E019321.Draft-Revisions.Pdf

BMJ Open: first published as 10.1136/bmjopen-2017-019321 on 29 December 2017. Downloaded from

BMJ Open is committed to open peer review. As part of this commitment we make the peer review history of every article we publish publicly available.

When an article is published we post the peer reviewers’ comments and the authors’ responses online. We also post the versions of the paper that were used during peer review. These are the versions that the peer review comments apply to.

The versions of the paper that follow are the versions that were submitted during the peer review process. They are not the versions of record or the final published versions. They should not be cited or distributed as the published version of this manuscript.

BMJ Open is an open access journal and the full, final, typeset and author-corrected version of record of the manuscript is available on our site with no access controls, subscription charges or pay- per-view fees (http://bmjopen.bmj.com).

If you have any questions on BMJ Open’s open peer review process please email [email protected]

http://bmjopen.bmj.com/ on September 30, 2021 by guest. Protected copyright. BMJ Open: first published as 10.1136/bmjopen-2017-019321 on 29 December 2017. Downloaded from BMJ Open

Efficacy and safety of Chimeric Antigen Receptor T- cell (CAR-T) therapy in patients with hematologic and solid malignancies: protocol for a systematic review and meta- analysis

For peer review only Journal: BMJ Open

Manuscript ID bmjopen-2017-019321

Article Type: Protocol

Date Submitted by the Author: 25-Aug-2017

Complete List of Authors: Grigor, Emma; University of Ottawa, School of Epidemiology, Public Health and Preventative Medicine; University of Ottawa , School of Epidemiology Fergusson, Dean; Ottawa Hospital Research Institute, Clinical Epidemiology ; University of Ottawa , School of Epidemiology and Public Health Haggar , Fatima; Ottawa Hospital Research Institute , Clinical Epidemiology ; Ottawa Hospital Research Institute , Blood and Marrow Transplant Program Kekre, Natasha; University of Ottawa , Faculty of Medicine ; Ottawa Hospital Research Institute, Cinical Epidemiology Program Atkins , Harold; University of Ottawa , Faculty of Medicine ; Ottawa http://bmjopen.bmj.com/ Hospital Research Institute , Blood and Marrow Transplant Program Shorr , Risa; Ottawa Hospital Holt , Robert ; Canada's Michael Smith Genome Sciences Centre Hutton, Brian; University of Ottawa, Ottawa, Ontario, Canada, Faculty of Medicine ; Ottawa Hospital Research Institute , Clinical Epidemiology Program Ramsay, Tim; The Ottawa Health Research Institute, Clinical Epidemiology Program ; University of Ottawa , School of Epidemiology and Public Health Seftel , Matthew; University of Winnipeg, Department of Hematology Jonker, Derek; Faculty of Medicine, University of Ottawa; Division of on September 30, 2021 by guest. Protected copyright. Medical Oncology, The Ottawa Hospital Daugaard , Mads; Vancouver Prostate Centre, Molecular Pathology & Cell Imaging Core Facility Thavorn, Kednapa; Institute for Clinical Evaluative Sciences, ICES @uOttawa; The Ottawa Hospital Research Institute, Presseau , Justin; Ottawa Hospital Research Institute, Clinical Epidemiology Program ; University of Ottawa , Faculty of Medicine Lalu, Manoj; Ottawa Hospital Research Institute, Clinical Epidemiology ; The Ottawa Hospital , Department of Anesthesiology and Pain Medicine

Primary Subject Oncology Heading:

Secondary Subject Heading: Haematology (incl blood transfusion), Pharmacology and therapeutics

Leukaemia < HAEMATOLOGY, Lymphoma < HAEMATOLOGY, Myeloma < Keywords: HAEMATOLOGY, Chimeric Antigen Receptor, CAR-T, Malignancy

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019321 on 29 December 2017. Downloaded from Page 1 of 38 BMJ Open

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31

32 http://bmjopen.bmj.com/ 33 34 35 36 37 38 39

40 on September 30, 2021 by guest. Protected copyright. 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019321 on 29 December 2017. Downloaded from BMJ Open Page 2 of 38

1 2 3 Efficacy and safety of Chimeric Antigen Receptor T-cell (CAR-T) therapy in patients with 4 5 6 hematologic and solid malignancies: protocol for a systematic review and meta-analysis 7 8 9 10 1,2,3,7 1,2,3,6 11 Emma J.M. Grigor, MSc (Candidate) , Dean Fergusson, PhD, MHA , Fatima Haggar, 12 3,4,5,7 1,3 1,4 13 MPH, PhD , Natasha Kekre, MD, MPH, FRCPC , Harold Atkins, MD , Risa Shorr, 14 15 MLIS9, FRCPCFor1,4,5, Robert peer A. Holt, PhD 10review, Brian Hutton, MSc, only PhD1,2,3, Tim Ramsay, PhD1,2,3, 16 17 11 1,14 18 Matthew Seftel, MD, MPH, MRCP, FRCPC , Derek Jonker, MD, FRCPC , Mads Daugaard, 19 20 MSc, PhD12, Kednapa Thavorn, MPharm, PhD2,3,13, Justin Presseau, MRes, PhD2,3, Manoj M. 21 22 Lalu, MD, PhD, FRCPC1,3,7,8 23 24 25 26 27 1Faculty of Medicine, University of Ottawa, Ontario, Ottawa, Canada; 28 29 2School of Epidemiology and Pubic Health, University of Ottawa, Ontario, Ottawa, Canada; 30 31 3

32 Clinical Epidemiology Program, Ottawa Hospital Research Institute, Ontario, Ottawa, Canada; http://bmjopen.bmj.com/ 33 34 4Blood and Marrow Transplant Program, Ottawa Hospital Research Institute, Ontario, Ottawa, 35 36 37 Canada; 38 5 39 Cancer Therapeutic Program, Ottawa Hospital Research Institute, Ontario, Ottawa, Canada;

40 on September 30, 2021 by guest. Protected copyright. 41 6Department of Surgery, University of Ottawa, Ontario, Ottawa, Canada; 42 43 7 44 Department of Anesthesiology, The Ottawa Hospital, Ontario, Ottawa, Canada; 45 46 8Regenerative Medicine Program, Ottawa Hospital Research Institute, Ontario, Ottawa, Canada; 47 48 9The Ottawa Hospital, Ottawa, Canada; 49 50 10 51 Michael Smith Genome Sciences Centre, British Columbia Cancer Agency, Vancouver, British 52 53 Columbia, Canada; 54 55 11Department of Hematology, University of Winnipeg, Manitoba, Canada; 56 57 58 59 60 1 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019321 on 29 December 2017. Downloaded from Page 3 of 38 BMJ Open

1 2 3 12Molecular Pathology & Cell Imaging Core Facility, Vancouver Prostate Centre, British 4 5 6 Columbia, Canada. 7 8 13Institute for Clinical and Evaluative Sciences, Ontario, Ottawa, Canada; 9 10 14 11 Division of medical oncology, The Ottawa Hospital, Ontario, Ottawa, Canada. 12 13 14 15 Emma J.M. GrigorFor [email protected] peer -review Dean Fergusson [email protected] only - Fatima 16 17 18 Haggar [email protected] - Natasha Kekre [email protected] - Harold Atkins [email protected] - Risa 19 20 Shorr [email protected] - Rob A. Holt [email protected] - Brian Hutton [email protected] - Tim 21 22 Ramsay Tim [email protected] - Matthew Seftel [email protected] - Derek Jonker 23 24 25 [email protected] - Mads Daugaard [email protected] - Kednapa Thavorn [email protected] 26 27 - Justin Presseau [email protected] 28 29 30 31

32 Dates of study: April 1, 2017 to November 1, 2017 http://bmjopen.bmj.com/ 33 34 Corresponding author: 35 36 37 Manoj M. Lalu 38 39 Ottawa Hospital - Civic campus

40 on September 30, 2021 by guest. Protected copyright. 41 1053 Carling Avenue, Room B307, Mail Stop 249 42 43 44 Ottawa, Ontario, Canada K1Y 4E9 45 46 Telephone: 613-761-4169 (Option #1) 47 48 Fax: 613-761-5209 49 50 51 Email: [email protected] 52 53 orcid.org/0000-0002-0322-382X 54 55 Word count (main text): 3931 56 57 58 59 60 2 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019321 on 29 December 2017. Downloaded from BMJ Open Page 4 of 38

1 2 3 4 5 6 ABSTRACT 7 8 Introduction: Patients with relapsed or refractory malignancies have a poor prognosis. 9 10 11 Immunotherapy with Chimeric Antigen Receptor T (CAR-T) cells redirects a patient’s immune 12 13 cells against tumor antigen. CAR-T cell therapy has demonstrated promise in treating patients 14 15 with several hematologicFor malignancies, peer including review acute B cell lymphoblasticonly leukemia and B cell 16 17 18 lymphomas. CAR-T cell therapy for patients with other solid tumors is also being tested. Safety 19 20 is an important consideration in CAR-T cell therapy given the potential for serious adverse 21 22 events, including death. Previous reviews on CAR-T cell therapy have been limited in scope and 23 24 25 methodology. Herein we present a protocol for a systematic review to identify CAR-T cell 26 27 interventional studies and examine the safety and efficacy of this therapy in patients with 28 29 hematology malignancies and solid tumors. 30 31

32 Methods and analysis: We will search MEDLINE, including In-Process and Epub Ahead of http://bmjopen.bmj.com/ 33 34 Print, EMBASE, and the Cochrane Central Register of Controlled Trials. Studies will be 35 36 37 screened by title, abstract, and full text independently and in duplicate. Studies that report 38 39 administering CAR-T of any chimeric antigen receptor construct targeting antigens in patients

40 on September 30, 2021 by guest. Protected copyright. 41 with hematologic malignancies and solid tumors will be eligible for inclusion. Outcomes to be 42 43 44 extracted will include complete response rate (primary outcome), overall response rate, overall 45 46 survival, relapse, and adverse events. A meta-analysis will be performed to synthesize the 47 48 prevalence of outcomes reported as proportions with 95% confidence intervals. The potential for 49 50 51 bias within included studies will be assessed using a modified Institute of Health Economics 52 53 tool. Heterogeneity of effect sizes will be determined using the Cochrane I2 statistic. 54 55 56 Ethics and Dissemination: The review findings will be submitted for peer-reviewed journal 57 58 59 60 3 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019321 on 29 December 2017. Downloaded from Page 5 of 38 BMJ Open

1 2 3 publication and presented at relevant conferences and scientific meetings to promote knowledge 4 5 6 transfer. 7 8 9 Registration: The protocol has been registered with the International Prospective Register of 10 11 Systematic Reviews (Awaiting registration number). 12 13 Word count (abstract): 300 14 15 For peer review only 16 STRENGTHS AND LIMITATIONS OF THIS STUDY 17 18 19 1. Our review will provide a comprehensive synthesis of the current literature and be the first 20 21 review of trials investigating chimeric antigen receptor T (CAR-T) cell therapy among 22 23 patients with solid tumors. 24 25 26 2. A major methodological limitation is that all eligible studies are expected to be single arm 27 28 and have no comparator group. This may increase risk of bias when interpreting results. 29 30 3. We provide a comprehensive plan to address limitations of meta-analysis of single arm 31

32 http://bmjopen.bmj.com/ 33 studies that includes use a modified Institute of Health Economics tool for assessing risk of 34 35 bias in single arm interventional studies. Our approach can serve as a model for future 36 37 38 systematic reviews assessing early-phase clinical data that is often single arm with no control 39

40 comparison. on September 30, 2021 by guest. Protected copyright. 41 42 43 44 45 INTRODUCTION 46 47 48 Among patients with relapsed and refractory malignancies, chimeric antigen receptor T (CAR-T) 49 50 cell therapy is a novel immunotherapy that has shown promise in both pre-clinical and early 51 52 clinical studies. This therapy allows for CARs directed against tumour associated antigens (e.g. 53 54 55 CD19, HER-2) to be introduced into a patient’s T cells; this serves to reprogram these cells to 56 57 target the patient’s tumor cells. A number of small clinical trials using anti-CD19 CAR-T cells in 58 59 60 4 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019321 on 29 December 2017. Downloaded from BMJ Open Page 6 of 38

1 2 3 hematologic malignancies have demonstrated sustained responses in patients with advanced 4 5 6 disease [1-5]. CAR-T cell therapy for solid malignancies has also identified a number of 7 8 potential cancer-specific targets and previous pre-clinical studies investigating efficacy and 9 10 11 feasibility show promise [6]. 12 13 14 15 In spite of some Forevidence of peer efficacy of CAR-T review cell therapy against only some malignancies, there are 16 17 18 a number of safety concerns that have been identified. Trials conducted by Juno Therapeutics 19 20 and Kite Pharma reported mortality among patients with hematologic malignancy treated with 21 22 CAR-T cell therapy [7, 8]. The Juno Therapeutics trial was closed after the death of five patients 23 24 25 from cerebral edema linked to CAR-T cell therapy [9]. Previous trials investigating CAR-T cell 26 27 therapy among patients with solid tumors have reported adverse events from treatment including 28 29 anaphylaxis and cardiac arrest [10]. Past studies have also reported other challenges in applying 30 31

32 CAR-T cell therapy to solid tumors, including a scarcity of tumor associated antigen targets, the http://bmjopen.bmj.com/ 33 34 potent immunosuppressive effects of the tumor microenvironment, and the limited trafficking of 35 36 37 CAR-T cells to tumor sites [11, 12]. 38 39

40 on September 30, 2021 by guest. Protected copyright. 41 Due to the variability and small size of clinical trials investigating CAR-T therapy there is a need 42 43 44 for a systematic review to evaluate its efficacy and safety. Although no systematic review 45 46 currently exists for CAR-T therapy among solid tumors, we have identified four publications that 47 48 self-identified as systematic reviews studying the efficacy and safety of CAR-T therapy for 49 50 51 patients with hematologic malignancies [13, 14-16]. However, there were significant limitations 52 53 in the scope and/or methodologies in these earlier reviews. Zhu et al. only considered CD19 54 55 targeted CAR-T therapies, did not report differences between adult and pediatric populations, 56 57 58 59 60 5 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019321 on 29 December 2017. Downloaded from Page 7 of 38 BMJ Open

1 2 3 and employed a non-systematic search strategy [14]. Anwer et al. only included allogeneic T- 4 5 6 cells, whereas most CAR-T cell therapy uses autologous cells [15]. Another systematic review 7 8 by Zhang et al. only considered CD19 CAR-T therapies and was published in journal controlled 9 10 11 by a predatory publisher [13, 17]. Finally, while titled as a systematic review, Holzinger et al. is 12 13 only a narrative review [16]. Given these limitations of previous publications, along with rapid 14 15 evolution of the ForCAR-T field, peer there is a need review for a current systematic only review, as we present here, 16 17 18 that adheres to rigorous, state-of-the-art methods and summarizes the findings among both solid 19 20 tumors and hematologic malignancies. 21 22 23 24 25 Our systematic review will clarify the determinants of efficacy and safety of CAR-T therapy and 26 27 identify gaps in current practice and knowledge. This will also be the first review to investigate 28 29 CAR-T therapy among patients with solid tumors. We expect the results from this clinical 30 31

32 systematic review will help inform the design of clinical trials. We also summarize our approach http://bmjopen.bmj.com/ 33 34 to appraise and analyze single arm interventional studies that are typically conducted for early 35 36 37 phase biotherapeutic trials; we believe this approach may be replicated for other systematic 38 39 reviews of early phase clinical data.

