Conceived and designed the experiments: TOC SVP DTK SJF CLJ RSF PK KJS RL KSC. Performed the experiments: TOC DTK RSF PK KJS RHL. Analyzed the data: TOC SVP SJF CLJ RSF PK KJS DT RHL FLT TP KSC. Contributed reagents/materials/analysis tools: DT TP. Wrote the paper: TOC SVP DTK SJF CLJ RSF KJS TP KSC.
¶ Members of the BforSMA Trial Group are listed in
See below for the Full List of Competing Interests.
The universal presence of a gene (SMN2) nearly identical to the mutated SMN1 gene responsible for Spinal Muscular Atrophy (SMA) has proved an enticing incentive to therapeutics development. Early disappointments from putative SMN-enhancing agent clinical trials have increased interest in improving the assessment of SMN expression in blood as an early “biomarker” of treatment effect.
A cross-sectional, single visit, multi-center design assessed SMN transcript and protein in 108 SMA and 22 age and gender-matched healthy control subjects, while motor function was assessed by the Modified Hammersmith Functional Motor Scale (MHFMS). Enrollment selectively targeted a broad range of SMA subjects that would permit maximum power to distinguish the relative influence of SMN2 copy number, SMA type, present motor function, and age.
SMN2 copy number and levels of full-length SMN2 transcripts correlated with SMA type, and like SMN protein levels, were lower in SMA subjects compared to controls. No measure of SMN expression correlated strongly with MHFMS. A key finding is that SMN2 copy number, levels of transcript and protein showed no correlation with each other.
This is a prospective study that uses the most advanced techniques of SMN transcript and protein measurement in a large selectively-recruited cohort of individuals with SMA. There is a relationship between measures of SMN expression in blood and SMA type, but not a strong correlation to motor function as measured by the MHFMS. Low SMN transcript and protein levels in the SMA subjects relative to controls suggest that these measures of SMN in accessible tissues may be amenable to an “early look” for target engagement in clinical trials of putative SMN-enhancing agents. Full length SMN transcript abundance may provide insight into the molecular mechanism of phenotypic variation as a function of SMN2 copy number.
Spinal muscular atrophy (SMA) is an autosomal recessive neuromuscular disorder that manifests across a wide range of severity. The cardinal clinical feature of SMA – diffuse skeletal muscle weakness – is a consequence of dysfunction or loss of alpha motor neurons. SMA is caused by loss-of-function mutations or deletions of the gene SMN1 (Gene ID = 6606). A wide range of disease severity can be partially attributed to the presence of variable copy number of a neighboring near-identical gene, SMN2 (Gene ID = 6607)
One impediment to efficient trial design is that the typical individual with SMA old enough to cooperate with motor function testing declines very slowly
The continuous spectrum of SMA phenotype severity is generally divided into three “Types” of SMA based upon the history of specific gross motor abilities achieved before the disease curtailed further developmental progress
The BforSMA study was designed to explore potential biomarkers of disease severity in accessible peripheral tissues – blood and urine – that may be of value in the execution of clinical trials. The cohort was selectively recruited to represent subjects with a wide range of SMA type, motor function, and age, who were then characterized clinically in a manner that would best power two independent projects. Results of the first project, an unbiased biomarker discovery effort, are presented in a companion paper (see companion paper, Finkel et al.
Important to this project was the recognition that age is an important potential confounder to be included in the statistical analysis. The power of this is enhanced by targeted recruitment of a clinical cohort in which SMA type and age are not highly correlated. At the outset of the study, the extent to which this broad distribution could be accomplished was unknown, given early childhood mortality of subjects with SMA Type I. Site investigators were encouraged to keep this goal in mind during subject recruitment.
