Statistical Analysis Plan (Final)

Title:

Trial Phase

II

PPD

PPD

PPD

1 Signature Page

BioStata Representative:

Biostatistician:

Merck Representative:

PPD

Medical Responsible:

Global Drug Safety Responsible:

Lead Statistical Programmer:

Biomarkers Lead:

4 Modification History

5 Purpose of the Statistical Analysis Plan

The purpose of this Statistical Analysis Plan (SAP) is to document technical and detailed specifications for the final analysis of data collected for protocol EMR700692_006. Results of the analyses described in this SAP will be included in the Clinical Trial Report (CTR). Additionally, the planned analyses specified in this SAP will be included in regulatory submissions or future

manuscripts. Any additional analyses performed to provide results for inclusion in the CTR, but not specified in this SAP, will be clearly identified in the CTR.

6

Summary of Clinical Trial Features

7 Sample Size/Randomization

1. A mean change from baseline at 2 years:

• -0.03 mm in the placebo group (i.e., 0 mcg over 2 years)
• -0.01 mm in the sprifermin 30 mcg once a year group (i.e., 180 mcg over 2 years)
• +0.01 mm in the sprifermin 30 mcg twice a year group (i.e., 360 mcg over 2 years)
• +0.01 mm in the sprifermin 100 mcg once a year group (i.e., 600 mcg over 2 years)
• +0.03 mm in the sprifermin 100 mcg twice a year group (i.e., 1200 mcg over 2 years)

2. A common SD (standard deviation) of 0.10 mm
3. A type-one error, , of the primary efficacy analysis set at two-sided
4. Normal distribution.

Figure 1 Expected Mean Changes in Cartilage Thickness (mm) in Total Knee at 2 Years, SD=0.10

• 55 evaluable subjects in each group (total 275 subjects) will ensure a power of for detecting a linear dose relationship using a linear trend test,
• 76 evaluable subjects in each group (total 380 subjects) will ensure a power of for detecting an overall treatment effect (meaning that at least one group differs from the other groups) as well.

8 Overview of Planned Analyses

• Reporting of the DBPC treatment phase: After the last subject's last visit of the DBPC treatment phase, the data collected until the end of the DBPC treatment phase will be cleaned and released, a Blind Data Review Meeting will be held to clarify potential remaining issues before the database can be partially locked (this terminology is used here to differentiate data coming from the DBPC treatment phase vs. extended follow-up phase, but all processes for formal database lock will be followed); if all participants of the Blind Data Review Meeting are in agreement
  that the database can be partially locked, then the database will be partially locked, the treatment code will be un-blinded, complete analyses of the DBPC phase will be performed. All data that is collected until week 104 (day 819:see table 3 for definition of target day and allowed windows) will be included in the statistical analysis of the partially locked database for the DBPC treatment phase. Each dataset will include variable to flag if a record will be included in analyses for the DBPC treatment phase. If records belong to days they will be part of the long term follow-up analysis. Further description of database lock handling appears in the Data Management plan. The SAP has to be approved prior to database partial lock and breaking the blind. This will be referred to as 2 -year analysis.
• Long term follow-up data will be analyzed and reported in addition to the DBPC treatment phase results. This will be referred to as 5 -year analysis and will include data from baseline (baseline, day 1) until end of extended follow-up period (Year 5). To minimize bias during the extended follow-up phase, all site staff including Investigators, study monitors, X-ray and MRI readers, and subjects will remain blinded during the extended follow-up phase. Subjects will maintain their subject number from treatment phase to follow-up. X-ray and MRI readers will remain blinded during the study and after the study completion.

Study Design

• In addition, a Year-3 analysis, repeating Year-2 analysis for primary and secondary efficacy endpoints as well as safety endpoints is planned.

8.1 Analysis of data from DBPC treatment phase

8.2 Analysis of data from the follow-up phase

8.3 Analysis of safety data during conduct of the trial

8.4 Interim analysis

9 Changes to the Planned Analyses in the Clinical Trial Protocol

Deletion:

Clarifications:

1. The protocol is not specific on what deviations are considered major and resulting in subjects who are excluded from the PP Analysis (Section 8.4 of the protocol). Section 10 provides more details for definition of major protocol deviations and the process to identify them.
2. The protocol is not specific on how data will be included in the extended follow-up period analysis (Section 8.5.1 of the protocol). Section 8.2 provides more details on what data will be included in this analysis. It also specifies that all data from the 5 years period will be included in this integrated analysis.
3. Clarification on time points, scored regions, scoring system and analyses for Whole Organ Magnetic Resonance Imaging Score (WORMS) and (Boston-Leeds Osteoarthritis Knee Score) BLOKS are provided in Section 16.5 (Section 8.5.5 of the protocol).

