Severe anti‐thymocyte globulin‐induced cytokine release syndrome in a renal transplant patient

Summary

Cytokine release syndrome is a systemic inflammatory response that can be triggered by a variety of factors such as infection or exposure to certain drugs, particularly novel T cell‐engaging immunotherapies. Severe cytokine release syndrome as a complication following treatment with anti‐thymocyte globulin, although recognised, is not well‐reported in the literature. We report the case of a 64‐year‐old man who developed catastrophic cytokine release syndrome after receiving anti‐thymocyte globulin during kidney transplantation. We highlight the importance of prompt recognition of severe cytokine release syndrome with strategies to aid survival in life‐threatening cases.

Introduction

Anti‐thymocyte globulin is a polyclonal T cell‐depleting antibody used as an induction immunosuppressive agent in renal transplantation, administered before reperfusion of the donor kidney to attenuate the immune response in recipients who have a high risk of rejection. Although anti‐thymocyte globulin was developed in the 1970s, its association with cytokine release syndrome (CRS) was first formally described in the 1990s when anti‐T cell antibody was introduced as an immunosuppressant for solid organ transplantation [1]. Cytokine release syndrome manifestations range from mild systemic symptoms such as fever, fatigue and headache, to a catastrophic systemic inflammatory response with cardiovascular collapse, disseminated intravascular coagulation, multi‐organ failure and death. Diagnosis may be challenging; symptoms can mimic several other systemic inflammatory disorders such as tumour lysis syndrome, septic shock and hypersensitivity reactions. The mainstay of treatment for CRS is supportive, however, the use of treatments such as corticosteroids and interleukin‐6 blockade with tocilizumab have been described [1].

We describe a case of life‐threatening CRS following administration of anti‐thymocyte globulin for a renal transplant, a rarely‐reported systemic inflammatory response to anti‐thymocyte globulin. Our case highlights the importance of early recognition of CRS following administration of a potentially causative agent. Further, we describe our supportive care strategy, including using off‐protocol continuous veno‐venous haemodiafiltration (CVVHDF), which in this instance was associated with an excellent clinical outcome. To our knowledge, the peri‐operative use of CVVHDF in the management of severe CRS has not been reported in the literature.

Report

A 64‐year‐old man with stage 5 chronic kidney disease secondary to IgA nephropathy underwent live donor kidney transplantation. Following an uneventful induction of general anaesthesia, 3 mg.kg⁻¹ of intravenous anti‐thymocyte globulin was given intra‐operatively over 4 h as an immunosuppressive induction agent. The surgery was notable for 2 l of blood loss but was otherwise uncomplicated. The patient successfully emerged from anaesthesia, his trachea was extubated and he was transferred to the post anaesthesia care unit with a mild metabolic acidosis, which was treated with intravenous fluids.

Due to worsening acidosis, hypotension and type 1 respiratory failure, a phenylephrine infusion and non‐invasive continuous positive airway pressure ventilation were commenced 3 h postoperatively. Further metabolic deterioration with uncontrollable hyperactive delirium necessitated transfer to the intensive care unit (ICU) for tracheal intubation and mechanical ventilation. Intravenous hydrocortisone and empiric antibiotic cover with piperacillin‐tazobactam were commenced and a diagnosis of CRS was made. Within 4 h, the vasopressor requirement rapidly escalated to noradrenaline at 4.48 mg.h⁻¹ and vasopressin at 2.4 units.h⁻¹. CVVHDF was initiated for anuria and severe metabolic acidosis, alongside LiDCO rapid cardiac output monitoring (LiDCO, London, UK). An ultrasound Doppler of the transplanted kidney confirmed adequate perfusion.

Despite aggressive supportive treatment, the patient experienced worsening metabolic acidaemia (pH 7.00; lactate 13 mmol.l⁻¹) even with citrate CVVHDF and intermittent sodium bicarbonate 8.4% infusions. He remained anuric in distributive circulatory shock. Ventilatory requirements were high, with FiO₂ of 0.9; inspiratory airway pressures of 23 cmH₂O; and positive end‐expiratory pressure (PEEP) of 10 cmH₂O on pressure control ventilation. The patient was sedated with propofol and fentanyl infusions and did not require continued neuromuscular blockade to achieve ventilator synchrony.

On the first postoperative day, a non‐blanching petechial rash was noted on the patient's lower limbs. His temperature was 37.7 °C. Acute liver failure had developed (alanine aminotransferase 1902 IU.l⁻¹) with grossly deranged coagulation indicative of disseminated intravascular coagulation (platelets of 25 × 10⁹.l⁻¹; D‐dimer 30,487 ng.ml⁻¹). A blood glucose level drop to 2.9 mmol.l⁻¹ prompted initiation of a glucose 20% infusion to maintain normoglycaemia. The patient required transfusion with packed red blood cells; fresh frozen plasma; platelets; and cryoprecipitate as his haemoglobin concentration dropped from 103 g.l⁻¹ to 69 g.l⁻¹ with no surgical or other overt cause of bleeding identified, platelet count dropped from 162 × 10⁹.l⁻¹ to 25 × 10⁹.l⁻¹, prothrombin time increased to 31 s and fibrinogen concentrations reduced from 2.5 to 0.8 g.l⁻¹.