To date, studies of SMN transcript and protein levels from amniocytes, skin fibroblasts, lymphocytes or peripheral blood mononuclear cells (PBMCs) from SMA subjects have reported variable results, but a general correlation of transcript or protein quantity with phenotype is identified only in individuals at the lower end of the SMA functional continuum
The protocol was developed by a combination of SMA patient advocates, experts in biostatistics and bioinformatics, basic scientists investigating the biology of SMA, drug metabolism and pathway experts, and clinical trial specialists. The protocol supported two related, but independent, objectives: (1) an unbiased protein, metabolite and transcript biomarker discovery analysis, and (2) the SMN-targeted biomarker analysis reported here, which intends to identify the relationship of quantitative measures of SMN expression to clinical features of SMA. A leading instrument for the measurement of gross motor function in SMA is the Modified Hammersmith Functional Motor Scale (MHFMS), and although this scale was initially designed and validated only for those with Type II SMA, it is suitable also to evaluate Type III individuals who have lost independent ambulation and to a lesser extent to some of those with only weak ability to walk
A cross-sectional, single visit, multi-center, exploratory study design was employed. Blood for SMN transcript and protein analysis and SMN2 copy number determination was collected. No therapeutic intervention occurred. There was special emphasis on enrolling children as subjects across a range phenotype severity. The full protocol is available in the
The administration and data coordination of the trial was conducted by the New England Research Institutes (NERI, Watertown, MA). A protocol committee was established to design the protocol and review potential violations as the trial progressed. Eighteen academic clinical sites headed by investigators with special expertise in the care of children with SMA participated in the study. A meeting of investigators and clinical evaluators from each study site was held to introduce them to the study protocol. All data entry was performed via a secure internet-based website. Sample collection, handling and chain of custody procedures were designed to ensure the best quality specimens for biomarker analysis. Samples were divided for specific analysis and shipped from study sites to the central lab, PPD, Inc. (Highland Heights, KT.) SMN transcripts were measured by Dr. F. Danilo Tiziano laboratory (Istituto di Genetica Medica, Università Cattolica S. Cuore, Roma, Italy), SMN2 copy number from blood were analyzed by Expression Analysis, Inc. (Durham, NC), and SMN protein analysis by an ELISA assay developed and run at Enzo Life Sciences (Ann Arbor, MI). The Project Director and Data Management group at NERI was responsible for data processing while their Biostatistics group handled the data analyses.
Institutional review board (IRB) approval for the protocol was obtained from each BforSMA clinical site before enrollment at that site and from the central Institutional Review Board, New England Research Institutes. Written informed consent for participation was obtained from the legal guardians of all subjects and assent for participation was obtained directly from subjects whenever applicable. This trial was registered with ClinicalTrials.gov with identifier NCT00756821.
The number and type of subjects enrolled was based upon the expected power necessary for the unbiased discovery project. The sample size had an 80% power at level 0.05 to detect an effect size of 0.6 standard deviations between SMA and controls, which is considered a moderate difference. An unbalanced number of subjects from the three types of SMA for targeted enrollment was selected (15 Type I, 45 Type II, and 40 Type III SMA subjects) in part due to the expectation that recruitment of eligible subjects with SMA Type I across the full age range would be difficult due to the high level of mortality in childhood
The ideal population would enable independent assessment of the contribution of SMN genotype, SMA type, present functional status, and age in biomarker characterization. The age range chosen excluded subjects less than 2 years of age – to avoid metabolic confounders of infancy, nascent developmental accomplishments, and incomplete cooperation with the functional outcome measure – and also excluded those over 12 years of age to limit the contribution of puberty-related confounders
Potential predictors of SMN protein, transcript and copy numbers included: age, gender, SMA diagnosis and type, and current level of function as measured by the MHFMS. Relationships among SMN protein, individual transcript types (listed below), and copy numbers were examined as well. One extreme outlier of SMN transcript ratios (SMN2-FL/SMN-Δ7 ratio, 26 standard deviations above the next highest value) was removed from analysis of that outcome alone. Predictors of continuous outcomes were assessed using analysis of covariance (ANCOVA) for categorical predictors and partial correlation for continuous predictors. As a significant effect of age was found for some outcomes, all models were controlled for age. Data was log-transformed to reduce the impact of skewed data. All analyses were performed with SAS v9.2 (SAS Institute, Cary, NC), and p<0.05 was considered statistically significant. By design, our study cohort included subjects whose function fell outside the range of MHFMS detection, and thus were scored the constrained maximum or minimum MHFMS value. Because ceiling and floor values will increase (positive) or decrease (negative) slopes of identified associations we performed the statistical analysis of SMN transcripts and protein to MHFMS comparisons by including and excluding those with border values.