40 on September 30, 2021 by guest. Protected copyright. 41 42 43 44 Protocol 45 46 Our review protocol is reported in accordance with the Preferred Reporting Items in Systematic 47 48 reviews and Meta-Analysis-Protocol guidelines (see supplemental research checklist) [18]. 49 50 51 52 53 Research objectives 54 55 56 57 58 59 60 6 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019321 on 29 December 2017. Downloaded from BMJ Open Page 8 of 38

1 2 3 We will review controlled and uncontrolled interventional studies of CAR-T cell therapy to 4 5 6 examine the safety and efficacy of this treatment in patients with relapsed or refractory 7 8 hematological malignancies and solid tumors. 9 10 11 12 13 METHODS AND ANALYSIS 14 15 Eligibility criteriaFor peer review only 16 17 18 Studies will be selected according to the eligibility criteria detailed in Table 1. Interventional 19 20 studies with and without comparators will be included. We anticipate that many of the included 21 22 studies will be single-arm interventional studies. Full text articles in any language will be 23 24 25 considered. Unpublished gray literature, abstracts, commentaries, letters, reviews, and editorials 26 27 will be excluded [19]. 28 29 30 31

32 http://bmjopen.bmj.com/ 33 34 35 36 37 38 39

40 on September 30, 2021 by guest. Protected copyright. 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 7 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019321 on 29 December 2017. Downloaded from Page 9 of 38 BMJ Open

1 2 3 Table 1. Population, Intervention, Comparison, Outcome, and Study design breakdown of study 4 5 6 eligibility criteria 7 8 Category Description of criteria 9 10 Population Patients with solid tumor or hematologic malignancies 11 12 13 14 Intervention Chimeric antigen receptor T cell therapy 15 For peer review only 16 17 Comparator(s) Studies with or without any comparator will be considered 18 19 20 21 Outcome(s) Primary outcome: 22 23 • Complete response 24 25 Secondary outcomes: 26 27 28 • Overall response (hematological) or objective response (solid) 29 30 • Progressive disease 31 • Relapse 32 http://bmjopen.bmj.com/ 33 • Overall survival 34 35 • Adverse events (infection, neurotoxicity, cytokine release syndrome, B- 36 37 cell aplasia, graft versus host disease, other types will be grouped by 38 39 organ system affected and severity)

40 on September 30, 2021 by guest. Protected copyright. 41 Tertiary outcomes: 42 43 • Health-related quality of life 44 45 • Health utility measures 46 47 • Patient experience 48 49 50 51 52 Study design Interventional: +/- controlled, +/- randomized 53 54 55 56 57 58 59 60 8 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019321 on 29 December 2017. Downloaded from BMJ Open Page 10 of 38

1 2 3 Information sources 4 5 6 We will search MEDLINE (OVID interface, including In-Process and Epub Ahead of Print), 7 8 EMBASE (OVID interface), and the Cochrane Central Register of Controlled Trials (Wiley 9 10 11 interface). Clinical trial registries will be searched to identify ongoing and completed trials. 12 13 Specifically, ClinicalTrials.gov and the International Prospective Register of Systematic Reviews 14 15 will be searchedFor to identify peerongoing or recently review completed trials only or systematic reviews. In order 16 17 18 to further ensure a comprehensive literature search, we will examine reference lists of included 19 20 studies or relevant reviews identified through the search. We will also contact key researchers in 21 22 the field of CAR-T to ensure relevant studies have been identified. Finally, we will circulate a 23 24 25 bibliography of included articles to the systematic review team for feedback. 26 27 28 29 Search strategy 30 31

32 Specific search strategies will be created in collaboration with a Health Sciences Librarian (RS) http://bmjopen.bmj.com/ 33 34 with expertise in the design of systematic searches. The literature search strategies will be 35 36 37 developed using key words related to CAR-T cell therapy as well as hematological malignancies 38 39 and solid tumors. Search strategies will use controlled vocabulary (e.g. Receptors, Antigen, T-

40 on September 30, 2021 by guest. Protected copyright. 41 Cell) and keywords (e.g. CAR-T). The syntax and subject headings used in the finalized 42 43 44 EMBASE strategy will be adapted to the other databases. A validated search filter for clinical 45 46 studies will be applied. Both qualitative and quantitative studies will be sought. No study design, 47 48 date or language limits will be imposed on the search. A Peer Review of Electronic Search 49 50 51 Strategy will be performed by a second librarian who is not associated with the project [20]. A 52 53 draft of the Medline (OVID interface) search strategy for hematologic and solid tumors is shown 54 55 in supplemental appendix 1. 56 57 58 59 60 9 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019321 on 29 December 2017. Downloaded from Page 11 of 38 BMJ Open

1 2 3 Study records 4 5 6 The literature search results will be uploaded to Distiller Systematic Review Software 7 8 (DistillerSR®, Evidence Partners, Ottawa). DistillerSR is a cloud-based software program that 9 10 11 provides transparent, reproducible, and audit-ready results necessary for accurate review. 12 13 14 15 Data collection Forprocess peer review only 16 17 18 Two review authors (EG, ML) will independently screen the titles and abstracts from the search 19 20 results using the pre-defined inclusion criteria presented in Table 1. A calibration test will be 21 22 performed to refine the screening question prior to formally commencing the screening process. 23 24 25 For all titles that appear to meet the inclusion criteria or where there is any uncertainty, we will 26 27 access the full text. Two review authors (EG, FH) will assess the eligibility of full reports. 28 29 Disagreement will be resolved through discussion with a third party member (DF, HA, ML, NK). 30 31

32 We will record the reasons for excluding studies. http://bmjopen.bmj.com/ 33 34 35 36 37 Data items 38 39 Standardized drafts of data extraction forms were designed to extract all information of interest

40 on September 30, 2021 by guest. Protected copyright. 41 from the screened studies in adherence with the Effective Practice and Organisation of Care 42 43 44 guidelines [21]. The drafts will be used to inform the construction of the online data abstraction 45 46 program (DistillerSR). Data will be extracted independently and in duplicate from each eligible 47 48 study (EG, FH). A calibration exercise will be conducted prior to formally starting data 49 50 51 abstraction. Demographic information, methodology, intervention details, and outcomes will be 52 53 recorded. Reviewers will resolve disagreements by discussion or by conferring with one of two 54 55 56 57 58 59 60 10 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019321 on 29 December 2017. Downloaded from BMJ Open Page 12 of 38

1 2 3 arbitrators (ML, DF), who will adjudicate to resolve disagreements. Where uncertainty is 4 5 6 identified, we will contact study authors for more information. 7 8 9 10 11 Study characteristics to be extracted will include the journal title, the first author, the inclusion 12 13 criteria (outlined in Table 1), patient characteristics (mean age, sex, malignancy diagnosis), trial 14 15 design, type andFor source ofpeer financial support, review publication status only from trial reports, and study 16 17 18 sample size. Study intervention characteristics to be extracted will include lymphodepletion 19 20 method (preconditioning agents), previous treatment (ablative, non-ablative), failed transplant, 21 22 co-morbidities, concomitant medications, and length of follow-up. CAR-T intervention 23 24 25 characteristics to be extracted will detail manufacturing and cell product characteristics, 26 27 including: fresh or frozen, T-cell origin, selection of T cell subsets, T cell expansion method 28 29 including cell culture duration, CAR target antigen, CAR antigen, CAR molecular structure (i.e. 30 31

32 affinity domain, hinge domain, transmembrane domain, co-stimulatory domain(s), signaling http://bmjopen.bmj.com/ 33 34 domain), transfection/transduction method, and the therapeutic regimen (CAR-T dose, 35 36 37 frequency, duration, route of administration). Absolute lymphocyte counts prior to CAR-T cell 38 39 therapy administration will also be recorded as this has useful information for patient eligibility

40 on September 30, 2021 by guest. Protected copyright. 41 of CAR-T cells. Among solid tumors, the tumour regression grade will be reported when 42 43 44 available. When necessary, we will obtain measures of central tendency and dispersion of data 45 46 by analyzing the figures and tables or by contacting the authors. Whenever possible, the results 47 48 from an intention to treat analysis will be used. 49 50 51 52 53 54 55 56 57 58 59 60 11 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019321 on 29 December 2017. Downloaded from Page 13 of 38 BMJ Open

1 2 3 4 5 6 Outcome justification and prioritization 7 8 Primary outcome 9 10 11 Complete response, our primary outcome, will be defined by type of disease: hematological 12 13 malignancies acute lymphocytic leukemia (ALL) or acute myeloid leukemia (AML) and solid 14 15 tumors. If completeFor response peer is not feasible, review secondary response only outcomes will be reported using 16 17 18 best overall response when available. Best overall response will be defined according to the 19 20 response evaluation criteria in solid tumors (RECIST) guidelines where patients will be assigned 21 22 into one of the following categories: Complete response, partial response, stable disease, 23 24 25 progression, or inevaluable for response [22]. Studies that recruit patients in complete remission 26 27 at the initiation of CAR-T cell therapy will not be included in the complete response data 28 29 reported. 30 31

32 ALL and AML http://bmjopen.bmj.com/ 33 34 For patients with ALL or AML, in studies that: 1) do not provide a definition for complete 35 36 37 response, it will be considered hematologic response; 2) report minimal residual disease, 38 39 response will be defined as molecular response. The sensitivity of the assay used will also be

40 on September 30, 2021 by guest. Protected copyright. 41 extracted for molecular response. 42 43 44 Solid tumors 45 46 For patients with solid tumors, target lesion and non-target lesion complete responses will be 47 48 defined as disappearance of all target lesions and non-target lesions, respectively, where non- 49 50 51 target lesions must be accompanies by normalization of tumor marker level as defined by 52 53 RECIST guidelines [22]. In patients with solid tumors, any pathological lymph nodes (among 54 55 target or non-target lesions) must decrease in the short axis to less than 10mm [22]. Furthermore, 56 57 58 59 60 12 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019321 on 29 December 2017. Downloaded from BMJ Open Page 14 of 38

1 2 3 in studies that report tumor regression grading of zero, response will be defined as pathological 4 5 6 response. 7 8 Secondary outcomes 9 10 11 Overall response, progression of disease, relapse, and adverse events are our secondary response 12 13 outcomes to be measured. 14 15 Overall responseFor or objective peer response review only 16 17 18 Overall response and objective response will be defined as the sum of partial or complete 19 20 responses in both hematologic malignancies and solid tumors, respectively. In hematologic 21 22 malignancies, partial response is considered when there has been a response to therapy but does 23 24 25 not meet the criteria for complete response. In target lesion evaluation for solid tumors, partial 26 27 response is defined as a 30% decrease in the sum of target lesion diameters (compared to 28 29 baseline measures) [22]. 30 31

32 Progressive disease http://bmjopen.bmj.com/ 33 34 Progressive disease in hematologic malignancies is considered when evidence of disease 35 36 37 increases in the peripheral blood or bone marrow, or progression or new extramedullary disease 38 39 is identified. In solid tumors progressive disease is defined as a relative increase in the sum of

40 on September 30, 2021 by guest. Protected copyright. 41 target lesions by 20% (smallest sum as reference), an absolute increase in target lesions by 5 mm, 42 43 44 as well as appearance of any new lesions [22]. In both hematologic and solid tumors, stable 45 46 disease is defined as not meeting criteria for partial response, complete response or progression. 47 48 Relapse 49 50 51 Relapse is defined as a patient who has a partial or complete response but then develops disease 52 53 progression. Studies that recruit patients in complete remission at the initiation of CAR-T cell 54 55 therapy will be descriptively reported in the proportion of the patients that relapse. For patients 56 57 58 59 60 13 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019321 on 29 December 2017. Downloaded from Page 15 of 38 BMJ Open

1 2 3 with lymphoma or chronic lymphocytic leukemia (CLL), response criteria are defined as per the 4 5 6 RECIST guidelines [22]. Lastly, if CLL is identified as circulating disease in the peripheral 7 8 blood and/or bone marrow only, the response criteria that is used for AML and ALL will be 9 10 11 employed. 12 13 Adverse events 14 15 Adverse events For secondary peer outcomes will review be used to evaluate only clinical safety of CAR-T cell 16 17 18 therapy. Adverse events are a measure of unplanned or undesired symptoms or diagnoses that 19 20 occur during the study, which were absent at baseline, or worsen over the course of the study 21 22 [25]. In the setting of CAR-T cells, adverse events of special interest include infection, 23 24 25 neurotoxicity, cytokine release syndrome, B-cell aplasia, and graft versus host disease. 26 27 Tertiary outcomes 28 29 Tertiary outcomes that will be extracted include overall survival, patient experience, health- 30 31

32 related quality of life, and health utility. http://bmjopen.bmj.com/ 33 34 Overall survival 35 36 37 We will define overall survival as the time from the start of treatment to the time of death from 38 39 any cause.

40 on September 30, 2021 by guest. Protected copyright. 41 Patient experience 42 43 44 Patient experience combines a number of different dimensions including patient satisfaction, 45 46 expectations, and outcomes that occur throughout the experience of clinical treatment [26]. 47 48 Health-related quality of life 49 50 51 Health-related quality of life is a multidimensional concept that describes an individual’s self- 52 53 perceived health status [27]. 54 55 Health utility 56 57 58 59 60 14 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019321 on 29 December 2017. Downloaded from BMJ Open Page 16 of 38

1 2 3 Health utility measures reflect the preference values that patients attach to their overall health 4 5 6 status. A utility value is the global measure of health status; it summarizes the effects of an 7 8 intervention into one value between 0 (equal to death) and 1 (equal to perfect health). Due to the 9 10 11 variety of measures for patient experience, health-related quality of life, and health utility used in 12 13 clinical trials, all reported indices will be considered. 14 15 For peer review only 16 17 18 Outcome follow-up periods 19 20 Early and durable response will be recorded among included studies. All time points will be 21 22 considered due to the anticipated variability in follow-up. The median duration of follow-up will 23 24 25 also be recorded for all studies. 26 27 28 29 Risk of bias assessment 30 31

32 Currently, no tool exists to assess the risk of bias for single arm interventional studies. To assess http://bmjopen.bmj.com/ 33 34 the risk of bias tool for single arm interventional studies, we have modified the Institute of 35 36 37 Health Economics (IHE) risk of bias tool for case series studies [29] and incorporated items from 38 39 the Cochrane Collaboration's tool for assessing risk of bias in randomized trials [30]. This

40 on September 30, 2021 by guest. Protected copyright. 41 modified IHE tool includes assessment of the study objective, design, study population, 42 43 44 intervention and co-interventions, outcome measures (i.e. blinding, incomplete outcome data 45 46 such as participants lost to follow-up and selective outcome reporting), statistical analysis, results 47 48 and conclusions, and conflicts of interest. Each item will be scored as high risk, moderate risk, or 49 50 51 low risk of bias. For each item, a score of three will indicate a low risk of bias, two moderate risk 52 53 of bias, and one highest risk of bias. A sum of items among each study will also be performed to 54 55 provide the overall appraisal score for each individual study. The overall risk of bias results from 56 57 58 59 60 15 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019321 on 29 December 2017. Downloaded from Page 17 of 38 BMJ Open

1 2 3 the quality assessment will be provided in a risk of bias graph using Review Manager 5.3 4 5 6 (London, UK). These judgments will be made independently by two review authors (FH, ML) 7 8 based on the judging criteria provided for the modified IHE risk of bias tool for interventional 9 10 11 study designs (see supplemental appendix 2). Disagreements will be resolved first by discussion 12 13 and then by consulting a third author (DF) for arbitration. 14 15 For peer review only 16 17 Meta-bias assessment (or Risk of bias across studies) 18 19 A recent study demonstrated that traditional funnel plots may be a potentially misleading tool to 20 21 22 assess publication bias in meta-analyses of proportion studies, particularly where low or high 23 24 event rates exist [32]. The same study suggested an alternative funnel plot using study sample 25 26 size on the vertical axis instead of log odds of the event rate may be a more accurate measure of 27 28 29 publication bias [32]. Therefore, our review will follow these recommendations to assess 30 31 publication bias and use an alternative funnel plot of sample size on the vertical axis and inverse

