All safety, tolerability, PK and PD data available up to the clinical cut-off date of 5 January 2018, plus data that became available prior to the database snapshot on 6 February 2018 were included in the Internal Monitoring Committee (IMC) review. The IMC was responsible for selecting the dose for Part 2 of the study (pivotal dose). An external Independent Data Monitoring Committee (iDMC) reviewed data from Part 1 and confirmed the dose-selection decision of the IMC. The dose for Part 2 selected by the IMC was a dose that: 1.Was judged to be safe and well-tolerated, based on all available safety data from Part 1 and as confirmed by the iDMC; 2. Resulted in an exposure at steady-state below the exposure cap (mean value) of AUC0-24h,ss 2000 ng*h/mL (adjusted for free-fraction, if required); 3. Resulted in an SMN protein increase that was expected to be clinically relevant.

All participants in Part 1 who received at least one dose of risdiplam and had available data at the time of the data snapshot for selecting the Part 2 dose.

The BSID-III is a standardized assessment commonly used to evaluate development across five domains in infants and young children aged between 1 and 42 months. The gross motor subscale of the BSID-III is evaluated based on the linear and hierarchical obtainment of motor skills, as seen in typically developing children. The gross motor scale consists of 72 items scored at 0 (unable to perform the activity) or 1 (criteria for item achieved). Item 22 is not considered achieved if the infant sits alone for less than 5 seconds before losing balance and falling over, or if the infant uses his or her arms to prop him- or herself up. The assessment was video recorded at study sites and reviewed/scored by two independent raters. 90% CI for one sample binomial was computed using Clopper-Pearson (exact) method.

The intent-to-treat (ITT) population is defined as all infants enrolled in Part 2 of the study, regardless of whether they received treatment or not.

The CHOP-INTEND instrument was developed to evaluate motor function in infants with SMA from the ages of 1.4 to 37.9 months. It consists of 16 items, where each item assesses a specific motor task graded on a scale of 0 to 4, where zero is no response and 4 is a complete response. The total score ranges from 0 to 64, with higher scores consistent with better motor function. 90% CI for one sample binomial was computed using Clopper-Pearson (exact) method.

The intent-to-treat (ITT) population is defined as all infants enrolled in Part 2 of the study, regardless of whether they received treatment or not.

The CHOP-INTEND instrument was developed to evaluate motor function in infants with SMA from the ages of 1.4 to 37.9 months. It consists of 16 items, where each item assesses a specific motor task graded on a scale of 0 to 4, where zero is no response and 4 is a complete response. The total score ranges from 0 to 64, with higher scores consistent with better motor function. 90% CI for one sample binomial was computed using Clopper-Pearson (exact) method.

The intent-to-treat (ITT) population is defined as all infants enrolled in Part 2 of the study, regardless of whether they received treatment or not.

The CHOP-INTEND instrument was developed to evaluate motor function in infants with SMA from the ages of 1.4 to 37.9 months. It consists of 16 items, where each item assesses a specific motor task graded on a scale of 0 to 4, where zero is no response and 4 is a complete response. The total score ranges from 0 to 64, with higher scores consistent with better motor function. 90% CI for one sample binomial was computed using Clopper-Pearson (exact) method.

The intent-to-treat (ITT) population is defined as all infants enrolled in Part 2 of the study, regardless of whether they received treatment or not.

The BSID-III is a standardized assessment commonly used to evaluate development across five domains in infants and young children aged between 1 and 42 months. The gross motor subscale of the BSID-III is evaluated based on the linear and hierarchical obtainment of motor skills, as seen in typically developing children. The gross motor scale consists of 72 items scored at 0 (unable to perform the activity) or 1 (criteria for item achieved). In this study the gross motor scale was assessed in a modified way compared with the standard administration. A total raw score was calculated by summing the item scores to give a maximum possible score of 72.

The intent-to-treat (ITT) population is defined as all infants enrolled in Part 2 of the study, regardless of whether they received treatment or not. Only participants for whom data were collected are included in the analysis.

The HINE was designed to evaluate infants between 2 months and 24 months of age. It is a simple and standardized instrument that includes 26 items assessing different aspects of neurological examinations, such as cranial nerves, posture, movements, tone, and reflexes. In this study, Module 2 of the HINE (HINE-2) was assessed. The HINE-2 evaluates 8 developmental milestones (head control, sitting, voluntary grasp, ability to kick, rolling, crawling, standing, and walking) scored on a 3, 4, or 5-point scale, with 0 indicating inability to perform a task and a score of 2, 3, or 4 (depending on the task) indicating full milestone development. The total score is calculated by summing the item scores to give a maximum possible score of 26. This measure represents subset numbers at Month 8 for head control, ability to kick and rolling milestones only.