Additions:

1. Analysis sets. The following additional analysis set is specified in Section 10. For the long term follow-up phase (following day 819), a separate analysis set is defined consisting of subjects from the ITT population who have at least one post-treatment quantitative MRI assessment or at least one WOMAC measurement in the follow-up period, i.e., the Second ITT (ITT2). The ITT2 Analysis Set will include all subjects from the ITT Analysis Set who have at least one MRI assessment available or one X-ray assessment available or one WOMAC measurement on or after day 820 .
2. Year-3 Analysis: A Year-3 Analysis is planned and described in Section 8.
3. Sensitivity analysis using LOCF. The LOCF method as imputation for missing data on post baseline observations will be used as sensitivity evaluation with potential to be utilized in future studies. This is described in Section 11.
4. Adverse events of special interest: Injection site reactions, Neoplasms, Musculoskeletal disorders will be identified using pre-specified rules and will be summarized (as specified in Section 17.1).
5. Key efficacy endpoints will be summarized by baseline biomarker subgroups (as specified in Section 16.4) and by clinical subgroups. Some of the subgroups (as specified in Section 10, last part), were not pre-specified in the protocol. For countries with low accrual, pooling of subjects will be done by similarity. As of September 2014 (when study was fully enrolled) there were only 11 randomized subjects from the USA and one from Poland. Therefore, for the primary analysis of ANOVA, Poland will be pooled with the smallest Eastern European country, Romania. Czech Republic and Estonia, as well as Hong Kong, are counted separately, and the USA will be pooled with Denmark as USA+ Western Europe. For subgroup analyses there will be 3 regions: (1) Eastern European, (2) USA+ Western Europe, and (3) Hong Kong. This is specified in sections 10 and 16.1. Among these subgroups, 3 subgroups have been identified to have higher priority and described in Section 10).
6. Responder variables: subjects with at least 10 -point improvement on WOMAC Total score, and subjects with change from baseline in total cartilage thickness and their combination (as specified in Section 16.5) are added.
7. Spearman's correlations between endpoints related to function and symptoms (e.g. WOMAC pain, WOMAC function, ICOAP, 20 meter walk test, KOOS symptoms, KOOS QoL, etc.) and structure related endpoints will be evaluated as specified in Section 16.5.

10 Analysis Sets

Screening Analysis Set

Intent-to-Treat Analysis Set (Full Analysis Set)

Modified Intention-to-Treat Analysis Set

Per Protocol Analysis Set

• Compliance with all entry criteria (fulfilling all inclusion criteria and no exclusion criteria)
• Absence of major clinical trial protocol deviations with respect to factors likely to affect the efficacy of treatment, such as non-missing and valid baseline and post-baseline measurement of the primary efficacy endpoint (quantitative MRI), where the nature of such deviations will be defined before or at the Blind Data Review Meeting (before breaking the blind) for the reporting of the treatment phase, in collaboration between the medical, clinical and statistical responsible person (see also Section 8).

- PPD

• PPD

Second Intention-to-Treat Analysis Set

Safety Analysis Set

Pharmacokinetic Analysis Set

PGX Analysis Set

Analysis of Subgroups for

Key subgroup prioritization:

• KL grade: Grade 2, Grade 3 (prognostic factor based on literature, and OAI anlysis)
• Cartilage metabolism biomarker at baseline: urinary CTX-II (creatinine adjusted), categories defined in section , identified in a post hoc analysis of the PoC study as a potential predictive biomarker)
• Joint tissue inflammatory biomarker at baseline: serum CRPM, categories defined in section 16.4 ( , identified in a post hoc analysis of the PoC study as a potential predictive biomarker)

KL grade (Grade 2, Grade 3 separately),
sex (Male, Female),
age ( \(<65\) years old, \(\geq 65\) years old),
body mass index (BMI; \(<30 kg / m 2, \geq 30 kg / m 2\) ),
pooled country (Eastern Europe [including Poland], US + Western EU [Denmark], and Hong
Kong),

Disease status at baseline (based on X-ray):

1. bilateral/unilateral ( OA in both knees/ OA in one knee only)
2. baseline medial mJSW in target knee ( )
3. baseline mal-alignment in target knee (<-3 or > 3 or [ ] degrees)

