Hypogammaglobulinemia during Rituximab Maintenance after Transplantation Is a Surrogate Marker for Disease Control in Patients with Mantle-Cell Lymphoma, an Analysis from the LyMa Trial

Introduction

Rituximab maintenance (RM) (375mg/m2 per infusion every 2 months for 3 years) in transplanted patients with mantle-cell lymphomas (MCL) prolongs disease control (LyMa trial, Le Gouill et al NEJM; NCT00921414). However, post-transplant RM might also induce long-term immune deficiency and thus increases risk of infection. To address these issues, we performed an ancillary pre-planned study based on the LyMa trial, a phase III trial that compared RM versus observation (Obs) after ASCT in MCL patients. We compared post-transplant immune-deficiency and its impact on PFS and OS in the RM vs Obs groups.

Method

All transplanted and randomized patients enrolled in the LyMa trial were eligible for the present study. The following data were collected during the post-ASCT period and monitored according to protocol procedure: febrile event, clinically documented infection, hospitalization for infection, neutropenia, hypogammaglobulinemia and T CD4 lymphocytes count. We also retrospectively collected the use of immune globulin (Ig) substitution. In the LyMa trial, patients were randomized between RM vs Obs after transplantation. To decipher the implication of ASCT or RM in immune recovery, treatment periods were divided in 4: < 6 months after randomization, from 6 to 12 months after randomization, from one to two year after randomization, and from 2 to 3 years after randomization (respectively periods A, B, C and D). Chi-square or Fisher's exact tests were used as appropriate to investigate differences between arms in each treatment period. For all tests, a two-sided p-value<0.05 was considered statistically significant

Results

240 patients were eligible, 120 in each arm. Patients' characteristics at diagnosis and inclusion were similar in the two arms. Number of hospitalizations due to infections was not statistically different in RM vs Obs in all periods. As previously shown, grade 3/4 infections incidence did not differ in the 2 arms. However, febrile events were more frequent in the RM arm (32 pts vs 11; 38 events vs 12) but this was statistically significant only in C and D periods; p=0,03 for the 2 periods. In all, 51 infections in 44 pts were reported in Obs vs 127 events in 82 pts in RM arm. This difference was also only statistically significant during the C period, p=0,001. Grade 4 neutropenia incidence and T CD4 count did not differ between the two arms in all tested periods. Hypogammaglobulinemia was statistically more frequent in RM during C and D periods (p=0,0001 and p< 0,0001, respectively). Mean level of gammaglobulinemia on D period was 6,50 g/L (range 0,6-11,7) in obs arm versus 4,99 (range 1,0-9,5) in RM arm (p< 0,0001). 36 pts in RM arm vs 10 pts in obs arm were substituted with Ig and the difference was statistically significant only in period D, p<0,0001.

Febrile and infectious episodes; neutropenia and T CD 4 lymphopenia did not modify PFS and OS. Patients with gammaglobulinemia < 6g/L in RM arm and in the whole cohort had longer PFS compared to pts who did not present hypogammaglobulinemia : 3-years PFS 93,2% vs 63,5% in RM arm HR = 0,294, 95% CI (0,113-0,767 and), p=0,01 and 3-years PFS 85,6% vs 63,6% in the whole cohort, HR adjusted on treatment arm=0,488 95% CI (0,287-0,830), p=0,008 . PFS was not modified by gammaglobulin level in the Obs arm and it did not modified OS in both arms.

We performed a multivariate analysis to determine which data were predictive of infectious events and delayed immune recovery (neutropenia, hypogamma, T CD 4 lymphopenia). This included all univariate parameters with p value < 0,2, among clinical and biological characteristics at diagnosis, response after induction and number of rituximab injections. Interestingly, among others expected parameters, complete response assessed by TDM was predictive of hypogamma with Odd Ratio 2,972 (1,263-6,994) p=0,0126. No value was predictive of neutropenia or T CD4 cytopenia.

Conclusion

As compared to observation, the use of post-transplant RM does not increase risk of neutropenia and T CD4 lymphopenia. However febrile and infectious events, hypogammaglobulinemia and Ig substitution are more frequent after one year post transplantation. Hypogamma < 6g/L is associated with longer PFS and complete morphologic response. This suggests that hypogammaglobulinemia could be a surrogate for disease response quality and duration. Our findings deserve to be confirmed.