Restoring T-cell Cytotoxicity To Correct Familial Hemophagocytic Lymphohistiocytosis

Hemophagocytic lymphohistiocytosis (HLH) is a devastating syndrome marked by runaway systemic inflammation that, if uncontrolled, leads to multi-organ failure and death. Primary or familial HLH (FHL) occurs in the presence of an underlying, recessively inherited predisposing genetic defect in immune function.^1,2  FHL typically manifests in the first six months of life. Viruses, including Epstein-Barr virus (EBV), are frequent triggers. Immunochemotherapy, usually the combination of dexamethasone and etoposide, is used to temporize hyperinflammation and bridge patients to allogeneic hematopoietic cell transplantation (HCT), without which FHL is fatal within months. Even so, the probability of survival after HCT is around 50%. Inadequate disease control prevents patients from reaching HCT and contributes to poor outcomes after HCT.³  Therapies that more effectively control FHL hyperinflammation could enable more patients to reach HCT and improve post-HCT survival.

Two-thirds of cases of FHL are caused by loss-of-function mutations in perforin (PRF1; FHL2) or Munc13-4 (UNC13D; FHL3). Perforin is a component of T-cell and natural killer (NK)–cell cytotoxic granules and causes osmotic lysis of target cells, while Munc13-4 is essential in priming cytolytic granule secretion prior to vesicle membrane fusion.⁴  Either defect removes the “off switch” built into immune responses. Effector, target, and antigen-presenting cells cannot be eliminated and remain in prolonged contact, leading to undampened cytokine production, lymphocyte proliferation, and macrophage activation. Consequently, immune-mediated, life-threatening multi-organ dysfunction occurs. Xun Li, MD, and colleagues hypothesized that it is possible to restore negative feedback to immune responses and control FHL by repairing the genetic defect in cytolytic granule function in a population of long-lived memory T cells adoptively transferred back to the patient.⁵ 

First, the authors modeled EBV-triggered FHL by introducing inducible EBV protein expression in a subset of B cells into mice with genetic perforin deficiency.^6,7  Inducing EBV protein expression rapidly led to fulminant lymphoproliferative disease and hyperinflammation recapitulating clinical features of FHL2, such as hepatosplenomegaly from expanded EBV protein-expressing B cells, cytopenias, and elevated transaminases, and lactate dehydrogenase. Next, they developed a repair for PRF1 using CRISPR-Cas9 to introduce a double-stranded DNA break, adeno-associated virus (AAV) to deliver a donor DNA template for homology-directed repair (HDR), and a small-molecule inhibitor of non-homologous end joining (M3814) to skew edits towards HDR. Repairing perforin-deficient mouse CD8⁺CD44^hi memory T cells restored their perforin expression and ability to respond to EBV protein–expressing B cells in vitro. In the mouse model, adoptive transfer of perforin-repaired T cells markedly reduced clinical and laboratory manifestations of EBV-triggered FHL2 and prevented death from lethal hyperinflammation triggered by induced EBV protein expression.

Moving from mouse to human, Dr. Li and colleagues devised two repair strategies for a patient with FHL2 with exon 3 PRF1 mutations: one employing a template with a repaired version of exon 3 and the other replacing exon 3 with full-length PRF1 cDNA with preceding self-cleaving T2A amino acid sequence to correct most exon 2 and 3 mutations. Edited T cells from a healthy third-party donor had similar perforin expression, proliferation, and in vitro cytolytic function compared to wild-type cells. Both repair strategies rescued FHL2 patient T cells’ ability to kill target cells, although the repaired exon 3 template was more efficient than full-length PRF1. Most repaired patient CD8+ T cells expressed CD45RA, CD62L, and CD95 without markers associated with T-cell inhibition, suggesting a T memory stem cell (T_SCM)–like phenotype. Similarly, repair of CD8+ T cells from a patient with FHL3 using a template with a repaired UNC13D exon 10 yielded T cells with restored cytotoxic function and T_SCM-like phenotype.