Abstract
Introduction
The use of programmed death-ligand 1 (PD-L1) checkpoint inhibitor therapy is expanding, although its adverse effects are not completely known. We report on a rare case of acute cytokine release syndrome related to pembrolizumab use in a patient with lung cancer.
Case report
A 79-year-old man with metastatic, PD-L1-positive, non-small-cell lung cancer developed a febrile condition associated with a systemic inflammatory response syndrome and suffered haemodynamic compromise four hours after the first intravenous administration of pembrolizumab. A thorough medical workup found no alternative cause and a grade 2 cytokine release syndrome (CRS) was diagnosed.
Discussion
Acute CRS after the administration of a PD-L1 inhibitor is infrequent but could be a fatal condition. Supportive treatment and, if necessary, corticosteroids should be considered.
Keywords
Introduction
According to the World Health Organization, over 18 million new cases of cancer were diagnosed worldwide in 2018. Lung cancer is the fourth leading cause of cancer in Switzerland, with an incidence of 23 per 100,000 inhabitants. 1
The tumour microenvironment is affected by the inflammatory reaction produced by the cancer. Immune cells are recruited to the cancer site and regulated by tumour-derived signals. In lung cancer, this interaction involves the programmed death-ligand 1 (PD-L1) proteins expressed by the tumour and the programmed cell death 1 (PD-1) proteins expressed by lymphocytes.2–5 This signalling pathway leads to the inhibition of T cells, causing the tumour to escape from physiological control and cancer cells to proliferate.6–8
The therapeutic rationale for checkpoint inhibitor immunotherapy is blocking the immune cell suppression signals induced by the interaction between the tumour microenvironment and lymphocytes. The use of antibody programmed cell-death 1 (anti-PD-1) and antibody programmed death-ligand 1 (anti-PD-L1) has therapeutic potential against non-small-cell lung cancers. 9
Pembrolizumab, a humanised monoclonal antibody against PD-L1, is a second-line treatment for use after lung cancer progression, despite first-line chemotherapy. 10 The use of immunotherapy with a combination of chemotherapy as a first-line treatment will most likely become the standard of care for non-small-cell lung cancer.
In recent years, haemodynamic compromise has been described after multiple intravenous doses of immunotherapy and secondary to a possible cytokine release syndrome (CRS). 11 , 12
We present the extremely rare case of a patient diagnosed with grade 2 CRS only four hours after his first intravenous dose of pembrolizumab for metastatic, non-small-cell lung cancer.
Case presentation
A 79-year-old patient and former smoker (50 pack-year units), with a history of chronic kidney disease, chronic moderate alcohol consumption, gastroesophageal reflux disease, high blood pressure and ischaemic heart disease, was admitted to the emergency department in March 2018 for left thoracic pain (see patient care timeline in Appendix 1). A chest computed tomography scan revealed bilateral pulmonary nodules and infiltrates and an osteolytic lesion of the sternum. A positron emission tomography-computed tomography scan confirmed the multiple pulmonary nodules, including a 3 cm infiltrate in the right upper lobe with a hypermetabolic pleural attachment, adenopathies (group 2 L, 4 L and pre-hilar), and osteolytic lesions (to the spine, pelvis, ribs and sternum) resulting in a pathological fracture of the ninth left rib. Histological analysis of the endobronchial lesions, via a bronchoscopy, revealed a non-small-cell lung cancer (PD-L1 expression of 1%, Kirsten rat sarcoma (KRAS) viral oncogene homologue-positive, anaplastic lymphoma kinase (ALK)-negative, c-ros oncogene 1 (ROS1)-negative, epidermal growth factor receptor (EGFR)-negative, and v-raf murine sarcoma viral oncogene homolog B (BRAF)-negative). Due to the low expression of PD-L1, a combination of chemotherapy with an intravenous carboplatin area under the curve (AUC) of 5 mg/ml/min (AUC 5) on day 1 and 21, and pemetrexed 500 mg/m2 on day 1 and 21, was quickly initiated in April 2018. Subcutaneous denosumab 120 mg was also administered due to the presence of symptomatic osteolytic bone lesions. After four cycles and four months after the initial diagnosis, the patient developed brain metastases, which were treated with whole-brain radiotherapy. The patient did not wish to resume chemotherapy and continued with denosumab alone.
