Virus-specific memory T cell responses unmasked by immune checkpoint blockade cause hepatitis

Hutchinson, James A.; Kronenberg, Katharina; Riquelme, Paloma; Wenzel, Jürgen J.; Glehr, Gunther; Schilling, Hannah-Lou; Zeman, Florian; Evert, Katja; Schmiedel, Martin; Mickler, Marion; Drexler, Konstantin; Bitterer, Florian; Cordero, Laura; Beyer, Lukas; Bach, Christian; Koestler, Josef; Burkhardt, Ralph; Schlitt, Hans J.; Hellwig, Dirk; Werner, Jens M.; Spang, Rainer; Schmidt, Barbara; Geissler, Edward K.; Haferkamp, Sebastian

doi:10.1038/s41467-021-21572-y

Download PDF

Article
Open access
Published: 04 March 2021

Virus-specific memory T cell responses unmasked by immune checkpoint blockade cause hepatitis

James A. Hutchinson ORCID: orcid.org/0000-0002-1199-4210¹,
Katharina Kronenberg¹,
Paloma Riquelme¹,
Jürgen J. Wenzel ORCID: orcid.org/0000-0001-8422-6581²,
Gunther Glehr ORCID: orcid.org/0000-0002-1495-9162³,
Hannah-Lou Schilling ORCID: orcid.org/0000-0002-5506-3506¹,
Florian Zeman⁴,
Katja Evert⁵,
Martin Schmiedel⁶,
Marion Mickler⁷,
Konstantin Drexler⁷,
Florian Bitterer¹,
Laura Cordero¹,
Lukas Beyer⁸,
Christian Bach⁹,
Josef Koestler²,
Ralph Burkhardt¹⁰,
Hans J. Schlitt ORCID: orcid.org/0000-0002-3874-0296¹,
Dirk Hellwig ORCID: orcid.org/0000-0002-3056-0143⁶,
Jens M. Werner¹,
Rainer Spang³,
Barbara Schmidt ORCID: orcid.org/0000-0003-0059-0349²,
Edward K. Geissler^1,11 &
…
Sebastian Haferkamp⁷

Nature Communications volume 12, Article number: 1439 (2021) Cite this article

12k Accesses
36 Citations
33 Altmetric
Metrics details

Subjects

This article has been updated

Abstract

Treatment of advanced melanoma with combined PD-1/CTLA-4 blockade commonly causes serious immune-mediated complications. Here, we identify a subset of patients predisposed to immune checkpoint blockade-related hepatitis who are distinguished by chronic expansion of effector memory CD4⁺ T cells (T_EM cells). Pre-therapy CD4⁺ T_EM cell expansion occurs primarily during autumn or winter in patients with metastatic disease and high cytomegalovirus (CMV)-specific serum antibody titres. These clinical features implicate metastasis-dependent, compartmentalised CMV reactivation as the cause of CD4⁺ T_EM expansion. Pre-therapy CD4⁺ T_EM expansion predicts hepatitis in CMV-seropositive patients, opening possibilities for avoidance or prevention. 3 of 4 patients with pre-treatment CD4⁺ T_EM expansion who received αPD-1 monotherapy instead of αPD-1/αCTLA-4 therapy remained hepatitis-free. 4 of 4 patients with baseline CD4⁺ T_EM expansion given prophylactic valganciclovir and αPD-1/αCTLA-4 therapy remained hepatitis-free. Our findings exemplify how pathogen exposure can shape clinical reactions after cancer therapy and how this insight leads to therapeutic innovations.

IFNα induces CCR5 in CD4+ T cells of HIV patients causing pathogenic elevation

Article Open access 19 March 2024

Hélène Le Buanec, Valérie Schiavon, … Robert C. Gallo

Restoration of HBV-specific CD8+ T-cell responses by sequential low-dose IL-2 treatment in non-responder patients after IFN-α therapy

Article Open access 05 November 2021

Dongyao Wang, Binqing Fu, … Haiming Wei

Circulating serum HBsAg level is a biomarker for HBV-specific T and B cell responses in chronic hepatitis B patients

Article Open access 04 February 2020

Jin Hyang Kim, Alip Ghosh, … Bhawna Poonia

Introduction

Treatment with anti-programmed cell death protein-1 (αPD-1) alone or combined with anti-cytotoxic T-lymphocyte-associated protein 4 (αCTLA-4) has greatly improved the prognosis of metastatic melanoma with a substantial fraction of patients achieving long-lasting partial or complete clinical responses^1,2,3,4,5,6. The principal limitation of αPD-1/αCTLA-4 dual therapy is the relatively high rate of immune-related complications, which commonly include colitis, hepatitis and thyroiditis, but may also include pneumonitis, myocarditis and encephalitis^7,8. Such adverse events range from mild reactions to life-threatening presentations, but are nevertheless clinically significant incidents that influence patient management^9,10. Avoidance of complications has led to a somewhat complicated treatment decision-making process (Fig. 1a) that seeks to balance individual risk and capacity to tolerate adverse reactions against the superior clinical responses achieved with dual αPD-1/αCTLA-4 therapy over αPD-1 monotherapy^11,12,13.

**Fig. 1: Individual predisposition to hepatitis after αPD-1/αCTLA-4 treatment.**

Our aetiological understanding of colitis, hepatitis and thyroiditis caused by immune checkpoint blockade remains incomplete¹⁴; however, these treatment-related conditions are clearly different from known autoimmune diseases^15,16. Such adverse reactions are usually attributed to generalised dysregulation of T cells^17,18 or autoantibodies¹⁵. Here, we explore a different view—namely, that individual patients are predisposed to particular treatment-related complications by virtue of pathogen-specific memory T cells present in their repertoire before immune checkpoint blockade is given.

Cytomegalovirus (CMV) seroprevalance amongst European adults is over 50%. Primary CMV infection usually presents during young adulthood as an acute febrile illness, after which life-long latency is established. The exact sites and mechanisms of CMV latency are not fully resolved, but myelomonocytic cells in blood and tissues are major reservoirs¹⁹. CMV reactivation characteristically follows immune compromise caused by immunosuppressive therapy or other viral infections²⁰. Clinical manifestations of CMV reactivation are variable, ranging from classic febrile illness to tissue-invasive disease presenting as colitis, hepatitis, pneumonitis or encephalitis²¹. Reactivation of CMV causes cellular injury through the cytopathic effect of viral replication, but more conspicuously through triggering T cell immunity. Pneumonitis in transplant recipients illustrates how CMV-reactive T cells provoke extensive, life-threatening inflammation despite low-level viral replication²². Even in anatomical locations from which T cells are normally excluded, such as the retina, T cell-mediated responses significantly contribute to CMV-related disease. Thus, low-level and compartmentalised reactivation of CMV can be responsible for serious acute immunopathology in humans^23,24,25.

Cycles of CMV latency and reactivation lead to a now well-recognised phenomenon of memory inflation, which is typified by failure of primed CMV-specific CD8⁺ T cell populations to contract and give rise to central memory T cells²⁶. Instead, a dominant pool of effector memory CD8⁺ T cells that does not become functionally exhausted is maintained²⁷. Such inflationary CD8⁺ T cells achieve high frequencies over time, sustained through repetitive stimulation by viral reactivation or long-term persistence of viral antigens²⁸. Memory inflation is described as an exclusive feature of CD8⁺ effector T cell responses, although durable expansion of CMV-specific CD4⁺ T cells can occur in transplant recipients²⁹.