Quantification of SMN1 and SMN2 copy number was conducted at Expression Analysis Inc., Durham, North Carolina by quantitative real-time Taqman PCR (qPCR)
Four specific transcripts were measured: (1) SMN2-Full Length (SMN2-FL), (2) SMN1-Full Length (SMN1-FL), (3) SMN transcript lacking exon 7 (SMN-Δ7), and (4) GAPDH, a housekeeping gene commonly used for normalization in SMN transcript analysis. Three combinations of these primary measurements were also considered of potential relevance: (5) Total SMN-FL (SMN1-FL+SMN2-FL), (6) Total SMN (SMN-FL+SMN-Δ7), and (7) a ratio of SMN2 transcripts that reflects exon 7 inclusion (SMN2-FL/SMN-Δ7). As SMN1-FL is only found in control subjects, the values of SMN2-FL and Total SMN-FL are the same in SMA subjects.
SMN2-FL and SMN1-FL levels in whole blood were evaluated by absolute real-time PCR
A total of 127 samples (SMA 105, control 22) were available for SMN ELISA measurements. Whole blood was collected into EDTA K2 tubes and approximately 4 mL was poured into CPT vacutainers (#362760 BD, Franklin Lakes, NJ). PBMCs were isolated by centrifugation at 1500 rpm within 2 hours of collection. PBMC samples were shipped at ambient temperature to PPD for further processing and frozen storage. Frozen samples were transferred to Enzo Life Sciences (Ann Arbor, MI) for SMN analysis. PBMCs were thawed in a 37°C water bath and viable cell hemocytometer counts, performed immediately prior to lysis, were used to determine the appropriate volume of lysis buffer (LB-11) needed to establish a consistent concentration of the cell suspension of 108 cells per milliliter. LB-11 containing 300 mM NaCl, 10% glycerol, 3 mM EDTA, 1 mM MgCl2, 20 mM b-glycerophosphate, 25 mM NaF and 1% Triton X-100 was used along with protease inhibitors PIC8340 (Sigma #P8340, St. Louis, MO) and phenylmethylsulphonyl fluoride (Sigma #P7626). The cell suspension was gently vortexed and placed on ice for 30 minutes. The cell lysate was transferred to a 1.5 mL centrifuge tube and was clarified by centrifugation for 10 minutes at 14,000 RCF, 4°C. The supernatant was transferred to a clean vial and either assayed immediately or stored at –70°C until use.
Recombinant human SMN1 was generated from full-length cDNA expressed in bacterial expression vectors and purified for use as a standard in the ELISA. The capture antibody Sigma anti-SMN clone 2B1 (#S2944) was coated at 100 uL onto Costar Stripwell (#92592, Lowell, MA) at 3.5 ug/mL. After overnight incubation at room temperature, the plate was blocked for 5 hours with 1% BSA in PBS. Cell lysate samples and recombinant hSMN1 or HeLa cell lysate standards were loaded at 100 uL per well. Standards were diluted in 2-fold dilutions or from 0.0625-4 ng/mL. Samples were incubated for one hour at room temperature, washed and then incubated with a detection antibody from ProteinTech (#11708-AP-1, Chicago, IL) at 2 ug/mL for one hour at room temperature. After washing, a peroxidase conjugated goat anti-rabbit IgG from Jackson Immunolabs (5-#035-144, West Grove, PA) was applied at 50 ng/mL to the plate and incubated for 30 minutes at room temperature. After washing, plates were developed with TMB substrate for 30 minutes at room temperature and the reaction stopped after 30 minutes with 1N HCl acid. Plates were then read on a spectrophotometer at 450 nm. Plates were sealed and gently shaken during all incubations, and dilutions of sample and standard were done in assay buffer (1% BSA, 0.1% Tween-20 in PBS).
