titrated down from their placebo? Sure, some on placebo experienced ARIA, though what happened then? They received a lower dose of placebo? Did placebos with ARIA also experience actual amyloid clearance or simply cognitive decline?
The issue about unblinding and placebo effect also need to be mentioned. I am unclear whether a placebo effect exists with AD. Is it not required for a placebo effect that the patient has sufficient mental ability? Does research demonstrate that placebo effects occur in AD?
Page 305, this is confusing some APOE 4+ who were in non-PV4 actually received 10 mg/kg dosing. Patients who migrated to PV-4 might have been selected for specific cognitive features. Very confusing. It shows how it is probably best not to go into some of these trial aspects too carefully.
Figure 23 is what we expect for the 301 high dose. As the trial continues and patients receive more high doses the CDRsb change improves for the high dose and the placebo regresses to its average which is largely stationary. The figure shows how 301 high dose was trending to significance at the last time point and this improvement appeared to be intensifying.
page 316, There is repeated mention made of how APOE- subgroup in 103 achieved a treatment benefit while in 302 it was the APOE + subgroup. Yet, the study populations were different in the two trials. 302 had earlier patients. Early APOE- patients might present particular problems in establishing true decline. Later in the disease process there van be a well-defined pattern of decline, yet at ~28 MMSE even with amyloid APOE- might pose particular problems in demonstrating consistent decline.
What is of interest is that even in the 103 the APOE+ showed CDRsb improvement of ~0.9. This is not completely out of line with what was seen in the 301 and 302 14 high doses. Perhaps this is a guideline for how epsilon 4s respond to aducan.
page 320, "A joint rank analysis of CDRsb change and survival gave an estimated week 78 difference in rank sums of 64.05 +/- 28.63, p=0.0256." Does this mean that patients with worse cognitive change had worse survival? If so, this highlights the importance that aducan be approved. Even early patients experience survival differences based upon cognitive changes? And earlier we saw that the 14 high dose patients in both 301 and 302 had better cognitive outcomes?
Page 322, looks at the different CDRsb changes by age. Once again it needs to be remembered that those groups that have not had much cognitive decline can not show much cognitive improvement. Perhaps this could have been better illustrated in some of the figures. For some patients, they might have only been able to show a 1 CDRsb improvement before hitting 0. Those who are younger might be especially sensitive to such improvements as it might allow them to return to work etc.. With those who are older, larger changes might relate more to regaining functional independence. It might have been better to simplify the trial design. Demonstrating larger cognitive gains in older patients would probably have been easier. In over 75 the CDRsb gain they saw was ~0.45, though it was 0.783 in 302.
Page 332, shows a 1.01 difference in 302 for the mild high group versus placebo. It reversed in 301 to -0.2774.
Re: Celebration Thread! Biogen is going to the FDA with Aducan.
Page 334, summary. They mention that the 20% of the pooled analysis was not met. However, it was explained in a previous section of the Document that there were aspects of the futility analysis that changed when considering the studies separately. For example, 302 by itself was quite a bit over 20% conditional power and that the 301 treatment effect for the high dose was rapidly evolving. The futility can be seen as a mistake. Why emphasize this? Stopping the trials was a mistake because they were demonstrating efficacy.
page 335, it is certainly true that it would have been greatly appreciated to have a cleaner data readout. There were a great many complications in the studies that do make it difficult to understand. However, more high doses does appear to have helped.
The more ARIA in the APOE + could explain the better cognitive performance relative to the APOE-. APOE+ had 3.7 higher dose titrations after PV-4 than APOE-.
page 338, it is very unclear why they introduce the correlation dialogue about the different cognitive measures. It is generally accepted that these different measures will be provided in AD trials. Combining them into a single measure is plausible, though it is not clear why this point is being even argued. Combining them through principal components as they are suggesting would likely increase the statistical significance that was reported for 302. With the tertiary measure, NPI, there clearly is an important dimension of behavior that is being captured by this test. Probably would have been best to have included this discussion before the conclusion.
Back to the correlation question on page 339. At the group level the correlation does seem strong. Perhaps there is a restriction of range problem involved. Why not see what the correlation of CDRsb change and amyloid change for all the patients (not just the high dose)?