Following oncological progression and the indication for second-line pembrolizumab treatment despite low PD-L1 expression (Tumour Proportion Score at 1%), intravenous pembrolizumab 200 mg was given on 9 November 2018, eight months after the original diagnosis. Simple medical treatment monitored using regular medical check-ups led to a stable clinical situation for his comorbidities for a long time. The first intravenous treatment using pembrolizumab, based on the usual protocol, was administered in the oncology department’s ambulatory unit. Premedication (paracetamol 1000 mg and clemastine 2 mg injection) was given before the intravenous administration of pembrolizumab (200 mg pembrolizumab in 100 ml sodium chloride 0.9% over 30 mins). After the intravenous treatment, ninety minutes of monitoring was uneventful and the patient returned home quickly.
Four hours later, however, the patient presented to the emergency department with fever. At admission, his temperature was 39.6°C, blood pressure 115/71 mmHg, pulse rate 91 beats/minute and oxygen saturation 91% on room air. Physical examination revealed a decreased left respiratory murmur with fine crackles on lung auscultation and confusion without other neurological symptoms. The electrocardiogram remained stable. Baseline laboratory testing revealed anaemia, thrombocytopenia, acute kidney disease improving global outcomes 1 renal failure, a procalcitonin value of 0.47 μg/l in the grey zone, normal leukocyte count and a slightly elevated high sensitivity troponin level, which stabilised after four hours. Chest X-ray revealed no pulmonary infection, urinalysis showed no white cells or bacteriuria, blood cultures were taken, and antibiotic treatment was not introduced initially.
During the night after admission, eight hours after pembrolizumab treatment, the patient remained feverish and developed confusion and dyspnoea with haemodynamic instability. Repeated fluid resuscitation with crystalloids and empiric antibiotic therapy with intravenous piperacillin/tazobactam were administered. Dexamethasone was also quickly initiated.
Blood and urine cultures were sterile, and antibiotic therapy stopped after seven days of treatment. During his stay, the patient did not present with renewed fever or signs of systemic infection. Further evaluation revealed no increases in inflammatory markers. After fluid resuscitation, supportive treatment and the cessation of hypotensive treatments, his blood pressure remained stable without the need for vasopressor therapy.
Confusion, dyspnea and renal failure regressed gradually. The patient was discharged after 18 days.
Following this hospitalisation, the patient declined any further oncological treatment and requested palliative care for symptom management. The patient’s course remained stable for six months. However, his condition declined in early April 2019, with a deterioration in cognitive function due to brain metastases and associated cerebral oedema. The patient was admitted to a palliative care unit and died in May 2019.
Discussion
We have presented an extremely rare case of CRS occurring only four hours after a first pembrolizumab treatment for non-small-cell lung cancer. Clinical, biological and imaging investigations reasonably excluded alternative diagnoses of shock (whether cardiac, infectious, anaphylactic or obstructive). The diagnosis of CRS was thus made clinically.