A total of 130 subjects were enrolled, including 17 Type I, 49 Type II, 42 Type III SMA, and 22 healthy control subjects. Enrollment figures slightly exceeded the original targets for each group due to replacement, per protocol, of subjects in whom specimens were of insufficient quantity (n = 7) or quality (n = 3). All 18 study sites contributed at least one subject over the course of 18 weeks concluding in March 2009. The age and gender distribution of those enrolled is presented in the second table of companion paper, Finkel et al.
Scores for the MHFMS were well-distributed by age across the enrollment cohorts. It should be noted that not all control individuals achieved a score of 40 on the scale, while all Type I SMA subjects were assigned scores of zero in the assessment.
SMA Type IN = 17 | SMA Type IIN = 49 | SMA Type IIIN = 42 | ControlN = 22 | p-value |
p-value |
|
|
0.14 | 0.84 | ||||
mean (SD) | 5.70 (3.54) | 6.55 (3.40) | 7.51 (3.11) | 6.95 (3.29) | ||
median [range] | 4.03 [2.39–12.66] | 6.49 [2.16–12.98] | 7.42[2.37–12.95] | 6.02 [2.16–12.59] | ||
0.73 | 0.64 | |||||
male | 10 (58.8%) | 26 (53.1%) | 20 (47.6%) | 10 (45.5%) | ||
female | 7 (41.2) | 23 (46.9%) | 22 (52.4%) | 12 (54.6%) | ||
|
<0.001 | <0.001 | ||||
mean (SD) | 0 (0) | 14.02 (10.55) | 34.1 (10.0) | 39.8 (0.7) | ||
median [range] | 0 [0–0] | 11 [0–36] | 40 |
40 |
ANOVA for continuous variables; Fisher exact test for categorical variables.
Legend: There were no significant differences in age or gender across the recruitment cohorts. The Modified Hammersmith Motor Function Scale differentiated between SMA subjects and controls and between Type I, II and III subjects.
Copy number of SMN2 varied in the SMA subjects from 1 to 6, and distributed with SMA type in a manner consistent with previous experience
SMN2 copy number is lower in Controls than it is in SMA subjects, in whom copy number is related to type.
SMA Type I | SMA Type II | SMA Type III | Control | |
|
|
|
|
|
mean (SD) | 2.5 (0.9) | 3.0 (0.7) | 3.5 (0.8) | 1.7 (1.1) |
median [range] | 2.1 [1.8–5.2] | 2.9 [1.8–5.2] | 3.2 [2.2–5.9] | 1.3 [0.0–4.2] |
|
|
|
|
|
mean (SD) | 10.1 (4.8) | 14.0 (7.3) | 14.7 (10.0) | 25.2 (19.6) |
median [range] | 10.6 [2.3–19.4] | 12.8 [1.6–33.1] | 12.4 [3.1–45.8] | 18.8 [5.0–86.2] |
|
|
|
|
|
mean (SD) | 38.3 (16.3) | 49.0 (18.6) | 58.1 (20.0) | 51.5 (22.3) |
median [range] | 32.1 [19.1–70.2] | 48.9 [17.3–94.8] | 54.2 [31.8–111.2] | 55.6 [15.5–91.7] |
|
|
|
|
|
mean (SD) | 38.3 (16.3) | 49.0 (18.6) | 58.4 (19.7) | 134.1 (65.1) |
median [range] | 32.1 [19.1–70.2] | 48.9 [17.3–94.8] | 54.2 [34.0–111.2] | 129.9 [40.9–259.8] |
|
|
|
|
|
mean (SD) | 140.6 (69.8) | 202.6 (73.0) | 200.4 (97.5) | 115.2 (64.2) |
median [range] | 133.5 [65.5–320.8] | 212.0 [54.4–316.9] | 188.9 [52.8–400.0] | 121.3 [16.2–236.9] |
|
|
|
|
|
mean (SD) | 178.8 (76.5) | 251.6 (82.1) | 258.8 (110.5) | 249.4 (98.3) |
median [range] | 163.1 [95.7–361.3] | 278.2 [78.3–367.8] | 247.4 [95.5–475.4] | 267.6 [63.9–448.8] |
|
|
|
|
|
mean (SD) | 0.31 (0.13) | 0.26 (0.09) | 34.1 (0.16) | 0.44 (0.15) |
median [range] | 0.33 [0.13–0.52] | 0.27 [0.09–0.45] | 0.28 [0.14–0.81] | 0.42 [0.22–0.68] |
|
|
|
|
|
mean (SD) | 4937.0 (1900.1) | 3662.0 (1870.4) | 3165.1 (757.5) | 4251.5 (2079.1) |
median [range] | 4535.0 [2428.6–8180.0] | 3298.5 [1450.0–7860.0] | 3054.5 [1580.0–4520.0] | 3570.0 [2230.0–9330.0] |
Mean and median values of SMN transcripts, protein and copy number for SMA Type I, II, III and Controls. Transcript values are presented as number of molecules per ng of total RNA (mol/ng).