32 http://bmjopen.bmj.com/ 33 34 of the standard error log odds in the horizontal axis. 35 36 37 38 Summary measures and synthesis of results 39

40 on September 30, 2021 by guest. Protected copyright. 41 We will perform a meta-analysis to synthesize the prevalence of outcomes reported. For patients 42 43 with hematologic malignancies, studies will be stratified by CD19 and non-CD19 targeted 44 45 antigens. Dichotomous outcomes will be reported as proportions with 95% confidence intervals 46 47 48 (CI). Continuous outcomes will be reported descriptively. A random effects model will be 49 50 employed using the DerSimonian and Laird random effects method in order to pool outcome 51 52 53 proportions (Comprehensive Meta-analysis 2.0, Englewood, USA). Continuity corrections will 54 55 be implemented in order to account for 0 and 100% event rates (0.5 was added to all cells for 56 57 trials with zero-events). Heterogeneity of effect sizes in the pooled proportions will be calculated 58 59 60 16 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019321 on 29 December 2017. Downloaded from BMJ Open Page 18 of 38

1 2 3 among included studies, for studies with n > 1, using the Cochrane I2 statistic. The following 4 5 2 6 thresholds are suggested to interpret the I statistic: 0–40% (low heterogeneity), 30–60% 7 8 (moderate heterogeneity), 50–90% (substantial heterogeneity), and 75–100% (considerable 9 10 11 heterogeneity) [31]. If there is considerable heterogeneity, sources of heterogeneity will 12 13 explored. 14 15 For peer review only 16 17 18 Subgroup Analyses 19 20 We will perform several a priori subgroup analyses to identify any subpopulations that may be 21 22 associated with different CAR-T therapy effectiveness. These analyses will include stratification 23 24 25 of studies based on the type of malignancy (i.e. non-Hodgkin’s lymphoma, chronic lymphocytic 26 27 leukemia, acute lymphocytic leukemia, metastatic breast cancer, etc.), pediatric versus adult 28 29 populations, interleukin-2 administration to cell and/or patient, lymphodepletion, T cell origin 30 31

32 (autologous versus allogeneic), T cell culture time, total cell dose, T cell persistence time, http://bmjopen.bmj.com/ 33 34 variability in T-cell culture time, dose and persistence time, fresh versus frozen CAR-T product 35 36 37 administered, and C19 CAR-T cells versus all other construct types. 38 39

40 on September 30, 2021 by guest. Protected copyright. 41 Reporting of Review 42 43 44 Our findings will be reported in agreement with the Preferred Reporting Items for Systematic 45 46 Reviews and Meta-analyses statement [33]. A completed copy of the checklist will be provided 47 48 as a supplementary document to the main report. 49 50 51 52 53 Confidence in cumulative estimate 54 55 56 57 58 59 60 17 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019321 on 29 December 2017. Downloaded from Page 19 of 38 BMJ Open

1 2 3 The quality of the treatment effects will be evaluated use the systematic and comprehensive 4 5 6 approach known as Grading of Recommendations, Assessment, Development and Evaluations 7 8 (GRADE). This approach is recognized as a highly effective method in terms of comparing the 9 10 11 treatment effectiveness and quality to clinical recommendations. The quality of evidence will be 12 13 assessed across the domains of risk of bias, consistency, directness, precision and publication 14 15 bias. Quality willFor be assigned peer as one of fourreview GRADE scores (0only to 4) reflecting high, moderate, 16 17 18 low, or very low quality evidence (34). High quality evidence reflects a high degree of 19 20 confidence in the estimate of effect whereas very low quality evidence indicates a high degree of 21 22 uncertainty regarding the estimate of effect. 23 24 25 26 27 List of abbreviations 28 29 ALL, acute lymphocytic leukemia; AML, acute myeloid leukemia; BioCanRx, Biotherapeutics 30 31

32 for Cancer Treatment; CAR-T, chimeric antigen receptor T cell; CD, cluster of differentiation; http://bmjopen.bmj.com/ 33 34 CLL, chronic lymphocytic leukemia; DistillerSR, Distiller systematic review; GRADE, grading 35 36 37 of recommendations, assessment, development and evaluations; IHE, Institute of Health 38 39 Economics; RECIST, response evaluation criteria in solid tumors.

40 on September 30, 2021 by guest. Protected copyright. 41 42 43 44 FUNDING 45 46 47 This study is supported by a grant (Grant reference number: FY17 / CSEI4) from Biotherapeutics 48 49 for Cancer Treatment (BioCanRx) a Canadian Network of Centres of Excellence. Funding will 50 51 support the collection of data, data management and analyses. BioCanRx will not be involved in 52 53 54 the design of the project’s protocol, analysis plan, collection of data, analyses, or interpretation 55 56 57 58 59 60 18 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019321 on 29 December 2017. Downloaded from BMJ Open Page 20 of 38

1 2 3 or publication of the study results. ML salary is supported by The Ottawa Hospital Anesthesia 4 5 6 Alternate Funds Association. 7 8 9 10 11 COMPETING INTERESTS 12 13 The authors have no declarations of conflicts of interest. 14 15 For peer review only 16 17 18 19 ETHICS AND DISSEMINATION 20 21 Ethics considerations 22 23 24 Not applicable. 25 26 27 28 29 30 Dissemination 31

32 http://bmjopen.bmj.com/ 33 The results of the study will be submitted for publication to a peer-reviewed journal and 34 35 presented at relevant national and international conferences and scientific meetings to promote 36 37 38 knowledge transfer. 39

40 on September 30, 2021 by guest. Protected copyright. 41 42 43 Consent for publication 44 45 46 Not applicable. 47 48 49 50 51 52 Availability of data and material 53 54 55 Not applicable. 56 57 58 59 60 19 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019321 on 29 December 2017. Downloaded from Page 21 of 38 BMJ Open

1 2 3 Amendments 4 5 6 If amendments are required for this protocol, date of each amendment will be provided with a 7 8 description and rationale for the change in this section. 9 10 11 12 13 Sources 14 15 This systematicFor review is peer funded by BioCanRx.review BioCanRx only is a Networks of Centres of 16 17 18 Excellence funded by the Government of Canada. 19 20 21 22 Sponsor 23 24 25 BioCanRx funded this research. 26 27 28 29 Role of sponsor 30 31

32 BioCanRx is funding this systematic review; funding will support the collection of data, data http://bmjopen.bmj.com/ 33 34 management and analyses. BioCanRx will not be involved in any other aspect of the project, 35 36 37 such as the design of the project’s protocol and analysis plan, the collection of data and analyses. 38 39 The funder will have no input on the interpretation or publication of the study results.

40 on September 30, 2021 by guest. Protected copyright. 41 42 43 44 AUTHORS' CONTRIBUTIONS 45 46 47 CRediT taxonomy was used to describe author contributions (see supplemental appendix 3). ML 48 49 is the guarantor. Conceptualization, ML, EG, and DF; Methodology, TR, RH, ML, NA, MS, HA, 50 51 52 EG, DF and BH; Writing - Original draft, EG; Resources, RS and RH; Writing - Review and 53 54 Editing, RS, RH, ML, NK, MS, KT, JP, HA, and DF; Supervision and Funding Acquisition, ML 55 56 and DF; Project Administration, EG. 57 58 59 60 20 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019321 on 29 December 2017. Downloaded from BMJ Open Page 22 of 38

1 2 3 4 5 6 Acknowledgements 7 8 Sarah Schlievert for administrative assistance. 9 10 11 12 13 REFERENCES 14 15 For peer review only 16 1) Jackson HJ, Rafiq S, Brentjens RJ. Driving CAR T-cells forward. Nat Rev Clin Oncol. 2016; 17 18 13: 370-383 19 20 21 2) Maude S, Barrett DM. Current status of chimeric antigen receptor therapy for 22 23 24 haematological malignancies. Br J Haematol. 2016; 172: 11-22 25 26 27 3) Lee DW, Kochenderfer JN, Stetler-Stevenson M, et al. T cells expressing CD19 chimeric 28 29 antigen receptors for acute lymphoblastic leukaemia in children and young adults: a phase 1 30 31 dose-escalation trial. Lancet. 2015; 385: 517-528

32 http://bmjopen.bmj.com/ 33 34 4) Davila ML, Riviere I, Wang X, et al. Efficacy and toxicity management of 19-28z CAR T 35 36 37 cell therapy in B cell acute lymphoblastic leukemia. Sci Transl Med. 2014;6:224ra225 38 39 5) Porter DL, Levine BL, Kalos M, Bagg A, June CH. Chimeric antigen receptor-modified T

40 on September 30, 2021 by guest. Protected copyright. 41 42 cells in chronic lymphoid leukemia. N Engl J Med. 2011; 365: 725-733 43 44 45 6) Yu S, Li A, Liu Q, et al. Chimeric Antigen Receptor T Cells: A Novel Therapy for Solid 46 47 Tumors. J Hematol Oncol. 10. Journal of Hematology & Oncology. 2017: 10. 48 49 50 7) Garde D. Juno pulls the plug on a once-promising cancer treatment. March 7, 2017. < 51 52 https://www.statnews.com/2017/03/01/juno-cancer-treatment/>. Accessed June 15, 2017. 53 54 55 8) Subject of Kite Drug Trial Dies, Sending Stock Down -- Market Talk. (2017, May 08). Dow 56 57 58 Jones Institutional News, p. Dow Jones Institutional News, Accessed June 15, 2017. 59 60 21 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019321 on 29 December 2017. Downloaded from Page 23 of 38 BMJ Open

1 2 3 9) Juno Therapeutics shuts down clinical trial of cancer drug after patient deaths. (2017, Mar 4 5 6 01). The Seattle Times, p. Technology, Accessed Jun 15, 2017. 7 8 9 10) Maus MV, Haas AR, Beatty GL, et al. T cells expressing chimeric antigen receptors can 10 11 cause anaphylaxis in humans. Cancer immunol research. 2013; 1: 26-31. 12 13 14 11) Zhang E, Xu H. A new insight in chimeric antigen receptor-engineered T cells for cancer 15 For peer review only 16 immunotherapy. J Hematol Oncol. 2017; 10: 1. 1. 17 18 19 12) Cartellieri M, Bachmann M, Feldmann A, et al. Chimeric antigen receptor-engineered T 20 21 22 cells for immunotherapy of cancer. J Biomed Biotechnol. 2010;2010:956304. 23 24 13) Zhang Tf, Cao L, Xie J, et al. Efficiency of CD19 Chimeric Antigen Receptor-modified T 25 26 27 Cells for Treatment of B Cell Malignancies in Phase I Clinical Trials: A Meta-analysis. 28 29 Oncotarget. 2015; 6: 32 30 31 32 14) Zhu Y, Tan Y, Ou R, et al. Anti CD19 Chimeric Antigen Receptor modified T Cells for B http://bmjopen.bmj.com/ 33 ‐ ‐ ‐ 34 35 36 cell Malignancies: A Systematic Review of Efficacy and Safety in Clinical Trials. Eur J 37 38 Haematol. 2016;96:4:389-96. 39

40 on September 30, 2021 by guest. Protected copyright. 41 15) Anwer F, Shaukat A, Zahid U, et al. Donor origin CAR T cells: graft versus malignancy 42 43 effect without GVHD, a systematic review. Immunotherapy. 2017; 9: 2.123–130. 44 45 46 16) Holzinger A, Barden M, Abken H. The Growing World of CAR T Cell Trials: A Systematic 47 48 Review. Cancer Immunol Immun. 2016; 65:12:1433-450. 49 50 51 17) Beall's List of Predatory. Journals and Publishers. (n.d.). < http://beallslist.weebly.com/>. 52 53 54 Accessed June 13, 2017. 55 56 18) Moher D, Shamseer L, Clarke M, et al. Preferred Reporting Items for Systematic Review 57 58 59 60 22 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019321 on 29 December 2017. Downloaded from BMJ Open Page 24 of 38

1 2 3 and Meta-Analysis Protocols (PRISMA-P) 2015 statement. Syst Rev. 2015; 4: 1. 1. doi: 4 5 6 10.1186/2046-4053-4-1 7 8 9 19) Fergusson D, Laupacis A, Salmi LR, Mcalister FA, Huet C. What Should Be Included in 10 11 Meta-analyses? An Exploration of Methodological Issues Using the ISPOT Meta-analyses. 12 13 International J Technol Assess. 2000; 16: 4. 1109-19. 14 15 For peer review only 16 20) Mcgowan J, Sampson M, Salzwedel DM, Cogo E, Foerster V, Lefebvre C. PRESS Peer 17 18 19 Review of Electronic Search Strategies: 2015 Guideline Statement. J Clin Epidemiol. 2016; 20 21 75: 40-46. 22 23 24 21) Effective Practice and Organisation of Care (EPOC). Data collection form. EPOC Resources 25 26 for review authors. Oslo: Norwegian Knowledge Centre for the Health Services; 2013. 27 28 29 Available at: http://epoc.cochrane.org/epoc-specific-resources-review-authors 30 31 22) Nishino M, Jagannathan J, Ramaiya N, Van Den Abbeele A. Revised RECIST Guideline 32 http://bmjopen.bmj.com/ 33 34 Version 1.1: What Oncologists Want to Know and What Radiologists Need to Know. Am J 35 36 Roentgenol. 2010; 195: 2. 281-9. 37 38 39 23) Choi H, Charnsangavej C, de Castro Faria S, Tamm EP, Benjamin RS, Johnson MM,

40 on September 30, 2021 by guest. Protected copyright. 41 Macapinlac HA, Podoloff DA. CT evaluation of the response of gastrointestinal stromal 42 43 44 tumors after imatinib mesylate treatment: a quantitative analysis correlated with FDG PET 45 46 findings. Am J Roentgenol. 2004; 183: 6. 1619-28. PubMed PMID: 15547201. 47 48 49 24) Hallek M, Cheson B, Catovsky D, et al. (2008). Guidelines for the diagnosis and treatment 50 51 of chronic lymphocytic leukemia: A report from the International Workshop on Chronic 52 53 54 Lymphocytic Leukemia updating the National Cancer Institute Working Group guidelines. 55 56 Blood. 1996; 111: 12. 5446-56. 57 58 59 60 23 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019321 on 29 December 2017. Downloaded from Page 25 of 38 BMJ Open

1 2 3 25) Brennan TA, Leape LL, Laird NM, et al. HH: Incidence of adverse events and negligence in 4 5 6 hospitalized patients. Results of the Harvard Medical Practice Study I. N Engl J Med. 1991; 7 8 324: 370-376. 9 10 11 26) Jenkinson C, Coulter A, Bruster S. The Picker Patient Experience Questionnaire: 12 13 Development and validation using data from in-patient surveys in five countries. Intern J for 14 15 For peer review only 16 Qual Health C. 2001; 14: 5. 353-358. 17 18 19 27) Santana M, Feeney D, and Institute of Health Economics. (n.d.). The importance of 20 21 measuring health related quality of life (IHE report). Edmonton, Alta.: Institute of Health 22 23 Economics. 24 25 26 28) Porrata LF, Gertz MA, Inwards DJ, et al. Early Lymphocyte Recovery Predicts Superior 27 28 29 Survival after Autologous Stem Cell Transplantation in Non-Hodgkin Lymphoma: A 30 31 Prospective Study. Biol Blood Marrow Tr. 2008; 14: 7. 807-16.