This study identifies pre-therapy expansion of CD4⁺ effector memory T cells (T_EM) in CMV-infected patients with metastatic melanoma as a prognostic marker of αPD-1/αCTLA-4-related hepatitis. Opting for αPD-1 monotherapy or prophylactic valganciclovir in CMV IgG⁺ patients with unresectable metastatic disease and pre-treatment expansion of CD4⁺ T_EM cells reduces their risk of hepatitis. We conclude that sub-clinical reactivation of CMV plays a previously unsuspected role in the immunopathogenesis of αPD-1/αCTLA-4-related hepatitis in a subgroup of patients.

Results

Clinical features do not predict αPD-1/αCTLA-4-related hepatitis

If immune-related complications of αPD-1/αCTLA-4 therapy resulted solely from treatment-induced immunological effects^30,31, we might expect hepatitis, colitis and thyroiditis to coincide often³² (Fig. 1b). However, no such associations were found in a cohort of 89 patients with metastatic melanoma treated with αPD-1/αCTLA-4 dual therapy (Fig. 1c, Supplementary Fig. 1 and Supplementary Tables 1–3). This raises the possibility that some individuals are predisposed to particular adverse reactions.

In this study, we focused on factors that predispose to αPD-1/αCTLA-4-related hepatitis. Except for patients who developed hepatitis being generally younger (Supplementary Table 4) no associations were found between incidence of hepatitis and pre-treatment clinical variables (Supplementary Tables 4 and 5). Conventional biochemical markers of liver dysfunction, systemic inflammation or tumour burden did not discriminate between patients who did or did not develop hepatitis (Fig. 1d–j). We found no associations between incidence of hepatitis and seropositivity for hepatitis B virus (HBV), hepatitis C virus (HCV) or hepatitis E virus (HEV; Fig. 1k–m). Incidence of hepatitis was not greater with more rounds of αPD-1/αCTLA-4 therapy; in fact, as immune-related complications are reason to withhold treatment, patients who developed hepatitis tended to receive fewer rounds of αPD-1/αCTLA-4 (Fig. 1n). Autoimmune hepatitis-related autoantibodies were not associated with risk of hepatitis (Supplementary Table 6). In short, case history and routine clinical investigations did not predict hepatitis after dual therapy in our study cohort.

CD4⁺ T_EM cell expansion predicts αPD-1/αCTLA-4-related hepatitis

We next profiled circulating leucocyte subsets immediately prior to start of dual therapy in a training set of n = 44 patients with unresectable metastatic melanoma (Supplementary Figs. 2–8). Comparing the frequency of 50 principal leucocyte subsets between patients who subsequently did (n = 18) or did not (n = 26) develop hepatitis revealed significant over-representation of CD4⁺ T_EM cells associated with hepatitis (Fig. 2a, b). This observation was confirmed in a validation set (n = 45) of similar patients (Fig. 2c). Baseline expansion of CD4⁺ T_EM cells was a fair discriminator (AUROC = 0.706) of patients who did or did not later develop hepatitis in the validation set (Fig. 2d). Strikingly, CD4⁺ T_EM cell frequencies in all patients with completely resected disease were normally distributed with no extreme cases of baseline CD4⁺ T_EM cell expansion, allowing us to set a threshold below which 99% of patients should fall (Fig. 2e). Applying this stringent cut-off to our validation cohort of patients with unresectable metastatic disease (Fig. 2f) allowed us to correctly predict hepatitis in 6 of 6 patients with CD4⁺ T_EM^≥21% (specificity = 100%) and correctly identify 6 of 20 patients who developed hepatitis (sensitivity = 30.0%). CD4⁺ T_EM^≥21% patients were not predisposed to colitis (Fig. 2g), thyroiditis or superior clinical responses to therapy (Supplementary Table 7). Hepatitis was not more severe in CD4⁺ T_EM^≥21% patients compared to CD4⁺ T_EM^<21% patients (Fig. 2h) and no difference in time-to-first presentation was observed (Fig. 2i). Hence, baseline CD4⁺ T_EM expansion is a specific, but relatively insensitive prognostic marker of individuals who are predisposed to hepatitis after αPD-1/αCTLA-4 therapy. Owing to its low sensitivity, classifying patients as CD4⁺ T_EM^<21% has limited clinical utility; however, specifically identifying cases at high-risk of hepatitis before treatment is useful from a research perspective because we can then prospectively study the immunological basis of their predisposition.

**Fig. 2: Circulating CD4⁺ T effector memory cell frequency predicts hepatitis after Nivolumab plus Ipilimumab treatment.**

Opting for αPD-1 monotherapy reduces risk of hepatitis

Although αPD-1 monotherapy achieves fewer sustained clinical responses in patients with metastatic melanoma, it causes fewer adverse reactions than αPD-1/αCTLA-4 dual therapy¹. Therefore, there is a clear rationale for avoiding αPD-1/αCTLA-4 dual therapy in patients at risk of hepatitis. Four CD4⁺ T_EM^≥21% patients with metastasic disease presented during our study, but were electively treated with αPD-1 monotherapy (Fig. 2j). Whereas 12 of 12 CD4⁺ T_EM^≥21% patients treated with αPD-1/αCTLA-4 dual therapy developed hepatitis, 3 of 4 CD4⁺ T_EM^≥21% patients who received αPD-1 monotherapy remained hepatitis free. Reducing the incidence hepatitis by opting for αPD-1 monotherapy hints that CTLA-4 blockade may be mechanistically important for development of hepatitis in CD4⁺ T_EM^≥21% patients³³.

CD4⁺ T_EM frequency is an independent predictor of hepatitis

We next investigated possible causes of baseline CD4⁺ T_EM expansion. To establish CD4⁺ T_EM expansion as an independent risk factor, we first sought to exclude possible confounding covariables. Baseline CD4⁺ T_EM^≥21% was not related to sex, age or body-mass index (Supplementary Table 7 and Fig. 3a, b). No significant correlation was observed between baseline CD4⁺ T_EM % and serum levels of C-reactive protein (CRP), aspartate aminotransferase (AST), alanine aminotransferase (ALT), gamma-glutamyl transferase (γ-GT) or total bilirubin (Fig. 3c–g). Because glucose metabolism provides a reliable measure of liver inflammation³⁴, we were able to further rule out liver inflammation prior to therapy as a predisposing cause of CD4⁺ T_EM^≥21% hepatitis by estimating hepatic glucose uptake in patients who underwent ¹⁸F-fluoro-2-deoxy-d-glucose (¹⁸F-FDG) PET/CT studies as part of routine tumour staging shortly before starting immunotherapy (Fig. 3h). This retrospective comparison showed ¹⁸F-FDG standard uptake ratios (SUR_mean liver) were higher in CD4⁺ T_EM^<21% patients than CD4⁺ T_EM^≥21% patients with hepatitis (Fig. 3i). Therefore, our biochemical and radiological investigations argue against pre-treatment liver inflammation as a common cause for baseline CD4⁺ T_EM expansion and treatment-related hepatitis.

**Fig. 3: CD4⁺ T_EM frequency is an independent predictor of αPD-1/αCTLA-4-related hepatitis.**

Because CD4⁺ T_EM expansion is principally a feature of metastatic melanoma (Fig. 2e), we next asked whether CD4⁺ T_EM% was somehow associated with tumour burden. No relationship was found between CD4⁺ T_EM% and the number of organs containing metastases, presence of hepatic metastases or serum lactate dehydrogenase (LDH) and protein-S100 serum levels (Fig. 3j–m and Supplementary Table 7). There was a tendency for CD4⁺ T_EM^≥21% patients to have received previous therapy of any type (including BRAFi/MEKi or Talimogen laherparepvec) but besides implying a longer disease course before starting dual therapy, we see no mechanistic interpretation because these drugs share no common pharmacological action (Fig. 3n). In short, we found no obvious clinical association to explain the predisposition of CD4⁺ T_EM^≥21% patients to hepatitis.