Pain medications [classified for 3 timing groups defined in Section 14]:

1. Pain medications (Yes, No) - any of 2, 3 or 4 defined below
2. Analgesics, Opiods in WHO ATC code group N02A (Yes, No)
3. Analgesics, Other analgesics and antipyretics in WHO ATC code group N02B (Yes, No)
4. Non-steroid anti-inflammatory and antirheumatic products in WHO ATC code group M01A (Yes, No)

Other subgroups:

11 General Specifications for Statistical Analyses

Summary Statistics

Baseline

For vital signs, which are measured at all visits, two mappings will be used, one relative to injections and one relative to target days.

Adjustment for Multiplicity

Missing Data

The number of subjects contributing data will decrease over time. In summary assessments by time-point within DBPC, the LOCF method on post baseline observations will be used as sensitivity evaluation, and will be defined as LOCF1 for the 2-year analysis, and LOCF2 for the

analysis of the extended follow-up. These will only be included in summary tables (not in proc mixed models) and are considered exploratory and will be utilized for phase III statistical consideration. In the remainder of the SAP, if not otherwise stated, LOCF refers to both LOCF1 and LOCF2. Details for derivation of LOCF1 and LOCF2 for corresponding endpoints are provided in Table 3, 3a, and 3b.

Listing of data

ANOVA models

proc mixed data=p3 method=reml;
    class grpn vs usubjid CNTRGR1N ;
    model &var. = grpn vs grpn*vs base CNTRGR1N / solution;
    repeated vs / subject=usubjid type=&type. r;
    lsmeans grpn / adjust=dunnett diff=control('1') cl;
    lsmeans grpn*vs / adjust=dunnett diff=control('1' 'Week 26') cl;
    lsmeans grpn*vs / adjust=dunnett diff=control('1' 'Week 52') cl;
    lsmeans grpn*vs / adjust=dunnett diff=control('1' 'Week 78') cl;
    lsmeans grpn*vs / adjust=dunnett diff=control('1' 'Week 104') cl;
    lsmeans grpn*vs / slice=vs;
    contrast 'Linear trend (overall)' grpn -2 -1 0 1 2;
    contrast 'Linear trend at Week 26' grpn -2 -1 0 1 2 grpn*vs -2 0 0 0 -1 0 0
000001000200 0;
    contrast 'Linear trend at Week 52' grpn -2 -1 0 1 2 grpn*vs 0 -2 0 0 0 -1 0
0000001000 2 0 0;
    contrast 'Linear trend at Week 78' grpn -2 -1 0 1 2 grpn*vs 0 0 -2 0 0 0 -1
0000000 1 0 0 0 2 0;
    contrast 'Linear trend at Week 104' grpn -2 -1 0 1 2 grpn*vs 0 0 0 -2 0 0 0
-10000000 1 0 0 0 2;

12 Disposition of Subjects and Discontinuations

• Total number of subjects screened (i.e. subjects who signed informed consent)
• Number and percentages of subjects discontinuing the trial prior to randomization divided by main reason for discontinuation (did not meet all eligibility criteria, withdrew consent, other), by country and site, and overall.
• Number and percentages of randomized subjects.
• Number and percentages of exposed subjects
• Number and percentages of randomized subjects who completed the DBPC treatment phase. For the 2-year analysis a completer is a subject with no date of study discontinuation before day 819.
• Number and percentages of randomized subjects who discontinued the trial after randomization, during DPBC i.e. before day 819, and during extended follow up phase (grouped by treatment) and main reason (AE, death, lost to follow-up, etc).
• Number and percentages of randomized subjects who discontinued treatment after randomization, during DPBC and during extended follow up phase (grouped by treatment) and main reason (AE, death, lost to follow-up, etc).
• Number of subjects in each analysis set and percentages of total enrolled
• Number and percentage of subjects completing the extended follow up phase.

12.1 Protocol Deviations

13 Demographics and Other Baseline Characteristics

13.1 Demographics

13.2 Medical History

13.3 Other Baseline Characteristics

• Alcohol and nicotine consumption
• Specification of target knee OA: (KL Grade 2 or 3, target knee location (Left or Right knee), medial and lateral mJSW, Anatomic axis angle (AAA) of knee alignment [in degrees calculated as 180 - ANGLE, where ANGLE is the value corresponding to xmtestcd = 'ANGLE'], response to Question 1 of the WOMAC pain index at screening, and time since diagnosis
• Contralateral knee: KL Grade ( ), medial and lateral mJSW
• Pain in other joints (yes/no, and if yes, how many joints and which joints specifically)
• All other attributes defined in subgroups analysis set in Section 10.