The intent-to-treat (ITT) population is defined as all infants enrolled in Part 2 of the study, regardless of whether they received treatment or not.

The HINE was designed to evaluate infants between 2 months and 24 months of age. It is a simple and standardized instrument that includes 26 items assessing different aspects of neurological examinations, such as cranial nerves, posture, movements, tone, and reflexes. In this study, Module 2 of the HINE (HINE-2) was assessed. The HINE-2 evaluates 8 developmental milestones (head control, sitting, voluntary grasp, ability to kick, rolling, crawling, standing, and walking) scored on a 3, 4, or 5-point scale, with 0 indicating inability to perform a task and a score of 2, 3, or 4 (depending on the task) indicating full milestone development. The total score is calculated by summing the item scores to give a maximum possible score of 26.

The intent-to-treat (ITT) population is defined as all infants enrolled in Part 2 of the study, regardless of whether they received treatment or not.

The HINE was designed to evaluate infants between 2 months and 24 months of age. It is a simple and standardized instrument that includes 26 items assessing different aspects of neurological examinations, such as cranial nerves, posture, movements, tone, and reflexes. In this study, Module 2 of the HINE (HINE-2) was assessed. The HINE-2 evaluates 8 developmental milestones (head control, sitting, voluntary grasp, ability to kick, rolling, crawling, standing, and walking) scored on a 3, 4, or 5-point scale, with 0 indicating inability to perform a task and a score of 2, 3, or 4 (depending on the task) indicating full milestone development. The total score is calculated by summing the item scores to give a maximum possible score of 26.

The intent-to-treat (ITT) population is defined as all infants enrolled in Part 2 of the study, regardless of whether they received treatment or not.

The HINE-2 evaluates 8 developmental milestones scored on a 3, 4, or 5-point scale, with 0 indicating inability to perform a task and a score of 2, 3, or 4 (depending on the task) indicating full milestone development. The total score is calculated by summing the item scores to give a maximum possible score of 26. For the motor milestone responder definition, an improvement in a motor milestone is defined as at least a 2-point increase in ability to kick (or maximal score) or a 1-point increase in head control, rolling, sitting, crawling, standing, or walking. Worsening is similarly defined as a 2-point decrease in ability to kick (or lowest score) or a 1-point decrease in head control, rolling, sitting, crawling, standing, or walking. Voluntary grasp is excluded from the definition. An infant is classified as a responder if more motor milestones show improvement than show worsening. 90% CI for one sample binomial was computed using Clopper-Pearson (exact) method.

The intent-to-treat (ITT) population is defined as all infants enrolled in Part 2 of the study, regardless of whether they received treatment or not.

The BSID-III is a standardized assessment commonly used to evaluate development across five domains in infants and young children aged between 1 and 42 months. The gross motor subscale of the BSID-III is evaluated based on the linear and hierarchical obtainment of motor skills, as seen in typically developing children. The gross motor scale consists of 72 items scored at 0 (unable to perform the activity) or 1 (criteria for item achieved). This specific measure included 6 milestones: Item 9 'Controls head while upright for 15 seconds', Item 14 'Rolls from side to back', Item 22 'Sits without support for 5 seconds', Item 30 'Crawls on stomach', Item 40 'Stands alone' and Item 42 'Walks alone'. Reported here is the percentage of participants for whom the reported item was the highest item achieved out of these six items by Months 12 and 24.

The intent-to-treat (ITT) population is defined as all infants enrolled in Part 2 of the study, regardless of whether they received treatment or not.

The BSID-III is a standardized assessment commonly used to evaluate development across five domains in infants and young children aged between 1 and 42 months. The gross motor subscale of the BSID-III is evaluated based on the linear and hierarchical obtainment of motor skills, as seen in typically developing children. The gross motor scale consists of 72 items scored at 0 (unable to perform the activity) or 1 (criteria for item achieved). Item 22 is not considered achieved if the infant sits alone for less than 5 seconds before losing balance and falling over, or if the infant uses his or her arms to prop him- or herself up. The assessment was video recorded at study sites and reviewed/scored by two independent raters. 90% CI for one sample binomial was computed using Clopper-Pearson (exact) method.