32 http://bmjopen.bmj.com/ 33 34 29) Guo B, Moga C, Harstall C, et al. A principal component analysis is conducted for a case 35 36 series quality appraisal checklist. J Clin Epidemiol. 2016; 69: 199 – 207. 37 38 39 30) Higgins JPT, Green S (editors). Cochrane Handbook for Systematic Reviews of

40 on September 30, 2021 by guest. Protected copyright. 41 Interventions Version 5.1.0 [updated March 2011]. The Cochrane Collaboration, 2011. 42 43 44 Available from www.handbook.cochrane.org. 45 46 47 31) Fergusson D, Laupacis A., Salmi LR, Mcalister FA, Huet C. What Should Be Included in 48 49 Meta-analyses? An Exploration of Methodological Issues Using the ISPOT Meta-analyses. 50 51 Intern J for Qual Health C. 2000; 16: 4. 1109-19. 52 53 54 32) Hunter, J.P., Saratzis, A., Sutton, A.J., et al. In meta-analyses of proportion studies, funnel 55 56 57 plots were found to be an inaccurate method of assessing publication bias. J Clin Epidemiol. 58 59 60 24 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019321 on 29 December 2017. Downloaded from BMJ Open Page 26 of 38

1 2 3 2014; 67: 8. 897-903. 4 5 6 33) Moher D, Liberati A, Tetzlaff J, Douglas GA. Preferred Reporting Items for Systematic 7 8 9 Reviews and Meta-analyses: The PRISMA Statement. Ann Intern Med. 2009;151:4. 264-9. 10 11 12 34) Guyatt G, Oxman A, Vist G, et al. GRADE: An emerging consensus on rating quality of 13 14 evidence and strength of recommendations. BMJ. 2008; 336: 7650: 924. 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31

32 http://bmjopen.bmj.com/ 33 34 35 36 37 38 39

40 on September 30, 2021 by guest. Protected copyright. 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 25 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019321 on 29 December 2017. Downloaded from Page 27 of 38 BMJ Open

1 2 3 Appendix 1. Representative Search Strategy 4 5 6 7 Hematologic malignancies: 8 9 10 Database: Embase Classic+Embase, EBM Reviews - Cochrane Central Register of 11 12 Controlled Trials, Ovid MEDLINE(R) Epub Ahead of Print, In-Process & Other Non- 13 14 15 Indexed Citations,For Ovid peerMEDLINE(R) review Daily and Ovid MEDLINE(R). only 16 17 18 Search Strategy: 19 20 21 1 ((chimeric antigen adj2 receptor*) and (therap* or treat* or immunity or 22 23 immunotherap* or cell*)).tw. 24 25 26 27 2 ((car adj3 t adj5 therap*) or (car adj3 t adj5 treat*)).tw. 28 29 30 3 (car adj3 t adj3 immunotherap*).tw. 31

32 http://bmjopen.bmj.com/ 33 4 (car therap* or (car adj2 t adj2 cell*)).tw. 34 35 36 5 ((modified or engineered) adj2 (t cell* or t lymphocyte*)).tw. 37 38 39 6 chimeric antigen receptor/

40 on September 30, 2021 by guest. Protected copyright. 41 42 43 7 or/1-6 44 45 46 8 (h?ematolog* cancer* or lymphoid malignanc* or b cell malignan* or h?ematolog* 47 48 neoplasm* or h?ematolog* malignanc* or lymphoma* or leuk?emi* or myeloma* or 49 50 51 nonhodgkin* or non hodgkin* or t cell malignan*).tw. 52 53 54 9 hematologic malignancy/ or exp lymphoma/ or exp leukemia/ or exp multiple 55 56 myeloma/ 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019321 on 29 December 2017. Downloaded from BMJ Open Page 28 of 38

1 2 3 Solid tumors: 4 5 6 7 Database: Embase Classic+Embase, EBM Reviews - Cochrane Central Register of 8 9 Controlled Trials, Ovid MEDLINE(R) Epub Ahead of Print, In-Process & Other Non- 10 11 Indexed Citations, Ovid MEDLINE(R) Daily and Ovid MEDLINE(R). 12 13 14 15 Search Strategy:For peer review only 16 17 18 1 ((chimeric antigen adj2 receptor*) and (therap* or treat* or immunity or 19 20 immunotherap* or cell*)).tw. 21 22 23 2 ((car adj3 t adj5 therap*) or (car adj3 t adj5 treat*)).tw. 24 25 26 27 3 (car adj3 t adj3 immunotherap*).tw. 28 29 30 4 (car therap* or (car adj2 t adj2 cell*)).tw. 31

32 http://bmjopen.bmj.com/ 33 5 ((modified or engineered) adj2 (t cell* or t lymphocyte*)).tw. 34 35 36 6 chimeric antigen receptor/ 37 38 39 7 or/1-6

40 on September 30, 2021 by guest. Protected copyright. 41 42 43 8 exp solid tumor/ 44 45 46 9 (solid adj (tumo?r* or malignan* or cancer)).tw. 47 48 49 10 exp breast cancer/ 50 51 52 11 exp colon cancer/ 53 54 55 56 12 exp rectum cancer/ 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019321 on 29 December 2017. Downloaded from Page 29 of 38 BMJ Open

1 2 3 13 colorectal cancer/ 4 5 6 7 14 exp kidney cancer/ 8 9 10 15 exp lung cancer/ 11 12 13 16 exp prostate cancer/ 14 15 For peer review only 16 17 exp pancreas cancer/ 17 18 19 20 18 exp bladder cancer/ 21 22 23 19 ((breast or lung or rect* or colorect* or colon or prostat* or renal or kidney or 24 25 bladder or pancrea*) adj2 (cancer* or neoplasm* or carcinoma* or tumo?r*)).tw. 26 27 28 20 exp skin cancer/ or cutaneous melanoma/ 29 30 31 21 (Skin adj2 (cancer or neoplasm*)).tw. 32 http://bmjopen.bmj.com/ 33 34 35 22 melanoma.tw. 36 37 38 23 brain cancer/ 39

40 on September 30, 2021 by guest. Protected copyright. 41 24 (brain adj2 (cancer or neoplasm* or tum?or*)).tw. 42 43 44 25 (glioma* or glioblastoma*).tw. 45 46 47 48 26 exp sarcoma/ 49 50 51 27 sarcoma.tw. 52 53 54 28 malignant mesothelioma/ 55 56 57 29 mesothelioma.tw. 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019321 on 29 December 2017. Downloaded from BMJ Open Page 30 of 38

1 2 3 30 liver cancer/ 4 5 6 7 31 ((liver or hepatic or Hepatocellular) adj2 (cancer* or neoplasm* or carcinoma* or 8 9 tumo?r*)).tw. 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31

32 http://bmjopen.bmj.com/ 33 34 35 36 37 38 39

1 2 3 Appendix 2. Modified Institute of Health Economics Tool 4 5 6 Question Text Answer Text 7 Was the hypothesis/ aim/ objective of the study stated? Yes 8 Yes = The hypothesis/aim/objective was reported (includes 9 patients, intervention and outcome). 10 11 Partial/ unclear = Only one or two components (patients, 12 intervention, or outcome) were included. 13 No = The hypothesis/aim/ objective was not reported. 14 Partial/ unclear 15 For peer review only 16 No 17 Was the study conducted prospectively? Yes 18 Yes = It was clearly stated that the study was conducted 19 prospectively. 20 Partial/ unclear = Unclear or no information was provided. 21 No = The study clearly stated it was a retrospective study. 22 23 Partial/ unclear 24 No 25 Were patients from more than one centre? Yes 26 For example, you can deduce single centre if they state "Data was 27 28 taken from the Sloan Memorial Research Centre". 29 Yes = Patients were from more than one centre (multicentre study). 30 Partial/ unclear = Unclear where the patients came from. 31 No = Patients were from one centre. 32 Partial/ unclear http://bmjopen.bmj.com/ 33 34 No 35 Were patients recruited consecutively? Yes 36 Note: Not based on previously published protocols. Must be stated 37 in this paper. 38 Yes = There was a clear statement or it was clear from the context 39 that the patients were recruited consecutively; or the study stated

40 on September 30, 2021 by guest. Protected copyright. 41 that all eligible patients were recruited. 42 Partial/ unclear = No information was provided about the method 43 used to recruit patients in the study. 44 No = The study clearly stated that patients were not recruited 45 46 consecutively; or the patients were recruited based on other criteria 47 such as access to intervention. 48 N/A = N of 1 study. 49 Partial/ unclear 50 No 51 52 N/A 53 Were the eligibility criteria (i.e. inclusion and exclusion criteria) for Yes 54 entry into the study clearly stated? 55 Note: Not based on previously published protocols. Must be stated 56 in this paper. 57 58 Yes = Both inclusion and exclusion criteria were reported. 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019321 on 29 December 2017. Downloaded from BMJ Open Page 32 of 38

1 2 3 Partial/ unclear = Either inclusion OR exclusion were reported. 4 5 No = Neither inclusion nor exclusion criteria were reported. 6 Partial/unclear 7 No 8 Were the characteristics of the patients included in the study Yes 9 10 described? 11 Relevant characteristics: age, sex, malignancy type, 12 lymphodepletion, previous treatment, concomitant treatments, co- 13 morbidities. 14 Yes = All of the relevant patient characteristics were reported. 15 For peer review only 16 Partial/ unclear = >= 1 of the relevant characteristics were reported. 17 No = None of the relevant characteristic were reported. 18 Partial/ unclear 19 No 20 Did patients enter the study at a similar point in the disease? Yes 21 22 Yes = The paper states that entering patients are in relapsed/ 23 refractory setting. 24 Partial/ unclear = There was no baseline information on patients' 25 characteristics to make a judgment. 26 No = There was a wide range in the severity of the disease and co- 27 28 morbidities of patients at baseline. 29 N/A = N of 1 study. 30 Partial/ unclear 31 No 32 N/A http://bmjopen.bmj.com/ 33 34 Was the intervention of interest described? Yes 35 Relevant characteristics: 36 T-cell origin, CD configuration (type (i.e. CD19), co-stimulatory 37 domain(s), dosage regimen (dose, frequency, duration). 38 Yes = All of the relevant characteristics of the intervention were 39 reported.

40 on September 30, 2021 by guest. Protected copyright. 41 Partial/unclear = Some of the relevant characteristics of the 42 intervention were reported. 43 No = None of the relevant characteristics of the intervention were 44 reported. 45 46 Partial/unclear 47 No 48 Were additional interventions clearly described? Yes 49 i.e. chemotherapy, HSCT, radiation. 50 Yes = All of the most relevant characteristics (type, dose, frequency 51 52 administration, duration) of the co-intervention were reported 53 Partial/ unclear = Some but not all of the most relevant 54 characteristics of the co-intervention were reported. 55 No = No information about the co-intervention was provided; or 56 only the name of the co-intervention was mentioned. 57 58 Partial/ unclear 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019321 on 29 December 2017. Downloaded from Page 33 of 38 BMJ Open

1 2 3 No 4 5 Were relevant outcome measures established a priori in the Yes 6 introduction or methods section? 7 Yes = All relevant outcome measures were stated. 8 Partial/ unclear = Some, but not all of the relevant outcome 9 measures were stated. 10 11 No = None of the relevant outcome measures were stated. 12 Partial/ unclear 13 No 14 Were outcome assessors blinded to the intervention that patients Yes 15 For peer review only 16 received? 17 i.e. Did the study have 'independent outcome assessors'? 18 Yes = The outcomes were assessed by individuals who were not 19 aware of patient intervention. 20 - SELECT where blinding is not necessary. i.e. Mortality is the 21 outcome. 22 23 - SELECT where the blinding to the outcome does not influence the 24 assessment. i.e. Response to CAR-T. 25 Partial/ unclear = The study did not report whether the outcome 26 assessors were aware of the intervention. 27 No = It was clearly stated or obvious from the context that 28 29 individuals assessing outcomes were unblinded. 30 Partial/ unclear 31 No 32 Were the relevant outcomes measured using appropriate objective Yes http://bmjopen.bmj.com/ 33 34 or subjective methods? 35 Yes = Complete response/remission and >=1 secondary outcomes 36 (i.e, overall response rate, non-relapse mortality, relapse, overall 37 survival, adverse events (infection, neurotoxicity, cytokine release 38 syndrome, B-cell aplasia, graft versus host disease, other types will 39 be grouped by organ system affected and severity)

40 on September 30, 2021 by guest. Protected copyright. 41 Partial/ unclear = >=1 secondary outcomes (listed in OUR protocol) 42 were reported 43 No = None of the outcomes listed in OUR protocol were reported. 44 45 Partial/ unclear 46 47 No 48 Were the relevant outcome measures made before and after the Yes 49 intervention? 50 Yes = The relevant outcome measures were made pre- and post- 51 intervention; or the baseline measurements were not possible (ex. 52 53 death). 54 Partial/ unclear = The study did not report when the outcome 55 measures were made. 56 No = The outcome measures were only made post-intervention. 57 Partial/ unclear 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019321 on 29 December 2017. Downloaded from BMJ Open Page 34 of 38

1 2 3 No 4 5 The study does not perform selective outcome reporting. Yes 6 Yes = The study protocol is registered and all of the study's pre- 7 specified outcomes of interest were stated in the methods section. 8 Partial/ unclear = Either study protocol registered or the study's pre- 9 specified outcomes of interest were stated in the methods section. 10 11 No = No study protocol registered and none of the study's pre- 12 specified outcomes of interest were stated in the methods section. 13 14 Partial/unclear 15 For peer review only 16 No 17 Were details of the statistical tests reported? Yes 18 Yes = The statistical tests were reported in the study. 19 Partial/unclear = Statistical tests only partially described or reported 20 elsewhere (e.g previous paper, or protocol). 21 No = The statistical tests were not described in the study. 22 23 N/A = N of 1 study. 24 Partial/unclear 25 No 26 N/A 27 28 Was follow-up period reported? Yes 29 Yes = follow-up information was reported. 30 No = Length of follow-up was not reported. 31 No

32 http://bmjopen.bmj.com/ 33 Did the study provide estimates of random variability in the data Yes 34 analysisof relevant outcomes? 35 Yes = Estimates of the random variability (ex. SE, SD, CI) were 36 reported for all relevant outcomes and/or could be calculated from 37 the raw data. 38 Partial/ unclear = Estimates if the random variability were reported 39 for some, but not all relevant outcomes.

40 on September 30, 2021 by guest. Protected copyright. 41 No = Estimates of the random variability were not reported for any 42 of the relevant outcomes. 43 N/A = N of 1 study. 44 Partial/unclear 45 46 No 47 N/A 48 Were the adverse events reported? Yes 49 Includes: Infection, neurotoxicity, cytokine release syndrome, B- 50 51 cell aplasia, and graft versus host disease. 52 Yes = All adverse events were reported. 53 Partial/ unclear = Unclear if all the adverse events were reported. 54 No = No information about adverse events reported. 55 Partial/ unclear 56 57 No 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019321 on 29 December 2017. Downloaded from Page 35 of 38 BMJ Open

1 2 3 Were both competing interests and sources of support for the study Yes 4 5 reported? 6 Yes = Both competing interests and sources of support (financial or 7 other) received for the study were reported; or the absence of 8 support was acknowledged. 9 Partial/ unclear = Either the competing interest or source of support 10 11 was reported. 12 No = Neither competing interests nor sources of support were 13 reported. 14 Partial/ unclear 15 For peer review onlyNo 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31

32 http://bmjopen.bmj.com/ 33 34 35 36 37 38 39

1 2 3 4 Appendix 3. Contributor roles taxonomy 5 6 Contributor Role1 Role Definition 7 8 Conceptualization Ideas; formulation or evolution of overarching research goals and aims. 9 Data Curation Management activities to annotate (produce metadata), scrub data and 10 maintain research data (including software code, where it is necessary for 11 interpreting the data itself) for initial use and later reuse. 12 13 Formal Analysis Application of statistical, mathematical, computational, or other formal 14 techniques to analyze or synthesize study data. 15 Funding ForAcquisition peer of the review financial support for only the project leading to this publication. 16 Acquisition 17 Investigation Conducting a research and investigation process, specifically performing the 18 19 experiments, or data/evidence collection. 20 Methodology Development or design of methodology; creation of models 21 Project Management and coordination responsibility for the research activity planning 22 Administration and execution. 23 24 Resources Provision of study materials, reagents, materials, patients, laboratory samples, 25 animals, instrumentation, computing resources, or other analysis tools. 26 Software Programming, software development; designing computer programs; 27 implementation of the computer code and supporting algorithms; testing of 28 existing code components. 29 30 Supervision Oversight and leadership responsibility for the research activity planning and 31 execution, including mentorship external to the core team.

