Authors: Tufts Center for the Study of Drug Development and Janssen Research & Development
R&D Senior Leadership Brief (excerpt of the summary)
Starting in 2012, the Tufts Center for the Study of Drug Development (Tufts CSDD) and Janssen Research & Development (JRD) collaborated on a pilot study to develop and test a model capable of predicting the likelihood of marketing approval of oncology NME's and NBE's.The study assessed the predictive power of select new compound characteristics. The result is a simple scoring algorithm (ANDI, Approved New Drug Index) to assist in the assignment of probability of approval using data after Phase II testing has been conducted.
link: http://csdd.tufts.edu/files/uploads/CSD ... e_Tool.pdf
A New Tool for Predicting Marketing Approval of Oncology Drugs (JNJ/Tufts University)
Forum rules
- Comments must be civil and on topic
- Back up claims with evidence/reasoning/sources (posting links is allowed)
- No commercials/harassment/spam
- Comments must be civil and on topic
- Back up claims with evidence/reasoning/sources (posting links is allowed)
- No commercials/harassment/spam
-
Fishermangents
- Site Admin
- Posts: 529
- Joined: Thu Dec 03, 2015 3:39 pm
-
Fishermangents
- Site Admin
- Posts: 529
- Joined: Thu Dec 03, 2015 3:39 pm
Re: A New Tool for Predicting Marketing Approval of Oncology Drugs (JNJ/Tufts University)
Phil Pecsok from Seeking Alpha said the following about this tool when applied to imetelstat:
======
This is THE algo that JANSSEN considered when it partnered with Geron.
Go to page 4. Note scores are 0-1-2 for each of 4 questions.
#1. IMET has 40% activity if you look at CR/PR/CI and higher if you note others such as SD / score = 2
#2. Number of patients: Consider the Mayo had 80 patients, a multi-center trial will certainly have double or 150-200. Regardless, it will be over 50. / Score = 2
#3. Worldwide number treated. Smaller is better (basically FDA getting Orphan diseases some help). Only 3000 in US and worldwide has to be less than 10,000 or total is less than 50,000 / Score = 2
#4. Phase 2 duration. No way to know, but we DO KNOW that results will be known in about 6 months. Virtually all remissions occur in the first 3-4 months. I can't see the trial going more than 12-18 months or under their mark of 21 months. / Score = 2
Total score = 8.
Now go to page 5.
score of 8 is 100% chance of approval. I will discount and say it is 75-90%, but also note that these are HIGH 2 point scores, none are close to be a score of 1 (I suppose duration is an educated guess). One could suggest that Geron/Janssen have 90-100% chance of getting this approved
=====
Do we challenge the way Phil has used the tool? If so, what should be different?
======
This is THE algo that JANSSEN considered when it partnered with Geron.
Go to page 4. Note scores are 0-1-2 for each of 4 questions.
#1. IMET has 40% activity if you look at CR/PR/CI and higher if you note others such as SD / score = 2
#2. Number of patients: Consider the Mayo had 80 patients, a multi-center trial will certainly have double or 150-200. Regardless, it will be over 50. / Score = 2
#3. Worldwide number treated. Smaller is better (basically FDA getting Orphan diseases some help). Only 3000 in US and worldwide has to be less than 10,000 or total is less than 50,000 / Score = 2
#4. Phase 2 duration. No way to know, but we DO KNOW that results will be known in about 6 months. Virtually all remissions occur in the first 3-4 months. I can't see the trial going more than 12-18 months or under their mark of 21 months. / Score = 2
Total score = 8.
Now go to page 5.
score of 8 is 100% chance of approval. I will discount and say it is 75-90%, but also note that these are HIGH 2 point scores, none are close to be a score of 1 (I suppose duration is an educated guess). One could suggest that Geron/Janssen have 90-100% chance of getting this approved
=====
Do we challenge the way Phil has used the tool? If so, what should be different?