Deviations from the usual relationship between copy number and SMA type further confirm that SMN2 copy number alone does not predict individual phenotype severity. SMN2 copy number was significantly lower in control subjects (p<0.001), with values similar to other reports
Subjects with each SMA type are broadly distributed across the age range, with the exception of type I SMA for whom there is some bias to younger age. As a consequence, SMN2 copy numbers are also broadly distributed. Values have been plotted with a small y-axis offset to avoid overlap of values.
In a regression model controlling for SMA diagnosis and type, SMN2-FL (p = 0.007), SMN-FL (p = 0.008) and SMN-total (p = 0.036) transcript levels decrease with age. SMN-Δ7 and GAPDH transcript levels did not change significantly with age (
The relationship of SMN transcripts to the three SMA types and control groups is presented in
A,C: SMN2-FL, SMN-FL and Total SMN transcripts generally increase with SMA Type. SMN-FL (A) is a sum of SMN1-FL (present only in healthy controls) and SMN2-FL. B: SMN-Δ7 expression levels are lower in Type I patients compared to other SMA Types but they are similar to that of Controls. D: Ratios of SMN2-FL to SMN- Δ7 differ between SMA Types and Controls, however differences between SMA Types are absent with the exception of Type II versus Type III patients. E: GAPDH transcript levels are elevated in SMA Type I and Controls relative to Type II and III patients.
A noteworthy finding is that within the SMA subject group there was no association of SMN2-FL, (Total SMN-FL) or any other transcript levels to SMN2 copy number. The only significant association between any SMN transcript and SMN2 copy number is that of SMN-Δ7 assessed in the whole study group combining SMA and control subjects (r = 0.31, p = 0.003).
SMN2-FL and SMN-Δ7 transcript levels in SMA subjects showed some modest relationships to MHFMS score when the transcripts were assessed in all SMA subjects, including those with a maximum or minimum MHFMS score value (r = 0.34, p = 0.009 for Type II+III only for SMN2-FL; r = 0.60, p = 0.0001 for Type III only for SMN-Δ7). When the analysis was confined to the 67 subjects having non-maximum or minimum MHFMS scores – to address the possible confounders associated with floor and ceiling effects – the relationship between SMN2-FL and SMN-Δ7 transcripts and MHFMS was lost.
As expected, SMN protein was significantly lower in SMA samples compared to control (
While SMN protein levels are lower in SMA relative to Control subjects, protein levels by SMA type are not statistically different from each other.
This is the first report in which SMN2 copy number, and transcript and protein measures of SMN expression in blood using the quantitative assay methodologies that reflect the current consensus in the field, have been applied prospectively to a large and broad cohort of subjects with SMA.
SMA group vs. control | Between SMA types | |
|
Yes | Yes |
|
No | Yes |
|
Yes | No |
The “Between SMA types” column includes the following comparisons: Type I vs Types II+III, Type I vs Type II, Type I vs Type III, Type II vs Type III.
SMN2 copynumber | SMN2-FLtranscript | MHFMS | |
|
N/A | No | No |
|
No | N/A | No |
|
Yes |
No | No |
The “SMN2 copy number” and “SMN-FL transcript” columns include the partial correlation analysis for SMA+Controls, SMA (all SMA patients), Type I, Type II, Type III, Controls. The “MHFMS” column includes the analysis for Type II, Type III and Type II+Type III with exclusion from the analysis of subjects having minimum or maximum scores.