32 Validation Verification, whether as a part of the activity or separate, of the overall http://bmjopen.bmj.com/ 33 replication/reproducibility of results/experiments and other research outputs. 34 35 Visualization Preparation, creation and/or presentation of the published work, specifically 36 visualization/data presentation. 37 Writing – Original Creation and/or presentation of the published work, specifically writing the 38 Draft Preparation initial draft (including substantive translation). 39 Writing – Review Preparation, creation and/or presentation of the published work by those from

40 on September 30, 2021 by guest. Protected copyright. 41 & Editing the original research group, specifically critical review, commentary or 42 revision – including pre- or post-publication stages. 43 1Author contributions based on contributor role taxonomy defined previously by Brand et 44 al. (50) 45

46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019321 on 29 December 2017. Downloaded from

1 1 : : : : : : : : 19 19 19 19 20 2 4

1 number(s) number(s) Page Page 4:1 44

4 2015 No

Yes Information reported reported Information

Systematic Reviews Systematic 5:15 55

5 2016

Systematic Reviews Systematic http://bmjopen.bmj.com/ BMJ Open details why this checklist was adapted - Moher D, Stewart L & Shekelle P: Shekelle L & D, Stewart Moher - adapted was checklist this why details tic review protocol submissions to BioMed Central journals from Table 3 in Moher D et al:et D Moher in 3 Table from journals Central BioMed to submissions protocol review tic on September 30, 2021 by guest. Protected copyright.

Systematic Reviews Systematic For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

For peer review only

If the protocol representsthe If anamendment of previouslya orcompleted published protocol, identify such andas changes;otherwise,list state for plan documenting important protocolamendments If registered, If theprovide name ofthe registry (e.g., PROSPERO) registrationand inthe number Provide Provide name,institutional affiliation, and e-mail address of protocol all authors; physicalprovide mailing address of corresponding author Checklist item item Checklist Abstract 3a 5a Indicate ofsources financial otheror for support the review 1b protocol the If for an is updateof a systematicreview,previous identify suchas 5b Provide name review for the and/orfunder sponsor 5c roles Describe of funder(s),sponsor(s), and/or institution(s), if any,developing in protocol the # #

P P 2015 Checklist P P 2015 Checklist P 2015 Checklist 4 2 -P -- 2015 Checklist

PRISMA INTRODUCTION sponsor/funder Role of Sponsor Sources Support Amendments Contributions 3b Describe contributions of authorsandprotocol the identifyguarantor review of the Contact PRISMA-PRISMAPRISMA This checklist has been adapted for use with systema with for use adapted been has checklist This Authors Registration Update Identification 1a thereport Identify a as protocol of a systematic review Title ADMINISTRATIVE INFORMATION Section/topic Section/topic Preferred reporting items for systematic review and meta-analysis protocols (PRISMA-P) 2015 statement. statement. 2015 (PRISMA-P) protocols and meta-analysis review systematic for items reporting Preferred Implementing PRISMA-P: recommendations for prospective authors. authors. prospective for recommendations PRISMA-P: Implementing

An Editorial from the Editors-in-Chief of of Editors-in-Chief the from Editorial An Page 37 of 38 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2017-019321 on 29 December 2017. Downloaded from

2 2 Page 38 of 16-17 16-17 15-16 11-15 10-11 10 10 10 9 9 7-8 7 5-6 number(s) number(s) Page Page

Yes Information reported reported Information

http://bmjopen.bmj.com/ BMJ Open on September 30, 2021 by guest. Protected copyright.

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml For peer review only Describe plannedDescribe method of extracting data reportsfrom (e.g., forms, piloting done independently, duplicate),in any processes for obtaining and dataconfirmingfrom investigators Describe anticipatedDescribe methods forassessing risk of bias of individual includingstudies, whether this be will done outcome at the level, study or or both; state this howinformation bewill in used data synthesis List and define and List which all outcomes for data will be sought, including prioritization of main and outcomes,additional rationale with Present draft draft Present of search strategy to be forused at least one database,electronic including planned thatlimits, such it could be repeated Describe allintendedDescribe information (e.g., sources databases,electronic contact withstudy authors, registers, trial other grey or with literature sources) planned dates of coverage List and define and List all variables for which bedata will sought (e.g.,PICO fundingitems, sources), any pre-planned data assumptionsand simplifications Specify the the Specify study (e.g.,characteristics PICO, study design, setting, time frame)reportand characteristics (e.g., years considered, language, status)publication to be used criteria as for fortheeligibility review Provide an explicit Provide an explicit statementthe of question(s) reviewthe addresswill reference with to participants, interventions, comparators, outcomesand (PICO) State theState process willthat be forselectingused twostudies (e.g., reviewers)independent through phase each ofthe review (i.e., screening, eligibility, inclusion and in meta-analysis) If data If are forappropriate quantitative describesynthesis, planned summary measures, methods of handling handling methodsanddata, of combining data from studies, including any planned exploration of Checklist item item Checklist 11c 11c 15a Describe undercriteria which study bedata willquantitatively synthesized 14 13 15b # # 9

8 10

12 7

6 the Describe rationalereview for the the in of context knownwhat already is

Synthesis DATA Risk of inbiasRisk studiesindividual Outcomes Outcomes and prioritization Data itemsData process Data collection Selection process 11b Data management 11a the Describe mechanism(s) will tothat used be recordsmanageandthroughout data the review STUDY RECORDSSTUDY Search strategySearch Information sources Eligibility criteriaEligibility METHODS Objectives Rationale Section/topic Section/topic

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2017-019321 on 29 December 2017. Downloaded from

3 3 18 16 16-17 17 number(s) number(s) Page Page

Yes Information reported reported Information

http://bmjopen.bmj.com/ BMJ Open on September 30, 2021 by guest. Protected copyright. , Kendall’s tau) 2

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml For peer review only Specify any Specify plannedof assessment meta-bias(es) (e.g., publication acrossbias studies,selective within reporting studies) Describe any Describe proposedanalyses additional (e.g., sensitivityor subgroupanalyses, meta- regression) consistency (e.g.,consistency Checklist item item Checklist 17 how Describe the of thestrength of body willevidence assessed(e.g., be GRADE) 15d quantitative If appropriate,synthesis notis describe thetype of summary planned 15c 15c # #

Confidence inConfidence evidencecumulative Meta-bias(es) Section/topic Section/topic

Page 39 of 38 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2017-019321 on 29 December 2017. Downloaded from BMJ Open

Efficacy and safety of Chimeric Antigen Receptor T- cell (CAR-T) therapy in patients with hematologic and solid malignancies: protocol for a systematic review and meta- analysis

For peer review only Journal: BMJ Open

Manuscript ID bmjopen-2017-019321.R1

Article Type: Protocol

Date Submitted by the Author: 11-Oct-2017

Complete List of Authors: Grigor, Emma; University of Ottawa, School of Epidemiology and Public Health; Ottawa Hospital Research Institute, Clinical Epidemiology Program, Blueprint Translational Research Group Fergusson, Dean; Ottawa Hospital Research Institute, Clinical Epidemiology Program, Blueprint Translational Research Group; University of Ottawa , School of Epidemiology and Public Health Haggar , Fatima; Ottawa Hospital Research Institute , Clinical Epidemiology Program, Blueprint Translational Research Group; Ottawa Hospital Research Institute , Blood and Marrow Transplant Program Kekre, Natasha; University of Ottawa , Faculty of Medicine ; Ottawa http://bmjopen.bmj.com/ Hospital Research Institute, Clinical Epidemiology Program, Blueprint Translational Research Group Atkins , Harold; University of Ottawa , Faculty of Medicine ; Ottawa Hospital Research Institute , Blood and Marrow Transplant Program Shorr , Risa; Ottawa Hospital Holt , Robert ; Canada's Michael Smith Genome Sciences Centre Hutton, Brian; University of Ottawa, Ottawa, Ontario, Canada, Faculty of Medicine ; Ottawa Hospital Research Institute , Clinical Epidemiology Program Ramsay, Tim; The Ottawa Health Research Institute, Clinical Epidemiology on September 30, 2021 by guest. Protected copyright. Program ; University of Ottawa , School of Epidemiology and Public Health Seftel , Matthew; University of Winnipeg, Department of Hematology Jonker, Derek; Faculty of Medicine, University of Ottawa; Division of Medical Oncology, The Ottawa Hospital Daugaard , Mads; Vancouver Prostate Centre, Molecular Pathology & Cell Imaging Core Facility Thavorn, Kednapa; Institute for Clinical Evaluative Sciences, ICES @uOttawa; The Ottawa Hospital Research Institute, Presseau , Justin; Ottawa Hospital Research Institute, Clinical Epidemiology Program, Blueprint Translational Research Group ; University of Ottawa , Faculty of Medicine Lalu, Manoj; Ottawa Hospital Research Institute, Clinical Epidemiology Program, Blueprint Translational Research Group; The Ottawa Hospital , Department of Anesthesiology and Pain Medicine

Primary Subject Oncology Heading:

Secondary Subject Heading: Haematology (incl blood transfusion), Pharmacology and therapeutics

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019321 on 29 December 2017. Downloaded from Page 1 of 38 BMJ Open

1 2 3 Leukaemia < HAEMATOLOGY, Lymphoma < HAEMATOLOGY, Myeloma < 4 Keywords: 5 HAEMATOLOGY, Chimeric Antigen Receptor, CAR-T, Malignancy 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31

32 http://bmjopen.bmj.com/ 33 34 35 36 37 38 39

32 Clinical Epidemiology Program, Blueprint Translational Research Group, Ottawa Hospital http://bmjopen.bmj.com/ 33 34 Research Institute, Ontario, Ottawa, Canada; 35 36 4 37 Blood and Marrow Transplant Program, Ottawa Hospital Research Institute, Ontario, Ottawa, 38 39 Canada;

40 on September 30, 2021 by guest. Protected copyright. 41 5Cancer Therapeutic Program, Ottawa Hospital Research Institute, Ontario, Ottawa, Canada; 42 43 6 44 Department of Surgery, University of Ottawa, Ontario, Ottawa, Canada; 45 46 7Department of Anesthesiology, The Ottawa Hospital, Ontario, Ottawa, Canada; 47 48 8Regenerative Medicine Program, Ottawa Hospital Research Institute, Ontario, Ottawa, Canada; 49 50 9 51 The Ottawa Hospital, Ottawa, Canada; 52 53 10Michael Smith Genome Sciences Centre, British Columbia Cancer Agency, Vancouver, British 54 55 Columbia, Canada; 56 57 58 59 60 1 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019321 on 29 December 2017. Downloaded from Page 3 of 38 BMJ Open

1 2 3 11Department of Hematology, University of Winnipeg, Manitoba, Canada; 4 5 12 6 Molecular Pathology & Cell Imaging Core Facility, Vancouver Prostate Centre, British 7 8 Columbia, Canada. 9 10 13 11 Institute for Clinical and Evaluative Sciences, Ontario, Ottawa, Canada; 12 14 13 Division of medical oncology, The Ottawa Hospital, Ontario, Ottawa, Canada. 14 15 For peer review only 16 17 18 Emma J.M. Grigor [email protected] - Dean Fergusson [email protected] - Fatima 19 20 Haggar [email protected] - Natasha Kekre [email protected] - Harold Atkins [email protected] - Risa 21 22 Shorr [email protected] - Rob A. Holt [email protected] - Brian Hutton [email protected] - Tim 23 24 25 Ramsay Tim [email protected] - Matthew Seftel [email protected] - Derek Jonker 26 27 [email protected] - Mads Daugaard [email protected] - Kednapa Thavorn [email protected] 28 29 - Justin Presseau [email protected] 30 31

32 http://bmjopen.bmj.com/ 33 34 Dates of study: April 1, 2017 to December 1, 2017 35 36 37 Corresponding author: 38 39 Manoj M. Lalu

40 on September 30, 2021 by guest. Protected copyright. 41 Ottawa Hospital - Civic campus 42 43 44 1053 Carling Avenue, Room B307, Mail Stop 249 45 46 Ottawa, Ontario, Canada K1Y 4E9 47 48 Telephone: 613-761-4169 (Option #1) 49 50 51 Fax: 613-761-5209 52 53 Email: [email protected] 54 55 orcid.org/0000-0002-0322-382X 56 57 58 59 60 2 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019321 on 29 December 2017. Downloaded from BMJ Open Page 4 of 38

1 2 3 4 5 6 ABSTRACT 7 8 Introduction: Patients with relapsed or refractory malignancies have a poor prognosis. 9 10 11 Immunotherapy with Chimeric Antigen Receptor T (CAR-T) cells redirects a patient’s immune 12 13 cells against the tumor antigen. CAR-T cell therapy has demonstrated promise in treating 14 15 patients with severalFor hematologic peer malignancies, review including acute only B cell lymphoblastic leukemia 16 17 18 and B cell lymphomas. CAR-T cell therapy for patients with other solid tumors is also being 19 20 tested. Safety is an important consideration in CAR-T cell therapy given the potential for serious 21 22 adverse events, including death. Previous reviews on CAR-T cell therapy have been limited in 23 24 25 scope and methodology. Herein we present a protocol for a systematic review to identify CAR-T 26 27 cell interventional studies and examine the safety and efficacy of this therapy in patients with 28 29 hematology malignancies and solid tumors. 30 31

32 Methods and analysis: We will search MEDLINE, including In-Process and Epub Ahead of http://bmjopen.bmj.com/ 33 34 Print, EMBASE, and the Cochrane Central Register of Controlled Trials from 1946 to 22 35 36 37 February 2017. Studies will be screened by title, abstract, and full text independently and in 38 39 duplicate. Studies that report administering CAR-T of any chimeric antigen receptor construct

40 on September 30, 2021 by guest. Protected copyright. 41 targeting antigens in patients with hematologic malignancies and solid tumors will be eligible for 42 43 44 inclusion. Outcomes to be extracted will include complete response rate (primary outcome), 45 46 overall response rate, overall survival, relapse, and adverse events. A meta-analysis will be 47 48 performed to synthesize the prevalence of outcomes reported as proportions with 95% 49 50 51 confidence intervals. The potential for bias within included studies will be assessed using a 52 53 modified Institute of Health Economics tool. Heterogeneity of effect sizes will be determined 54 55 using the Cochrane I2 statistic. 56 57 58 59 60 3 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019321 on 29 December 2017. Downloaded from Page 5 of 38 BMJ Open

1 2 3 Ethics and Dissemination: The review findings will be submitted for peer-reviewed journal 4 5 6 publication and presented at relevant conferences and scientific meetings to promote knowledge 7 8 transfer. 9 10 11 Registration: The protocol has been registered with the International Prospective Register of 12 13 Systematic Reviews (CRD42017075331). 14 15 For peer review only 16 Word count (abstract): 298 17 18 STRENGTHS AND LIMITATIONS OF THIS STUDY 19 20 21 1. A major methodological limitation is that all eligible studies are expected to be single arm 22 23 and have no comparator group. This may increase risk of bias when interpreting results. 24 25 26 2. We provide a comprehensive plan to address limitations of meta-analysis of single arm 27 28 studies that includes use a modified Institute of Health Economics tool for assessing risk of 29 30 bias in single arm interventional studies. Our approach can serve as a model for future 31