Rapid progressors could be removed by looking at different parts of the distribution. For example, what was the average cDRsb change in the first 25%ile of the high dose. Mean response does give the rapid responders a great deal of leverage because they can have as much effect on the mean as ~25 patients.
103 had a much larger effect size than 302 though it also treated a population of AD patients.
page 341, "A p-value <0.05 doesn't necessarily reflect a clinically meaningful effect especially if there would be connotation f disease modification but the actual evidence supporting that is and there is a failed study calling hat p-value into question."
This is true and yet earlier in the document a statistically significant result was interpreted as clinically relevant in the context of AD. What is also of interest is that we now have seen the 103 study evolve out to at least 4 years and the benefit appears to grow and grow. It was noted quite some time ago on this thread that in fact at the expected time of the FDA's decision for aducan, those on placebo would be expected to have reached total disability with MMSE ~0. The trend line for the high dose suggests that they might still be mentally competent. Such a stark difference would curtail much of the number crunching in the Briefing Document: The apparent differences would be so large as to realistically preclude denying aducan treatment. The optics would not be good.
It is not so much that the clinical effects need to be that large within the clinical trial treatment time frame but only that delay in progression can be seen. The 40%-50% slowing that was even suggested in Appendix 2 would accumulate out over time to large delays in progression. There are also other approaches that could already be used in combination immediately upon aducan approval. aducan could be the mainline goto that stopped progression and then there would be addons that would amplify the effect-- those on forum are quite familiar with these other approaches.
Wowha, "Based on a Bayesian meta analysis of study 301 and 302 the posterior probability of the alternative is 0.62 and a corresponding 0.38 probability for the null hypothesis". page 342. This is moving close to the beta. This is what patients want to know. What is the probability that aducan actually works. This is important. The FDA could decline to approve aducan when it works: type 1 error. The FDA is in the don't make a type 1 error business. 302 posted 1.2% on the topline primary. I am not entirely sure though whether posterior probability of the alternative is the same as beta. It's pretty close. Would also be better if it were strictly 302 because 301 was something of a muddle. Better still would be combination beta of high dose 301 and 302. 38% probability for the null seems off.
Appendix 2 has issues that I am unsure how to rebut, They seem almost unfalsifiable. Chasing after all the small subgroup issues seems largely futile. The correlation issue possibly can be adequately rebutted by moving to the group level correlation. Nonetheless
the 301 and 302 high dose figure that was provided really did not show a striking relationship between CDRsb and amyloid clearance. I would greatly like to see the complete dataset. Showing only glimpses of the data can hide so much. Providing a 2 or 3 dimensional view can greatly move things into focus.
page 335, it is certainly true that it would have been greatly appreciated to have a cleaner data readout. There were a great many complications in the studies that do make it difficult to understand. However, more high doses does appear to have helped.
The more ARIA in the APOE + could explain the better cognitive performance relative to the APOE-. APOE+ had 3.7 higher dose titrations after PV-4 than APOE-.
page 338, it is very unclear why they introduce the correlation dialogue about the different cognitive measures. It is generally accepted that these different measures will be provided in AD trials. Combining them into a single measure is plausible, though it is not clear why this point is being even argued. Combining them through principal components as they are suggesting would likely increase the statistical significance that was reported for 302. With the tertiary measure, NPI, there clearly is an important dimension of behavior that is being captured by this test. Probably would have been best to have included this discussion before the conclusion.
Back to the correlation question on page 339. At the group level the correlation does seem strong. Perhaps there is a restriction of range problem involved. Why not see what the correlation of CDRsb change and amyloid change for all the patients (not just the high dose)?
Rapid progressors could be removed by looking at different parts of the distribution. For example, what was the average cDRsb change in the first 25%ile of the high dose. Mean response does give the rapid responders a great deal of leverage because they can have as much effect on the mean as ~25 patients.
103 had a much larger effect size than 302 though it also treated a population of AD patients.
page 341, "A p-value <0.05 doesn't necessarily reflect a clinically meaningful effect especially if there would be connotation f disease modification but the actual evidence supporting that is and there is a failed study calling hat p-value into question."