14 Previous or Concomitant Medications/Procedures

15 Treatment Compliance and Exposure

• Duration of exposure date of last injection - date of first injection +1
• Total dose sum of (number of dose injections * )
• Listing of treatment exposure including time between injections within a cycle and time between last injection of previous cycle and first injection of next cycle will be provided.

• Mean number of injections by cycle and overall;
• Number of subjects (and percentage) receiving injections; and
• Number of subjects (and percentage) with treatment compliance. Subjects who have (actual total dose)/(planned total dose)*100 within 75-100% will be considered treatment compliant. Treatment compliance will not be summarized for the placebo group.
• Listing of subjects receiving wrong medication will be provided.

16 Endpoint Evaluation

16.1 Primary Efficacy Endpoint Analyses

1. Average Cartilage Thickness (Total Volume divided by Total Surface Area):
   (MFTC.VC+LFTC.VC)/(cMF.tAB+cLF.tAB+MT.tAB+LT.tAB)
2. Total Cartilage Thickness (sum of cartilage thickness in medial and lateral compartment):

Summary statistics

• Cartilage thickness in the femorotibial joint as evaluated by quantitative MRI
• absolute change from baseline in cartilage thickness by quantitative MRI
• percentage change from baseline in cartilage thickness by quantitative MRI

Primary analyses of linear dose-relationship and treatment effect

Sensitivity analysis using Per Protocol Analysis Set

Sensitivity analysis regarding fixed effects and covariates in the model

16.2 Secondary Efficacy Endpoint Analyses

Secondary endpoints related to structure:

Cartilage thickness

1. Medial
   a. MFTC.VC/(cMF.tAB+MT.tAB), primary
   b. MFTC_ThCtAB_aMe
2. Lateral:
   a. LFTC.VC/(cLF.tAB+LT.tAB), primary
   b. LFTC_ThCtAB_aMe

Cartilage Volume

Joint Space Width

Secondary endpoints related to function and symptoms:

WOMAC Total Score

WOMAC Sub Scores

WALK test

Patient Global Assessment (PGA)

16.3 Pharmacokinetic Endpoint Analyses

Pharmacokinetics (PK)

Pharmacokinetics based on concentrations measured in the serum

Pharmacokinetics based on concentrations measured in the synovial fluid

Occurrence of antibodies against FGF18 (FGF18ab)

16.4

Biomarker subgroups

PGX markers

Biochemical markers of joint tissue inflammation

Biochemical markers of cartilage metabolism

16.5 Other Endpoint Analyses

Outcome Measures in Rheumatology

proc genmod data=qs2 desc ;
    class grpn CNTRGR1N;
    model aval = grpn CNTRGR1N / dist = bin link = logit type3;
    estimate '30\mug x2 vs Placebo' grpn -1 1 0 0 0 / exp ;
    estimate '30\mug x4 vs Placebo' grpn -1 0 1 0 0 / exp ;
    estimate '100ug x2 vs Placebo' grpn -1 0 0 1 0 / exp ;

    estimate '100μg x4 vs Placebo' grpn -1 0 0 0 1 / exp ;
run ;

Patient Global Impression of Change (PGIC)

Knee Outcome in Osteoarthritis Score (KOOS)

• Symptoms (items S1 to S5);
• Quality of life (items Q1 to Q4).

The Medical Outcomes Survey Short Form 36 (MOS SF-36) Score

• general health (GH, based on 5 items [Q1,Q11a, Q11b, Q11c, Q11d]),
• physical function (PF, based on 10 items [Q3a-Q3j]),
• physical role limitations (RP, based on 4 items [Q4a-Q4d]),
• bodily pain (BP, based on 2 items [Q7, Q8]),
• vitality (VT, based on 4 items [Q9a, Q9e, Q9g, Q9i]),
• social function (SF, based on 2 items [Q6, Q10]),
• emotional role limitations (RE, based on 3 items [Q5a-Q5c])
• mental health (MH, based on 5 items [Q9b, Q9c, Q9d, Q9f, Q9h]).
  and a total score (based on 36 items [all above+Q2]).
  Each scale is directly transformed into a scale on the assumption that each question carries equal weight. All of the scales are scored so that the least health has a value of 0 and the greatest health has a value of 100 .