32 http://bmjopen.bmj.com/ 33 systematic reviews assessing early-phase clinical data that is often single arm with no control 34 35 comparison. 36 37 38 39

40 INTRODUCTION on September 30, 2021 by guest. Protected copyright. 41 42 43 Among patients with relapsed and refractory malignancies, chimeric antigen receptor T (CAR-T) 44 45 cell therapy is a novel immunotherapy that has shown promise in both pre-clinical and early 46 47 48 clinical studies. This therapy allows for CARs directed against tumour associated antigens (e.g. 49 50 CD19, HER-2) to be introduced into a patient’s T cells; this serves to reprogram these cells to 51 52 target the patient’s tumor cells. A number of small clinical trials using anti-CD19 CAR-T cells in 53 54 55 hematologic malignancies have demonstrated sustained responses in patients with advanced 56 57 disease [1-5]. CAR-T cell therapy for solid malignancies has also identified a number of 58 59 60 4 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019321 on 29 December 2017. Downloaded from BMJ Open Page 6 of 38

1 2 3 potential cancer-specific targets and previous pre-clinical studies investigating efficacy and 4 5 6 feasibility show promise [6]. 7 8 9 10 11 In spite of some evidence of efficacy of CAR-T cell therapy against some malignancies, there are 12 13 a number of safety concerns that have been identified. Trials conducted by Juno Therapeutics 14 15 and Kite PharmaFor reported mortalitypeer among review patients with hematologic only malignancy treated with 16 17 18 CAR-T cell therapy [7, 8]. The Juno Therapeutics trial was closed after the death of five patients 19 20 from cerebral edema linked to CAR-T cell therapy [9]. Previous trials investigating CAR-T cell 21 22 therapy among patients with solid tumors have reported adverse events from treatment including 23 24 25 anaphylaxis and cardiac arrest [10]. Past studies have also reported other challenges in applying 26 27 CAR-T cell therapy to solid tumors, including a scarcity of tumor associated antigen targets, the 28 29 potent immunosuppressive effects of the tumor microenvironment, and the limited trafficking of 30 31

32 CAR-T cells to tumor sites [11, 12]. http://bmjopen.bmj.com/ 33 34 35 36 37 Due to the variability and small size of clinical trials investigating CAR-T therapy there is a need 38 39 for a systematic review to evaluate its efficacy and safety. Although no systematic review

40 on September 30, 2021 by guest. Protected copyright. 41 currently exists for CAR-T therapy among solid tumors, we have identified four publications that 42 43 44 self-identified as systematic reviews studying the efficacy and safety of CAR-T therapy for 45 46 patients with hematologic malignancies [13, 14-16]. However, there were significant limitations 47 48 in the scope and/or methodologies in these earlier reviews. Zhu et al. only considered CD19 49 50 51 targeted CAR-T therapies, did not report differences between adult and pediatric populations, 52 53 and employed a non-systematic search strategy [14]. Anwer et al. only included allogeneic T- 54 55 cells, whereas most CAR-T cell therapy uses autologous cells [15]. Another systematic review 56 57 58 59 60 5 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019321 on 29 December 2017. Downloaded from Page 7 of 38 BMJ Open

1 2 3 by Zhang et al. only considered CD19 CAR-T therapies and was published in journal controlled 4 5 6 by a predatory publisher [13, 17]. Finally, while titled as a systematic review, Holzinger et al. is 7 8 only a narrative review [16]. Given these limitations of previous publications, along with rapid 9 10 11 evolution of the CAR-T field, there is a need for a current systematic review, as we present here, 12 13 that adheres to rigorous, state-of-the-art methods and summarizes the findings among both solid 14 15 tumors and hematologicFor malignancies. peer review only 16 17 18 19 20 Our systematic review will clarify the determinants of efficacy and safety of CAR-T therapy and 21 22 identify gaps in current practice and knowledge. This will also be the first review to investigate 23 24 25 CAR-T therapy among patients with solid tumors. We expect the results from this clinical 26 27 systematic review will help inform the design of clinical trials. We also summarize our approach 28 29 to appraise and analyze single arm interventional studies that are typically conducted for early 30 31

32 phase biotherapeutic trials; we believe this approach may be replicated for other systematic http://bmjopen.bmj.com/ 33 34 reviews of early phase clinical data. 35 36 37 38 39 Protocol

40 on September 30, 2021 by guest. Protected copyright. 41 Our review protocol is reported in accordance with the Preferred Reporting Items in Systematic 42 43 44 reviews and Meta-Analysis-Protocol guidelines (see supplemental research checklist) [18]. 45 46 47 48 Research objectives 49 50 51 We will review controlled and uncontrolled interventional studies of CAR-T cell therapy to 52 53 examine the safety and efficacy of this treatment in patients with relapsed or refractory 54 55 hematological malignancies and solid tumors. 56 57 58 59 60 6 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019321 on 29 December 2017. Downloaded from BMJ Open Page 8 of 38

1 2 3 4 5 6 METHODS AND ANALYSIS 7 8 Eligibility criteria 9 10 11 Studies will be selected according to the eligibility criteria detailed in Table 1. Interventional 12 13 studies with and without comparators will be included. We anticipate that many of the included 14 15 studies will be For single-arm peer interventional review studies. Full text articlesonly in any language will be 16 17 18 considered. Unpublished gray literature, abstracts, commentaries, letters, reviews, and editorials 19 20 will be excluded [19]. 21 22 23 24 25 26 27 28 29 30 31

32 http://bmjopen.bmj.com/ 33 34 35 36 37 38 39

1 2 3 Information sources 4 5 6 We will search MEDLINE (OVID interface, including In-Process and Epub Ahead of Print), 7 8 EMBASE (OVID interface), and the Cochrane Central Register of Controlled Trials (Wiley 9 10 11 interface) from 1946 to 22 February 2017 and plan to update our search prior to submission for 12 13 publication. Clinical trial registries will be searched to identify ongoing and completed trials. 14 15 Specifically, ClinicalTrials.govFor peer and the International review Prospective only Register of Systematic Reviews 16 17 18 will be searched to identify ongoing or recently completed trials or systematic reviews. In order 19 20 to further ensure a comprehensive literature search, we will examine reference lists of included 21 22 studies or relevant reviews identified through the search. Authors of the studies included in the 23 24 25 present review will be contacted for review of reported outcomes. Finally, we will circulate a 26 27 bibliography of included articles to the systematic review team for feedback. 28 29 30 31

32 Search strategy http://bmjopen.bmj.com/ 33 34 Specific search strategies will be created in collaboration with a Health Sciences Librarian (RS) 35 36 37 with expertise in the design of systematic searches. The literature search strategies will be 38 39 developed using key words related to CAR-T cell therapy as well as hematological malignancies

40 on September 30, 2021 by guest. Protected copyright. 41 and solid tumors. Search strategies will use controlled vocabulary (e.g. Receptors, Antigen, T- 42 43 44 Cell) and keywords (e.g. CAR-T). The syntax and subject headings used in the finalized 45 46 EMBASE strategy will be adapted to the other databases. A validated search filter for clinical 47 48 studies will be applied. Both qualitative and quantitative studies will be sought. No study design, 49 50 51 date or language limits will be imposed on the search. A Peer Review of Electronic Search 52 53 Strategy will be performed by a second librarian who is not associated with the project [20]. A 54 55 56 57 58 59 60 9 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019321 on 29 December 2017. Downloaded from Page 11 of 38 BMJ Open

1 2 3 draft of the Medline (OVID interface) search strategy for hematologic and solid tumors is shown 4 5 6 in supplemental appendix 1. 7 8 9 10 11 Study records 12 13 The literature search results will be uploaded to Distiller Systematic Review Software 14 15 (DistillerSR®, EvidenceFor Partners, peer Ottawa). review DistillerSR is a cloud-based only software program that 16 17 18 provides transparent, reproducible, and audit-ready results necessary for accurate review. 19 20 21 22 Data collection process 23 24 25 Two review authors (EG, ML) will independently screen the titles and abstracts from the search 26 27 results using the pre-defined inclusion criteria presented in Table 1. A calibration test will be 28 29 performed to refine the screening question prior to formally commencing the screening process. 30 31

32 For all titles that appear to meet the inclusion criteria or where there is any uncertainty, we will http://bmjopen.bmj.com/ 33 34 access the full text. Two review authors (EG, FH) will assess the eligibility of full reports. 35 36 37 Disagreement will be resolved through discussion with a third party member (DF, HA, ML, NK). 38 39 We will record the reasons for excluding studies. In the case of screening eligibility of non-

40 on September 30, 2021 by guest. Protected copyright. 41 English full-text articles, Ottawa Hospital Employees with fluency in the article languages will 42 43 44 first be contacted for assistance determining article eligibility. If the article meets the eligibility 45 46 criteria for inclusion in the review, a verbatim translation using the scientific translation services 47 48 at the Ottawa Hospital will be used. 49 50 51 52 53 Data items 54 55 56 57 58 59 60 10 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019321 on 29 December 2017. Downloaded from BMJ Open Page 12 of 38

1 2 3 Standardized drafts of data extraction forms were designed to extract all information of interest 4 5 6 from the screened studies in adherence with the Effective Practice and Organisation of Care 7 8 guidelines [21]. The drafts will be used to inform the construction of the online data abstraction 9 10 11 program (DistillerSR). Data will be extracted independently and in duplicate from each eligible 12 13 study (EG, FH). A calibration exercise will be conducted prior to formally starting data 14 15 abstraction. DemographicFor information,peer methodology, review intervention only details, and outcomes will be 16 17 18 recorded. Reviewers will resolve disagreements by discussion or by conferring with one of two 19 20 arbitrators (ML, DF), who will adjudicate to resolve disagreements. Where uncertainty is 21 22 identified, we will contact study authors for more information. 23 24 25 26 27 Study characteristics to be extracted will include the journal title, the first author, the inclusion 28 29 criteria (outlined in Table 1), patient characteristics (mean age, sex, malignancy diagnosis), trial 30 31

32 design, type and source of financial support, publication status from trial reports, and study http://bmjopen.bmj.com/ 33 34 sample size. Study intervention characteristics to be extracted will include lymphodepletion 35 36 37 method (preconditioning agents), previous treatment (ablative, non-ablative), failed transplant, 38 39 co-morbidities, concomitant medications, and length of follow-up. CAR-T intervention

40 on September 30, 2021 by guest. Protected copyright. 41 characteristics to be extracted will detail manufacturing and cell product characteristics, 42 43 44 including: fresh or frozen, T-cell origin, selection of T cell subsets, T cell expansion method 45 46 including cell culture duration, CAR target antigen, CAR antigen, CAR molecular structure (i.e. 47 48 affinity domain, hinge domain, transmembrane domain, co-stimulatory domain(s), signaling 49 50 51 domain), transfection/transduction method, and the therapeutic regimen (CAR-T dose, 52 53 frequency, duration, route of administration). Absolute lymphocyte counts prior to CAR-T cell 54 55 therapy administration will also be recorded as this has useful information for patient eligibility 56 57 58 59 60 11 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019321 on 29 December 2017. Downloaded from Page 13 of 38 BMJ Open

1 2 3 of CAR-T cells. Among solid tumors, the tumour regression grade will be reported when 4 5 6 available. When necessary, we will obtain measures of central tendency and dispersion of data 7 8 by analyzing the figures and tables or by contacting the authors. Whenever possible, the results 9 10 11 from an intention to treat analysis will be used. 12 13 14 15 For peer review only 16 17 18 19 20 21 22 Outcome justification and prioritization 23 24 25 Primary outcome 26 27 Complete response, our primary outcome, will be defined by type of disease: hematological 28 29 malignancies acute lymphocytic leukemia (ALL) or acute myeloid leukemia (AML) and solid 30 31

32 tumors. If complete response is not feasible, secondary response outcomes will be reported using http://bmjopen.bmj.com/ 33 34 best overall response when available. Best overall response will be defined according to the 35 36 37 response evaluation criteria in solid tumors (RECIST) guidelines where patients will be assigned 38 39 into one of the following categories: Complete response, partial response, stable disease,

40 on September 30, 2021 by guest. Protected copyright. 41 progression, or inevaluable for response [22]. Studies that recruit patients in complete remission 42 43 44 at the initiation of CAR-T cell therapy will not be included in the complete response data 45 46 reported. 47 48 ALL and AML 49 50 51 For patients with ALL or AML, in studies that: 1) do not provide a definition for complete 52 53 response, it will be considered hematologic response; 2) report minimal residual disease, 54 55 56 57 58 59 60 12 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019321 on 29 December 2017. Downloaded from BMJ Open Page 14 of 38

1 2 3 response will be defined by any response criteria, including molecular, morphological, and 4 5 6 immunological. The sensitivity of the assay used will also be extracted for molecular response. 7 8 Solid tumors 9 10 11 For patients with solid tumors, target and non-target lesions recorded at measurement baseline 12 13 will be defined according to the RECIST guidelines. Target lesions may include up to five 14 15 lesions which willFor likely include peer lesions withreview the longest diameters. only Non-target lesions will be 16 17 18 inclusive of all other lesions (or disease targets), including pathological lymph nodes. Target 19 20 lesion and non-target lesion complete responses will be defined as disappearance of all target 21 22 lesions and non-target lesions, respectively, where non-target lesions must be accompanied by 23 24 25 normalization of tumor marker level as defined by RECIST guidelines [22]. In patients with 26 27 solid tumors, any pathological lymph nodes (among target or non-target lesions) must decrease 28 29 in the short axis to less than 10mm [22]. Furthermore, in studies that report tumor regression 30 31

32 grading of zero, response will be defined as pathological response. http://bmjopen.bmj.com/ 33 34 Secondary outcomes 35 36 37 Overall response, progression of disease, relapse, and adverse events are our secondary response 38 39 outcomes to be measured.