This is true and yet earlier in the document a statistically significant result was interpreted as clinically relevant in the context of AD. What is also of interest is that we now have seen the 103 study evolve out to at least 4 years and the benefit appears to grow and grow. It was noted quite some time ago on this thread that in fact at the expected time of the FDA's decision for aducan, those on placebo would be expected to have reached total disability with MMSE ~0. The trend line for the high dose suggests that they might still be mentally competent. Such a stark difference would curtail much of the number crunching in the Briefing Document: The apparent differences would be so large as to realistically preclude denying aducan treatment. The optics would not be good.
It is not so much that the clinical effects need to be that large within the clinical trial treatment time frame but only that delay in progression can be seen. The 40%-50% slowing that was even suggested in Appendix 2 would accumulate out over time to large delays in progression. There are also other approaches that could already be used in combination immediately upon aducan approval. aducan could be the mainline goto that stopped progression and then there would be addons that would amplify the effect-- those on forum are quite familiar with these other approaches.
Wowha, "Based on a Bayesian meta analysis of study 301 and 302 the posterior probability of the alternative is 0.62 and a corresponding 0.38 probability for the null hypothesis". page 342. This is moving close to the beta. This is what patients want to know. What is the probability that aducan actually works. This is important. The FDA could decline to approve aducan when it works: type 1 error. The FDA is in the don't make a type 1 error business. 302 posted 1.2% on the topline primary. I am not entirely sure though whether posterior probability of the alternative is the same as beta. It's pretty close. Would also be better if it were strictly 302 because 301 was something of a muddle. Better still would be combination beta of high dose 301 and 302. 38% probability for the null seems off.
Appendix 2 has issues that I am unsure how to rebut, They seem almost unfalsifiable. Chasing after all the small subgroup issues seems largely futile. The correlation issue possibly can be adequately rebutted by moving to the group level correlation. Nonetheless
the 301 and 302 high dose figure that was provided really did not show a striking relationship between CDRsb and amyloid clearance. I would greatly like to see the complete dataset. Showing only glimpses of the data can hide so much. Providing a 2 or 3 dimensional view can greatly move things into focus.
Re: Celebration Thread! Biogen is going to the FDA with Aducan.
Appendix 2 became too fixated on the subgroups and the statistical aspects of the trials that it was unable to recognize the efficacy signals that were sent. This is a problem when the number analysis is removed from the clinical intuition gained during a trial.
For example, the phase 103 has continued with extended dosing for at least 4 years and the high dose patients have clearly done much better than expected. Appendix 2 never even mentioned this. It only considered the results to as of June 2020(?), yet even the October 2018 look at 103 had ongoing updates. There was also surprisingly no mention of the long term extension for the 301 and 302. 1623 aducan patients from 301 and 302 received at least 1 long term extension dose. 792 of those on 10 mg/kg also received at least one long term extension dose. 319 were reported to have received over 2 years of 10 mg/kg dosing. After 2 or more years of dosing the cognitive benefit probably continued to expand. At 78 weeks perhaps there can be a debate, though not so much after years of dosing. The brain then has the space it needs to heal. Co-treatments that are already available and been tested in clinical trials could help from there.
After hundreds of posts to this thread I should have also been more aware of these patients. An update about these patients could easily end the debate. For whatever reason the advisory committee also apparently were not given an update on the LTE.
With potentially >2,000 patients soon to be back on site receiving high dose aducan treatment there should soon be little doubt about the efficacy of the treatment.
For example, the phase 103 has continued with extended dosing for at least 4 years and the high dose patients have clearly done much better than expected. Appendix 2 never even mentioned this. It only considered the results to as of June 2020(?), yet even the October 2018 look at 103 had ongoing updates. There was also surprisingly no mention of the long term extension for the 301 and 302. 1623 aducan patients from 301 and 302 received at least 1 long term extension dose. 792 of those on 10 mg/kg also received at least one long term extension dose. 319 were reported to have received over 2 years of 10 mg/kg dosing. After 2 or more years of dosing the cognitive benefit probably continued to expand. At 78 weeks perhaps there can be a debate, though not so much after years of dosing. The brain then has the space it needs to heal. Co-treatments that are already available and been tested in clinical trials could help from there.
