On 2015 Mar 11, Alexis Clapin commented:

Congratulation for your work.

I would like to draw your attention on one of the four selected studies: MSCRG 1999 study(1). This study compared whole brain atrophy in relapsing-remitting MS patients treated with either Avonex or a placebo. One hundred forty patients are selected for this comparison. These patients participated to the initial MSCRG 1996 study (2), which was stopped earlier than planned. Among the 301 patients included, 172 were followed up for two years. The 140 patients are thus drawn from these 172 patients.

When reviewing the MSCRG 1996 study, the FDA analysed the efficacy of Avonex on the relapse rate for different subgroups of patients to check if, as Biogen said, there was a lag time between initiation of treatment and onset of clinical benefit. FDA reviewers demonstrated (3)that there was no lag time but that patients followed up for two years were different from those followed up for less than two years. During the first year of the study, Avonex patients followed up for less than two years experienced a +29% increase of the relapse rate versus placebo while patients followed up for at least two years experienced a -29% decrease versus placebo. During the second year, relapse rate did not change (-32%). FDA finally concluded that patients followed for two years were different from those followed up for less than two years, but did not provide a comparison of the subgroups.

The group of patients followed up for two years is thus a very odd subgroup of the randomised patients in the initial MSCRG 1996 study (2); obviously there are favouring the positive evaluation of the product.

For further explanation of the bias in the MSCRG 1996 study, you can see this analysis (4) on the MSCRG 1996 study (2).

The Cochrane collaboration recently considered that Avonex had a negative benefit/risk ratio (5). Their risk of bias analysis for the MSCRG (1996) (2) is more negative than the one you provide for the MSCRG 1999 study (1). To conclude, I am not sure that you can consider MSCRG 1999 (2) patients study as an unbiased subgroup of the initial MSCRG 1996 study (1). It is just a comparison of a group of best responders to Avonex to a group of worst responders to placebo.

(1)http://www.ncbi.nlm.nih.gov/pubmed/10563615 (2)http://www.ncbi.nlm.nih.gov/pubmed/8602746<br> (3)http://www.fda.gov/downloads/Drugs/DevelopmentApprovalProcess/HowDrugsareDevelopedandApproved/ApprovalApplications/TherapeuticBiologicApplications/ucm086056.pdf (4)http://www.ncbi.nlm.nih.gov/pubmed/23662092<br> (5)http://www.ncbi.nlm.nih.gov/pubmed/23744561

On 2015 Mar 11, Alexis Clapin commented:

Congratulation for your work.

I would like to draw your attention on one of the four selected studies: MSCRG 1999 study(1). This study compared whole brain atrophy in relapsing-remitting MS patients treated with either Avonex or a placebo. One hundred forty patients are selected for this comparison. These patients participated to the initial MSCRG 1996 study (2), which was stopped earlier than planned. Among the 301 patients included, 172 were followed up for two years. The 140 patients are thus drawn from these 172 patients.

When reviewing the MSCRG 1996 study, the FDA analysed the efficacy of Avonex on the relapse rate for different subgroups of patients to check if, as Biogen said, there was a lag time between initiation of treatment and onset of clinical benefit. FDA reviewers demonstrated (3)that there was no lag time but that patients followed up for two years were different from those followed up for less than two years. During the first year of the study, Avonex patients followed up for less than two years experienced a +29% increase of the relapse rate versus placebo while patients followed up for at least two years experienced a -29% decrease versus placebo. During the second year, relapse rate did not change (-32%). FDA finally concluded that patients followed for two years were different from those followed up for less than two years, but did not provide a comparison of the subgroups.

The group of patients followed up for two years is thus a very odd subgroup of the randomised patients in the initial MSCRG 1996 study (2); obviously there are favouring the positive evaluation of the product.

For further explanation of the bias in the MSCRG 1996 study, you can see this analysis (4) on the MSCRG 1996 study (2).

The Cochrane collaboration recently considered that Avonex had a negative benefit/risk ratio (5). Their risk of bias analysis for the MSCRG (1996) (2) is more negative than the one you provide for the MSCRG 1999 study (1). To conclude, I am not sure that you can consider MSCRG 1999 (2) patients study as an unbiased subgroup of the initial MSCRG 1996 study (1). It is just a comparison of a group of best responders to Avonex to a group of worst responders to placebo.

(1)http://www.ncbi.nlm.nih.gov/pubmed/10563615 (2)http://www.ncbi.nlm.nih.gov/pubmed/8602746<br> (3)http://www.fda.gov/downloads/Drugs/DevelopmentApprovalProcess/HowDrugsareDevelopedandApproved/ApprovalApplications/TherapeuticBiologicApplications/ucm086056.pdf (4)http://www.ncbi.nlm.nih.gov/pubmed/23662092<br> (5)http://www.ncbi.nlm.nih.gov/pubmed/23744561