Types of studies.

We will include randomized clinical trials (RCTs) with a concurrent control group and compare Pola-BR with other interventions for the treatment of R/R DLBCL. We will exclude non-randomized trials, single group before-and-after trials, case series, and case reports. We will also exclude cohort studies because they are not suitable for assessing the effectiveness of Pola-BR. In addition, we will exclude non-original studies, such as narrative reviews, systematic reviews, and scoping reviews.

Types of participants.

We will enroll adult patients (18 years and older) who have a confirmed diagnosis of R/R DLBCL, regardless of gender, race, ethnicity, disease severity, or stage. We will exclude patients with other types of lymphoma, such as follicular lymphoma, mantle cell lymphoma, or Hodgkin lymphoma, or contraindications to Pola-BR or any of the comparator interventions.

Interventions.

This review will include studies evaluating the experimental intervention of Pola-BR for the treatment of R/R DLBCL. Polatuzumab vedotin acts as an antibody-drug conjugate against CD79b, while Bendamustine is a chemotherapy agent, and rituximab is a monoclonal antibody directed against CD20. Eligible studies will include any dosage, duration, frequency, or route of administration of Pola-BR consistent with the manufacturer’s recommendations or accepted clinical practice guidelines.

Comparators.

Comparative interventions will be considered as follows: no intervention or placebo, standard treatments such as rituximab monotherapy, rituximab in combination with chemotherapy, or stem cell transplantation, and other experimental interventions that include novel drugs, immunotherapies, or gene therapies.

Outcomes of interest.

Studies will be included if they reported at least one of the following outcomes:

Primary outcome.

• OS: It will measure the length of time from randomization (when participants are assigned to a treatment group) until death from any cause.

Secondary outcomes.

• PFS: This outcome will assess the time from randomization to disease progression (worsening), relapse (return of the disease), or death from any cause. PFS provides an indication of how long treatment delays disease progression.
• AEs: This category will record any adverse or unintended effects that participants experience as a result of treatment. The severity of these events will be assessed using the Common Terminology Criteria for Adverse Events, a standardized system for classifying adverse events in clinical trials.

Studies that do not assess the primary endpoint or explicitly aim to assess a prespecified endpoint will be excluded from this review.

Database search.

The following electronic databases will be searched from inception to April 2024: Medline/PubMed, Cochrane Library, Web of Science, Scopus, EMBASE, ProQuest, EU Clinical Trials Register, WHO International Clinical Trials Registry Platform, and ClinicalTrials.gov. Search strategies are tailored to the specific syntax and functionality of each database.

Gray literature search.

The database search will be supplemented by gray literature searches, including unpublished or non-indexed sources. Relevant sources include:

1. Research reports from organizations such as NICE, ASCO, and ESMO.
2. Conference proceedings and abstracts from oncology meetings (e.g., ASH, EHA, ICML, ISPOR, SMDM, HTA).
3. Theses and dissertations from repositories such as ProQuest Dissertations.
4. Reference lists of eligible studies and relevant systematic reviews.
5. Manual search of three prominent hematological malignancy journals: Cancers, British Journal of Hematology, and Leukemia and Lymphoma, to ensure comprehensive literature coverage. Experts in the field will also be contacted to identify additional or ongoing studies that are not publicly available.

Search syntax.

The detailed search strategy for PubMed is provided in Table 1.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. PubMed search strategy using population and intervention elements of the PICO framework, January 2000 to April 2024.

https://doi.org/10.1371/journal.pone.0308247.t001

Search documentation and management.

The search process will be documented and reported using the PRISMA flow chart. Details recorded will include the date, database, search terms, and number of records retrieved for each search, as well as any changes or updates to the strategy. Reference management software like Mendeley will be used to store, organize, and deduplicate records, which will be exported to screening software like Rayyan to facilitate the screening and selection process. A Comprehensive search will be carried out across all databases by the end of April 2024.

For the Web of Science and the Cochrane Library, we will use the integrated date filters in the databases themselves.

Objective.

This meta-analysis aims to synthesize existing evidence on the efficacy and safety of combination therapy with Pola-BR for the treatment of DLBCL. The analysis will be carried out on the condition that at least five studies are included to ensure a reliable statistical analysis [25].

Statistical methods.

The synthesis will use either a fixed-effects or a random-effects model based on the degree of methodological heterogeneity observed in the included studies. For the fixed effects model, the inverse variance method will be used, while for the random effects model, the Der Simonian and Laird method will be used.

Effect measures.

• Dichotomous Outcomes: Risk ratios will serve as the effect measure.
• Continuous Outcomes: Mean Differences will be used for outcomes measured on identical scales, while Standardized Mean Differences will be used for outcomes measured on different scales.
• Primary and other time-to-event outcomes: These outcomes will be analyzed using hazard ratios as an effect measure for time-to-event data.

Heterogeneity assessment.

Heterogeneity between the studies will be quantified using the I² statistic with the following thresholds:

• Low: 0–25%
• Moderate: 25–50%
• Substantial: 50–75%
• High: 75–100%

Additionally, the p-value of the Cochran Q test will be reported to further evaluate the statistical heterogeneity.

Subgroup analyses.

Subgroup analyzes will examine potential sources of heterogeneity, focusing on:

1. Study Design: The impact of different study designs on treatment effect estimates.
2. Age: Variations in prognosis and treatment response in DLBCL patients according to age.
3. Gender: Influence of sex on the efficacy and toxicity of chemotherapeutic regimens.

