Tocilizumab for Antibody Mediated Rejection Treatment in Lung Transplantation

Abstract

Tocilizumab (TCZ), an IL-6 inhibitor, has shown promise in the treatment of donor-specific antibodies (DSA) and chronic antibody-mediated rejection (AMR) in renal transplant recipients. However, its use in lung transplantation has not been described. This retrospective case-control study compared AMR treatments containing TCZ in nine bilateral lung transplant recipients to 18 patients treated for AMR without TCZ. Treatment with TCZ resulted in more clearance of DSA, lower recurrence of DSA, lower incidence of new DSA, and lower rates of graft failure when compared to those treated for AMR without TCZ. The incidence of infusion reactions, elevation in transaminases and infections were similar between the two groups. These data support a role for TCZ in pulmonary AMR and establish preliminary evidence to design a randomized controlled trial of IL-6 inhibition for the management of AMR.

Development of de novo donor-specific antibodies (DSA) against human leukocyte antigens (HLA) following lung transplantation has been reported to occur in 13–56% of recipients and has been associated with an increased rate of, and shorter time to, chronic lung allograft dysfunction (CLAD), increased mortality, and antibody-mediated rejection (AMR).^1–4 The IL-6 inhibitor tocilizumab (TCZ) has been described for the treatment of AMR in renal transplant recipients^5,6 but the use of TCZ for AMR in lung transplantation has not been reported.

Adult bilateral lung transplant recipients after 1/1/2011 with HLA DSA and diagnosed AMR were retrospectively reviewed for inclusion into the present study. Nine patients received an AMR treatment regimen containing TCZ (8 mg/kg monthly with a maximum dose of 800 mg per dose) and were compared to 18 historically treated AMR patients who received treatment regimens without TCZ (Fig. 1A). Patients were included in the control arm (non-TCZ) if they had a bilateral lung transplant and received a combination of therapies for AMR in the setting of positive HLA DSA. Patients were followed for 12 months after AMR onset or until patient death if it occurred sooner than 12 months after treatment. AMR treatment was assessed for efficacy through HLA DSA mean fluorescence intensity (MFI) changes following treatment, as well as FEV₁ changes from immediately prior to AMR treatment to the end of follow-up. Graft failure was also compared between TCZ and non-TCZ treated patients. AMR was classified into definite, probable, and possible based on the 2016 ISHLT criteria. DSA clearance was defined as two consecutive negative HLA DSA screens. DSA recurrence was a positive HLA DSA screen following DSA clearance, and new DSA was defined as DSA to a different HLA epitope than the HLA epitope that was positive at the time of AMR treatment. Progression was defined as the need for salvage AMR treatment or if the patient died due to respiratory failure within 12 months following AMR onset. Graft failure was defined as retransplant or death. Safety parameters included infusion reactions, liver function test (LFT) abnormalities (i.e., elevations in transaminases to >5x the upper limit of normal), and infectious complications. This study was approved by the local institutional review board.

Figure 1.

CONSORT flow diagram of study exclusions (Panel A), outcomes of donor-specific antibodies up to 12 months after AMR treatment completion (Panel B), average FEV₁ trend from the start of AMR treatment up to 12 months after AMR treatment completion (Panel C), Kaplan-Meier graph of graft failure in the 12 months after AMR treatment completion, P=0.213 via log-rank test (Panel D).

Thirty-three patients were diagnosed with AMR following lung transplantation. Exclusion criteria is outlined in Figure 1A. Two patients were excluded due to death shortly after AMR diagnosis before receipt of any therapies and one patient only received IVIg which had been started pre-AMR diagnosis for DSA; IVIg is the cornerstone of treatment for AMR at our institution and thus the patient was excluded given the aim of the study was to compare AMR treatments. One non-TCZ patients did not have repeat DSA (due to loss of follow-up shortly after AMR treatment). The average time from transplant to AMR treatment was 756 ± 499 days in non-TCZ treated patients and 402 ± 341 days in TCZ treated patients (P=0.068). Patient demographics can be found in Supplementary Table S1. Most patients in each treatment group were on maintenance immunosuppression with a calcineurin inhibitor, antimetabolite, and steroid at the time of AMR diagnosis. The average MFI of the immunodominant DSA at the time of AMR diagnosis was similar between groups (7763 in TCZ group compared to 8199 in non-TCZ group, P=0.878).