The Numeric Rating Scale (NRS) for pain score

Pain in other joints

The Intermittent and Constant OA Pain (ICOAP) score

• If a region is graded " -9 " (Unscorable) or " 0.1 " (Progression unscorable, coded as "NA" in the data), then the region's grade is set to missing. No imputation for missing data for WORMS will be done.
• If a region is graded " 0.5 " (Worsening (within grade), coded as "+") or "- 0.5 " (Improvement (within grade), coded as "-"), then the region's grade is set to the last available existing grade within the previous visits, for the purpose of calculating change from baseline on numeric scale.

Cartilage:

Meniscal Extrusion:

Meniscal Tear:

Baseline protein markers and/or genetic markers associated with response to treatment or disease progression (response assessed by MRI and/or questionnaire)

Relationship between dosing (dose and regimen), cartilage structure, and clinical scores as a function of time

Subgroup analysis

• Cartilage thickness in the total femorotibial joint
• Cartilage thickness in the medial and lateral compartments
• WOMAC total score
• WOMAC pain and function sub scores
• 20 meter walk test
• Lateral and medial JSW (mJSW) in target knee

Additional analyses for primary and secondary endpoints

17 Safety Evaluation

17.1 Adverse Events

17.1.1 All Adverse Events

• Overview of all AEs
• TEAE by SOC and PT
• TEAE by severity
• TEAE by relationship
• Local TEAE by SOC and PT
• Local TEAE by severity
• Local TEAE by relationship
• Systemic TEAE by SOC and PT
• Systemic TEAE by severity
• Systemic TEAE by relationship
• Any serious adverse events (SAEs) further specified in Section 17.2.2
• Any non-serious AEs (at least in any treatment group) by SOC and PT
• Most frequent TEAEs (at least in any treatment group) by PT only
• Treatment emergent Injection Site Reactions by PT only (as described in Section 17.1)
• Treatment emergent Musculoskeletal disorders by PT only
• Treatment emergent Neoplasms by PT only

17.1.2 Adverse Events Leading to Discontinuation

• TEAE leading to discontinuation
• Local TEAE leading to discontinuation
• Systemic TEAE leading to discontinuation

17.2 Deaths, Other Serious Adverse Events, and Other Significant Adverse Events

17.2.1 Deaths

17.2.2 Serious Adverse Events

• All SAEs
• All treatment emergent SAEs
• All treatment emergent SAEs leading to Death
• Local treatment emergent SAEs
• Systemic treatment emergent SAEs
• Serious TEAEs by relationship

17.2.3 Acute Inflammatory Reactions (AIRs)

• Pain VAS (last available value before i.a.injection)

• Number and percentage of subjects with self-reported synovial fluid effusion within 3 days following i.a. injection, for each cycle at Week 1, 2 and 3 (Yes, No);
• Number and percentage of subjects with increase of pain by 30 mm on a 100 mm VAS from pre-dose to post-dose within 3 days following i.a. injection, for each cycle at Week 1, 2 and 3 (Yes, No);
• Number and percentage of subjects experiencing an AIR, for Week 1, 2, 3 and cycle for each cycle (Yes, No).
• Number and percentage of subjects experiencing an AIR, overall (at any of the 12 injections) (Yes, No).
• Number and percentage of subjects experiencing single, multiple AIRs, or Not experiencing any AIR, overall (at any of the 12 injections).
• Number of total AIRs observed at each injection, cycle and overall (number of events, not subjects)

17.3 Clinical Laboratory Evaluation

• White blood cell (WBC) count
• Lymphocytes (absolute count, %)
• Monocytes (absolute count, %)
• Neutrophils (absolute count, %)
• Eosinophils (absolute count, %)
• Basophils (absolute count, %)
• Red blood cells
• Hematocrit
• Hemoglobin
• Platelets

• Aspartate aminotransferase (AST)
• Alanine aminotransferase (ALT)
• Alkaline phosphatase (ALP)
• Total protein
• Total bilirubin
• Creatine kinase (CK)
• Creatinine
• Calcium
• Phospate
• Sodium
• Potassium
• Magnesium
• Urea (Blood urea nitrogen (BUN))
• Glucose (not fasting)
• Chloride
• Albumin
• High-sensitivity C-reactive protein (hsCRP)
• Erythrocyte sedimentation rate (ESR)

• Dipstick parameters
• pH
• Specific gravity
• Protein
• Glucose
• Ketones
• Bilirubin
• Blood
• Urobilinogen
• Leukocytes

• Binding antibodies
• Neutralizing antibodies

• Partial thromboplastin time (PTT)
• International Normalized Ratio (INR)

• baseline vs largest post-baseline value
• baseline vs lowest post-baseline value

17.4 Vital Signs

• sitting systolic and diastolic blood pressure
• heart rate
• body temperature

• Maximal shifts (change in categories [increase and decrease]), this will be done both relative to injections and relative to target days following either table 3a or table 3b.

