Development of hepatitis triggered by SARS-CoV-2 vaccination in patient with cancer during immunotherapy: a case report

Patients with cancer have a well-known higher risk of severe COVID-19 [1]. After the emergence of the COVID-19 pandemic, caused by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), a large number of vaccines were in development. A recent paper reviewed cancer patients’ immune response to the main approved COVID-19 vaccines: the two mRNA vaccines (mRNA-1273 and the Pfizer-BioNTech vaccine) and the viral vector-based vaccines (ChAdOx1 nCov-19/AZD1222 and Ad26.COV2.S) [2]. In Italy, only the two mRNA vaccines have been approved for immunocompromised subjects, including the patients with cancer [3].

Expert consensus advocates for COVID-19 vaccination in the patients with cancer [4–6], although there are still many unclear issues about its safety and efficacy in terms of the magnitude and durability of the humoral and cell-mediated immune response in this frail population. A retrospective cohort study of 373 patients with cancer at a tertiary cancer center in London collected data of both solicited and unsolicited adverse events after at least one dose of COVID-19 vaccine. The authors showed that patients undergoing immune checkpoint inhibitor (ICI) treatment (15.3%) are less at risk of developing any vaccine-related adverse event (AE; odds ratio [OR]: 0.495; 95% CI: 0.256–00.958; p = 0.0037) [7]. The safety of COVID-19 vaccines in the patients with cancer, demonstrated in this study, is similar to the results of another observational study (SOAP-02) [8]: intriguingly, one patient during immunotherapy developed grade 4 transaminitis of undetermined cause 3 weeks after Pfizer/BioNTech COVID-19 vaccine.

In contrast, immune-related AEs (irAEs) caused by ICIs are better recognized. Usually, irAEs occur within few weeks after initiation of ICIs, but they have been documented 1  year after discontinuation of the therapy [9]. Skin AEs (rash, pruritus and vitiligo) are the most frequent irAEs (34–45%) [10], followed by colitis, hepatitis, pneumonitis and immune-related endocrinopathies [9]. A systematic review and meta-analysis of 15 studies showed that the incidence of all-grades diarrhea was 13.7% with PD-1 inhibitors and 35.4% with CTLA-4 inhibitors, and the incidence of all-grades colitis was 1.6 and 8.8%, respectively [11].

The incidence of all-grade immune-related acute hepatitis ranges between 4 and 9% of patients treated with anti-CTLA-4 and 18% of patients who received a combination of anti-PD-1 and anti-CTLA-4, whereas liver IRAEs are less frequent with anti-PD-1 alone, with an incidence of 1–4% [12].

Case report

The present report describes a case of hepatitis and colitis that occurred soon after a 52-year-old woman on immunotherapy received her first Pfizer-BioNTech COVID-19 vaccine dose (Figure 1). She had been diagnosed with lung adenocarcinoma with bone metastases; molecular analyses revealed that the tumor was negative for EGFR mutations and ALK gene rearrangements and that <1% of the tumor cells expressed PD‐L1. Accordingly, first‐line treatment with carboplatin, pemetrexed and pembrolizumab was started on 25 February 2021. Concurrent with the detection of cancer, hepatitis B virus (HBV) infection was diagnosed: HBsAg seropositive, HBV-DNA <20 IU/ml, HBeAg seronegative and anti-HBe positive, with normal liver function test. She received entecavir as antiviral prophylaxis before starting chemo-immunotherapy. Her medical history was unremarkable. No specific personal or family history suggested exposure to the major hepatotropic viruses.

After three cycles of therapy, the patient received the first dose of Pfizer-BioNTech COVID-19 vaccine without immediate side effects. After 10 days, she presented with >10 nonbloody bowel movements a day. The main laboratory data are summarized in Table 1. Abdominal ultrasound was unremarkable. Colon biopsies revealed microscopic colitis (lymphocytic subtype) with a concomitant eosinophilic infiltration, and liver histology showed only portal inflammation, without evidence of piecemeal necrosis and lobular hepatitis. Neither hepatocyte resetting nor plasma cell infiltrate was observed. Bile ducts were normal, and no cholestatic reaction was present.

The Revised Original Score for autoimmune hepatitis pretreatment [13] was 2.

The corticosteroid treatment with high-dose prednisone (1 mg/kg) was started with the rapid normalization of liver enzymes and the improvement of diarrhea. This case was notified to the Italian Sanitary Authority (Agenzia Italiana del Farmaco [AIFA]), and the patient resumed only chemotherapy. The main laboratory data at the resolution of irAEs are summarized in Table 2.

Notably, 3 weeks after the first dose, the patient developed a positive response in terms of both humoral and cell-mediated response. Despite the absence of a second dose of vaccine, the immune response remained sustained after 21 days of follow-up (Table 3).