Table 1 shows the comparison in DSA, AMR treatment, AMR classification, and DSA outcomes between the two groups. Average follow-up time was 261 days in non-TCZ treated patients and 211 days in TCZ treated patients (P=0.409). Three TCZ patients (33.3%) and three non-TCZ patients (16.7%) met criteria for CLAD at the time of AMR diagnosis. There were more HLA DSA rechecks in non-TCZ treated patients (median of six versus two in TCZ treated patients). In TCZ treated patients, the average number of monthly TCZ doses received was six (range 2 – 12). TCZ treated patients had more favorable DSA clearance, less DSA recurrence and development of new DSA, and less progression but these differences were not statistically significant. Graft failure occurred in 11.1% of TCZ treated patients compared to 50.0% of non-TCZ patients (P=0.049); when the patient who was excluded because of no repeat DSA due to death shortly after AMR treatment was included in the graft failure analysis, failure rates were 11.1% of TCZ patients compared to 52.6% of non-TCZ patients (P=0.036). Time to graft failure did not differ between groups (Fig. 1D). No patients underwent retransplant in the year following AMR diagnosis. The average FEV₁ and FVC at the time of AMR treatment was 2.31 ± 0.86 and 2.87 ± 0.97 liters in non-TCZ treated patients compared to 2.31 ± 0.86 and 2.80 ± 0.81 liters in TCZ treated patients (P=0.689 for FEV₁ and P=0.857 for FVC). At the end of follow-up, average FEV₁ and FVC was 1.45 ± 0.83 and 2.08 ± 0.88 in the non-TCZ treated patients compared to 1.24 ± 0.81 and 1.90 ± 0.78 liters in TCZ treated patients (P=0.686 for FEV₁ and P=0.818 for FVC). DSA outcomes and FEV₁ trend is depicted in Fig. 1B and 1C respectively. The FVC trend from AMR diagnosis to the end of follow-up is found in Supplementary Figure S1.

Table 1.

Tocilizumab (TCZ) and non-TCZ containing treatment regimen characteristics and HLA DSA outcomes.

Patient	Regimen	ISHLT AMR Classification	HLA DSA Present	Outcomes
TCZ-1	TCZ + IVIg	Probable	DR	DSA cleared in 69 days
TCZ-2	TCZ + IVIg	Probable	DQ	MFI decrease after 1 TCZ dose
TCZ-3	TCZ + IVIg	Probable	DQ	DSA cleared in 56 days
TCZ-4	TCZ + CFZ + IVIg	Definite	A + C + DR + DQ	MFI decrease after 1 TCZ dose
TCZ-5	TCZ + ECP + IVIg	Probable	DQ	DSA MFI decrease after 2 TCZ doses
TCZ-6	TCZ + rATG + CFZ	Definite	DR	DSA cleared in 261 days
TCZ-7	TCZ + rATG + CFZ + IVIg	Probable	DQ	DSA cleared in 24 days
TCZ-8	TCZ + rATG + CFZ + IVIg	Probable	DR + DQ	DSA cleared in 165 days
TCZ-9	TCZ + rATG + CFZ + IVIg	Probable	DR + DQ	DSA cleared in 39 days, rebound 77 days later
Non-TCZ -1	RTX + IVIg	Probable	DQ	DSA cleared in 32 days, rebound 44 days later
Non-TCZ -2	RTX + IVIg	Definite	DQ	DSA cleared in 83 days, rebound 22 days later
Non-TCZ -3	RTX + IVIg	Definite	B	New DSA 139 days after treatment
Non-TCZ -4	RTX + IVIg	Definite	DR	DSA cleared in 156 days
Non-TCZ -5	rATG + IVIg	Probable	DR	DSA MFI decreased
Non-TCZ -6	rATG + IVIg	Probable	DQ + DP	DSA cleared in 144 days
Non-TCZ -7	CFZ + IVIg	Probable	DQ	DSA cleared in 42 days
Non-TCZ -8	RTX + rATG + IVIg	Probable	DR + DQ	DSA MFI increased
Non-TCZ -9	RTX + rATG + IVIg	Definite	DQ	DSA MFI decreased
Non-TCZ -10	RTX + rATG + IVIg	Possible	DQ	DSA MFI decreased
Non-TCZ -11	RTX + BTZ + IVIg	Probable	DP	DSA cleared in 14 days
Non-TCZ -12	RTX + PLEX + IVIg	Probable	DQ	DSA MFI increased
Non-TCZ -13	RTX + BTZ + PLEX + IVIg	Probable	A + DR + DQ	DSA MFI increased
Non-TCZ -14	RTX + CFZ + PLEX + IVIg	Probable	DP	New DSA 5 days and 140 days after treatment
Non-TCZ -15	RTX + rATG + BTZ + PLEX	Probable	DR	DSA MFI increased
Non-TCZ -16	RTX + rATG + PLEX + IVIg	Definite	A	DSA cleared in 30 days, rebound 18 days later, new DSA 236 days after clearance
Non-TCZ -17	RTX + BTZ + PLEX + IVIg	Probable	A + DQ	DSA cleared in 11 days, rebound 209 days later, new DSA 339 days after clearance
Non-TCZ -18	RTX + rATG + CFZ + PLEX + IVIg	Probable	DQ	No change in DSA MFI

