Hemophagocytic lymphohistiocytosis following pembrolizumab and bevacizumab combination therapy for cervical cancer: a case report and systematic review

Cancer immunotherapy using immune-checkpoint inhibitors (ICIs) has emerged as a cornerstone treatment for various cancers. However, by eliminating the physiological inhibitory control that moderates T-cell activation, ICIs may cause T-cell hyperactivation and immune-related adverse events (irAEs) [1], one rare but serious irAE is hemophagocytic lymphohistiocytosis (HLH). HLH is a severe condition caused by immune activation and dysregulation, potentially leading to excessive pro-inflammatory cytokine secretion, rapid tissue destruction, multi-organ failure, and death [2]. Although the incidence of ICI-associated HLH is very low, with the World Health Organization's VigiBase pharmacovigilance database reporting only 5.7% of all HLH cases as potentially ICI-associated [3], the Society for Immunotherapy of Cancer (SITC) includes HLH in lethal irAE but does not provide specific treatment recommendations [4].

In the phase I/II CheckMate 358 trial (NCT02488759) [5] and KEYNOTE-158 (NCT02628067) [6], large clinical trials confirming the safety of ICIs in cervical cancer, ICIs have been widely used for cervical cancer treatment. However, no serious hematological adverse reactions due to ICIs in cervical cancer have been identified recently, and we report the first case of HLH attributed to ICIs in cervical cancer. This finding has not been previously documented in the literature or pharmacovigilance databases.

To better understand the biomarkers of ICI-associated HLH, clinical features, treatment strategies, characteristics of deceased cases, and the potential impact of HLH on tumor outcomes, we gathered all cases that met the HLH-2004 [7] or HScore > 169 criteria [8]. (Supplemental data 1) for ICI-related HLH. In total, we analyzed 52 confirmed ICI-related HLH cases by comparing our collection with data from two WHO pharmacovigilance databases. Our review revealed an increasing occurrence of ICI-related HLH in various cancer types and ICI classifications, a high number of cases with a history of autoimmune disease in HLH occurrence, autoimmune encephalitis as a significant cause of death, and the potential for an exceptional oncologic response to treatment if recovery from HLH is achieved.

Hemophagocytic lymphohistiocytosis (HLH) is a severe, hyperinflammatory syndrome triggered by aberrantly activated macrophages and cytotoxic T lymphocytes (CTLs) [51]. In adults, it frequently arises in contexts like untreated hematologic malignancies, chronic rheumatic diseases, or conditions of immunosuppression. The pathogenesis of HLH involves excessive activation of CTLs and their consequential depletion of IL-2, potentially leading to regulatory T cell (Treg) dysfunction [52, 53]. This cascade is pivotal in HLH development. Integral to ICI-based cancer immunotherapy is the simultaneous inhibition of Tregs and activation of CTLs [54]. ICIs enhance T-cell activation and proliferation while impairing Treg cell functionality. By inhibiting immune checkpoint molecules, ICIs prevent tumor cells from evading immune detection [54]. This same mechanism, however, disrupts peripheral T-cell tolerance, fostering the rapid diversification and clonal expansion of potentially toxic cells, which can culminate in hyperinflammation and autoimmunity [55, 56]. Furthermore, in HLH, CTL activation is perpetuated by a macrophage/monocyte expansion loop [51]. Studies in mouse and human cancers indicate that tumor-associated macrophage PD-1 expression inversely correlates with its phagocytic efficacy against tumor cells. Accordingly, PD-1/PD-L1 blockade has been observed to enhance macrophage-mediated phagocytosis of tumor cells [57]. These studies offer insights into the specific mechanisms of HLH associated with immunotherapy and confirm that immunotherapy-related HLH is a genuine and serious complication.

ICI becomes increasingly popular in treating various cancers, the number of immune-related adverse event (irAE) reports has also grown exponentially [58]. For instance, there were only 38 cases of ICI-associated HLH in VigiBase before 2019, but the latest study showed 177 cases in VigiBase. Our brief review suggests that ICI-associated HLH can occur across a broader range of patient populations and cancer types, including the first-ever reported case in cervical cancer, which we found outside the database. The latest VigiBase data on age at onset aligns with the findings from our literature collection, indicating a median age of onset around 60 years. Furthermore, very young onset cases were also found in both our systematic review and the database, with patients as young as two years old. In our analysis, melanoma and lung cancer have emerged as the most common cancer types associated with ICI-related HLH. The predominance of these cancers in HLH cases can be linked to the early approval of ICIs for their treatment, leading to a more extensive pool of data and reported cases. Therapies like pembrolizumab, nivolumab, and their combination with ipilimumab, predominantly used in melanoma and lung cancer, have been well-documented in this regard. However, as the spectrum of approved ICIs expands, an increase in HLH incidence is being observed in other cancers. This trend reflects the evolving landscape of ICI usage and its associated irAEs. Notable among these are various types of acute myeloid leukemia (AML) [45], as well as previously undetected cases in squamous cell skin cancer [44], Kaposi sarcoma [35] and other cancer types. While melanoma and lung cancer currently have a higher number of HLH cases due to their early inclusion in ICI therapy, the rising number of HLH cases in cancers like AML, squamous cell skin cancer, and Kaposi sarcoma, as seen with newer ICIs like camrelizumab and avelumab, indicates a broader oncological concern. Identifying specific cancer types or ICI agents that are more prone to inducing HLH remains a challenge, necessitating further research to clarify these associations and improve our understanding and management of ICI-related HLH across different cancer scenarios.

