Table 3 Intercurrent events for ALS-FRS-R co-primary endpoint

A. Death before commencing randomised treatment

A hypothetical strategy [22] will be used assuming that deaths do not occur. In this case, only ALS-FRS-R at baseline may be available, which will be included in the analysis. If the participant did not record any ALS-FRS-R values, then the participant will be excluded from the analysis. This event is expected to be extremely rare

B. MND complication/admission. Event may include long-lasting complications such as cognitive decline leading to permanent discontinuation

The impact of this event will be minimised in the trial design through the use of telephone follow-up. In the event it does occur, the treatment policy strategy [22] will be used

C. Stopping riluzole during study

Riluzole affects ALS outcomes, they decline 12% slower than on no treatment. Audit data suggest that 15–20% of participants will stop riluzole during the course of MND-SMART. This event will be handled using a treatment policy strategy

D. Starting riluzole during study

As riluzole treatment is generally initiated close to diagnosis it is expected that few participants (< 10%) will start riluzole during MND-SMART. This event will therefore be handled by a treatment policy strategy

E. Deterioration in participant condition

Deterioration will occur frequently, and may vary across randomised treatment group. The trajectory of decline is expected to be observable in the ALS-FRS-R values and therefore a treatment policy strategy will be used

F. Concomitant illness

The treatment policy strategy will be used as per event B

G. Death before end of follow-up resulting in truncated data on ALS-FRS-R

A mortality rate of 25–30% is expected within 2 years of randomisation. ALS-FRS-R scores prior to death will likely reflect the imminent death. Here, a hypothetical strategy will be used assuming the participant has not died. Separate consideration of deaths not due to MND is not required as these will be rare

H. Discontinuation of memantine due to intolerance or adverse event

These discontinuations will mostly occur early, in about 10% of participants, who will then continue in study via telephone follow-up. The treatment policy perspective of MND-SMART means that the treatment policy strategy can be used to handle this event

I. Discontinuation of trazodone due to intolerance or adverse event

These discontinuations will mostly occur early, in about 15% of participants, who will then continue in the study via telephone follow-up. Will be addressed in the same way as Event H: treatment policy

J. Discontinuation of memantine due to lack of efficacy

The treatment policy strategy will be used, as this event is expected to be rare: perceived lack of efficacy withdrawal rare in literature (2% of withdrawals) as is progression (3% of withdrawals)

K. Discontinuation of trazodone due to lack of efficacy

The treatment policy strategy will be used, as this event is expected to be rare: perceived lack of efficacy withdrawal rare in literature (2% of withdrawals) as is progression (3% of withdrawals)

L. Participant entering end of life care resulting in truncated data

This will generally happen shortly before end of life, and therefore only one or two ALS-FRS-R measurements will be missing. The ALS-FRS-R scores prior to withdrawal will likely reflect deterioration in condition and a hypothetical strategy will be used assuming that end-of-life care or death does not occur

M. Use of memantine in placebo group

The treatment policy strategy will be used, as this event is expected to be extremely rare

N. Use of trazodone in placebo group

The treatment policy strategy will be used, as this event is expected to be extremely rare

O. Implementation of life extending treatment—gastrostomy

Enteral feeding—affects quality of life rather than survival and occurs in 10–20% of patients. Gastrostomy extends life by about 3 months based on data in Gorrie et al. [23]. As the trial is taking a treatment policy perspective, this event will be handled by using a treatment policy strategy

P. Implementation of life extending treatment—ventilation

Trachaeostomy extends life but is very rare. Non-invasive ventilation (NIV) via a mask is more common, occurring in 5% of patients soon after diagnosis and overall 10–20% would be expected over 18 months. NIV does not affect whether ALS-FRS-R can be measured. This event will be dealt with in the same way as Event O: treatment policy