CD4⁺ T_EM cell expansion is associated with chronic T cell activation

Although CD4⁺ T cell counts were not different between CD4⁺ T_EM^<21% and CD4⁺ T_EM^≥21% patients (Supplementary Table 7), there were significantly fewer circulating naïve T cells in CD4⁺ T_EM^≥21% patients (Fig. 4a). This reduction was complemented by increased absolute numbers of CD4⁺ T_EM and T_EMRA cells (Fig. 4b, c) whereas central memory T cell (T_CM) counts tended to be lower in CD4⁺ T_EM^≥21% patients (Fig. 4d). We next defined other leucocyte subsets that correlated with CD4⁺ T_EM % in both our training and validation cohorts (Fig. 4e). Most covariant populations were T cells, particularly CD4⁺ T cell subsets. Changes in the CD4⁺ T cell compartment of T_EM^≥21% patients were consistent with chronic activation, particularly downregulation of CD27 and upregulation of CD57, HLA-DR and CD279 expression (Fig. 4f–i). Similar differences were detected in CD8⁺ T cells (Supplementary Fig. 9). Notably, CD4⁺ T_EM^≥21% patients tended to have higher frequencies of CD160⁺ CD244⁺ exhausted CD8⁺ T cells than CD4⁺ T_EM^<21% patients (Fig. 4j).

**Fig. 4: CD4⁺ T_EM cell enrichment is associated with chronic activation and increased numbers of effector memory CD4⁺ T cells.**

Most adults are persistently infected with Torque Teno Virus (TTV), a non-enveloped, circular, single-stranded DNA virus of the Anellovirus family that causes no apparent disease. T cell immunity controls TTV replication in healthy people, whereas TTV copy number in immunocompromised individuals is increased; therefore, TTV load in plasma is a useful marker of immune competence. Despite accumulating chronically activated and exhausted T cells, TTV load in CD4⁺ T_EM^≥21% versus CD4⁺ T_EM^<21% patients was not different, implying that CD4⁺ T_EM^≥21% patients are not generally immunocompromised (Fig. 4k).

CD4⁺ T_EM cell expansion correlates with immunity to CMV

To briefly summarise, baseline expansion of CD4⁺ T_EM cells identifies a subset of patients with metastasic melanoma who are predisposed to αPD-1/αCTLA-4-related hepatitis. These patients show signs of persistent or recurrent CD4⁺ and CD8⁺ T cell responses³⁵, but are otherwise clinically and immunologically unremarkable. It is usually very difficult to isolate factors responsible for such a non-specific immunological picture³⁶. Fortunately, a surprising pattern emerged from our dataset: CD4⁺ T_EM^≥21% patients mostly presented between September and February (Fig. 5a). This seasonal influence was apparent over all 4 years of the study, leading us to think that CD4⁺ T_EM expansion might be driven by a pathogen. Active infections with hepatitis viruses, including HBV and HCV, are relative contraindications to αPD-1/αCTLA-4 therapy (Supplementary Table 7), so we were forced to consider possible involvement of other hepatotropic viruses, such as HEV and herpesgroup viruses.

**Fig. 5: CD4⁺ T_EM cell expansion correlates with immunity to cytomegalovirus.**

Latent CMV infection is prevalent amongst European adults and involves many organs, including liver. Compartmentalised reactivation of CMV infection occurs in immunocompromised individuals and those with concomitant viral upper respiratory tract infections²⁰. In our patients, we found a strong association between serum CMV-specific IgG antibody levels and CD4⁺ T_EM status (Fig. 5b). Furthermore, T cell reactivity against CMV pp65 peptides in an IFN-γ enzyme-linked immunospot (ELISPOT) assay was tightly associated with CD4⁺ T_EM status (Fig. 5c). These results led us to hypothesise a causal relationship between compartmentalised, sub-clinical CMV reactivation in liver and chronic enrichment of CD4⁺ T_EM cells in blood.

Because baseline CD4⁺ T_EM cell expansion was mainly a feature of CMV IgG⁺ patients (Fig. 5d), we next asked whether incorporating CMV IgG status could improve our predictive model. When considering only CMV IgG⁺ cases, CD4⁺ T_EM % was a more discriminatory marker of hepatitis (Fig. 5e). Therefore, using our training set filtered for CMV IgG⁺ cases, we revised our cut-off to CD4⁺ T_EM ≥ 16% in CMV IgG⁺ cases in order to achieve a better compromise between specificity and sensitivity (Supplementary Fig. 10). The new cut-off performed well in the validation set with a correct classification rate of 81.0% (positive predictive value (PPV) = 85.7% and negative predictive value (NPV) = 78.6%) compared to 42.9% under the no-information model.

For descriptive purposes, we can consider the distribution of patients from the combined training and validation sets according to CMV IgG and CD4⁺ T_EM status (Fig. 5f). Our predictive model explains the occurrence of hepatitis in 34 of 89 (38.2%) of all patients (red boxes). Two of 17 (12%) CMV IgG⁺ CD4⁺ T_EM^≥16% patients (green box) did not develop hepatitis: one of these received four rounds of αPD-1/αCTLA-4 therapy and registered no complications; however, the other received only 1 round of αPD-1/αCTLA-4 before treatment with prednisolone from weeks 2 to 7 for colitis. We speculate 4 of 23 CMV IgG⁺ CD4⁺ T_EM^<16% patients who were incorrectly predicted as hepatitis-negative (pink box) were sampled at an early stage of CD4⁺ T_EM cell expansion and our test was not sensitive enough to discriminate the change. Our model does not explain why 19 of 49 CMV IgG⁻ patients developed hepatitis (blue box) but we believe these represent an aetiologically distinct subset.

CMV-reactive CD4⁺ T cells are enriched in CD4⁺ T_EM ^high patients

We next asked whether CMV-reactive CD4⁺ T cells were enriched in patients with baseline CD4⁺ T_EM expansion. Peripheral blood mononuclear cells (PBMC) were stimulated in culture for 18 h with CMV lysates before IFN-γ, TNFα, IL-17, IL-4 and CD69 expression was measured by flow cytometry (Fig. 6a and Supplementary Figs. 11 and 12). To detect hyporesponsive CMV-specific T cells, paired cultures were also stimulated in the presence of neutralising antibodies against PD-1 and CTLA-4. In contrast to CMV IgG⁻ and CMV IgG⁺ CD4⁺ T_EM^<16% patients, we readily detected IFN-γ-producing, CMV-reactive CD4⁺ T cells in CMV IgG⁺ CD4⁺ T_EM^≥16% patients (Fig. 6b). This difference implies a higher number of circulating IFN-γ-producing, CMV-reactive CD4⁺ T cells (Fig. 6c). Proportions of CMV-reactive CD4⁺ T cells expressing IL-17, IL-4 or only TNFα were too low to quantify accurately.

