AUTHORS
Matthew D'Ambra, BA, Leah Desrochers2, Rene Bermea3, Sunil Poudel, PhD4, Ming Lo5, Nicholas Crossland6, E. Lake Potter, PhD7, Mario Roederer8, Bruce Blazar, MD, PhD9, Leslie Kean, MD, PhD10, Victor Tkachev, PhD11
AFFILIATIONS
1 Massachusetts General Hospital, Charlestown, Massachusetts, United States
2 Massachusetts General Hospital, Charlestown, Massachusetts, United States
3 Massachusetts General Hospital, Charlestown, Massachusetts, United States
4 Massachusetts General Hospital, Charlestown, Massachusetts, United States
5 Boston University Chabanian & Avedisian School of Medicine, Boston, Massachusetts, United States
6 Boston University Chabanian & Avedisian School of Medicine, Boston, Massachusetts, United States
7 National Institute of Allergy and Infectious Diseases, Bethesda, United States
8 National Institute of Allergy and Infectious Diseases, Bethesda, United States
9 University of Minnesota, Minneapolis, United States
10 Boston Children's Hospital, Boston, United States
11 Massachusetts General Hospital, Charlestown, Massachusetts, United States
ABSTRACT
Early post-alloHCT prophylaxis with the α4β7-integrin antagonist vedolizumab (VDZ) protects from lower gastrointestinal (GI) aGVHD (Chen, Nat Med, 2024) but demonstrates limited efficacy as a 2nd line therapy for steroid-refractory (SR) GI aGVHD (Floisand, Bone Marrow Trans., 2021). To investigate the differential susceptibility to VDZ when used as GI aGVHD prophylaxis vs. treatment, we used the Serial Intravenous Staining (SIVS) technique (Tkachev, Sci Transl Med, 2021), which leverages the in vivo administration of uniquely fluorescent-labeled anti-CD45 antibodies 5 minutes before organ harvest to discriminate between intra- and extravascular compartments, and at 48 hours (h), 24 h, and 6 h before organ harvest to measure the speed of leukocyte immigration into tissues. The proportion of tissue leukocytes labeled by each SIVS pulse was quantified by flow cytometry, plotted over time, and summarized by the area under curve (AUC) as an integrated measure of cumulative tissue entry over the labeling window (48h).
Using this metric, we confirmed that aGVHD occurring after MHC-mismatched alloHCT in rhesus macaques increased the speed of migration of donor CD45RA-CCR7- CD4 and CD8 T effector-memory (Tem) cells to barrier organs, including lungs, and the small and large intestine. Next, we compared the effect of VDZ on T cell migration under homeostatic conditions and during active aGVHD. We first looked at the α4β7 blockade effect in healthy control (HC) animals by administering a rhesus-specific primatized VDZ (pVDZ) at a single dose of 50 mg/kg, which completely blocks α4β7 in other rhesus studies (Pereira, Clin. Immunol. 2009). Organs were harvested for analysis on day+6 (n=3). Compared to untreated controls (n=4), pVDZ reduced migration of both CD4 and CD8 T cells to the colon (CD4: 3.2-fold, CD8: 4.7-fold, both p<0.05), ileum (CD4: 3.3-fold; CD8: 2.9-fold, p=0.06), and jejunum (CD4: 3.9-fold, p=0.06; CD8: 5.8-fold, p=0.05). Unexpectedly, pVDZ also reduced CD8 T cell migration to the lung (1.8-fold, p<0.05) and the bronchoalveolar space (2.2-fold, p<0.05), where T cell migration has previously been considered α4β7/Madcam-1-independent. In the peripheral blood, pVDZ increased the proportion of α4β7+ T cells labeled at all SIVS times (48h+24h+6h+), representing a circulation-confined subset with low tissue entry (CD4 Tcm: 46.7±6.7 vs 1.1±0.5; CD8 Tcm: 38.7±7.7 vs 4.9±0.7; CD4 Tem: 55.1±4.9 vs 5.2±0.9; CD8 Tem: 45.8±10.7 vs 21.2±3.7, all p<0.05), without affecting intravascular retention of α4β7- T cells. Therefore, in homeostatic conditions, a4b7 blockade reduced the entry of T cells to multiple organs, while increasing the dwelling time of T cells in the vasculature.
Next, we administered pVDZ to alloHCT recipients on day+1, followed by assessment of T cell migration on day+7, during active aGVHD (n=4). Unlike the profound effect of pVDZ on T cell trafficking into both small and large intestines in HCs, after alloHCT, pVDZ diminished the entry of T cells only to the colon (CD4: 2.2-fold, CD8: 2.2-fold vs. untreated alloHCT recipients (n=4), p<0.05 for both), including the migration of Tem cells – the major gut-infiltrating subset during aGVHD. Even during pVDZ treatment, donor T cells were able to establish residency within the gut, as measured by donor chimerism levels (jejunum CD4: 21.1±8.2 in pVDZ vs 18.0±3.1 in untreated alloHCT cohort; CD8: 32.2±7.7 vs 28.6±9.9. Ileum CD4: 29.9±9.4 vs 18.0±1.4; CD8: 42.8±9.4 vs 22.5±2.7. Colon CD4: 37.1±9.5 vs 20.6±4.7; CD8: 59.6±13.2 vs 55.6±14.7, p>0.05 for all). Consistent with only a moderate reduction of T cell entry rate and unaffected donor chimerism in the gut T cell compartment, pVDZ treatment did not reduce histologic GVHD pathology scores in these GVHD-target organs, nor in the severity of aGVHD symptoms. In contrast to HCs, pVDZ-treated alloHCT recipients with decreased migration of T cells to the gut unexpectedly had increased entry of CD8 T cells to the lung. Also surprising was that during aGVHD pVDZ did not result in retention of blood T cells within the vasculature.
We identified a potential mechanism that likely limits the clinical success of VDZ in aGVHD: T cells can utilize α4β7-independent migration mechanisms, which allows them to bypass α4β7/Madcam-1 barriers and infiltrate the inflamed GI tract and other organs. These findings underscore the ongoing need to fully characterize the diverse pathways used by T cells to invade the gut during GI aGVHD.
DISCLOSURES
Victor Tkachev, PhD: Regatta Bio - Current equity holder in private company