After hundreds of posts to this thread I should have also been more aware of these patients. An update about these patients could easily end the debate. For whatever reason the advisory committee also apparently were not given an update on the LTE.
With potentially >2,000 patients soon to be back on site receiving high dose aducan treatment there should soon be little doubt about the efficacy of the treatment.
Re: Celebration Thread! Biogen is going to the FDA with Aducan.
You do not have the required permissions to view the files attached to this post.
Re: Celebration Thread! Biogen is going to the FDA with Aducan.
You do not have the required permissions to view the files attached to this post.
Re: Celebration Thread! Biogen is going to the FDA with Aducan.
You do not have the required permissions to view the files attached to this post.
Re: Celebration Thread! Biogen is going to the FDA with Aducan.
Yeah! This is great news!
I am happy that I have been able to work my way through the Briefing Document and develop understanding of the perspectives presented in it. Now I can explore other aducan related materials.
Embark! That is not in the imperative but the exclamative!
This is great!
I had seen the clinical trials.gov link for a continuation phase 3, though on the link they never actually mentioned cognitive end points. I was worried that perhaps they for some reason decided not to pursue cognitive endpoints; that would make no sense though sometimes you never know. The first url below and the first 2 posts above detail the Embark study. They have included a whole bunch of cognitive endpoints. All the tests from Emerge and Engage and also included MoCa. Perhaps they could have also added in RBANS.
https://investors.biogen.com/static-fil ... 769c455d75
https://www.clinicaltrials.gov/ct2/show ... w=2&rank=1
They are going to let the phase 3 mature for 2 more years of treatment.
One of the slides (with the map of the world) speaks of 550 randomized patients. The randomized part is a mystery to me as I thought all the patients were assigned to 10 mg/kg?
Biogen has sooo much unreleased patient results from Emerge and Engage and now Embark. It truly is startling.
The deck is stacked.
If the FDA reports back in March that they need more information; no problem-- back up the truck and unload.
They want data -- we got data.
Given the nature of an open label trial potentially the cognitive findings could be sent for analysis immediately after being recorded; foregoing the months of delay that can occur when part of a blinded trial.
78 weeks is now understood to be somewhat premature to expect results in an AD trial.
Amyloid needs to be cleared -- that takes about 18 months -- ARIA needs to clear up ...
After the amyloid is gone and the treatment continues, the brain has a better chance for some recovery room.
The next time measured time point is at 2 years.
By next March there could be quite a substantial number of patients who had reached 18-24 months and achieved a
negative amyloid determination. Perhaps >1000? This should move the numbers very far past the goal line. Possibly even the low dose arms from the phase 3s will have moved to negative amyloid readings and cleared significance.
Did the advisory meeting even make note of all of this data? The Briefing Document mentioned over 1600 patients who rotated into the long term extension; over 300 high dose patients who were on treatment for more than 2 years... That was only the original Long-Term Extension. My understanding is that patients completed the phase 3 and then went to LTE. When the trials were declared futile (March 2019) these LTE patients would have no longer received treatment. Then Embark started up in March 2020, aducan patients that enrolled were all moved up to high dose (after titration). Now that the COVID vaccine is rolling out, there could be a great deal more aducan patients that could soon be redosed with aducan. There is a lot of unreported treatment results.
Another piece of great news: Biogen's press releases report that Biogen applied for a new drug submission in Japan on December 9th. The Japanese patients in the aducan phase 3 trials had a better than average treatment response: Their CDRsb scores improved by ~1.4. That is quite a bit. While the error bar is quite wide, and the patient number was probably smallish, though it certainly makes you wonder. It does seem somewhat unlikely, though perhaps there is some specific environmental or genetic feature of Japan that makes it more responsive to aducan treatment. The press release notes that there is a massive massive burden of dementing illness in Japan. It is almost difficult to imagine-- 8.6 million with AD or MCI out of ~125 million? A dementia crisis! Given the seemingly large treatment effect and the very large unmet need, the sociopolitical force to approve there could be substantial.
