Rapid Identification of Cytokine Release Syndrome after Haploidentical Peripheral Blood Stem Cell Transplantation and Successful Therapy with Tocilizumab

To the Editors

Cytokine release syndrome (CRS) is a well-documented toxicity associated with immunotherapies, particularly chimeric antigen receptor (CAR)-modified T-cells, and tocilizumab, a monoclonal antibody directed against the interleukin-6 receptor, has been effectively utilized for the management of CRS associated with CAR-T therapy (1, 2). A febrile syndrome attributed to cytokine elaboration by proliferating T-cells has similarly been reported in the setting of haploidentical peripheral blood stem cell transplant (PBSCT) (3, 4). Here we report a case of grade 3 CRS emerging as a complication of haploidentical PBSCT, successfully treated with tocilizumab.

The patient was a 42-year-old man with primary refractory stage IVB Hodgkin lymphoma with involvement of the bone and pleura and suspected involvement of the lung parenchyma. He had been diagnosed more than 2 years prior to admission, with disease resistant to multiple lines of therapy including ABVD, brentuximab, and nivolumab, but he had finally achieved a partial remission with everolimus. He was admitted for hematopoietic stem cell transplantation from his haploidentical brother; as the donor was deemed ineligible for bone marrow harvest, a PBSC graft was planned. Graft-versus-host disease prophylaxis was to consist of post-transplant cyclophosphamide in addition to tacrolimus and mycophenolate mofetil. Pre-transplant comorbidities included a mild transaminitis and severe deficits on pulmonary function testing (FEV1 56% predicted, adjusted DLCO 75% predicted), consistent with an extraparenchymal defect likely secondary to known pleural disease; this was possibly exacerbated by a human metapneumovirus infection 1 month preceding transplant. The patient was asymptomatic on hospital admission.

The preparative regimen consisted of melphalan 140 mg/m2 (day –7), thiotepa 5 mg/kg (day –6), and fludarabine 40 mg/m2/day for 4 days (day –5 through –2). On day 0, the patient received the haploidentical PSBC graft containing 5 × 10⁶ CD34⁺ cells/kg and 2.06 × 10⁸ CD3⁺ cells/kg without event. Approximately 12 hours after allograft infusion, the patient became febrile with associated mild tachycardia and new diarrhea, but he was normotensive with adequate oxygenation on room air. Broad infectious evaluation with blood and urine cultures, Legionella and Pneumococcal urine antigens, respiratory viral PCR, beta-D-glucans, serum galactomannan, stool C. difficile PCR, and stool viral PCR was negative. Chest X-ray showed a new left base consolidation with associated trace left pleural effusion. Antibiotics were modified to vancomycin, piperacillin/tazobactam, and azithromycin. Throughout day +1 and into day +2, however, the patient remained febrile to 39.5° C. He eventually developed worsening tachycardia to the 120s, relative hypotension to SBP 90s, tachypnea with respiratory rate in the 20s, and desaturation to 87%. Mental status was intact. There was new acute kidney injury with serum creatinine of 1.9 mg/dL (from 0.8 mg/dL one day prior). A grade 1 transaminitis, present on admission, persisted but remained stable. Repeat cultures were negative, but antibiotics were again broadened. Serum C-reactive protein (CRP) was found to be 15 mg/dL, and interleukin-6 (IL-6) measured at 1,113.60 pg/mL. Given suspicion for CRS, the patient received tocilizumab 4 mg/kg on day +2. Corticosteroids were not used as it is standard to avoid them prior to giving cyclophosphamide in this transplant setting. On days +3 and +4, he received cyclophosphamide 50 mg/kg/day as planned. He initially remained febrile, required up to 5 L of supplemental oxygen, and was briefly hypotensive to the 80s/50s. Blood pressure responded to intravenous hydration, and the patient continued to receive supportive management. By the evening of day +3, he had defervesced, with normalization of hemodynamics and decreasing oxygen requirement. A second dose of tocilizumab was not given. CRP peaked at 20.22 mg/dL but swiftly declined (Figure 1). The patient was weaned off oxygen on day +12 and remained afebrile and stable on room air for the remainder of his hospitalization. Renal function returned to baseline by day +8; transaminitis transiently worsened to grade 3 but also eventually improved to normal limits on day +18. Neutrophil engraftment occurred on day +24. At outpatient follow-up 1 month after transplant, CRP remained stable at 4.06 mg/dL, and IL-6 had declined to 41.6 pg/mL.

Figure 1.

Serial C-reactive protein measurements following administration of tocilizumab for the treatment of CRS in a patient receiving haploidentical PBSCT.

As the use of haploidentical transplantation increases, issues regarding the finer points of this approach – including optimal conditioning, graft source, and management of its unique toxicities – will merit ongoing thought. Details regarding the incidence, clinical presentation, and management of CRS including the use of tocilizumab during haploidentical PBSCT represents an important area for further discourse (5, 6).

We employed a reduced intensity preparative regimen that has previously been described by Ciurea and colleagues, although almost exclusively with bone marrow grafts (7). It is possible the combination of a PBSC graft and the increased mucosal damage engendered by higher intensity conditioning set the stage for our patient’s CRS. There are instances, however, where a marrow graft cannot be procured due to donor ineligibility or preference. Used more frequently in the haploidentical setting and shown to produce low rates of GVHD and non-relapse mortality, the regimen of fludarabine, cyclophosphamide, and 200 cGy of total body irradiation has been utilized extensively with haploidentical marrow transplant by investigators at Johns Hopkins (8) and with haploidentical PBSC grafts elsewhere (9), without CRS or other undue toxicity, but concerns remain about the relatively high rates of relapse associated with this reduced intensity strategy. Investigators at Northside Hospital in Atlanta have paired haploidentical PBSC grafts with an ablative regimen of fludarabine, busulfan, and cyclophosphamide and have indeed observed a febrile syndrome in the majority of patients early after graft infusion (3). Supporting the role of cytokine release in this syndrome, fevers have reportedly resolved in most cases after post-transplant cyclophosphamide administration. In the current example, the time course of clinical improvement, with resolution of fevers approximately 24 hours after tocilizumab and only 12 hours after the patient’s first dose of cyclophosphamide, suggests that the former played a larger role in the immediate term.

Consideration of CRS including standardized procedures for the measurement of inflammatory cytokine levels will allow for the rapid diagnosis and management of this potentially severe complication of haploidentical transplantation, particularly when PBSC grafts are required. Possible alternative diagnoses, particularly infection, should be thoroughly ruled out. Other potential items on the differential diagnosis for the patient’s constellation of symptoms, such as pre-engraftment syndrome, were unlikely based on the timing of onset. Of note, the Northside authors have not reported the associated hemodynamic instability or end-organ dysfunction that can accompany CRS; aside from the use of a different conditioning regimen, it is unclear if other factors contributed to our patient’s more fulminant course, and overall, patient-specific characteristics that may predispose to transplant-associated CRS have not been characterized. In addition to his underlying comorbidities, our patient had previously received nivolumab, and it remains to be seen whether prior exposure to checkpoint blockade places patients at higher risk of systemic inflammatory complications including CRS. In a preliminary report by Merryman et al, there was 32% incidence of febrile syndrome early after allogeneic HSCT in patients who had previsouly been treated with checkpoint inhibitors (10). Long-term consequences of tocilizumab use early post-transplant also remain unknown. Given the critical role of haploidentical transplantation in expanding access to potentially curative therapy for patients with relapsed/refractory or high-risk hematologic malignancies, it will be important to address such questions in future investigation.