The intent-to-treat (ITT) population is defined as all infants enrolled in Part 2 of the study, regardless of whether they received treatment or not.

The BSID-III is a standardized assessment commonly used to evaluate development across five domains in infants and young children aged between 1 and 42 months. The gross motor subscale of the BSID-III is evaluated based on the linear and hierarchical obtainment of motor skills, as seen in typically developing children. The gross motor scale consists of 72 items scored at 0 (unable to perform the activity) or 1 (criteria for item achieved). Item 26 is not considered achieved if the infant sits alone for less than 30 seconds before losing balance and falling over, or if the infant uses his or her arms to prop him- or herself up. The assessment was video recorded at study sites and reviewed/scored by two independent raters. 90% CI for one sample binomial was computed using Clopper-Pearson (exact) method.

The intent-to-treat (ITT) population is defined as all infants enrolled in Part 2 of the study, regardless of whether they received treatment or not.

The BSID-III is a standardized assessment commonly used to evaluate development across five domains in infants and young children aged between 1 and 42 months. The gross motor subscale of the BSID-III is evaluated based on the linear and hierarchical obtainment of motor skills, as seen in typically developing children. The gross motor scale consists of 72 items scored at 0 (unable to perform the activity) or 1 (criteria for item achieved). Reported here is the percentage of participants who achieved item 40 'Stands alone' at Month 24. The assessment was video recorded at study sites and reviewed/scored by two independent raters. 90% CI for one sample binomial was computed using Clopper-Pearson (exact) method.

The intent-to-treat (ITT) population is defined as all infants enrolled in Part 2 of the study, regardless of whether they received treatment or not.

The BSID-III is a standardized assessment commonly used to evaluate development across five domains in infants and young children aged between 1 and 42 months. The gross motor subscale of the BSID-III is evaluated based on the linear and hierarchical obtainment of motor skills, as seen in typically developing children. The gross motor scale consists of 72 items scored at 0 (unable to perform the activity) or 1 (criteria for item achieved). Reported here is the percentage of participants who achieved item 42 'Walks alone' at Month 24. The assessment was video recorded at study sites and reviewed/scored by two independent raters. 90% CI for one sample binomial was computed using Clopper-Pearson (exact) method.

The intent-to-treat (ITT) population is defined as all infants enrolled in Part 2 of the study, regardless of whether they received treatment or not.

The median time to event was estimated using Kaplan-Meier methodology. 90% CI was estimated using the method of Brookmeyer and Crowley.

The intent-to-treat (ITT) population is defined as all infants enrolled in Part 2 of the study, regardless of whether they received treatment or not.

Permanent ventilation is defined as >/=16 hours of non-invasive ventilation per day or intubation for >21 consecutive days in the absence of, or following the resolution of, an acute reversible event; or tracheostomy. Permanent ventilation events were reviewed and confirmed by an independent adjudication committee. The median time to event was estimated using Kaplan-Meier methodology. 90% CI was estimated using the method of Brookmeyer and Crowley.

The intent-to-treat (ITT) population is defined as all infants enrolled in Part 2 of the study, regardless of whether they received treatment or not.

Permanent ventilation is defined as >/=16 hours of non-invasive ventilation per day or intubation for >21 consecutive days in the absence of, or following the resolution of, an acute reversible event; or tracheostomy. Permanent ventilation events were reviewed and confirmed by an independent adjudication committee. The median time to event was estimated using Kaplan-Meier methodology. 90% CI was estimated using the method of Brookmeyer and Crowley.

The intent-to-treat (ITT) population is defined as all infants enrolled in Part 2 of the study, regardless of whether they received treatment or not.

Permanent ventilation is defined as >/=16 hours of non-invasive ventilation per day or intubation for >21 consecutive days in the absence of, or following the resolution of, an acute reversible event; or tracheostomy. Permanent ventilation events were reviewed and confirmed by an independent adjudication committee. Kaplan-Meier methodology was used to estimate the percentages. 90% CI was computed using the complementary log-log transformation.

The intent-to-treat (ITT) population is defined as all infants enrolled in Part 2 of the study, regardless of whether they received treatment or not.

Kaplan-Meier methodology was used to estimate the percentages. 90% CI was computed using the complementary log-log transformation.

The intent-to-treat (ITT) population is defined as all infants enrolled in Part 2 of the study, regardless of whether they received treatment or not.