Despite his profound clinical deterioration and increasing concern that the situation was becoming unsalvageable, a repeat Doppler ultrasound of his liver and transplanted kidney was normal. The mainstay of subsequent management was goal‐directed fluid resuscitation using cadiac output monitoring and optimisation of CVVHDF. Settings were manually altered from a dialysate rate of 750–1500 ml.h⁻¹ and a replacement rate of 600–1500 ml.h⁻¹, resulting in an effluent increase from 33 to 51 ml.kg⁻¹.h⁻¹. The blood pump speed remained at 180 ml.h⁻¹. The goal of this CVVHDF strategy was to increase both small and large molecular clearance alongside ultrafiltration of larger mediators and cytokines. We used a Prismaflex ST150 membrane (Baxter, Illinois, USA).

The patient achieved clinical stability by day 2 in ICU; his physiology rapidly recovered leading to de‐escalation of multi‐organ support and tracheal extubation 24 h thereafter. Continuous venovenous haemodiafiltration was required until postoperative day 6 and he was stepped down to ward‐based care on postoperative day 10. He required intermittent haemodialysis twice in the following week for uraemia (peak urea 52.5 mmol.l⁻¹; creatinine 632 μmol.l⁻¹). Kidney biopsies demonstrated 100% infarction of the renal cortex and acute tubular necrosis but no evidence of rejection. Despite this, a renal ultrasound scan showed good perfusion and he was discharged home on postoperative day 28 with gradual recovery of renal function (creatinine 130 μmol.l⁻¹ 6 weeks postoperatively).

Discussion

Cytokine release syndrome, a rare systemic inflammatory response, is a complication increasingly seen with the rise in use of T cell‐engaging immunotherapies [1]. The common terminology criteria for adverse events has defined CRS as “a disorder characterised by nausea, headache, tachycardia, hypotension, rash and shortness of breath; caused by the release of cytokines from the cells” [2]. Cytokine release syndrome is graded into five categories:

• Grade 1 – Fever with or without constitutional symptoms

• Grade 2 – Hypotension responding to fluids; hypoxia responding to < 40% FiO₂

• Grade 3 – Hypotension managed with one vasopressor; hypoxia requiring > 40% FiO₂

• Grade 4 – Life‐threatening consequences; urgent intervention indicated

• Grade 5 – Death

Cytokine release syndrome triggered by known precipitants, predominantly monoclonal antibodies and chimeric antigen receptor adoptive T cell therapies with chimeric antigen receptors has been well documented in the literature [1]. The manufacturer of anti‐thymocyte globulin notes adverse immune‐mediated reactions including anaphylaxis and CRS. Given the known cause and effect in these situations, CRS is relatively well‐understood; why some patients demonstrate such a dramatic effect compared with others remains a mystery. A few case reports have described severe reactions after receiving anti‐thymocyte globulin with varying survival outcomes following resuscitation [3, 4]. However, as far as we are aware, there have been no documented cases of such severe multi‐organ failure related to CRS following anti‐thymocyte globulin for renal transplant that have resulted in a good patient outcome. Despite life‐threatening cardiovascular, respiratory, renal and hepatic compromise, we persevered with aggressive supportive treatment due to the potential reversibility of the suspected diagnosis. Our management with off‐protocol CVVHDF to facilitate increased clearance of cytokines may have helped achieve this, although the evidence in this area is lacking. Increasing the dialysate and replacement rates leads to increased effluent rate which increases clearance of both small and large molecules. Dialysis effectively aids small molecule clearance, and ultrafiltration removes larger molecules such as inflammatory cytokines.

The administration of anti‐thymocyte globulin under general anaesthesia in this instance was guided by our transplant unit's previous experience using anti‐thymocyte globulin, including original involvement studying the incidence of acute rejection and delayed graft function [5]. In this particular instance, we administered a dose of 3 mg.kg⁻¹, which was greater than the doses described by our unit's studies [5] but well below the 6 mg.kg⁻¹ single dose reported by the Stevens et al. for prophylaxis against rejection [6].

The management of CRS is largely supportive and optimal clinical management remains uncertain but is evolving. To date, biological models, expert opinion and retrospective analyses have largely guided management. More recently, observations that interleukin‐6 levels were elevated in patients with CRS following chimeric antigen receptors therapy [7] led to the consideration of interleukin‐6 blockade with monoclonal antibodies against interleukin‐6 (siltuximab) and its receptor (tocilizumab) in the treatment of CRS. Administration of these agents was shown to be an effective treatment with rapid resolution of CRS symptoms in subsequent studies [7]. In August 2017, concurrently with tisagenlecleucel, tocilizumab received Food and Drug Administration (USA) approval for treatment of CRS and is fast becoming the gold standard for initial treatment of severe CRS in patients receiving chimeric antigen receptors T cells [8]. To date, there are no trials or case reports on the use of tocilizumab in anti‐thymocyte globulin‐related CRS and hence it was not used on our patient. Very early use of cytokine adsorption therapy might have been beneficial to potentially reduce interleukin‐6 levels and improve haemodynamic stability [9, 10].

Our case outlines catastrophic anti‐thymocyte globulin‐induced CRS causing rapid and profound clinical deterioration followed by an equally rapid recovery and remarkable renal graft survival. We highlight the importance of prompt recognition of severe CRS and that perseverance with full supportive management for at least 48 h can have a favourable outcome. Optimising support over the extremely challenging period of acute derangements may be the key to optimising management until there is further evidence for specific treatment. Clinicians should have a high index of suspicion for diagnosis of CRS in the presence of clinical symptoms in a susceptible patient, and should seek early and continued multidisciplinary input.