**Fig. 6: CMV-reactive CD4⁺ T cells are enriched in patients with CD4⁺ T_EM cell expansion.**

Valganciclovir reverses αPD-1/αCTLA-4-related hepatitis

Strong associations between T cell immunity against CMV, expansion of CD4⁺ T_EM and risk of treatment-related hepatitis imply an underlying pathogenic mechanism; however, to demonstrate a causal relationship, we needed to establish that inhibiting CMV replication was effective in treating or preventing hepatitis. Valganciclovir is an anti-viral agent licensed for therapy and prophylaxis of CMV infections in immunocompromised patients. It is a widely used and safe drug with a well-known and broadly acceptable profile of adverse effects. Therefore, we next gave valganciclovir treatment to two patients with αPD-1/αCTLA-4-related hepatitis.

The first patient was a comparatively fit 54-year-old man with a new diagnosis of metastatic melanoma (Fig. 7a). Staging CT revealed multiple lung and bone metastases, but no metastasis to liver or lymph nodes. At baseline, the patient was CMV-IgG⁺ and CD4⁺ T_EM = 18.7% (Fig. 7b). Nivolumab plus Ipilimumab were given at 0, 3 and 6 weeks. A fourth round of αPD-1/αCTLA-4 at week 9 was omitted owing to joint pain and the patient received two rounds of Denosumab (αRANKL). Nivolumab monotherapy was administered in weeks 12, 16, 20 and 24. Staging CT at week 14 showed a partial clinical response with significant regression of pulmonary metastases; hence, checkpoint blockade was of clear clinical benefit.

**Fig. 7: Case 1: valganciclovir treatment of late-onset αPD-1/αCTLA-4-related hepatitis.**

The patient received a dose of αPD-1 monotherapy in week 32. Four weeks later, he presented in clinic with grade 4 hepatitis. Strong reactions against pp65 and IE-1 peptides were measured by ELISPOT. Accordingly, the patient was treated with 900 mg/day valganciclovir plus prednisolone. Liver function tests (LFTs) returned to normal, so over the subsequent 2 weeks, prednisolone was weaned and then valganciclovir was withdrawn. Seven days later, the patient returned to clinic with deranged LFTs. Valganciclovir and prednisolone were both reinstated, leading to a rapid normalisation of LFTs. Two days later, a for-cause liver biopsy was performed (Fig. 7c–e). The tissue was essentially normal, although a few dying hepatocytes and isolated lymphocytes were observed in the portal areas. Immunohistochemical staining for CMV and nested CMV PCR from fixed biopsy material were negative. The patient remained clinically stable for 9 days, so he was again weaned from prednisolone by the start of week 43; however, on this occasion, 900 mg/day valganciclovir treatment was continued. After complete withdrawal of prednisolone, while still receiving valganciclovir, the patient’s LFTs remained stable for a further 4 weeks. At the start of week 47, valganciclovir was withdrawn for a second time. Subsequently, the patient did not develop hepatitis and required no further steroids. The patient received further doses of αPD-1 monotherapy in weeks 54 and 58 without signs of recurrent liver injury. To us, this suggests the T cell antigens driving hepatitis in this patient were absent after valganciclovir treatment. In our clinical opinion, despite the absence of detectable virus, the patient’s fast recovery and unusual relapse of hepatitis are best explained by introduction, withdrawal and reintroduction of valganciclovir.

The second patient was a 49-year-old man who had an uncomplicated first year after αPD-1/αCTLA-4 and had been receiving Encorafenib/Binimetinib therapy for 7 weeks when he developed hepatitis in week 67 (Fig. 8a, b). Hepatitis failed to resolve despite withdrawal of B-RAFi/MEKi and escalating doses of prednisolone. 900 mg/day valganciclovir was introduced at the start of week 72 and a for-cause biopsy was performed on the same day (Fig. 8b). Four days later, the patient started a course of 1 mg/day mycophenolate mofetil. Strikingly, AST, ALT and γ-GT levels declined consistently from the day after valganciclovir was given, although total bilirubin continued to rise. Starting valganciclovir treatment appeared to be temporally related to increased reactivity in CMV IE-1 ELISPOT and a small decline in CD4⁺ T_EM frequency.

**Fig. 8: Case 2: valganciclovir treatment of late-onset αPD-1/αCTLA-4-related hepatitis.**

Liver biopsy revealed dense infiltration by lymphocytes, eosinophils and neutrophils with an accompanying ductal reaction (Fig. 8c–e). Extensive centrilobular hepatocyte necrosis was observed, which was also associated with inflammatory infiltration. These histopathological signs of significant liver damage could have been consistent with autoimmunity or viral infection. Both histological staining for CMV and CMV PCR from fixed biopsy material were negative. Although we were unable to detect virus, it is our clinical impression that markers of liver damage responded to valganciclovir treatment.

Valganciclovir prophylaxis prevents αPD-1/αCTLA-4-related hepatitis

We next asked whether valganciclovir prophylaxis could prevent αPD-1/αCTLA-4-related hepatitis in predisposed individuals. The 50th, 75th and 90th percentiles of time-to-hepatitis in CMV-IgG⁺ patients were 24, 52 and 84 days, respectively; therefore, we treated four CMV IgG⁺ CD4⁺ T_EM^≥16% patients receiving αPD-1/αCTLA-4 therapy with prophylactic valganciclovir for between 4 and 23 weeks (Fig. 9a). Two of these patients registered mild, transient transaminitis that resolved without treatment, so were not diagnosed with hepatitis. Hence, we observed no hepatitis in the four valganciclovir-treated patients, whereas 15 of 17 (88.2%) CMV IgG⁺ CD4⁺ T_EM^≥16% patients treated with αPD-1/αCTLA-4 and no valganciclovir prophylaxis developed clinical hepatitis (Fig. 9b). To us, these four cases are consistent with the idea that CMV plays an important aetiological role in development of hepatitis after αPD-1/αCTLA-4 therapy in a specific subset of patients.

Discussion

Our study seeks to explain why a subset of patients with advanced melanoma is predisposed to hepatitis following immune checkpoint blockade. In our model, metastatic disease promotes compartmentalised CMV reactivation in latently infected individuals, possibly triggered by an unknown winter virus infection; in turn, CMV stimulates expansion of CD4⁺ T_EM cells prior to immunotherapy. When αPD-1/αCTLA-4 therapy is given, these T_EM cells might inflict liver injury directly or promote bystander T cell responses. We propose CMV-driven T cell responses are a single, but salient example of a general mechanism responsible for hepatitis after checkpoint blockade because other viruses or toxic agents could also drive chronic CD4⁺ T cell activation before therapy. We speculate anti-viral immunity might also explain predisposition to other adverse reactions following αPD-1/αCTLA-4 therapy, such as colitis or pneumonitis³⁷. The remarkable overlap between organs affected by tissue-invasive CMV disease and those susceptible to immune-related adverse reactions has not escaped our attention^38,39.

Being able to identify patients predisposed to hepatitis opens various possibilities for avoiding or preventing treatment-related hepatitis⁴⁰. Three of four CD4⁺ T_EM^≥21% patients assigned to αPD-1 monotherapy instead of αPD-1/αCTLA-4-treatment did not develop hepatitis as predicted; therefore, baseline CD4⁺ T_EM expansion could be taken as a relative contraindication to dual therapy. However, a more promising strategy is to suppress occult CMV reactivation. In our study, valganciclovir prophylaxis apparently prevented hepatitis in four of four CMV IgG⁺ CD4⁺ T_EM^≥16% patients. This approach must be refined in terms of dosing and duration of anti-viral therapy in randomised, controlled clinical trials, but could potentially achieve the superior outcomes of αPD-1/αCTLA-4 dual therapy over αPD-1 monotherapy with fewer complications.