17.5 Physical Examination

17.6 ECG Evaluation

18 References

1. Tubach F , Ravaud P , Baron G et al. Evaluation of clinically relevant changes in patient reported outcomes in knee and hip osteoarthritis: the minimal clinically important improvement. Annals of Rheumatic Diseases. 2005; 64:29-33.
2. Independent Data Monitoring Committee Charter version 2, July 2014.
3. Attur, M. et al. Radiographic severity of knee osteoarthritis is conditional on interleukin 1 receptor antagonist gene variations. Annals of Rheumatic Diseases. 2010; 69, 856-861.
4. Wu X et al. IL-1 receptor antagonist gene as a predictive biomarker of progression of knee osteoarthritis in a population cohort. Osteoarthritis and Cartilage 2013; 21, 930-938.
5. Siebuhr et al., Identification and characterization of osteoarthritis patients with inflammation derived tissue turnover. Osteoarthritis and Cartilage 2014; (22) 44-50.
6. Sharif et al., A 5-yr longitudinal study of type IIA collagen synthesis and total type II collagen degradation in patients with knee osteoarthritis- association with disease progression. Rheumatology 2007;46: 938-943.
7. Kraus et al., Application of biomarkers in the development of drugs intended for the treatment of osteoarthritis: Osteoarthritis and Cartilage 2011; (19) 515-542.
8. Gudmann et al., Cartilage turnover reflected by metabolic processing of type II Collagen: A novel marker of anabolic function in chondrocytes. International Journal of Molecular Sciences. 2014; 15:18789-18803.
9. Pham T, van der Heijde D, Altman RD, Anderson JJ, Bellamy N, Hochberg M, et al. OMERACT-OARSI initiative: Osteoarthritis Research Society International set of responder criteria for osteoarthritis clinical trials revisited. Osteoarthritis Cartilage 2004; 12:389-99.
10. Medical Outcomes Study Short Form-36 General Health Survey User's Guide, version 2, 2000.
11. Measure of Intermittent and Constant Osteoarthritis Pain User's Guide, version 3, 2007.
12. Guermazi A, Roemer FW, Haugen IK, Crema MD and Hayashi D. MRI-based semi quantitative scoring of joint pathology in osteoarthritis. Nature Reviews Rheumatology. 2013; 9(4):236-51.
13. Altman RD, Gold GE. Atlas of individual radiographic features in osteoarthritis, revised. Osteoarthritis Cartilage 2007; 15 (Suppl A): A1-56.

19 Appendices

19.1 List of endpoints

Group	Endpoint	Treatment Phase	Follow-up (integrated)	Reference
		Baseline to LOCF1	Baseline to LOCF2
Trial Subjects	Disposition	Descriptive	Descriptive	12
	Protocol Deviations	Descriptive	Descriptive	12.1
Demographics and other baseline characteristics	Demographics	Descriptive	Descriptive	13.1
	Medical History	Descriptive	Descriptive	13.2
	Baseline Characteristics	Descriptive	Descriptive	13.3
Concomitant Medication and procedures	Previous medication and procedures	Descriptive	Descriptive	14
	Concomitant medication and procedures	Descriptive	Descriptive	14
	Pain medication	Descriptive	Descriptive	14
Exposure and treatment compliance	Exposure	Descriptive	Descriptive	15
	Compliance	Descriptive	Descriptive	15
Efficacy	Change from baseline in average cartilage thickness in the total femorotibial joint as evaluated by MRI in the target knee	Primary: Descriptive Analysis Sensitivity analysis (PP) Model check	Exploratory: Descriptive Analysis Model check	16.1
	WOMAC total score in the target knee at baseline and postbaseline	Secondary endpoint Descriptive Analysis Sensitivity analysis (PP) Subgroup analysis	Exploratory: Descriptive Analysis Model check	16.2
	WOMAC sub scores in the target knee at baseline and postbaseline	Secondary endpoint Descriptive Analysis Sensitivity analysis (PP) Subgroup analysis	Exploratory: Descriptive Analysis Model check	16.2