Abbreviations: BTZ: bortezomib, CFZ: carfilzomib, ECP: extracorporeal photopheresis, IVIg: intravenous immunoglobulins, MFI: mean fluorescence intensity, PLEX: plasmapheresis, rATG: rabbit antithymocyte globulin, RTX: rituximab, TCZ: tocilizumab

Post-treatment infections occurred in 61% of non-TCZ treated patients and 56% of TCZ treated patients (Supplementary Table S2). The median number of organisms isolated post-treatment was higher in non-TCZ treated patients (three) when compared to TCZ treated patients (one). Pre-medications of acetaminophen, diphenhydramine, and methylprednisolone were given prior to any dose of TCZ, rituximab, IVIg, and rabbit antithymocyte globulin; there were no instances of infusion reactions. Elevations in transaminases occurred in two non-TCZ treated patients and two TCZ treated patients and resolved without intervention.

Targeting IL-6, mainly through antagonizing its effects on B cell stimulation, increased plasmablast formation, and decreased T regulatory cell differentiation, holds promise as a therapeutic option for AMR.⁷ However, the use of IL-6 inhibitors has not been previously described in lung transplantation. In the present study, TCZ treated patients were more likely to clear DSA than non-TCZ treated patients and were less likely to have DSA recurrence or develop new DSA. Furthermore, TCZ treated patients were less likely to have graft failure in the 12 months following AMR onset. Although these differences did not reach statistical significance, the small sample size likely limited the study’s statistical power. Of note, 44.4% of TCZ treated patients remain on monthly TCZ therapy. These data highlight the difficulty in treating AMR to resolution, with the majority of patients progressing to death or salvage AMR treatment in addition to experiencing a continued fall in average FEV₁. TCZ was well tolerated; there were no instances of infusion reactions. There were similar proportions of patients who developed infections, although less organisms were identified in the 12 months following TCZ treatment when compared to those treated without TCZ. Similarly, elevations in transaminases were uncommon and similar between groups.

Treatment of DSA typically involves a multimodal, patient-specific approach that may target removal of circulating antibodies via plasma exchange (PLEX), interference of F_c binding of DSA to graft endothelium with intravenous immunoglobulins, depletion of B cells with anti-CD20 monoclonal antibodies, depletion of plasma cells with proteasome inhibitors, depletion of T cells to prevent B cell activation with antithymocyte globulin, and treatment of inflammation with corticosteroids.⁸ These multimodal approaches are warranted given the numerous avenues by which DSA inflict graft damage. However, they make the assessment of individual medications difficult, which is an inherent limitation of this study. Specifically, whereby tocilizumab treated individual patients fared better with DSA and graft outcomes but also received other treatments, our data are suggestive but not conclusive that the improved outcomes post-AMR are associated with the addition of tocilizumab.

There were notable differences between TCZ and non-TCZ treated patients. At our institution, use of PLEX has decreased over time due to anecdotal lack of benefit and because it may delay administration of monoclonal antibody medications which are removed by PLEX. No patients who received TCZ received plasma exchange whereas 38.9% of non-TCZ recipients received PLEX. Given that there was more clearance with less recurrence and lower incidence of new DSA in the TCZ arm, it is unlikely that higher rates of PLEX in the non-TCZ arm was a significant confounder to beneficial DSA outcomes. In general, more AMR therapies were given to non-TCZ treated patients. Treatment for AMR also occurred earlier in the post-transplant course in TCZ treated patients (402 days versus 756 days, P=0.068); however, both groups had an average time from transplant to AMR of over a year from transplant. There is no standard definition of early or late AMR and the impact of time to AMR on outcomes has not been elucidated in the lung transplant population but is an important area for future research. In addition, our limited sample size with non-standard follow-up times for repeat HLA DSA and PFT testing and the retrospective study design detracts from the ability to make definitive conclusions about the safety and efficacy of TCZ for the treatment of AMR in the lung transplant population. A future prospective study in pulmonary transplant patients with AMR could address these limitations by standardizing medication regimens with uniform repeat DSA and PFT testing after treatment. Dosing of TCZ at 8 mg/kg monthly is derived from the rheumatoid arthritis population¹⁰ but it is unknown if this is the optimal dose for treatment of pulmonary AMR. However, the data demonstrates TCZ results in numerically lower graft failure rates, higher clearance of DSA and less recurrence or new DSA development after treatment, which is promising. A randomized clinical trial should be considered to fully elucidate the effect of TCZ on treatment of AMR in lung transplant recipients.

Non-Standard Abbreviations Used:

AMR

antibody mediated rejection

BTZ

bortezomib

CFZ

carfilzomib

CLAD

chronic lung allograft dysfunction

DSA

donor specific antibody

ECP

extracorporeal photopheresis

HLA

human leukocyte antigen

IVIg

intravenous immunoglobulin

LFT

liver function test

MFI

mean fluorescence intensity

PLEX

plasmapheresis

rATG

rabbit antithymocyte globulin

RTX

rituximab

TCZ

tocilizumab