We used the Naranjo Adverse Drug Reaction (ADR) probability scale to assess the causality of the first intravenous pembrolizumab treatment, rating our patient’s case as a ‘’Probable’’ ADR, with a total score of 6 points. 13
Although pembrolizumab has changed the prognosis for many patients with non-small-cell lung cancer, associated immune-related adverse events (IRAEs) can affect virtually any body system, with dramatic morbidity. Due to pembrolizumab’s mechanism of action, IRAEs are a risk and might affect any organ. 14 Among possible IRAEs, CRS is defined as a massive release of pro-inflammatory cytokines via the interferon-gamma pathway. 15 Due to the release of interleukin-1 (IL-1), IL-6, IL-8 and IL-10, CRS can cause fever, nausea, vomiting, diarrhoea, tachypnoea and tachycardia, hypotension, seizures, confusion, delirium and hallucinations. 16 Symptoms appear suddenly after exposure to the triggering factor. Diagnosis remains primarily clinical, based on scientific recommendations after the exclusion of differential diagnoses.16,17
There are several known risk factors for developing IRAEs with pembrolizumab. Treatment duration and drug interactions may play a major role.18,19 However, Shimizu et al. found no associations between dose and toxicity, particularly in solid tumours. 19 Although a literature review found potential IRAEs in all neoplasia treated with immunotherapy, 20 the occurrence of IRAEs is significantly higher in non-small-cell lung cancer and melanomas. 21 Male sex appears to be a risk factor for some IRAEs 22 and a protective factor for others. 23 No studies have shown associations between gender and CRS. Besides, age does not predispose patients to IRAEs. 24 The safety of checkpoint inhibitors among the elderly, compared to younger patients, has been demonstrated previously 25 , 26 and was confirmed in a 2019 meta-analysis. 14
The patient presented here had a stage 3 chronic kidney disease and heart disease. Organ failure does not affect the pharmacokinetics or pharmacodynamics of pembrolizumab according to the manufacturer's preclinical studies. In their retrospective analysis, Kanz et al. confirmed that chronic organ failure (cardiac, renal and hepatic) did not appear to be a confounding precipitating factor. 27 , 28 Nevertheless, pre-existing organ dysfunctions could decompensate during the treatment.29–31
Our patient developed acute CRS after the first intravenous dose of pembrolizumab. A preclinical animal study demonstrated an increase in pro-inflammatory markers consecutive to a single dose of pembrolizumab in humanised primates with normal cells: these increases in inflammatory markers could explain the manifestations of CRS. 32
Despite our patient’s low level of PD-L1 expression (Tumour Proportion Score at 1%), pembrolizumab is still recommended, with a favourable overall survival rate. 33 It remains unclear whether the patient's low PD-L1 level played a role in the manifestation of this IRAE. However, preclinical studies conducted in humanised cynomolgus macaques found no CRS at high concentrations of pembrolizumab, despite the total absence of tumour cells with PD-L1 markers. 32 In addition, a retrospective study published in 2020 identified a high rate of PD-L1 expression as a statistically significant factor in the development of an IRAE, but not the low Tumour Proportion Score, as in the present case. 34
Management of CRS depends on its severity, divided into four grades according to clinical symptoms and paraclinical parameters. 15 Mild CRS are usually treated symptomatically. In some cases of severe CRS, intravenous fluids are necessary, and the use of corticosteroids or anti-interleukin 2 antibodies may be helpful to reduce symptoms and consequences. 35 The clinical course is generally favourable. According to the US National Cancer Institute’s clinical criteria, our patient presented with grade 2 CRS. 36 He received symptomatic treatment supplemented with intravenous crystalloids to correct the circulatory failure.
The sequential organ failure assessment (SOFA) score was developed for grading mortality risk among patients admitted to intensive care units, 37 but it has also shown itself to be useful among elderly patients admitted to emergency departments for prognosis assessment. 38 , 39 Our patient’s SOFA score was 6, which equates to a five-fold increased risk of in-hospital mortality. 40 Recent studies have evaluated the SOFA score’s prognostic value for mortality in oncology patients admitted to emergency departments with systemic inflammatory syndrome in the absence of infection. They observed mortality rates for oncology patients with a SOFA score > 2 at admission of between 27.9% and 47.8%. 41 , 42
Acute CRS may develop during treatment with pembrolizumab, even after a single dose. Emergency physicians, oncologists and general practitioners need to be aware of this rare side effect. Treatment is usually supportive, but may also include corticosteroid therapy in severe cases.
Footnotes
Acknowledgements
We thank Dr. Christopher Richard for his careful review and sound advice. Thanks to the department of oncology of the Réseau hospitalier neuchâtelois – La Chaux-de-Fonds, for their oncological contribution to the patient's history and acute care. We also thank the medical imaging department for the performance and interpretation of the different radiological examinations.
Author’s contributions
All the authors have contributed to manuscript writing and approved its final content in the present form.
Availability of data and materials
The data analysed during the current case report are available from the corresponding author on reasonable request.
Consent of publication
The family of the patient consented to reporting the present case.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Ethical approval
The patient consented to all reported procedure and treatments, according to local regulations and Ethics authority.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
Appendix 1