I am happy that I have been able to work my way through the Briefing Document and develop understanding of the perspectives presented in it. Now I can explore other aducan related materials.
Embark! That is not in the imperative but the exclamative!
This is great!
I had seen the clinical trials.gov link for a continuation phase 3, though on the link they never actually mentioned cognitive end points. I was worried that perhaps they for some reason decided not to pursue cognitive endpoints; that would make no sense though sometimes you never know. The first url below and the first 2 posts above detail the Embark study. They have included a whole bunch of cognitive endpoints. All the tests from Emerge and Engage and also included MoCa. Perhaps they could have also added in RBANS.
https://investors.biogen.com/static-fil ... 769c455d75
https://www.clinicaltrials.gov/ct2/show ... w=2&rank=1
They are going to let the phase 3 mature for 2 more years of treatment.
One of the slides (with the map of the world) speaks of 550 randomized patients. The randomized part is a mystery to me as I thought all the patients were assigned to 10 mg/kg?
Biogen has sooo much unreleased patient results from Emerge and Engage and now Embark. It truly is startling.
The deck is stacked.
If the FDA reports back in March that they need more information; no problem-- back up the truck and unload.
They want data -- we got data.
Given the nature of an open label trial potentially the cognitive findings could be sent for analysis immediately after being recorded; foregoing the months of delay that can occur when part of a blinded trial.
78 weeks is now understood to be somewhat premature to expect results in an AD trial.
Amyloid needs to be cleared -- that takes about 18 months -- ARIA needs to clear up ...
After the amyloid is gone and the treatment continues, the brain has a better chance for some recovery room.
The next time measured time point is at 2 years.
By next March there could be quite a substantial number of patients who had reached 18-24 months and achieved a
negative amyloid determination. Perhaps >1000? This should move the numbers very far past the goal line. Possibly even the low dose arms from the phase 3s will have moved to negative amyloid readings and cleared significance.
Did the advisory meeting even make note of all of this data? The Briefing Document mentioned over 1600 patients who rotated into the long term extension; over 300 high dose patients who were on treatment for more than 2 years... That was only the original Long-Term Extension. My understanding is that patients completed the phase 3 and then went to LTE. When the trials were declared futile (March 2019) these LTE patients would have no longer received treatment. Then Embark started up in March 2020, aducan patients that enrolled were all moved up to high dose (after titration). Now that the COVID vaccine is rolling out, there could be a great deal more aducan patients that could soon be redosed with aducan. There is a lot of unreported treatment results.
Another piece of great news: Biogen's press releases report that Biogen applied for a new drug submission in Japan on December 9th. The Japanese patients in the aducan phase 3 trials had a better than average treatment response: Their CDRsb scores improved by ~1.4. That is quite a bit. While the error bar is quite wide, and the patient number was probably smallish, though it certainly makes you wonder. It does seem somewhat unlikely, though perhaps there is some specific environmental or genetic feature of Japan that makes it more responsive to aducan treatment. The press release notes that there is a massive massive burden of dementing illness in Japan. It is almost difficult to imagine-- 8.6 million with AD or MCI out of ~125 million? A dementia crisis! Given the seemingly large treatment effect and the very large unmet need, the sociopolitical force to approve there could be substantial.
Last edited by J11 on Mon Dec 14, 2020 7:10 pm, edited 1 time in total.
Re: Celebration Thread! Biogen is going to the FDA with Aducan.
One point I want to return to from the Briefing Document is the question of correlation between amyloid removal and change in CDRsb. The low correlation was presented as being unexpected: yet, it really wasn't. The last 2 decades of AD research has found this low correlation. Reducing amyloid near the symptomatic stage has not been seen to be strongly related to improving cognitive health. In fact, strangely amyloid tends to naturally decline somewhat as the disease progresses. If AD were a simple matter of reducing amyloid, then there would have been a cure decades ago. It is understood that amyloid is related to the neuropathology (aducan is an anti-Amyloid antibody), though in the later stages its role becomes less direct.
Re: Celebration Thread! Biogen is going to the FDA with Aducan.
You do not have the required permissions to view the files attached to this post.