Correlation was only found for the SMA group (r = 0.33, p = 0.001) and Type II group (r = 0.41, p = 0.008).
Our findings confirm the data from previous blood sample studies that show lower SMN-FL transcripts and protein in the most severely affected subjects with SMA Type I
There was no relationship between any measure of SMN expression and motor function as measured by the MHFMS within the subset of subjects having non-maximum or non-minimum scores. One important caveat to these findings is that this functional scale was designed for children with SMA Type II. In this study a MHFMS score was given to all subjects (SMA Type I assigned a score of “0”), but only subjects with SMA Type II and those with Type III SMA who are no longer able to walk achieved scores that were inside the minimum and maximum boundary values of the scale. There is, at present, no validated instrument of functional motor assessment that encompasses the entire range of SMA into a single scale. Comparisons of blood derived measures of SMN expression to subject motor function were thus necessarily constrained to a subgroup of subjects. Previous findings
SMN2 copy number is grossly related to the disease severity but has only modest predictive value for individual patients. SMN transcript levels were reduced in patients compared to controls in a manner graded by SMA type. The differences in SMN-FL abundance relative to SMA type (
A notable finding is that SMN2-FL (Total SMN-FL) transcript abundance correlates with SMA type, but not with SMN2 copy number, but SMA type and SMN2 copy number are themselves highly correlated. A similar lack of correlation between SMN transcript and SMN2 copy number was found previously
The lack of strong correlation between SMN protein levels, and SMN2 copy number or transcript levels in SMA patients suggests that there may be significant modulation of SMN at the posttranscriptional level. This regulation could be different for different cell types and tissues and it may also vary by age. Moreover, it should be pointed out that PBMCs may not fully reflect SMN expression levels in more disease relevant cells like motor neurons, muscle, or both. There is a large body of evidence that suggests that other genetic modifiers of SMN can significantly impact SMA phenotype in animal models and humans
The absence of an association of SMN transcripts or protein to motor function as assessed by MHFMS might be expected, as SMA likely progresses due to downstream cellular consequences of diminished SMN abundance. This is a robust negative finding, as the power of this study cohort was aided by the successful recruitment of a wide and evenly distributed range of SMA type, SMN genotype, motor function and age. Nevertheless, these findings do not preclude further exploring measures of SMN expression in blood as a biomarker of early pharmacodynamic assessment of
Further studies are also needed to evaluate the potential value of blood-derived measures of SMN expression to SMA therapeutics. The first will be to understand better its performance as a clinical measure. An important extension of this initial cross-sectional analysis will be a longitudinal analysis of individual patients, across the spectrum of SMA type and age, to evaluate the stability of measures of SMN expression over time. The extent to which measures of SMN transcript or protein values in blood are stable in repeated within-subject measurements, as suggested by preliminary findings
Dr. Crawford serves on the medical and scientific advisory boards of Families of SMA, the medical board of the Muscular Dystrophy Association, and has served as frequent ad-hoc advisor to the Scientific Advisory Board of the SMA Foundation. He has received research support for clinical studies from Families of SMA, SMA Foundation, and the Ataxia Telangiectasia Children’s Project.; Dr. Paushkin is an employee of the SMA Foundation.; Dr. Kobayashi is an employee of the SMA Foundation.; Ms. Forrest was an employee of the SMA Foundation during the time of the study.; Ms. Joyce was an employee of the SMA Foundation during the time of the study.; Dr. Finkel receives commercial research support from PTC Therapeutics, Santhera Pharmaceuticals, and Genzyme Corp. and non-profit research support from the SMA Foundation and support from the NIH/NIAMS and NIH/NINDS, accepted travel stipends as part of grants from PTC Therapeutics and the SMA Foundation, reviewed and prepared a report for Adibi legal proceeding, spends 50% of his professional time carrying out