Permanent ventilation is defined as >/=16 hours of non-invasive ventilation per day or intubation for >21 consecutive days in the absence of, or following the resolution of, an acute reversible event; or tracheostomy. Permanent ventilation events were reviewed and confirmed by an independent adjudication committee. Kaplan-Meier methodology was used to estimate the percentages. 90% CI was computed using the complementary log-log transformation.

The intent-to-treat (ITT) population is defined as all infants enrolled in Part 2 of the study, regardless of whether they received treatment or not.

RP measures the degree of synchrony between abdominal and thoracic cage-driven breathing. The weakness of intercostal muscles leads to asynchrony of the thorax with the diaphragm, resulting in inefficient and paradoxical breathing patterns. The degree of synchrony between the movement of the chest wall and abdomen during the respiratory cycle can be expressed as the phase angle between the two compartments and measured by placing two RP bands around the thorax and abdomen. In paradoxical breathing, the phase angle is reversed compared with the normal ventilation cycle. A phase angle of 0° indicates perfect in-phase movement, while a value of 180° indicates completely out-of-phase movement between the two compartments. In this measure, 8 or more valid breaths were used to determine the phase angle at each visit; the calculation was not performed if fewer than 8 valid breaths had been measured. 90% CI for one sample binomial was computed using Clopper-Pearson (exact) method.

The intent-to-treat (ITT) population is defined as all infants enrolled in Part 2 of the study, regardless of whether they received treatment or not.

90% CI for one sample binomial was computed using Clopper-Pearson (exact) method.

The intent-to-treat (ITT) population is defined as all infants enrolled in Part 2 of the study, regardless of whether they received treatment or not.

Able to feed orally includes participants fed orally and participants fed via a combination of oral and tube feeding. 90% CI for one sample binomial was computed using Clopper-Pearson (exact) method.

The intent-to-treat (ITT) population is defined as all infants enrolled in Part 2 of the study, regardless of whether they received treatment or not.

An adverse event (AE) is any untoward medical occurrence in a subject administered a pharmaceutical product and which does not necessarily have to have a causal relationship with the treatment. An adverse event can therefore be any unfavorable and unintended sign (including an abnormal laboratory finding, for example), symptom, or disease temporally associated with the use of a pharmaceutical product, whether or not considered related to the pharmaceutical product. Preexisting conditions which worsen during a study are also considered as adverse events.

All infants who received at least one dose of study medication (risdiplam) were included in the safety population.

An adverse event (AE) is any untoward medical occurrence in a subject administered a pharmaceutical product and which does not necessarily have to have a causal relationship with the treatment. An adverse event can therefore be any unfavorable and unintended sign (including an abnormal laboratory finding, for example), symptom, or disease temporally associated with the use of a pharmaceutical product, whether or not considered related to the pharmaceutical product. Preexisting conditions which worsen during a study are also considered as adverse events.

All infants who received at least one dose of study medication (risdiplam) were included in the safety population.

An adverse event (AE) is any untoward medical occurrence in a subject administered a pharmaceutical product and which does not necessarily have to have a causal relationship with the treatment. An adverse event can therefore be any unfavorable and unintended sign (including an abnormal laboratory finding, for example), symptom, or disease temporally associated with the use of a pharmaceutical product, whether or not considered related to the pharmaceutical product. Preexisting conditions which worsen during a study are also considered as adverse events.

All infants who received at least one dose of study medication (risdiplam) were included in the safety population.

Anthropometric examination included weight, height, head circumference and chest circumference.

All infants who received at least one dose of study medication (risdiplam) were included in the safety population. Only participants for whom data were collected are included in the analysis.

Anthropometric examination included weight, height, head circumference and chest circumference.

All infants who received at least one dose of study medication (risdiplam) were included in the safety population. Only participants for whom data were collected are included in the analysis.

Reported here is the maximum observed concentration throughout the observation period.

All participants with at least one time point with a risdiplam concentration measurement were included in the PK analysis data set.

The AUC was derived by modelling & simulation via the population pharmacokinetics (PPK) method based on the Year 5 pre-dose concentration measurement only. Hence, no other timepoints have been reported.

All participants with at least one time point with a risdiplam concentration measurement were included in the PK analysis data set. Only participants for whom data were collected are included in the analysis.

Ctrough at Year 5 was the last predose sample collected from each participant who had at least 1400 days of risdiplam treatment duration.

All participants with at least one time point with a risdiplam concentration measurement were included in the PK analysis data set. Only participants for whom data were collected are included in the analysis.

All participants with at least one time point with a protein measurement were included in the pharmacodynamic (PD) analysis data set. Only participants for whom data were collected are included in the analysis.