Because baseline CD4⁺ T_EM expansion was observed in patients with metastatic disease, but not in patients with fully resected melanoma, it is possible that disseminated melanoma promotes low-level CMV reactivation, perhaps through compromised T cell immunity, which then stimulates virus-specific CD4⁺ T cell expansion. The prominent association between CD4⁺ T_EM expansion and CMV (in contrast to other herpesgroup virus infections) might simply reflect differences in tissue tropism and, we speculate, the capacity of compartmentalised CMV reactivation in liver to drive especially strong T cell responses. Seasonal variation in the incidence of CD4⁺ T_EM^high cases suggests that environmental factors influence expansion and contraction of the CD4⁺ T_EM cell pool^41,42. In particular, we suspect common winter viruses somehow promote CMV reactivation in liver.

Chronic, low-level exposure to viral antigens preferentially expands T_EM over T_CM^{43,44,45,46,47,48,49}, so CMV-reactive CD4⁺ T cell expansion might be sustained by very little CMV replication. On-going T cell reactions against CMV in CD4⁺ T_EM^high patients are likely to suppress CMV replication almost completely. Our inability to detect CMV in baseline serum samples from CD4⁺ T_EM^high patients or liver biopsies from patients with high-grade hepatitis is a challenge for our theory; however, it may simply be beyond our technical capability to detect very low levels of compartmentalised viral reactivation. In our view, the strong correlation between CMV-specific immunity and development of hepatitis, as well as the apparent effectiveness of valganciclovir, suggests a causal role for CMV in the aetiology of αPD-1/αCTLA-4-related hepatitis.

We do not presently know whether CMV-specific CD4⁺ T_EM cells inflict liver injury directly or provoke bystander T cell responses⁵⁰. Administration of αPD-1 or αCTLA-4 has been associated with CMV reactivation in colon, so it is probable that CMV also reactivates in liver during dual therapy⁵¹. In two patients with high-grade hepatitis, we observed liver inflammation subsided immediately after starting valganciclovir. In our view, the likely presence of viral antigen and apparent responsiveness to valganciclovir is consistent with a direct pathological effect of CMV-specific T cells, although we accept this will be hard to prove in patients.

Better understanding the contribution of anti-viral T cell immunity to checkpoint blockade-related adverse reactions leads us to unexpected solutions for preventing serious complications of cancer immunotherapy. From a clinical perspective, the immunological mechanisms described illustrate how patients’ medical histories, especially prior exposure to pathogens, impact their experience of cancer and response to therapy.

Methods

Study approval and patient management

This study involving human research participants was performed in accordance with the Declaration of Helsinki and all applicable German and European laws and ethical standards. Specifically, specimens were obtained from patients with Stage III or IV melanoma participating in a single-centre observational clinical trial authorised by the Ethics Committee of the University of Regensburg (approval 16-101-0125) and registered with clinicaltrials.gov (NCT04158544). All participants gave full, informed written consent. The first reported patient was recruited in October 2016 and the last reported patient was recruited in July 2020. Patients received standard-of-care treatment according to local guidelines (Fig. 1a). Stage IV patients with unresectable metastatic disease who received first- or second-line checkpoint inhibitor therapy were initially treated with Nivolumab (αPD-1; Bristol-Myers Squibb) and Ipilimumab (αCTLA-4; Bristol-Myers Squibb) for four cycles, and thereafter with Nivolumab maintenance therapy (3 mg/kg at 3-week intervals). Those patients with complete resection of Stage III melanoma who received adjuvant first-line checkpoint inhibitor therapy were treated for up to 1 year with Pembrolizumab (αPD-1; MSD) or Nivolumab.

Clinical flow cytometry

Detailed step-by-step protocols for preparation and analysis (including specimen gating strategies) of clinical samples by flow cytometry are available through Protocol Exchange⁵². In brief, peripheral blood samples were collected into ethylenediaminetetraacetic acid (EDTA)-vacutainers by venepuncture and then delivered to the immune monitoring laboratory at ambient temperature. Pre-analytical samples were stored for up to 4 h at 4 °C until processing. Whole blood was stained with DuraClone reagents (Duraclone IM Phenotyping Basic Tube, B53309; Duraclone IM T cell Subsets Tube, B53328; Duraclone IM TCRs Tube, B53340; Duraclone IM Treg Tube, B53346; Duraclone IM B cells Tube, B53318; Duraclone IM Dendritic Cells Tube, B53351; all from Beckman Coulter, Krefeld, Germany). Data were recorded with a Navios^TM cytometer running Cytometry List Mode Data Acquisition and Analysis Software version 1.3 (Beckman Coulter). Blinded analyses were performed by an experienced operator using Kaluza version 2.1 according to gating strategies illustrated in Supplementary Figs. 2–8.

Detection of CMV-reactive T cells by flow cytometry

Detailed step-by-step protocols are available through Protocol Exchange⁵³. Briefly, frozen PBMC from patients were thawed and resuspended in RPMI medium (ThermoFisher, Schwerte, Germany) supplemented with 5% heat-inactivated pooled, male-only human AB serum (ZKT Tübingen, Germany), 100 U/ml penicillin, 100 μg/ml streptomycin and 2 mM Glutamax (ThermoFisher). Cells were plated in 96-well round-bottom plates at 1 × 10⁶ cells in 150 µl/well then incubated at 37 °C with 5% CO₂. After 6 h, cultures were stimulated with 5 µg/ml CMV Lysate (Origene) and αCD28/αCD49 for 18 h. In some conditions, cultures were treated with neutralising monoclonal antibodies against PD-1 (MAB10864, R&D) plus CTLA‐4 (MAB3254, R&D) at 5 µg/ml each or mouse IgG1 isotype control (MAB002, R&D) at a final concentration of 10 µg/ml. During the last 4 h of cell culture, protein secretion was blocked with Protein Transport Inhibitor Cocktail (ThermoFisher) containing Brefeldin A und Monensin. Cytokine production was then assessed by flow cytometry. Dead cells were excluded with ViaKrome-808 Fixable Viability Dye (Beckman Coulter). Staining was performed in Cell Staining Buffer (BioLegend) with 10% FcR Block (Miltenyi). Cells were fixed and permeabilised using the eBioscience IC Fixation Buffer and Permeabilization Buffer (eBiosciences, ThermoFisher). Data were recorded with a Cytoflex LX^TM cytometer running CytoExpert Software (Beckman Coulter) and analysed using Kaluza version 2.1 (Beckman Coulter) according to the gating strategy shown in Supplementary Fig. 11.

¹⁸F-FDG PET/CT imaging

Hepatic glucose metabolism as a measure of liver inflammation³⁴ was analysed retrospectively in all 32 patients who underwent pretreatment ¹⁸F-FDG positron emission tomography and computed tomography (PET-CT) imaging as part of routine tumour staging shortly before starting immunotherapy. After overnight fasting, whole-body PET acquisitions covering at least the trunk were started 60–90 min after i.v. injection of ¹⁸F-FDG (3 MBq per kg body weight) on a Biograph Sensation 16 LSO or Biograph mCT40 Flow PET/CT scanner (both Siemens Healthcare, Germany) and lasted 18–24 min depending on patient size. The same area was covered by a low-dose CT scan (tube current <50 mAs, tube voltage 120 kV) if no contrast agents were used. Otherwise 130 ml of Accupaque^TM 300 (GE Healthcare) was applied as intravenous contrast agent with consecutive full-dose CT acquisition (120 kV, <100 mAs). PET images (slice thickness 5 mm) were corrected for random coincidences, decay, scatter, and attenuation and reconstructed iteratively using vendor parameter presets. PET and CT images were checked for breathing artifacts and scaled to standardised uptake values (SUV). SUV_mean Blood was measured in a thread-like region-of-interest (ROI) manually set into the aorta and SUV_mean Liver in a spherical ROI in the right liver lobe using the program ROVER (Version v3.0.35, Helmholtz-Zentrum Dresden-Rossendorf, Germany). SUR was calculated as with correction for the distribution time T of ¹⁸F-FDG to the reference time T₀ of 75 min post injection⁵⁴: SUR = (T₀/T) × (SUV_mean liver / SUV_mean blood).

