Abstract

   Gene and cell therapies pose safety concerns due to potential
   insertional mutagenesis by viral vectors. We introduce MELISSA, a
   regression-based statistical framework for analyzing Integration Site
   (IS) data to assess insertional mutagenesis risk, by estimating and
   comparing gene-specific integration rates and their impact on clone
   fitness. We characterized the IS profile of a lentiviral vector on
   Mesenchymal Stem Cells (MSCs) and compared it with that of
   Hematopoietic Stem and Progenitor Cells (HSPCs). We applied MELISSA to
   published IS data from patients enrolled in gene therapy clinical
   trials, successfully identifying both known and novel genes that drive
   changes in clone growth through vector integration. MELISSA offers a
   quantitative tool to bridge the gap between IS data and safety and
   efficacy evaluation, facilitating the generation of comprehensive data
   packages supporting Investigational New Drug (IND) and Biologics
   License (BLA) applications and the development of safe and effective
   gene and cell therapies.

   Subject terms: Statistical methods, Mesenchymal stem cells, Data
   mining, Haematopoietic stem cells
     __________________________________________________________________

   Viral vector integration can affect the safety of gene and cell
   therapies. Here, authors introduce MELISSA, a regression-based
   statistical tool that quantifies integration site risks and clone
   growth effects, aiding the safety evaluation of therapies in both
   research and clinical settings.

Introduction

   Gene therapy using Hematopoietic Stem and Progenitor Cells (HSPC)
   modified with viral vectors has emerged as a promising approach for
   treating rare monogenetic diseases^[42]1,[43]2. This strategy involves
   introducing functional copies of therapeutic genes into patients’
   HSPCs, enabling them to produce the missing or malfunctioning protein
   and restore normal cellular function. Due to their efficient gene
   delivery capabilities, viral vectors have also found extensive
   application in novel cell therapies such as Chimeric Antigen Receptor
   (CAR)-T immunotherapies, where Lenti/Retroviral Vectors are used to
   deliver CAR genes in a patient’s T cell.

   The capability for viral vectors to integrate their cargo DNA into
   unpredictable locations within the host genome carries a certain risk
   of Insertional Mutagenesis (IM), defined as the disruption or
   dysregulation of a gene caused by a vector insertion. IM can lead to
   enhanced cell growth due to the activation of oncogenes or the
   disruption of tumor suppressor genes by the integrated DNA^[44]3–[45]6.
   Clonal dominance underscores the risk of a particular clone or a small
   subset of clones of transduced cells gaining a growth advantage,
   potentially skewing the clonal composition of the repopulating cells
   towards monoclonal or oligoclonal configurations that pose additional
   long-term safety and efficacy concerns, including oncogenic
   transformation^[46]7–[47]9.

   Regulatory agencies like the US FDA require rigorous preclinical safety
   and efficacy evaluations and 15 years of IM monitoring for patients
   treated with genetically modified HSPCs^[48]10. Recently, the US FDA
   extended monitoring to individuals receiving BCMA- or CD19-directed
   autologous CAR T cell immunotherapies. A comprehensive risk-benefit
   analysis is crucial for evaluating the potential for IM, yet the
   scientific community still lacks standardized indices, metrics, and
   methods to clearly distinguish between safe and unsafe integration
   profiles.

   The association of the integrome, namely the distribution of IS across
   the host genome, with genomic annotation data has elucidated the role
   of chromatin conformation, transcriptional activity, and the cellular
   genome 3D nuclear organization on IS frequency^[49]11–[50]15. These
   insights, combined with the analysis of oncogenic transformation
   mechanisms observed in clinical trials, enabled the generation of novel
   and safer viral vectors for HSPC gene therapies.

   Using IS analysis to monitor the clonal composition over time,
   scientists can gain insights into the diversity and dynamics of the
   engrafting cell population, assess the long-term efficacy, and the risk
   of clonal dominance^[51]7,[52]16–[53]18. From a safety perspective, IS
   analyses serve as the primary screening tool for identifying potential
   risks associated with uncontrolled clonal expansion by detecting
   abundant clones based on their relative contribution and mapping IS
   location within the host cell genome. Conventionally, the
   identification of abundant clones has been based on predefined
   percentage thresholds calculated on an individual sample basis.
   Leveraging IS information from multiple datasets to estimate the
   integrome and clonal contributions can improve the characterization of
   gene targeting preferences and associated risks and enhance the