Group	Endpoint	Treatment Phase	Follow-up (integrated)	Reference
		Baseline to LOCF1	Baseline to LOCF2
	20-meter WALK test	Secondary endpoint Descriptive Analysis	Descriptive	16.2
	PGA	Secondary endpoint Descriptive	Descriptive	16.2
	JSW on X-ray in the target knee at screening or baseline - and posttreatment	Secondary endpoint Descriptive Analysis Sensitivity analysis (PP) Subgroup analysis	Exploratory: Descriptive Analysis Model check	16.2
	Regional Average Cartilage thickness on MRI	Secondary endpoints Descriptive Analysis	Descriptive	16.2
	Total and Regional Cartilage Volume	Secondary endpoint Descriptive Analysis	Descriptive	16.2
PK	Synovial fluid levels of sprifermin/FGF-18	Descriptive	NA	16.3
	Serum levels of sprifermin/FGF-18	Descriptive	NA	16.3
PK/Safety	Occurrence of antibodies	Descriptive	NA	16.3 and 17
Other endpoints	OMERACT-OARSI score for the target knee	Descriptive Analysis	Descriptive	16.5
	PGIC	Descriptive	Descriptive	16.5
	KOOS Qol Score	Descriptive	Descriptive	16.5
	KOOS Symptom index	Descriptive	Descriptive	16.5
	MOS SF-36 Score	Descriptive	Descriptive	16.5
	NRS Pain Score for the target and contralateral knee	Descriptive	Descriptive	16.5
	Pain in other joints	Descriptive	Descriptive	16.5

Group	Endpoint	Treatment Phase	Follow-up (integrated)	Reference
		Baseline to LOCF1	Baseline to LOCF2
	ICOAP Score for the target knee	Descriptive	Descriptive	16.5
	Semi quantitative MRI endpoints (WORMS/BLOKS) for the target knee	Descriptive Analysis	Descriptive	16.5
	Biomarkers	Descriptive for PGX and biochemical biomarkers. Other biomarkers will be defined in the BSAP [to be finalized prior to unblinding]		16.4
	Modeling	Will be defined in Modeling SAP
	Subgroup analysis (including biomarkers subgroups)	Descriptive Analysis		16.4 and 16.5
	Additional analyses (changes and in addition to those specified in the protocol)	Descriptive Analysis	Descriptive	16.5
Safety	Adverse Events	Descriptive	Descriptive	17.1
	Death	Descriptive	Descriptive	17.2
	AIR	Descriptive Analysis	NA	17.2.3
	Clinical laboratory parameters	Descriptive	Descriptive	17.3
	Vital signs	Descriptive	Descriptive	17.4
	Physical Examination	Descriptive	Descriptive	0
	ECG	Descriptive	Descriptive	17.6

19.2 List of permitted pain medication

19.4 Handling of TEAE timing

19.5 List of key statistical outputs

19.5.1 General Information Tables

1. Disposition of Patients by Treatment - ITT Analysis Set
2. Major and Minor Protocol Deviations - ITT Analysis Set
3. Demographics by Treatment - ITT Analysis Set
4. Baseline Target and Contralateral Knee Characteristics by Treatment - ITT Analysis Set
5. Treatment Compliance by Treatment - ITT Analysis Set
6. Exposure to Sprifermin - ITT Analysis Set

19.5.2 Efficacy Tables

1. Repeated Measurement ANCOVA on Absolute 2 Years Change from Baseline Average Cartilage Thickness in the Total Femorotibial Joint - mITT Analysis Set
2. Repeated Measurement ANCOVA on Absolute 2 Years Change from Baseline Average Cartilage Thickness in the Medial Femorotibial Compartment - mITT Analysis Set
3. Repeated Measurement ANCOVA on Absolute 2 Years Change from Baseline Average Cartilage Thickness in the Lateral Femorotibial Compartment - mITT Analysis Set
4. Repeated Measurement ANCOVA on Absolute 2 Years Change from Baseline Average Cartilage Thickness [mm] in the Total Femorotibial Joint Transformed as Ranked Data - mITT Analysis Set
5. Repeated Measurement ANCOVA on Absolute 2 Years Change from Baseline Average Cartilage Thickness in the Medial Femorotibial Compartment Transformed as Ranked Data - mITT Analysis Set
6. Repeated Measurement ANCOVA on Absolute 2 Years Change from Baseline Average Cartilage Thickness in the Lateral Femorotibial Compartment Transformed as Ranked Data - mITT Analysis Set
7. Summary Statistics by Visit, Relative and Absolute Change from Baseline in Cartilage Thickness [ mm ] of the Total Femorotibial Joint, by Treatment - mITT Analysis Set
8. Summary Statistics by Visit, Relative and Absolute Change from Baseline in Cartilage Thickness [mm] of the Medial Femorotibial Compartment, by Treatment - mITT Analysis Set
9. Summary Statistics by Visit, Relative and Absolute Change from Baseline in Cartilage Thickness [ mm ] of the Lateral Femorotibial Compartment, by Treatment - mITT Analysis Set
10. Repeat 1-3 and 7-9 for PP Analysis set.

