The article reflects a well-planned and executed work, which justifies its interest for NEJM. It is a phase 3 trial in a very specific subgroup of ALS patients, carriers of a mutation in a gene (SOD1). This limits the external validity of its conclusions but homogenizes the sample. ALS is a multifactorial disease and this makes it difficult - in my opinion, it is the cause of so much therapeutic failure - for a pathway focused on a specific therapeutic target to be successful. It should be noted that, if we look at the different genes involved in familial forms of ALS, virtually all cellular/metabolic functions may be involved. Therefore, the trial has limited external validity but high internal validity.  

108 patients have been enrolled with 72 patients in the active arm [those who received treatment and not placebo], and these have been distributed between slowly (33) and rapidly (39) progressing patients, which is well done but somewhat small numbers for a 28-week progression. They were mostly taking Riluzole and a few Edaravone [other ALS treatments] simultaneously.  

There is a significant difference in the time of illness from onset (8.3 months for those with rapid progression and 39.6 for those with slow progression). This difference is normal, and the time for the fast ones is short, which is fine because I am persuaded that any therapy will work better or only work in early stages (unfortunately, by the time symptoms debut the disease has been evolving for some time). In other words, the design seems correct and feasible to replicate with more cases.  

As for the results, there are significant differences in some biomarkers of neuronal damage (neurofilaments), but at week 28 the [functional] difference of those treated is not statistically significant in the group of those who progress rapidly. Many patients (95) are still in the 236-week open-label extension [the study continued without a placebo group], and it seems that putting the treatment early generates a smaller drop in the functional assessment scale. It also allows us to venture that perhaps in a cut-off at 52 weeks the differences could be increased by introducing mortality as a secondary variable of analysis, but, a priori, the trial fails in the essential aspect, which is to avoid disease progression.  

Another issue to highlight is that in this disease, as in other neurodegenerative diseases, it is one thing to improve biomarkers and another to recover or halt the progression of the disease. By analogy, annulling the causes of the fire is not the same as putting it out, and this in turn is different from recovering the burnt forest.  

The age of the patients is approximately 50 +/- 12 years. The ideal would be to do a trial in presymptomatic patients. For example, you could do a trial in presymptomatic carriers with the mutation and measure conversion rate to clinical as well as biomarkers.  

Finally, and this is positive, the rate of side effects of intrathecal administration [in the space where the cerebrospinal fluid is located] with 15 ml injections (8 doses in total) during 24 weeks is reasonable (4 related events in the treated group, which is 7%, although there are 7 cases with pleocytosis [increased cells in the cerebrospinal fluid] and 6 with increased proteins, which could suggest a myelitic process). This would make it necessary in successive studies to include an MRI to monitor it.  

Company press releases are rarely not optimistic. I believe that the outcome of the trial today is negative. Overall it is an interesting study, but it needs follow-up and possibly a design in presymptomatic patients if some doubts about these myelitis are clarified.