Routine clinical investigations

Routine biochemical and haematological investigations were performed by an in-house accredited diagnostic laboratory (Institute of Clinical Chemistry and Laboratory Medicine, UKR). Virological investigations were conducted by an in-house accredited laboratory (Institute of Clinical Microbiology and Hygiene, UKR) following routine diagnostic procedures. CMV-reactive T cells were quantified by IFN-γ ELISPOT (Lophius, Germany) according to the manufacturer’s instructions in an accredited laboratory (UKR).

Statistics, data transformations and visualisation

Significance tests and curve-fitting were performed with GraphPad Prism 6.04 (GraphPad Software, Inc., La Jolla, USA) or SPSS® version 25 (IBM Analytics, New York, USA). As indicated, data were analysed using a two-tailed Mann–Whitney (M.W.) test, two-tailed Fisher’s exact test (F.E.), D’Agostino & Pearson’s normality test (K2), one-way or two-way ANOVA, two-tailed t-tests, log-rank survival analysis, receiver-operator characteristic (ROC) analysis or calculating pairwise Pearson correlations (R²). Where stated in the figure legends, p-values were adjusted for multiple comparison using Benjamini–Hochberg (B.H.) correction or Bonferroni correction, but otherwise no adjustment was made.

Reporting summary

Further information on research design is available in the Nature Research Reporting Summary linked to this article.

Data availability

Data supporting the findings of this study are available within the article and its Supplementary information files, or are available from the corresponding author subject to legal restrictions governing use of personal information. Source data are provided with this paper.

