Pre-transplant serum levels of claudin-3 as a predictive biomarker for intestinal GvHD and non-relapse mortality risk after allogeneic transplantation

Acute graft-versus-host disease (GvHD) is a major complication following allogeneic hematopoietic stem cell transplant (allo-HSCT) that contributes to high non-relapse mortality (NRM). The gastrointestinal system is most commonly involved. Gastrointestinal GvHD (GI GvHD) is difficult to treat and is the greatest cause of GvHD-related mortality.¹ Yet, identifying patients at risk of developing GI GvHD remains a challenge. Although some risk factors like chemotherapy and certain infections have shown to contribute, there is no widely accepted method for identifying high-risk patients. The availability of reliable pre-transplant biomarkers would help predict the risk of GI GvHD.

Shernan Holtan, University of Minnesota, Minneapolis, US, and colleagues performed a retrospective study investigating the association of pre-transplant gut barrier and inflammation biomarkers with the incidence of GI GvHD and 1-year NRM.²  Below is a summary of an abstract from the study presented during the 61^st ASH annual meeting.

Methods

• Pre-transplant serum samples collected from 528 patients undergoing first allo-HSCT
• Least absolute shrinkage and selection operator (Lasso) multivariate regression analysis of:
  • Gut barrier biomarkers: regenerating islet-derived 3a (REG3a), citruline, claudin-3, epidermal growth factor (EGF), lipopolysaccharide
  • Inflammation biomarkers: amphiregulin (AREG), suppressor of tumorgenicity 2 (ST2), T-cell immunoglobulin and mucin-domain containing-3 (TIM-3)
  • Confounding clinical factors included patient age, donor type, disease risk, GvHD prophylaxis (cyclosporine or tacrolimus/methotrexate or prednisone cyclosporine or tacrolimus/mycophenolate mofetil vs. sirolimus/ mycophenolate mofetil), and comorbidities

Patient Characteristics

• The median age was 41 years (<1–75 years)
• The majority of patients were male (61%)
• Cord blood was the most common source of cells (45%), followed by peripheral blood stem cells (28%) and bone marrow (27%)
• Over half of patients received reduced-intensity conditioning or nonmyeloablative conditioning regimens (54%), a third (32%) had myeloablative total body irradiation, and the remaining patients (14%) had myeloablative non-total body irradiation
• The majority of patients were classified as intermediate risk (53%), followed by non-malignant (27%), high-risk (11%), low-risk (9%) and very high-risk (1%)
• Many patients had a comorbidity index of 0 (49%), and the remaining patients had an index of 1–2 (27%) or 3+ (23%)

Results

• Lasso regression analysis incorporating clinical parameters revealed the correlation of
  • claudin-3, age, and an intermediate disease risk index with risk of GI GvHD
  • claudin-3, unrelated donor source, TIM3, and double umbilical cord blood graft with higher NRM risk
  • EGF was reversely correlated with NRM risk
• Only Claudin-3, a component of tight junctions highly expressed in colon, was associated with both GI GvHD and NRM (Table 1)

Table 1. GI GvHD or NMR biomarkers

Factor	Hazard ratio
GI GvHD biomarkers Claudin-3 AREG	1.16 0.99
NRM biomarkers TIM-3 Claudin-3 Reg3a EGF	1.59 1.38 1.04 0.99

• Claudin-3 threshold levels
  • 24.5 pg/ml for GI GvHD; 31% of patients with higher pre-transplant levels had increased incidence of GvHD versus 19% of those with lower levels
  • 27.1 pg/ml for 1-year NRM; 24% of patients with higher pre-transplant levels had increased risk of 1-year NRM versus 12% of those with lower levels
• Factors associated with lower pre-transplant claudin-3 levels
  • Female gender (p= 0.01)
  • Age < 50 years (p< 0.01)
  • Non-malignant disease (p< 0.01)
  • Lack of comorbidities (p< 0.01)

Conclusion

Claudin-3 is a potential pre-transplant biomarker of increased risk of GI GvHD and NRM in patients undergoing allo-HSCT. The data suggest that elevated pre-transplant levels may indicate higher susceptibility to the damage caused by chemotherapy, radiotherapy, and infections, resulting in GvHD. Further studies are needed to explore whether lower intensity treatment options would minimize the risk of GvHD in those patients.