X-ray mJSW:

11. Absolute Change (ANCOVA) from Baseline up to 2 years in in the Medial Femorotibial Compartment ITT Analysis Set
12. Absolute Change (ANCOVA) from Baseline up to 2 years in in the Lateral Femorotibial Compartment ITT Analysis Set
13. Summary Statistics by Visit, Relative and Absolute Change from Baseline in in the Medial Femorotibial Compartment, by Treatment - ITT Analysis Set
14. Summary Statistics by Visit, Relative and Absolute Change from Baseline in in the Lateral Femorotibial Compartment, by Treatment - ITT Analysis Set
15. Repeat 11-14 for PP Analysis set

WOMAC:

16. Summary Statistics by Visit and Change in WOMAC Total Score from Baseline to 2 Years, by Treatment - ITT Analysis Set
17. Summary Statistics by Visit and Change in WOMAC Pain Sub Score from Baseline to 2 Years, by Treatment ITT Analysis Set
18. Summary Statistics by Visit and Change in WOMAC Function Sub Score from Baseline to 2 Years, by Treatment - ITT Analysis Set
19. Absolute Change (ANCOVA) from Baseline up to 2 years in WOMAC Total Score - ITT Analysis Set
20. Absolute Change (ANCOVA) from Baseline up to 2 years in WOMAC Pain Sub Score - ITT Analysis Set
21. Absolute Change (ANCOVA) from Baseline up to 2 years in WOMAC Function Sub Score - ITT Analysis Set
22. Repeat 16-21 for PP Analysis set
23. Concomitant Pain Medication by Treatment - ITT Analysis Set

19.5.3 Safety Tables

1. Overview of AEs, by Treatment - Safety Analysis Set
2. Summary of TEAEs Leading to Discontinuation of Treatment - Safety Analysis Set
3. Summary of SAEs - Safety Analysis Set
4. Summary of Most Frequent TEAEs (at Least in Any Treatment Group) - Safety Analysis Set
5. Summary of Treatment Emergent Injection Site Reactions - Safety Analysis Set
6. Summary of Patient Reported Acute Inflammatory Reactions by Cycle and Overall, by Treatment - Safety Analysis Set

19.5.4 Figures

1. Mean and CI in Absolute Change from Baseline Over Visits by Treatment, Cartilage Thickness [mm] in the Total Femorotibial Joint - mITT Analysis Set
2. Mean and CI in Absolute Change from Baseline Over Visits by Treatment, Cartilage Thickness in the Medial Femorotibial Compartment - mITT Analysis Set
3. Mean and CI in Absolute Change from Baseline Over Visits by Treatment, Cartilage Thickness [ mm ] in the Lateral Femorotibial Compartment - mITT Analysis Set
4. Mean and CI in Absolute Change from Baseline Over Visits by Treatment, in the Medial Femorotibial Compartment - ITT Analysis Set
5. Mean and CI in Absolute Change from Baseline Over Visits by Treatment, in the Lateral Femorotibial Compartment - ITT Analysis Set
6. Mean and CI in Absolute Change from Baseline Over Visits by Treatment, WOMAC Total Score in the Target Knee - ITT Analysis Set
7. Mean and CI in Absolute Change from Baseline Over Visits by Treatment, WOMAC Pain Sub Score in the Target Knee - ITT Analysis Set
8. Mean and in Absolute Change from Baseline Over Visits by Treatment, WOMAC Function Sub Score in the Target Knee - ITT Analysis

19.5.5 Data Listings

1. Adverse Events Leading to Discontinuation of Treatment - Safety Analysis Set
2. Serious Adverse Events - Safety Analysis Set
3. AIR Reported in Subject Diary - Safety Analysis Set