40 on September 30, 2021 by guest. Protected copyright. 41 Overall response or objective response 42 43 44 Overall response and objective response will be defined as the sum of partial or complete 45 46 responses in both hematologic malignancies and solid tumors, respectively. In hematologic 47 48 malignancies, partial response is considered when there has been a response to therapy but does 49 50 51 not meet the criteria for complete response. In target lesion evaluation for solid tumors, partial 52 53 response is defined as a 30% decrease in the sum of target lesion diameters (compared to 54 55 baseline measures) [22]. 56 57 58 59 60 13 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019321 on 29 December 2017. Downloaded from Page 15 of 38 BMJ Open

1 2 3 Progressive disease 4 5 6 Progressive disease in hematologic malignancies is considered when evidence of disease 7 8 increases in the peripheral blood or bone marrow, or progression or new extramedullary disease 9 10 11 is identified. In solid tumors, progressive disease is defined as a relative increase in the sum of 12 13 target lesions by 20% (smallest sum as reference), an absolute increase in target lesions by 5 mm, 14 15 as well as appearanceFor of anypeer new lesions review [22]. In both hematologic only and solid tumors, stable 16 17 18 disease is defined as not meeting criteria for partial response, complete response or progression. 19 20 Relapse 21 22 Relapse is defined as a patient who has a partial or complete response but then develops disease 23 24 25 progression. Studies that recruit patients in complete remission at the initiation of CAR-T cell 26 27 therapy will be descriptively reported in the proportion of the patients that relapse. For patients 28 29 with lymphoma or chronic lymphocytic leukemia (CLL), response criteria are defined as per the 30 31

32 RECIST guidelines [22]. Lastly, if CLL is identified as circulating disease in the peripheral http://bmjopen.bmj.com/ 33 34 blood and/or bone marrow only, the response criteria that is used for AML and ALL will be 35 36 37 employed. 38 39 Adverse events

40 on September 30, 2021 by guest. Protected copyright. 41 Adverse events secondary outcomes will be used to evaluate clinical safety of CAR-T cell 42 43 44 therapy. Adverse events are a measure of unplanned or undesired symptoms or diagnoses that 45 46 occur during the study, which were absent at baseline, or worsen over the course of the study 47 48 [25]. In the setting of CAR-T cells, adverse events of special interest include infection, 49 50 51 neurotoxicity, cytokine release syndrome, B-cell aplasia, and graft versus host disease. 52 53 Tertiary outcomes 54 55 56 57 58 59 60 14 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019321 on 29 December 2017. Downloaded from BMJ Open Page 16 of 38

1 2 3 Tertiary outcomes that will be extracted include overall survival, patient experience, health- 4 5 6 related quality of life, and health utility. 7 8 Overall survival 9 10 11 We will define overall survival as the time from the start of treatment to the time of death from 12 13 any cause. 14 15 Patient experienceFor peer review only 16 17 18 Patient experience combines a number of different dimensions including patient satisfaction, 19 20 expectations, and outcomes that occur throughout the experience of clinical treatment [26]. 21 22 Health-related quality of life 23 24 25 Health-related quality of life is a multidimensional concept that describes an individual’s self- 26 27 perceived health status [27]. 28 29 Health utility 30 31

32 Health utility measures reflect the preference values that patients attach to their overall health http://bmjopen.bmj.com/ 33 34 status. A utility value is the global measure of health status; it summarizes the effects of an 35 36 37 intervention into one value between 0 (equal to death) and 1 (equal to perfect health). Due to the 38 39 variety of measures for patient experience, health-related quality of life, and health utility used in

40 on September 30, 2021 by guest. Protected copyright. 41 clinical trials, all reported indices will be considered. 42 43 44 45 46 Outcome follow-up periods 47 48 Early and durable response will be recorded among included studies. All time points will be 49 50 51 considered due to the anticipated variability in follow-up. The median duration of follow-up will 52 53 also be recorded for all studies. 54 55 56 57 58 59 60 15 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019321 on 29 December 2017. Downloaded from Page 17 of 38 BMJ Open

1 2 3 Risk of bias assessment 4 5 6 Currently, no tool exists to assess the risk of bias for single arm interventional studies. To assess 7 8 the risk of bias tool for single arm interventional studies, we have modified the Institute of 9 10 11 Health Economics (IHE) risk of bias tool for case series studies [29] and incorporated items from 12 13 the Cochrane Collaboration's tool for assessing risk of bias in randomized trials [30]. This 14 15 modified IHE For tool includes peer assessment review of the study objective, only design, study population, 16 17 18 intervention and co-interventions, outcome measures (i.e. blinding, incomplete outcome data 19 20 such as participants lost to follow-up and selective outcome reporting), statistical analysis, results 21 22 and conclusions, and conflicts of interest. Each item will be scored as high risk, moderate risk, or 23 24 25 low risk of bias. For each item, a score of three will indicate a low risk of bias, two moderate risk 26 27 of bias, and one highest risk of bias. A sum of items among each study will also be performed to 28 29 provide the overall appraisal score for each individual study. The overall risk of bias results from 30 31

32 the quality assessment will be provided in a risk of bias graph using Review Manager 5.3 http://bmjopen.bmj.com/ 33 34 (London, UK). These judgments will be made independently by two review authors (FH, ML) 35 36 37 based on the judging criteria provided for the modified IHE risk of bias tool for interventional 38 39 study designs (see supplemental appendix 2). Disagreements will be resolved first by discussion

40 on September 30, 2021 by guest. Protected copyright. 41 and then by consulting a third author (DF) for arbitration. 42 43 44 45 Meta-bias assessment (or Risk of bias across studies) 46 47 48 A recent study demonstrated that traditional funnel plots may be a potentially misleading tool to 49 50 assess publication bias in meta-analyses of proportion studies, particularly where low or high 51 52 53 event rates exist [32]. The same study suggested an alternative funnel plot using study sample 54 55 size on the vertical axis instead of log odds of the event rate may be a more accurate measure of 56 57 publication bias [32]. Therefore, our review will follow these recommendations to assess 58 59 60 16 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019321 on 29 December 2017. Downloaded from BMJ Open Page 18 of 38

1 2 3 publication bias and use an alternative funnel plot of sample size on the vertical axis and inverse 4 5 6 of the standard error log point estimate in the horizontal axis. 7 8 9 10 11 Summary measures and synthesis of results 12 13 We will perform a meta-analysis to synthesize the prevalence of outcomes reported. For patients 14 15 with hematologicFor malignancies, peer studies willreview be stratified by only CD19 and non-CD19 targeted 16 17 18 antigens. Dichotomous outcomes will be reported as proportions with 95% confidence intervals 19 20 (CI). Continuous outcomes will be reported descriptively. A random effects model will be 21 22 employed using the DerSimonian and Laird random effects method in order to pool outcome 23 24 25 proportions (Comprehensive Meta-analysis 2.0, Englewood, USA). Continuity corrections will 26 27 be implemented in order to account for 0 and 100% event rates (0.5 was added to all cells for 28 29 trials with zero-events). Heterogeneity of effect sizes in the pooled proportions will be calculated 30 31 2

32 among included studies, for studies with n > 1, using the Cochrane I statistic. The following http://bmjopen.bmj.com/ 33 34 thresholds are suggested to interpret the I2 statistic: 0–40% (low heterogeneity), 30–60% 35 36 37 (moderate heterogeneity), 50–90% (substantial heterogeneity), and 75–100% (considerable 38 39 heterogeneity) [31]. If there is considerable heterogeneity, sources of heterogeneity will be

40 on September 30, 2021 by guest. Protected copyright. 41 explored. 42 43 44 45 46 Subgroup Analyses 47 48 We will perform several a priori subgroup analyses to identify any subpopulations that may be 49 50 51 associated with different CAR-T therapy effectiveness. These analyses will include stratification 52 53 of studies based on the type of malignancy (i.e. non-Hodgkin’s lymphoma, chronic lymphocytic 54 55 leukemia, acute lymphocytic leukemia, metastatic breast cancer, etc.), pediatric versus adult 56 57 58 59 60 17 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019321 on 29 December 2017. Downloaded from Page 19 of 38 BMJ Open

1 2 3 populations, interleukin-2 administration to cell and/or patient, lymphodepletion, T cell origin 4 5 6 (autologous versus allogeneic), T cell culture time, total cell dose, T cell persistence time, 7 8 variability in T-cell culture time, dose and persistence time, fresh versus frozen CAR-T product 9 10 11 administered, and C19 CAR-T cells versus all other construct types. 12 13 14 15 Reporting of ReviewFor peer review only 16 17 18 Our findings will be reported in agreement with the Preferred Reporting Items for Systematic 19 20 Reviews and Meta-analyses statement [33]. A completed copy of the checklist will be provided 21 22 as a supplementary document to the main report. 23 24 25 26 27 Confidence in cumulative estimate 28 29 The quality of the treatment effects will be evaluated use the systematic and comprehensive 30 31

32 approach known as Grading of Recommendations, Assessment, Development and Evaluations http://bmjopen.bmj.com/ 33 34 (GRADE). This approach is recognized as a highly effective method in terms of comparing the 35 36 37 treatment effectiveness and quality to clinical recommendations. The quality of evidence will be 38 39 assessed across the domains of risk of bias, consistency, directness, precision and publication

40 on September 30, 2021 by guest. Protected copyright. 41 bias. Quality will be assigned as one of four GRADE scores (0 to 4) reflecting high, moderate, 42 43 44 low, or very low quality evidence (34). High quality evidence reflects a high degree of 45 46 confidence in the estimate of effect whereas very low quality evidence indicates a high degree of 47 48 uncertainty regarding the estimate of effect. 49 50 51 52 53 List of abbreviations 54 55 56 57 58 59 60 18 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019321 on 29 December 2017. Downloaded from BMJ Open Page 20 of 38

1 2 3 ALL, acute lymphocytic leukemia; AML, acute myeloid leukemia; BioCanRx, Biotherapeutics 4 5 6 for Cancer Treatment; CAR-T, chimeric antigen receptor T cell; CD, cluster of differentiation; 7 8 CLL, chronic lymphocytic leukemia; DistillerSR, Distiller systematic review; GRADE, grading 9 10 11 of recommendations, assessment, development and evaluations; IHE, Institute of Health 12 13 Economics; RECIST, response evaluation criteria in solid tumors. 14 15 For peer review only 16 17 18 FUNDING 19 20 This study is supported by a grant (Grant reference number: FY17 / CSEI4) from Biotherapeutics 21 22 23 for Cancer Treatment (BioCanRx) a Canadian Network of Centres of Excellence. Funding will 24 25 support the collection of data, data management and analyses. BioCanRx will not be involved in 26 27 28 the design of the project’s protocol, analysis plan, collection of data, analyses, or interpretation 29 30 or publication of the study results. ML salary is supported by The Ottawa Hospital Anesthesia 31

32 Alternate Funds Association. http://bmjopen.bmj.com/ 33 34 35 36 37 38 COMPETING INTERESTS 39

40 The authors have no declarations of conflicts of interest. on September 30, 2021 by guest. Protected copyright. 41 42 43 44 45 46 ETHICS AND DISSEMINATION 47 48 Ethics considerations 49 50 51 Not applicable. 52 53 54 55 56 57 Dissemination 58 59 60 19 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019321 on 29 December 2017. Downloaded from Page 21 of 38 BMJ Open

1 2 3 The results of the study will be submitted for publication to a peer-reviewed journal and 4 5 6 presented at relevant national and international conferences and scientific meetings to promote 7 8 knowledge transfer. 9 10 11 12 13 14 Consent for publication 15 For peer review only 16 17 Not applicable. 18 19 20 21 22 23 Availability of data and material 24 25 Not applicable. 26 27 28 29 30 31 Amendments

32 http://bmjopen.bmj.com/ 33 If amendments are required for this protocol, date of each amendment will be provided with a 34 35 description and rationale for the change in this section. 36 37 38 39

40 Sources on September 30, 2021 by guest. Protected copyright. 41 42 43 This systematic review is funded by BioCanRx. BioCanRx is a Networks of Centres of 44 45 Excellence funded by the Government of Canada. 46 47 48 49 50 Sponsor 51 52 BioCanRx funded this research. 53 54 55 56 57 Role of sponsor 58 59 60 20 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019321 on 29 December 2017. Downloaded from BMJ Open Page 22 of 38

1 2 3 BioCanRx is funding this systematic review; funding will support the collection of data, data 4 5 6 management and analyses. BioCanRx will not be involved in any other aspect of the project, 7 8 such as the design of the project’s protocol and analysis plan, the collection of data and analyses. 9 10 11 The funder will have no input on the interpretation or publication of the study results. 12 13 14 15 For peer review only 16 AUTHORS' CONTRIBUTIONS 17 18 19 CRediT taxonomy was used to describe author contributions (see supplemental appendix 3). ML 20 21 is the guarantor. Conceptualization, ML, EG, and DF; Methodology, TR, RH, ML, NK, MS, HA, 22 23 EG, DF and BH; Writing - Original draft, EG; Resources, RS and RH; Writing - Review and 24 25 26 Editing, RS, RH, ML, MD, NK, MS, KT, JP, HA, FH, DJ, and DF; Supervision and Funding 27 28 Acquisition, ML and DF; Project Administration, EG. 29 30 31

32 http://bmjopen.bmj.com/ 33 Acknowledgements 34 35 Sarah Schlievert for administrative assistance. 36 37 38 39

40 REFERENCES on September 30, 2021 by guest. Protected copyright. 41 42 43 1) Jackson HJ, Rafiq S, Brentjens RJ. Driving CAR T-cells forward. Nat Rev Clin Oncol. 2016; 44 45 46 13: 370-383 47 48 2) Maude S, Barrett DM. Current status of chimeric antigen receptor therapy for 49 50 51 haematological malignancies. Br J Haematol. 2016; 172: 11-22 52 53 54 55 56 57 58 59 60 21 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019321 on 29 December 2017. Downloaded from Page 23 of 38 BMJ Open

1 2 3 3) Lee DW, Kochenderfer JN, Stetler-Stevenson M, et al. T cells expressing CD19 chimeric 4 5 6 antigen receptors for acute lymphoblastic leukaemia in children and young adults: a phase 1 7 8 dose-escalation trial. Lancet. 2015; 385: 517-528 9 10 11 4) Davila ML, Riviere I, Wang X, et al. Efficacy and toxicity management of 19-28z CAR T 12 13 cell therapy in B cell acute lymphoblastic leukemia. Sci Transl Med. 2014;6:224ra225 14 15 For peer review only 16 5) Porter DL, Levine BL, Kalos M, Bagg A, June CH. Chimeric antigen receptor-modified T 17 18 19 cells in chronic lymphoid leukemia. N Engl J Med. 2011; 365: 725-733 20 21 22 6) Yu S, Li A, Liu Q, et al. Chimeric Antigen Receptor T Cells: A Novel Therapy for Solid 23 24 Tumors. J Hematol Oncol. 10. Journal of Hematology & Oncology. 2017: 10. 25 26 27 7) Garde D. Juno pulls the plug on a once-promising cancer treatment. March 7, 2017. < 28 29 https://www.statnews.com/2017/03/01/juno-cancer-treatment/>. Accessed June 15, 2017. 30 31

32 8) Subject of Kite Drug Trial Dies, Sending Stock Down -- Market Talk. (2017, May 08). Dow http://bmjopen.bmj.com/ 33 34 Jones Institutional News, p. Dow Jones Institutional News, Accessed June 15, 2017. 35 36 37 9) Juno Therapeutics shuts down clinical trial of cancer drug after patient deaths. (2017, Mar 38 39 01). The Seattle Times, p. Technology, Accessed Jun 15, 2017.