clinical studies, and serves on the medical advisory board of DuchenneConnect and Families of SMA and on the scientific advisory board of PTC Therapeutics. A member of his immediate family receives commercial research support from Merck Pharmaceuticals, has received license fee payments from Southern Biotechnology Associates, Upstate Pharmaceuticals, and Santa Cruz Biotechnology, receives research funding from the NIH (grant #s: AR058606, 1R21 AI078387, 1R21 AI078387-S1, 1R41 AI071927, R01 AI063623, 1U19AI082726, T32; pending grant #s: 1R21, AR059466, T32, AR059650), holds 6 patents or pending patents, contributes to other clinical research as a local co-investigator or PI in studies funded by the NIH and the UK, is the editor of Janeway Textbook of Immunology and Arthritis Research and Therapy, and devotes 33% of her professional time to clinical studies in her practice.; Dr. Kaufmann is an employee of the federal government and has no disclosures. Prior to August 2009, Dr. Kaufmann was an employee of Columbia University and received research support from the NIH, the SMA Foundation, PTC, Santhera Pennwest and the DoD. This report is based on Dr. Kaufmann’s work at Columbia University and is not related to the National Institutes of Health.; Dr. Swoboda serves on the scientific advisory boards of Families of SMA, the Pediatric Neurotransmitter Disorders Foundation, California Stem Cell, Inc., and the Alternating Hemiplegia of Childhood Foundation (AHCF). She serves as an ad-hoc reviewer for the Muscular Dystrophy Association (MDA) and NIH. She has accepted research funds for consultation for Biomarin Pharmaceuticals and Shire, Inc. She receives or has received in the past year grant funding from Families of SMA, MDA, FightSMA, AHCF and NIH (R01-HD054599; ARRA 5-R01 HD054599-04, and 1-R01-HD69045 from NICHD).; Dr. Tiziano reports no disclosures.; Dr. Lomastro reports no disclosures.; Dr. Li is employed by New England Research Institutes (NERI). NERI was paid by the sponsoring organization of this study to coordinate this study and perform additional analyses.; Dr. Trachtenberg is employed by New England Research Institutes (NERI). NERI was paid by the sponsoring organization of this study to coordinate this study and perform additional analyses.; Dr. Plasterer was employed by BG Medicine (BGM) from July 2001-March 2010. BGM was paid by the sponsoring organization of this study for sample and statistical analysis.; Dr. Chen is an employee of the SMA Foundation.; This does not alter our adherence to all the PLoS ONE policies on sharing data and materials.
(TIF)
(DOCX)
(DOC)
(DOC)
(DOC)
Use of the “Modified Hammersmith Motor Scale©” was by permission of the Families of Spinal Muscular Atrophy. This scale was developed by Kristin J. Krosschell (Northwestern University), Jo Anne Maczulski (Independent Consultant), Thomas O. Crawford (Johns Hopkins University), Dr. Charles Scott (Children’s Hospital of Philadelphia), Dr. Kathryn Swoboda (University of Utah) and other clinicians of the Project Cure SMA Clinical Trial Network, with funding from Families of SMA.
This study acknowledges the collaboration of the International Spinal Muscular Atrophy Patient Registry (Indiana University), supported by the Patient Advisory Group of the International Coordinating Committee for SMA Clinical Trials which includes Families of SMA, FightSMA, Muscular Dystrophy Association, SMA Foundation, and other SMA advocacy groups.
Authors extend their appreciation to SMA Foundation members Brett Chung for assistance with data analysis, Dr. Maryann Martone for comments and discussion on the manuscript, and Kathleen McCarthy and Mackensie Yore for assistance in editing the manuscript.
The authors also thank Families of SMA, FightSMA, and the Muscular Dystrophy Association for their help in subject recruitment, Margaret Bell (NERI) for acting as Project Manager for the BforSMA Study, Kiara Donahoo (Enzo Life Sciences) for analysis of PBMCs by ELISA, Drs. David Jacoby (Repligen Corporation) and Robert McBurney (BG Medicine) for their contributions to the study design, Drs. Wendy Chung (Columbia University) and Louise Simard (University of Manitoba) for SMN copy number analysis, and Dr. Mustafa Sahin (Children’s Hospital, Boston) for his service as Medical Monitor of the study.
Lastly, the authors thank the co-investigators and evaluators of the Pilot Study of Biomarkers for Spinal Muscular Atrophy (BforSMA) Trial Study Group (List S1) for conducting the clinical study at their respective clinical sites.