In cases of substantial heterogeneity (I² > 50%), additional post-hoc subgroup analyses will be considered based on factors such as disease stage or specific intervention characteristics (e.g., dose, duration, route of administration).

Meta-regression analyses.

Meta-regression will examine the associations between study-level covariates and effect size estimates using a random-effects model with restricted maximum likelihood estimation. The significance level for meta-regression will be adjusted based on the number of studies:

• Fewer than 10 studies: p-value < 0.15
• 10–20 studies: p-value < 0.10
• More than 20 studies: p-value < 0.05

Publication bias assessment.

Potential publication bias will be assessed using funnel plots and statistical tests:

• Funnel Plots: Visual inspection of plot symmetry.
• Begg’s [26] and Egger’s [27] tests: A p-value < 0.1 will indicate significant publication bias.

The trim-and-fill method, a type of sensitivity analysis, [28] will estimate the number of missing studies and adjusts the pooled effect estimate accordingly.

Sensitivity analyses.

Sensitivity analyzes will evaluate the robustness of the overall results:

• Leave-One-Out Method: One study will be excluded at a time to assess its impact on the overall results [29].
• Model comparison: Various statistical models, weighting methods, effect measures, and heterogeneity thresholds will be employed to observe any changes in results.

Quality analysis.

The quality of the included studies will be assessed, and their relationships with effect estimates will be examined:

• Quality Categories: Studies will be categorized as high, moderate, or low quality based on risk of bias.
• Comparative Analysis: Effect estimates of different quality categories will be compared, with the p-value and I² statistic reported for each quality analysis.

Software and data handling.

STATA software version 14.2 software will be used to perform the meta-analysis and meta-regression analyses, manage data extraction, and assess quality. Any missing data or outliers in the primary studies will be documented along with the methods used to process them.

Interpretation and limitations.

The results will be interpreted taking into account the size and direction of the pooled effect estimates, the precision and width of the confidence intervals, the statistical significance (p-values), the clinical significance of effect sizes, and heterogeneity measures (the I² statistic and Cochran’s Q test). Implications for clinical practice and policy making, gaps in the evidence, and directions for future research will be discussed. Limitations and potential biases, such as: selective reporting, quality of primary studies, and generalizability of findings, will be acknowledged.

This protocol is intended to provide a rigorous framework for evaluating the efficacy of Pola-BR in the treatment of R/R DLBCL, thereby contributing valuable insights to clinicians and researchers in the field.

Approach overview.

The certainty of evidence for each outcome will be evaluated employing the Grading of Recommendations Assessment, Development, and Evaluation (GRADE) approach [30]. This method assesses the certainty of evidence based on multiple factors, including study design, risk of bias, inconsistency, indirectness, imprecision, and other relevant considerations.

Certainty ratings.

After assessing these factors, the certainty of evidence for each outcome will be categorized as “high,” “moderate,” “low,” or “very low.” A summary of results will be constructed, presenting these certainty ratings along with the effect estimates and confidence intervals for each outcome. This comprehensive approach ensures a systematic and transparent assessment of the certainty of evidence, and facilitates informed decision-making and interpretation of study results.

Protocol amendment procedure.

If amendments to the protocol are necessary, we will document and report those amendments as follows:

1. Circumstances for amendments: Potential amendments may arise due to changes in the research question, the availability of new evidence, or methodological considerations identified during the review process.
2. Documentation: Any amendments will be documented in a log, including a description of the amendment, the rationale, the reason and the date of the change.
3. Reporting: Amendments will be reported in the final manuscript of the systematic review clearly distinguishing between pre-planned and post hoc analyses.

By adhering to these guidelines, we aim to ensure transparency and reproducibility in our systematic review and meta-analysis processes.

Study characteristics.

The included studies differ in design, population, interventions and measured outcomes. These studies generally include a series of RCTs to evaluate the efficacy and safety of combination therapy with Pola-BR for the treatment of R/R DLBCL. Key characteristics such as sample size, study duration, follow-up period, and specific inclusion and exclusion criteria are described in detail to understand the context and applicability of the results of each study.

Treatment efficacy.

The qualitative synthesis highlights the OS and PFS outcomes reported in the studies. While some studies demonstrate a significant improvement in OS and PFS with Pola-BR compared to other treatment regimens, others indicate variability based on patient demographics, disease stage, and previous treatments. The narrative integrates these findings to provide a holistic view of the performance of Pola-BR in different subgroups of patients.

Safety and adverse events.

Safety profiles and AEs associated with Pola-BR are critically examined. The severity and frequency of AEs will be compared across studies to identify consistent patterns or significant concerns. The qualitative synthesis also discusses how these AEs impact the overall tolerability of the therapy.

Heterogeneity and contextual factors.

The heterogeneity of the study results will be examined through a detailed examination of potential sources, such as differences in study design, patient populations, interventions, and outcome measures. The synthesis examines how these factors contribute to different efficacy and safety outcomes. Contextual factors such as the setting of the studies (e.g., geographical location, healthcare system) are also considered to understand the broader applicability of the findings.

Overall interpretation.

The qualitative synthesis integrates findings from individual studies to provide a comprehensive account of the use of Pola-BR in the treatment of R/R DLBCL. It identifies trends, common results and outliers and provides a nuanced interpretation that takes into account the complexity and variability of the data. This approach ensures that, even without meta-analysis, meaningful and actionable insights are gained that inform clinical practice and future research.