40 on September 30, 2021 by guest. Protected copyright. 41 42 43 10) Maus MV, Haas AR, Beatty GL, et al. T cells expressing chimeric antigen receptors can 44 45 cause anaphylaxis in humans. Cancer immunol research. 2013; 1: 26-31. 46 47 48 11) Zhang E, Xu H. A new insight in chimeric antigen receptor-engineered T cells for cancer 49 50 immunotherapy. J Hematol Oncol. 2017; 10: 1. 1. 51 52 53 12) Cartellieri M, Bachmann M, Feldmann A, et al. Chimeric antigen receptor-engineered T 54 55 cells for immunotherapy of cancer. J Biomed Biotechnol. 2010;2010:956304. 56 57 58 59 60 22 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019321 on 29 December 2017. Downloaded from BMJ Open Page 24 of 38

1 2 3 13) Zhang Tf, Cao L, Xie J, et al. Efficiency of CD19 Chimeric Antigen Receptor-modified T 4 5 6 Cells for Treatment of B Cell Malignancies in Phase I Clinical Trials: A Meta-analysis. 7 8 Oncotarget. 2015; 6: 32 9 10 11 14) Zhu Y, Tan Y, Ou R, et al. Anti CD19 Chimeric Antigen Receptor modified T Cells for B 12 ‐ ‐ ‐ 13 14 15 cell Malignancies:For A Systematicpeer Review review of Efficacy and only Safety in Clinical Trials. Eur J 16 17 Haematol. 2016;96:4:389-96. 18 19 20 15) Anwer F, Shaukat A, Zahid U, et al. Donor origin CAR T cells: graft versus malignancy 21 22 effect without GVHD, a systematic review. Immunotherapy. 2017; 9: 2.123–130. 23 24 25 16) Holzinger A, Barden M, Abken H. The Growing World of CAR T Cell Trials: A Systematic 26 27 Review. Cancer Immunol Immun. 2016; 65:12:1433-450. 28 29 30 17) Beall's List of Predatory. Journals and Publishers. (n.d.). < http://beallslist.weebly.com/>. 31

32 http://bmjopen.bmj.com/ 33 Accessed June 13, 2017. 34 35 18) Moher D, Shamseer L, Clarke M, et al. Preferred Reporting Items for Systematic Review 36 37 and Meta-Analysis Protocols (PRISMA-P) 2015 statement. Syst Rev. 2015; 4: 1. 1. doi: 38 39 10.1186/2046-4053-4-1

40 on September 30, 2021 by guest. Protected copyright. 41 42 43 19) Fergusson D, Laupacis A, Salmi LR, Mcalister FA, Huet C. What Should Be Included in 44 45 Meta-analyses? An Exploration of Methodological Issues Using the ISPOT Meta-analyses. 46 47 International J Technol Assess. 2000; 16: 4. 1109-19. 48 49 50 20) Mcgowan J, Sampson M, Salzwedel DM, Cogo E, Foerster V, Lefebvre C. PRESS Peer 51 52 53 Review of Electronic Search Strategies: 2015 Guideline Statement. J Clin Epidemiol. 2016; 54 55 75: 40-46. 56 57 58 59 60 23 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019321 on 29 December 2017. Downloaded from Page 25 of 38 BMJ Open

1 2 3 21) Effective Practice and Organisation of Care (EPOC). Data collection form. EPOC Resources 4 5 6 for review authors. Oslo: Norwegian Knowledge Centre for the Health Services; 2013. 7 8 Available at: http://epoc.cochrane.org/epoc-specific-resources-review-authors 9 10 11 22) Nishino M, Jagannathan J, Ramaiya N, Van Den Abbeele A. Revised RECIST Guideline 12 13 Version 1.1: What Oncologists Want to Know and What Radiologists Need to Know. Am J 14 15 For peer review only 16 Roentgenol. 2010; 195: 2. 281-9. 17 18 19 23) Choi H, Charnsangavej C, de Castro Faria S, Tamm EP, Benjamin RS, Johnson MM, 20 21 Macapinlac HA, Podoloff DA. CT evaluation of the response of gastrointestinal stromal 22 23 tumors after imatinib mesylate treatment: a quantitative analysis correlated with FDG PET 24 25 26 findings. Am J Roentgenol. 2004; 183: 6. 1619-28. PubMed PMID: 15547201. 27 28 29 24) Hallek M, Cheson B, Catovsky D, et al. (2008). Guidelines for the diagnosis and treatment 30 31 of chronic lymphocytic leukemia: A report from the International Workshop on Chronic

32 http://bmjopen.bmj.com/ 33 Lymphocytic Leukemia updating the National Cancer Institute Working Group guidelines. 34 35 36 Blood. 1996; 111: 12. 5446-56. 37 38 39 25) Brennan TA, Leape LL, Laird NM, et al. HH: Incidence of adverse events and negligence in

40 on September 30, 2021 by guest. Protected copyright. 41 hospitalized patients. Results of the Harvard Medical Practice Study I. N Engl J Med. 1991; 42 43 324: 370-376. 44 45 46 26) Jenkinson C, Coulter A, Bruster S. The Picker Patient Experience Questionnaire: 47 48 49 Development and validation using data from in-patient surveys in five countries. Intern J for 50 51 Qual Health C. 2001; 14: 5. 353-358. 52 53 54 27) Santana M, Feeney D, and Institute of Health Economics. (n.d.). The importance of 55 56 measuring health related quality of life (IHE report). Edmonton, Alta.: Institute of Health 57 58 59 60 24 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019321 on 29 December 2017. Downloaded from BMJ Open Page 26 of 38

1 2 3 Economics. 4 5 6 28) Porrata LF, Gertz MA, Inwards DJ, et al. Early Lymphocyte Recovery Predicts Superior 7 8 9 Survival after Autologous Stem Cell Transplantation in Non-Hodgkin Lymphoma: A 10 11 Prospective Study. Biol Blood Marrow Tr. 2008; 14: 7. 807-16. 12 13 14 29) Guo B, Moga C, Harstall C, et al. A principal component analysis is conducted for a case 15 For peer review only 16 series quality appraisal checklist. J Clin Epidemiol. 2016; 69: 199 – 207. 17 18 19 30) Higgins JPT, Green S (editors). Cochrane Handbook for Systematic Reviews of 20 21 22 Interventions Version 5.1.0 [updated March 2011]. The Cochrane Collaboration, 2011. 23 24 Available from www.handbook.cochrane.org. 25 26 27 31) Fergusson D, Laupacis A., Salmi LR, Mcalister FA, Huet C. What Should Be Included in 28 29 Meta-analyses? An Exploration of Methodological Issues Using the ISPOT Meta-analyses. 30 31 Intern J for Qual Health C. 2000; 16: 4. 1109-19. 32 http://bmjopen.bmj.com/ 33 34 32) Hunter, J.P., Saratzis, A., Sutton, A.J., et al. In meta-analyses of proportion studies, funnel 35 36 37 plots were found to be an inaccurate method of assessing publication bias. J Clin Epidemiol. 38 39 2014; 67: 8. 897-903.

40 on September 30, 2021 by guest. Protected copyright. 41 42 33) Moher D, Liberati A, Tetzlaff J, Douglas GA. Preferred Reporting Items for Systematic 43 44 Reviews and Meta-analyses: The PRISMA Statement. Ann Intern Med. 2009;151:4. 264-9. 45 46 47 34) Guyatt G, Oxman A, Vist G, et al. GRADE: An emerging consensus on rating quality of 48 49 50 evidence and strength of recommendations. BMJ. 2008; 336: 7650: 924. 51 52 53 54 55 56 57 58 59 60 25 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019321 on 29 December 2017. Downloaded from Page 27 of 38 BMJ Open

1 2 3 Appendix 1. Representative Search Strategy 4 5 6 7 Hematologic malignancies: 8 9 10 Database: Embase Classic+Embase, EBM Reviews - Cochrane Central Register of 11 12 Controlled Trials, Ovid MEDLINE(R) Epub Ahead of Print, In-Process & Other Non- 13 14 15 Indexed Citations,For Ovid peerMEDLINE(R) review Daily and Ovid MEDLINE(R) only. 16 17 18 Search Strategy: 19 20 21 1 ((chimeric antigen adj2 receptor*) and (therap* or treat* or immunity or 22 23 immunotherap* or cell*)).tw. 24 25 26 27 2 ((car adj3 t adj5 therap*) or (car adj3 t adj5 treat*)).tw. 28 29 30 3 (car adj3 t adj3 immunotherap*).tw. 31

32 http://bmjopen.bmj.com/ 33 4 (car therap* or (car adj2 t adj2 cell*)).tw. 34 35 36 5 ((modified or engineered) adj2 (t cell* or t lymphocyte*)).tw. 37 38 39 6 chimeric antigen receptor/

32 http://bmjopen.bmj.com/ 33 5 ((modified or engineered) adj2 (t cell* or t lymphocyte*)).tw. 34 35 36 6 chimeric antigen receptor/ 37 38 39 7 or/1-6

32 http://bmjopen.bmj.com/ 33 34 35 36 37 38 39

1 2 3 Appendix 2. Modified Institute of Health Economics Tool 4 5 6 Question Text Answer Text 7 Was the hypothesis/ aim/ objective of the study stated? Yes 8 Yes = The hypothesis/aim/objective was reported (includes 9 patients, intervention and outcome). 10 11 Partial/ unclear = Only one or two components (patients, 12 intervention, or outcome) were included. 13 No = The hypothesis/aim/ objective was not reported. 14 Partial/ unclear 15 For peer review only 16 No 17 Was the study conducted prospectively? Yes 18 Yes = It was clearly stated that the study was conducted 19 prospectively. 20 Partial/ unclear = Unclear or no information was provided. 21 22 No = The study clearly stated it was a retrospective study. 23 Partial/ unclear 24 No 25 Were patients from more than one centre? Yes 26 For example, you can deduce single centre if they state "Data was 27 28 taken from the Sloan Memorial Research Centre". 29 Yes = Patients were from more than one centre (multicentre study). 30 Partial/ unclear = Unclear where the patients came from. 31 No = Patients were from one centre. 32 Partial/ unclear http://bmjopen.bmj.com/ 33 34 No 35 Were patients recruited consecutively? Yes 36 Note: Not based on previously published protocols. Must be stated 37 in this paper. 38 Yes = There was a clear statement or it was clear from the context 39 that the patients were recruited consecutively; or the study stated

1 2 3 4 Partial/ unclear = Either inclusion OR exclusion were reported. 5 No = Neither inclusion nor exclusion criteria were reported. 6 Partial/unclear 7 No 8 Were the characteristics of the patients included in the study Yes 9 10 described? 11 Relevant characteristics: age, sex, malignancy type, 12 lymphodepletion, previous treatment, concomitant treatments, co- 13 morbidities. 14 Yes = All of the relevant patient characteristics were reported. 15 For peer review only 16 Partial/ unclear = >= 1 of the relevant characteristics were reported. 17 No = None of the relevant characteristic were reported. 18 Partial/ unclear 19 No 20 Did patients enter the study at a similar point in the disease? Yes 21 22 Yes = The paper states that entering patients are in relapsed/ 23 refractory setting. 24 Partial/ unclear = There was no baseline information on patients' 25 characteristics to make a judgment. 26 No = There was a wide range in the severity of the disease and co- 27 28 morbidities of patients at baseline. 29 N/A = N of 1 study. 30 Partial/ unclear 31 No 32 N/A http://bmjopen.bmj.com/ 33 34 Was the intervention of interest described? Yes 35 Relevant characteristics: 36 T-cell origin, CD configuration (type (i.e. CD19), co-stimulatory 37 domain(s), dosage regimen (dose, frequency, duration). 38 Yes = All of the relevant characteristics of the intervention were 39 reported.

1 2 3 4 No 5 Were relevant outcome measures established a priori in the Yes 6 introduction or methods section? 7 Yes = All relevant outcome measures were stated. 8 Partial/ unclear = Some, but not all of the relevant outcome 9 measures were stated. 10 11 No = None of the relevant outcome measures were stated. 12 Partial/ unclear 13 No 14 Were outcome assessors blinded to the intervention that patients Yes 15 For peer review only 16 received? 17 i.e. Did the study have 'independent outcome assessors'? 18 Yes = The outcomes were assessed by individuals who were not 19 aware of patient intervention. 20 - SELECT where blinding is not necessary. i.e. Mortality is the 21 22 outcome. 23 - SELECT where the blinding to the outcome does not influence the 24 assessment. i.e. Response to CAR-T. 25 Partial/ unclear = The study did not report whether the outcome 26 assessors were aware of the intervention. 27 No = It was clearly stated or obvious from the context that 28 29 individuals assessing outcomes were unblinded. 30 Partial/ unclear 31 No 32 Were the relevant outcomes measured using appropriate objective Yes http://bmjopen.bmj.com/ 33 34 or subjective methods? 35 Yes = Complete response/remission and >=1 secondary outcomes 36 (i.e, overall response rate, non-relapse mortality, relapse, overall 37 survival, adverse events (infection, neurotoxicity, cytokine release 38 syndrome, B-cell aplasia, graft versus host disease, other types will 39 be grouped by organ system affected and severity)

1 2 3 4 No 5 The study does not perform selective outcome reporting. Yes 6 Yes = The study protocol is registered and all of the study's pre- 7 specified outcomes of interest were stated in the methods section. 8 Partial/ unclear = Either study protocol registered or the study's pre- 9 specified outcomes of interest were stated in the methods section. 10 11 No = No study protocol registered and none of the study's pre- 12 specified outcomes of interest were stated in the methods section. 13 14 Partial/unclear 15 For peer review only 16 No 17 Were details of the statistical tests reported? Yes 18 Yes = The statistical tests were reported in the study. 19 Partial/unclear = Statistical tests only partially described or reported 20 elsewhere (e.g previous paper, or protocol). 21 22 No = The statistical tests were not described in the study. 23 N/A = N of 1 study. 24 Partial/unclear 25 No 26 N/A 27 28 Was follow-up period reported? Yes 29 Yes = follow-up information was reported. 30 No = Length of follow-up was not reported. 31 No

1 2 3 4 Were both competing interests and sources of support for the study Yes 5 reported? 6 Yes = Both competing interests and sources of support (financial or 7 other) received for the study were reported; or the absence of 8 support was acknowledged. 9 Partial/ unclear = Either the competing interest or source of support 10 11 was reported. 12 No = Neither competing interests nor sources of support were 13 reported. 14 Partial/ unclear 15 For peer review onlyNo 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31

32 http://bmjopen.bmj.com/ 33 34 35 36 37 38 39

1 1 : : : : : : : : 20 18-20 18-20 19 20-21 2 4

1 number(s) number(s) Page Page 4:1 44

4 2015 No

Yes Information reported reported Information

Systematic Reviews Systematic 5:15 55

5 2016

Systematic Reviews Systematic For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

For peer review only

P P 2015 Checklist P P 2015 Checklist P 2015 Checklist 4 2 -P -- 2015 Checklist

2 2 Page 38 of 16-17 16-17 15-16 12-15 10-11 10 10 10 9 9 7-8 8 4-6 number(s) number(s) Page Page

Yes Information reported reported Information

http://bmjopen.bmj.com/ BMJ Open on September 30, 2021 by guest. Protected copyright.

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml For peer review only Describe plannedDescribe method of extracting data reportsfrom (e.g., forms, piloting doneindependently, duplicate),in any processes for obtaining and dataconfirmingfrom investigators Describe anticipatedDescribe methods forassessing risk of bias of individual includingstudies, whether this be will done outcome at the level, study or or both; state this howinformation bewill in used data synthesis List and define and List which all outcomes for data will be sought, including prioritization of main and outcomes,additional rationale with Present draft draft Present of search strategy to be forused at least one database,electronic including planned thatlimits, such it could be repeated Describe allintendedDescribe information (e.g., sources databases,electronic contact withstudy authors, registers, trial other grey or with literature sources) planned dates of coverage List and define and List all variables for which bedata will sought (e.g.,PICO fundingitems, sources), any pre-planned data assumptionsand simplifications Specify the the Specify study (e.g.,characteristics PICO, study design, setting, time frame)reportand characteristics (e.g., years considered, language, status)publication to be used criteria as for fortheeligibility review Provide an explicit Provide an explicit statementthe of question(s) reviewthe addresswill reference with to participants, interventions, comparators, outcomesand (PICO) State theState process willthat be forselectingused twostudies (e.g., reviewers)independent through phase each ofthe review (i.e., screening, eligibility, inclusion and in meta-analysis) If data If are forappropriate quantitative describesynthesis, planned summary measures, methods of handling handling methodsanddata, of combining data from studies, including any planned exploration of Checklist item item Checklist 11c 11c 15a Describe undercriteria which study bedata willquantitatively synthesized 14 13 15b # # 9

8 10

12 7

6 the Describe rationalereview for the the in of context knownwhat already is

3 3 18 16 16-17 17 number(s) number(s) Page Page

Yes Information reported reported Information

http://bmjopen.bmj.com/ BMJ Open on September 30, 2021 by guest. Protected copyright. , Kendall’s tau) 2

Confidence inConfidence evidencecumulative Meta-bias(es) Section/topic Section/topic

Page 39 of 38 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60