Change history

19 May 2021
Open Access funding information has been added to this article.

References

Larkin, J. et al. Five-year survival with combined nivolumab and ipilimumab in advanced melanoma. N. Engl. J. Med. 381, 1535–1546 (2019).
Robert, C. et al. Pembrolizumab versus ipilimumab in advanced melanoma (KEYNOTE-006): post-hoc 5-year results from an open-label, multicentre, randomised, controlled, phase 3 study. Lancet Oncol. 20, 1239–1251 (2019).
Article CAS PubMed Google Scholar
Robert, C. et al. Nivolumab in previously untreated melanoma without BRAF mutation. N. Engl. J. Med. 372, 320–330 (2015).
Article CAS PubMed Google Scholar
Postow, M. A. et al. Nivolumab and ipilimumab versus ipilimumab in untreated melanoma. N. Engl. J. Med. 372, 2006–2017 (2015).
Article PubMed PubMed Central Google Scholar
Wolchok, J. D. et al. Overall survival with combined nivolumab and ipilimumab in advanced melanoma. N. Engl. J. Med. 377, 1345–1356 (2017).
Article CAS PubMed PubMed Central Google Scholar
Hodi, F. S. et al. Nivolumab plus ipilimumab or nivolumab alone versus ipilimumab alone in advanced melanoma (CheckMate 067): 4-year outcomes of a multicentre, randomised, phase 3 trial. Lancet Oncol. 19, 1480–1492 (2018).
Article CAS PubMed Google Scholar
Postow, M. A., Sidlow, R. & Hellmann, M. D. Immune-related adverse events associated with immune checkpoint blockade. N. Engl. J. Med. 378, 158–168 (2018).
Article CAS PubMed Google Scholar
Schadendorf, D. et al. Efficacy and safety outcomes in patients with advanced melanoma who discontinued treatment with nivolumab and ipilimumab because of adverse events: a pooled analysis of randomized phase II and III trials. J. Clin. Oncol. 35, 3807–3814 (2017).
Article CAS PubMed PubMed Central Google Scholar
Thompson, J. A. et al. Management of immunotherapy-related toxicities, version 1.2019. J. Natl Compr. Canc Netw. 17, 255–289 (2019).
Article CAS PubMed Google Scholar
Teufel, A. et al. Management of immune related adverse events induced by immune checkpoint inhibition. Cancer Lett. 456, 80–87 (2019).
Article CAS PubMed Google Scholar
Larkin, J. et al. Combined nivolumab and ipilimumab or monotherapy in untreated melanoma. N. Engl. J. Med. 373, 23–34 (2015).
Article PubMed PubMed Central Google Scholar
Wolchok, J. D. et al. Nivolumab plus ipilimumab in advanced melanoma. N. Engl. J. Med. 369, 122–133 (2013).
Article CAS PubMed PubMed Central Google Scholar
Ruotsalainen, J. & Tuting, T. Live or let die: t cell survival in cancer immunotherapy. Immunity 50, 280–282 (2019).
Article CAS PubMed Google Scholar
Perez-Ruiz, E. et al. Prophylactic TNF blockade uncouples efficacy and toxicity in dual CTLA-4 and PD-1 immunotherapy. Nature 569, 428–432 (2019).
Article ADS CAS PubMed Google Scholar
Young, A., Quandt, Z. & Bluestone, J. A. The balancing act between cancer immunity and autoimmunity in response to immunotherapy. Cancer Immunol. Res. 6, 1445–1452 (2018).
Article CAS PubMed PubMed Central Google Scholar
Johnson, D. B. et al. Ipilimumab therapy in patients with advanced melanoma and preexisting autoimmune disorders. JAMA Oncol. 2, 234–240 (2016).
Article PubMed Google Scholar
Oh, D. Y. et al. Immune toxicities elicted by CTLA-4 blockade in cancer patients are associated with early diversification of the T-cell repertoire. Cancer Res. 77, 1322–1330 (2017).
Article MathSciNet CAS PubMed Google Scholar
Tivol, E. A. et al. Loss of CTLA-4 leads to massive lymphoproliferation and fatal multiorgan tissue destruction, revealing a critical negative regulatory role of CTLA-4. Immunity 3, 541–547 (1995).
Article CAS PubMed Google Scholar
Krishna, B. A. et al. Targeting the latent cytomegalovirus reservoir with an antiviral fusion toxin protein. Nat. Commun. 8, 14321 (2017).
Article ADS CAS PubMed PubMed Central Google Scholar
Peghin, M. et al. Epidemiology and immediate indirect effects of respiratory viruses in lung transplant recipients: a 5-year prospective study. Am. J. Transplant. 17, 1304–1312 (2017).
Article CAS PubMed Google Scholar
Lancini, D., Faddy, H. M., Flower, R. & Hogan, C. Cytomegalovirus disease in immunocompetent adults. Med. J. Aust. 201, 578–580 (2014).
Article PubMed Google Scholar
Coussement, J. et al. When polymerase chain reaction does not help: cytomegalovirus pneumonitis associated with very low or undetectable viral load in both blood and bronchoalveolar lavage samples after lung transplantation. Transpl. Infect. Dis. 18, 284–287 (2016).
Article CAS PubMed PubMed Central Google Scholar
Almanan, M. et al. Tissue-specific control of latent CMV reactivation by regulatory T cells. PLoS Pathog. 13, e1006507 (2017).
Article PubMed PubMed Central Google Scholar
Andrei, G. et al. Persistent primary cytomegalovirus infection in a kidney transplant recipient: multi-drug resistant and compartmentalized infection leading to graft loss. Antiviral Res. 168, 203–209 (2019).
Article CAS PubMed Google Scholar
Luck, S. E., Emery, V. C., Atkinson, C., Sharland, M. & Griffiths, P. D. Compartmentalized dynamics of cytomegalovirus replication in treated congenital infection. J. Clin. Virol. 82, 152–158 (2016).
Article CAS PubMed Google Scholar
Pardieck, I. N., Beyrend, G., Redeker, A. & Arens, R. Cytomegalovirus infection and progressive differentiation of effector-memory T cells. F1000Res 7 (2018).
Marchi, E., Lee, L. N. & Klenerman, P. Inflation vs. exhaustion of antiviral CD8 + T-cell populations in persistent infections: two sides of the same coin? Front. Immunol. 10, 197 (2019).
Article CAS PubMed PubMed Central Google Scholar
Gordon, C. L. et al. Induction and maintenance of CX3CR1-intermediate peripheral memory CD8(+) T cells by persistent viruses and vaccines. Cell Rep. 23, 768–782 (2018).
Article CAS PubMed PubMed Central Google Scholar
Makwana, N. et al. CMV drives the expansion of highly functional memory T cells expressing NK-cell receptors in renal transplant recipients. Eur. J. Immunol. 47, 1324–1334 (2017).
Article CAS PubMed Google Scholar
Wei, S. C., Duffy, C. R. & Allison, J. P. Fundamental mechanisms of immune checkpoint blockade therapy. Cancer Discov. 8, 1069–1086 (2018).
Article PubMed Google Scholar
Tumeh, P. C. et al. PD-1 blockade induces responses by inhibiting adaptive immune resistance. Nature 515, 568–571 (2014).
Article ADS CAS PubMed PubMed Central Google Scholar
Fujii, T. et al. Incidence of immune-related adverse events and its association with treatment outcomes: the MD Anderson Cancer Center experience. Invest. New Drugs 36, 638–646 (2018).
Article CAS PubMed Google Scholar
Herbst, R. S. et al. Predictive correlates of response to the anti-PD-L1 antibody MPDL3280A in cancer patients. Nature 515, 563–567 (2014).
Article ADS CAS PubMed PubMed Central Google Scholar
Verloh, N. et al. In vivo confirmation of altered hepatic glucose metabolism in patients with liver fibrosis/cirrhosis by (18)F-FDG PET/CT. EJNMMI Res. 8, 98 (2018).
Article PubMed PubMed Central Google Scholar
Khan, O. et al. TOX transcriptionally and epigenetically programs CD8(+) T cell exhaustion. Nature 571, 211–218 (2019).
Article CAS PubMed PubMed Central Google Scholar
Im, S. J. et al. Defining CD8+ T cells that provide the proliferative burst after PD-1 therapy. Nature 537, 417–421 (2016).
Article ADS CAS PubMed PubMed Central Google Scholar
Caselli, E. et al. Virologic and immunologic evidence supporting an association between HHV-6 and Hashimoto’s thyroiditis. PLoS Pathog. 8, e1002951 (2012).
Article CAS PubMed PubMed Central Google Scholar
Del Castillo, M. et al. The spectrum of serious infections among patients receiving immune checkpoint blockade for the treatment of melanoma. Clin. Infect. Dis. 63, 1490–1493 (2016).
Article PubMed PubMed Central Google Scholar
Franklin, C. et al. Cytomegalovirus reactivation in patients with refractory checkpoint inhibitor-induced colitis. Eur. J. Cancer 86, 248–256 (2017).
Article PubMed Google Scholar
Wilkinson, R. W. & Leishman, A. J. Further advances in cancer immunotherapy: going beyond checkpoint blockade. Front. Immunol. 9, 1082 (2018).
Article PubMed PubMed Central Google Scholar
Contreras, N. A. et al. Life-long control of cytomegalovirus (CMV) by T resident memory cells in the adipose tissue results in inflammation and hyperglycemia. PLoS Pathog. 15, e1007890 (2019).
Article MathSciNet CAS PubMed PubMed Central Google Scholar
Wang, Z. et al. Paradoxical effects of obesity on T cell function during tumor progression and PD-1 checkpoint blockade. Nat. Med. 25, 141–151 (2019).
Article CAS PubMed Google Scholar
Marzo, A. L. et al. Initial T cell frequency dictates memory CD8+ T cell lineage commitment. Nat. Immunol. 6, 793–799 (2005).
Article CAS PubMed PubMed Central Google Scholar
Wherry, E. J. et al. Lineage relationship and protective immunity of memory CD8 T cell subsets. Nat. Immunol. 4, 225–234 (2003).
Article CAS PubMed Google Scholar
Williams, M. A. & Bevan, M. J. Shortening the infectious period does not alter expansion of CD8 T cells but diminishes their capacity to differentiate into memory cells. J. Immunol. 173, 6694–6702 (2004).
Article CAS PubMed Google Scholar
van Faassen, H. et al. Reducing the stimulation of CD8+ T cells during infection with intracellular bacteria promotes differentiation primarily into a central (CD62LhighCD44high) subset. J. Immunol. 174, 5341–5350 (2005).
Article PubMed Google Scholar
Jelley-Gibbs, D. M. et al. Repeated stimulation of CD4 effector T cells can limit their protective function. J. Exp. Med. 201, 1101–1112 (2005).
Article CAS PubMed PubMed Central Google Scholar
Jelley-Gibbs, D. M. et al. Persistent depots of influenza antigen fail to induce a cytotoxic CD8 T cell response. J. Immunol. 178, 7563–7570 (2007).
Article CAS PubMed Google Scholar
Sarkar, S. et al. Strength of stimulus and clonal competition impact the rate of memory CD8 T cell differentiation. J. Immunol. 179, 6704–6714 (2007).
Article CAS PubMed Google Scholar
Rosato, P. C. et al. Virus-specific memory T cells populate tumors and can be repurposed for tumor immunotherapy. Nat. Commun. 10, 567 (2019).
Article ADS CAS PubMed PubMed Central Google Scholar
Medicines and Healthcare Products Regulatory Agency. Nivolumab (Opdivo): reports of cytomegalovirus (CMV) gastrointestinal infection or reactivation. Drug Safety Update 13: No. 3. https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/840565/Oct-2019-PDF.pdf (2019).
Kronenberg, K., Riquelme, P. & Hutchinson, J. A. Standard protocols for immune profiling of peripheral blood leucocyte subsets by flow cytometry using DuraClone IM reagents. Protocol Exchange https://doi.org/10.21203/rs.3.pex-757/v1 (2021).
Riquelme, P. & Hutchinson, J. A. Standard protocols for detection of human CMV-specific T cells Protocol Exchange https://doi.org/10.21203/rs.3.pex-758/v1 (2021).
Hofheinz, F. et al. An investigation of the relation between tumor-to-liver ratio (TLR) and tumor-to-blood standard uptake ratio (SUR) in oncological FDG PET. EJNMMI Res. 6, 19 (2016).
Article PubMed PubMed Central Google Scholar

Download references

Acknowledgements

The authors gratefully acknowledge the financial support of Bristol-Myers Squibb and the BMS Foundation (Award FA-19-009) in the form of unrestricted research grants. This study was partly supported by grants from the German Research Foundation to S.H. (DFG HA 8481/1-1) and J.A.H. (DFG HU 1838/1-1; DFG HU 1838/2-1). L.C. is a Marie Skłodowska-Curie Research Fellow affiliated with INsTRuCT and receives funding from the European Union’s Horizon 2020 research and innovation programme (Award 860003). The authors thank Dr. Michael Kapinsky of Beckman Coulter GmbH for advice and support with clinical flow cytometry. The authors also thank Prof. Tobias Pukrop and staff of the Interdisciplinary Center for drug Tumor Therapy (ICT) for supporting collection of clinical samples. This work would not have been possible without the outstanding technical support of Erika Ostermeier, Marina Sorokina, Veronika Menath, Anke Hofmann and Joachim Schweimer.

