The Risk Factors for IgG Hypogammaglobulinemia after Allogeneic Hematopoietic Stem Cell Transplantation and Its Impact on Transplant Outcomes

Introduction: The reconstitution of the immune system after allogeneic hematopoietic stem cell transplantation (allo-HCT) depends on multiple factors such as the conditioning regimen, age, stem cell source, and graft-versus-host disease (GVHD). Patient with chronic GVHD (cGVHD) can remain B-cell deficient even 1 year after allo-HCT. Although hypogammaglobulinemia post-transplant has been suggested as a poor prognostic factor for survival and transplant-related mortality (TRM), very little data has been published in the recent decade in the field of reconstitution of B-cell repertoires and hypogammaglobulinemia, particularly after the introduction of peripheral blood stem cell allo-HCT and in vivo T-cell depletion. This study aimed to identify the risk factors for hypogammaglobulinemia and evaluate the association between hypogammaglobulinemia and transplant outcomes in adult allo-HCT population.

Methods: We retrospectively reviewed 339 consecutive patients who received allo-HCT between 2009 and 2012 at the Princess Margaret Cancer Centre, Toronto, Canada. At lest one measurement of immunoglobulin level was available in 157 patients between 3 months and 1 year post allo-HCT. IgG hypogammaglobulinemia (hypo-IgG) was defined as IgG <7g/L. The Kaplan-Meier method was used for OS, and cumulative incidences considering competing risks were calculated for NRM and relapse. The independent student’s t-test was used to compare mean IgG levels in each subgroup. The Χ² test was used to identify the risk factors for hypo-IgG in univariate analysis and binary logistic regression was used in multivariate analysis. Multivariate analysis for OS was performed using the time-dependent cox proportional hazard model with cGVHD as a time-dependent covariate. Fine-Gray proportional hazard regression for competing events were used for NRM in multivariate analysis.

Results: The mean values of IgG were 6.67±0.41g/L (n=103) at 3 months, 6.93±0.53g/L (n=94) at 6 months, and 8.34±0.49g/L (n=136) at 1 year post allo-HCT. The proportions of patients with hypo-IgG (<7g/L) among those with available IgG level at the select time-points were 59.2% at 3 months, 59.6% at 6 months and 44.9% at 1 year.

Patients with lymphoid malignancies showed a lower IgG level at 3 months than those with other diseases (5.12g/L vs 7.25gL, p=0.041). Non-T-cell depletion (non-TCD) was associated with lower IgG levels at 6 months and at 1 year: 5.83g/L vs 8.98g/L at 6 months (p=0.004) and 6.97g/L vs 10.94g/L at 1 year (p<0.001). The presence of previous acute GVHD (aGVHD) grades 2-4 at 6 months was associated with a lower IgG level at 6 months (5.21g/L vs 8.66 g/L, p=0.001), but cGVHD at 6 months was not. However, a lower level of IgG at 1 year was observed among patients who developed cGVHD by 1 year than those who did not (7.79 g/L vs 10.38g/L, p=0.031).

The proportion of patients with hypo-IgG at 6 months was significantly higher in the lymphoid malignancies group (78% vs 56%, p=0.049), in the non-TCD group (71% vs 39%, p=0.003), in patients with aGVHD grades 2-4 (81% vs 38%, p<0.001) and with cGVHD (71% vs 50%, p=0.056). Binary logistic regression identified the following variables as independent risk factors for hypo-IgG at 6 months; non-TCD (hazard ratio (HR) of 3.61, p=0.16), aGVHD grades 2-4 (HR 8.45, p<0.001) and cGVHD (HR 3.31, p=0.025).

Overall survival at 2 years post allo-HCT was significantly lower in the group with hypo-IgG (n=56) than those with a normal IgG level (n=38) at 6 months (54.5% vs 86.6%, p=0.001). NRM at 2 years was also significantly higher in the hypo-IgG group than in the normal IgG group at 6 months (44.0% vs 3.6%, p<0.001). There was no difference in the relapse rate at 2 years between the two groups.

Multivariate analysis demonstrated that hypo-IgG at 6 months (HR 6.10, p=0.006) and aGVHD grade 2-4 (HR 3.23, p=0.31) were adverse prognostic factors while reduced-intensity conditioning (HR 0.27, p=0.028) and time-dependent cGVHD (HR 0.194, p=0.001) were associated with better OS. Furthermore, cGVHD was associated with lower NRM (HR 0.27, p=0.004) and hypo-IgG at 6 months was associated with higher NRM (HR 20.0, p=0.004).

Conclusion: A significant number of patients remain hypogammaglobulinemic at 6 months and even 1 year post allo-HCT. Non-TCD and acute and chronic GVHD were identified as risk factors for hypogammaglobulinemia. Hypogammaglobulinemia at 6 months was found to adversely affect OS and NRM.