Our case involved the combination of bevacizumab for anti-angiogenic therapy, and prior to this, there were two reported cases of HLH resulting from ICI combined with anti-angiogenic therapy. One of these previous cases suggested that the development of HLH was closely linked to anti-angiogenic therapy [26]. Several studies using the FDA adverse event reporting system (FAERS) database have demonstrated that bevacizumab combined with PD-1 monoclonal antibody increases the risk of serious adverse effects such as fever, physical condition deterioration, thrombocytopenia, bone marrow failure, and neutropenia in oncology patients [59, 60]. Regarding specific mechanisms, VEGF can inhibit T cell function, increase Tregs and MDSCs, and hinder the differentiation and activation of DCs [61]. However, the number of HLH cases due to ICI combined with anti-vascular therapy remains low, and further confirmation of the relevance is needed.

In our case, the patient had a history of psoriasis, and our case review found that 20.5% of patients with ICI-related HLH had a history of autoimmune disease. Recent studies have demonstrated that pre-existing autoantibodies (including antithyroid [62], antinuclear [63] and other autoimmune-related antibodies [64]) are strong biomarkers for irAE, and autoimmune diseases such as rheumatoid arthritis and psoriasis have been shown to confer an elevated risk of irAEs [65]. We suggest that, in HLH as in other irAEs, patients with autoantibodies and autoimmune diseases are likely to be at a higher risk of developing the disease.

Among the symptoms exhibited by the patient, fever remains the most common. Interestingly, in terms of laboratory tests, there is a difference from our case compared to most of the reported cases, as they often experienced a drop in platelets as the first symptom. However, our case showed granulocyte deficiency as the initial abnormal laboratory test, and the platelet count remained normal from the onset of HLH to the end. No laboratory test results similar to ours have been found in the cases collected so far.

Regarding the treatment of HLH, patients (n = 25 or 48%) used steroids alone or in combination with antibiotics. Typically, patients with HLH refractory to steroids are treated with immunosuppressive drugs such as cyclophosphamide, or with etoposide, a drug that may even be used as a front-line treatment in immunosuppressed patients [66]. IL-6 inhibition with specific anti-IL-6 receptor antibodies, such as tocilizumab, has proven highly effective against cytokine release syndrome [67]. One study suggested that IL-6 blockade administered alongside ICIs can ameliorate irAEs while enhancing the antitumoral effect of ICIs [68]. In the 10 cases we collected where tocilizumab was used, only 1 death occurred, which may indicate the effectiveness of tocilizumab in treating HLH. However, confirmation through larger-scale case information is needed.

The prognosis of ICI-associated HLH is relatively favorable, with only 15.3% mortality in the pooled cases, which is much better than the 41% mortality reported in other types of HLH [2]. However, among the four patients who developed immune-associated encephalitis, we found three deaths, suggesting that immune-associated encephalitis is a significant factor contributing to a poor prognosis.

ICI-associated HLH is undoubtedly a challenging complication that clinicians strive to avoid. However, we observed that patients who recover from it may experience improved tumor control. In recent years, numerous studies have demonstrated that irAEs may lead to a better prognosis for tumor treatment [58], but the link between high-grade irAEs and tumor prognosis remains poorly understood. A retrospective study based on a small sample showed that patients with grade 3 or higher irAEs had longer overall survival (OS) than those with grade 1 or 2 irAEs [69]. Additionally, the results of a recent study, which combined data from three large clinical trials, revealed that grade 3–4 irAEs had a favorable prognostic effect over time compared to patients without irAEs [70]. According to our summary data, ICI-associated HLH tended to have a very positive effect on tumor outcome, with varying degrees of tumor regression and stabilization in 80% of cases with available prognosis information. This effect was reported to persist after a period of follow-up in 8 cases, which aligns with our own case suggesting that recovery from ICI-associated HLH can have a significant impact on tumor tissue. This phenomenon was first proposed in 2016 by M Takeshita et al. [10], who discovered the coincidence of immunotherapy-associated hemophagocytic syndrome and rapid tumor regression. However, the exact mechanism still requires further investigation, as the number of cases is small. Nonetheless, the exact mechanism behind this observation warrants further investigation, given the limited number of cases currently available for study. It is crucial to continue researching this topic, taking into account that the primary goal of medical professionals is to avoid complications like ICI-associated HLH and provide the most effective and safe treatment for patients.

In conclusion, ICI-associated HLH may occur in any ICI and in any cancer, and combined anti-vascular therapy and the presence of a history of autoimmune disease may increase the incidence of HLH, with complete recovery in most cases of HLH after steroid-based therapy, and may have a favourable impact on tumour outcome.

HLH, a relatively understudied complication, tends to occur more frequently in patients with a history of autoimmune diseases. Our findings indicate that recovery from HLH during treatment may correlate with improved survival outcomes, underlining the need for further research in this area.