Funding

Open Access funding enabled and organized by Projekt DEAL.

Author information

Authors and Affiliations

Department of Surgery, University Hospital Regensburg, Regensburg, Germany
James A. Hutchinson, Katharina Kronenberg, Paloma Riquelme, Hannah-Lou Schilling, Florian Bitterer, Laura Cordero, Hans J. Schlitt, Jens M. Werner & Edward K. Geissler
Institute of Clinical Microbiology and Hygiene, University Hospital Regensburg, Regensburg, Germany
Jürgen J. Wenzel, Josef Koestler & Barbara Schmidt
Institute of Functional Genomics and Statistical Bioinformatics, University of Regensburg, Regensburg, Germany
Gunther Glehr & Rainer Spang
Center for Clinical Studies, University Hospital Regensburg, Regensburg, Germany
Florian Zeman
Institute of Pathology, University Hospital Regensburg, Regensburg, Germany
Katja Evert
Department of Nuclear Medicine, University Hospital Regensburg, Regensburg, Germany
Martin Schmiedel & Dirk Hellwig
Department of Dermatology, University Hospital Regensburg, Regensburg, Germany
Marion Mickler, Konstantin Drexler & Sebastian Haferkamp
Institute of Radiology, University Hospital Regensburg, Regensburg, Germany
Lukas Beyer
Department of Medicine V, University Hospital Erlangen, Erlangen, Germany
Christian Bach
Institute of Clinical Chemistry and Laboratory Medicine, University Hospital Regensburg, Regensburg, Germany
Ralph Burkhardt
Personalised Tumour Therapy, Fraunhofer Institute for Experimental Medicine and Toxicology, Regensburg, Germany
Edward K. Geissler

Authors

James A. Hutchinson
View author publications
You can also search for this author in PubMed Google Scholar
Katharina Kronenberg
View author publications
You can also search for this author in PubMed Google Scholar
Paloma Riquelme
View author publications
You can also search for this author in PubMed Google Scholar
Jürgen J. Wenzel
View author publications
You can also search for this author in PubMed Google Scholar
Gunther Glehr
View author publications
You can also search for this author in PubMed Google Scholar
Hannah-Lou Schilling
View author publications
You can also search for this author in PubMed Google Scholar
Florian Zeman
View author publications
You can also search for this author in PubMed Google Scholar
Katja Evert
View author publications
You can also search for this author in PubMed Google Scholar
Martin Schmiedel
View author publications
You can also search for this author in PubMed Google Scholar
Marion Mickler
View author publications
You can also search for this author in PubMed Google Scholar
Konstantin Drexler
View author publications
You can also search for this author in PubMed Google Scholar
Florian Bitterer
View author publications
You can also search for this author in PubMed Google Scholar
Laura Cordero
View author publications
You can also search for this author in PubMed Google Scholar
Lukas Beyer
View author publications
You can also search for this author in PubMed Google Scholar
Christian Bach
View author publications
You can also search for this author in PubMed Google Scholar
Josef Koestler
View author publications
You can also search for this author in PubMed Google Scholar
Ralph Burkhardt
View author publications
You can also search for this author in PubMed Google Scholar
Hans J. Schlitt
View author publications
You can also search for this author in PubMed Google Scholar
Dirk Hellwig
View author publications
You can also search for this author in PubMed Google Scholar
Jens M. Werner
View author publications
You can also search for this author in PubMed Google Scholar
Rainer Spang
View author publications
You can also search for this author in PubMed Google Scholar
Barbara Schmidt
View author publications
You can also search for this author in PubMed Google Scholar
Edward K. Geissler
View author publications
You can also search for this author in PubMed Google Scholar
Sebastian Haferkamp
View author publications
You can also search for this author in PubMed Google Scholar

Contributions

J.A.H. designed the immune monitoring studies and other experiments, analysed the results and wrote the manuscript; K.K. organised and performed immune monitoring studies; P.R. designed and performed the experiments; J.J.W., J.K. and B.S. performed the virological analyses; K.E. provided expert histopathological descriptions; M.S., L.B. and D.H. performed the radiological and molecular imaging analyses; G.G., F.Z. and R.S. provided expert statistical and bioinformatics advice; L.C. quality-checked the manuscript; C.B. and R.B. were responsible for the routine clinical laboratory analyses; M.M. and K.D. provided the biological samples; H.L.S. and F.B. performed the experiments; H.J.S. provided infrastructural support; E.K.G., J.M.W. and S.H. provided material support, patient samples and critical feedback throughout the project. All authors approved the final version of the manuscript.

Corresponding author

Correspondence to James A. Hutchinson.

Ethics declarations

Competing interests

S.H. declares he has received consulting fees and speaker’s honoraria from BMS and Merck Sharp & Dohme (MSD). All other authors declare no competing interests.

Additional information

Peer review information Nature Communications thanks the anonymous reviewer(s) for their contribution to the peer review of this work. Peer reviewer reports are available.

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary Information

Peer Review File

Reporting Summary

Source data

Source Data

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and permissions

About this article

Cite this article

Hutchinson, J.A., Kronenberg, K., Riquelme, P. et al. Virus-specific memory T cell responses unmasked by immune checkpoint blockade cause hepatitis. Nat Commun 12, 1439 (2021). https://doi.org/10.1038/s41467-021-21572-y

Download citation

Received: 28 October 2019
Accepted: 29 January 2021
Published: 04 March 2021
DOI: https://doi.org/10.1038/s41467-021-21572-y

This article is cited by

Clinical and translational attributes of immune-related adverse events
- Karijn P. M. Suijkerbuijk
- Mick J. M. van Eijs
- Alexander M. M. Eggermont
Nature Cancer (2024)
Effectiveness of tacrolimus in a case of immune checkpoint inhibitor-induced hepatotoxicity that was refractory to steroids and mycophenolate mofetil
- Kosuke Sato
- Jun Inoue
- Atsushi Masamune
Clinical Journal of Gastroenterology (2023)
Immune profile of patients with peritoneal carcinomatosis selected for CRS-HIPEC therapy
- Julia Kleber
- Jordi Yang Zhou
- Jens M. Werner
Cancer Immunology, Immunotherapy (2023)
Toxicity-specific peripheral blood T and B cell dynamics in anti-PD-1 and combined immune checkpoint inhibition
- Mick J. M. van Eijs
- Rik J. Verheijden
- Jeanette H. W. Leusen
Cancer Immunology, Immunotherapy (2023)
Incidence and severity of immune-related hepatitis after dual checkpoint therapy is linked to younger age independent of herpes virus immunity
- Zhen Zhang
- David Rafei-Shamsabadi
- Bertram Bengsch
Journal of Translational Medicine (2022)

Comments

By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.