For this reason, it was decided to omit the second dose of COVID-19 vaccine and to recheck her humoral and cell-mediated response after 3 and 6 months with the evidence of the persistence of only humoral response (Figure 2).

Discussion

As previously reported, some cases of liver toxicity with the characteristics of autoimmune hepatitis have been described after the administration of COVID-19 vaccine. Lodato and colleagues [14] reported severe cholestatic hepatitis developed in a 43-year-old woman a few days after the second dose of Pfizer-BioNTech COVID-19 vaccine. The liver biopsy demonstrated the presence of eosinophil infiltrate with the absence of autoantibodies, but the patient had a dramatic response to steroid treatment, similar to autoimmune hepatitis. Other analogue cases were described by Bril [15] and Londoño [16] – in a 35-year-old woman in her third month postpartum 1 week after the first dose of Pfizer-BioNTech COVID-19 vaccine and in a 41-year-old woman after her first dose of mRNA-1273 vaccine, respectively. They were apparently in good health condition. Antinuclear antibody was positive in both cases, whereas the anti-smooth-muscle antibody was positive and serum IgG level was raised only in the latter case. Clayton-Chubb et al. [17] described a case of vaccine-induced autoimmune hepatitis (AIH) after COVID-19 vaccination in a 36-year-old Iraqi-born man without apparent confounding factors. In contrast to the previous cases, this patient had received the Oxford-AstraZeneca vaccine, an adenovirus-based vaccine. Chow et al. [18] recently conducted a systematic search of the literature about the documented cases of AIH following COVID-19 vaccination [14–17,19–30]. They reported that 29 patients received mRNA vaccines and three patients received the Oxford-AstraZeneca vaccine. None of these patients had solid tumors, and only 10 had a history of liver disease [18]. More recently, Pinazo-Bandera et al described two new cases of AIH related to COVID-19 vaccination [31] and suggested that regulatory authorities should include this potential AE on the label of COVID-19 vaccines. A summary of the reported cases of vaccine-induced AIH is shown in Table 4

These cases indicate the need to recognize and promptly treat this unusual side effect irrespective of the mechanism of action of the vaccines.

The present case is unusual for the co-occurrence of two immune-mediated reactions: mixed eosinophilic/lymphocytic colitis and alanine aminotransferase (ALT) elevation. We have performed a complex diagnostic pathway to evaluate HBV reactivation, autoimmune hepatitis and a drug-induced liver injury related to ICI administration.

First, the authors hypothesized that the HBV reactivation was induced by anti-PD-1/PD-L1. The blocking of the PD-1/PD-L1 axis may lead to the destruction of hepatocytes with the release of previously latent virus into the circulation and promote proliferation of Tregs with consequent increased immunosuppression, and hence the reactivation of HBV [32]. The patient was chronically infected with HBV, but the HBV-DNA remained unquantifiable, so this hypothesis was discarded.

Second, immune-mediated toxicity was considered. Colitis is the second most commonly reported AE with ICI administration, and the symptoms typically develop from 6 to 8 weeks from the start of treatment; median onset of transaminase elevation is approximately 6–14 weeks after starting ICIs [33]. The pathogenesis of ICI-induced hepatitis is not well understood. Two recent papers point out the immunological mechanism in animal models [34,35]: macrophages and neutrophils are mediators and effectors of aberrant inflammation in TH1-promoting immunotherapy, suggesting distinct mechanisms of toxicity and antitumor immunity.

What role might the vaccine have played?

The mRNA vaccine strongly stimulates innate immunity by the immunostimulatory properties of mRNA, which triggers intracellular innate sensors, including Toll-like receptors 3 and 7 and components of the inflammasome, resulting in the production of interferon I and other pro-inflammatory cytokines and chemokines [36]. This mechanism is the basis of the immunologic activation leading to the neutrophil-driven liver damage, and it has been recognized as a probable effector of immune-mediated hepatitis following ICI administration. Ultimately, the clinical evolution was consistent with an immune-mediated side effect of ICIs, but the vaccination may have triggered such toxicity.

In a cohort study, none of the 134 patients enrolled who had received two doses of the BNT162b2 COVID-19 vaccine reported any severe irAE [37]. Because both ICIs and COVID-19 vaccines stimulate the immune response, it has been hypothesized that these vaccines may increase the incidence of the immune-related AEs with ICI treatment. To date, there are no data demonstrating a direct answer.

Conclusion

This case report may represent the first to describe a complex interplay between an immune-mediated reaction triggered by vaccination and PD-L1 co-administration. It may be interesting for clinicians involved in the management of cancer patients who receive COVID-19 vaccination. The main limit of this report is that the diagnosis was one of exclusion; it is therefore not possible to define the exact role played by the various drugs (ICIs and COVID-19 vaccine). Importantly, this report should not deter individuals from getting vaccinated but only increase awareness of possible pharmacological interactions.