与免疫检查点抑制剂有关的肝损伤

StephenW

[email protected]

6:27 AM (11 hours ago)

[tr][/tr][/table][table]
to hiv, nataphcv, HCV/HIV, undefined, Natap

www.natap.org
Liver damage related to immune checkpoint inhibitors


Abstract

Recently, immune checkpoint inhibitors are becoming one of the key agents of systemic treatment of cancer. The anti-cancer mechanism of this type of agent is totally different from that of conventional therapies; blockade of regulatory receptors and ligand of immune checkpoint molecules arose anti-tumor immunity with durable response. However, owing to its unique action to host immune system, immune checkpoint inhibitors sometimes induce immune-related adverse events (irAEs) which has not been observed for conventional chemotherapies. It has been reported that irAEs are manageable by discontinuation of immune checkpoint inhibitors and corticosteroid. However, severe irAEs might lead to the unsuccessful management of cancer treatment. It is conceivable that irAEs during the treatment of immune checkpoint blockade might mimic the autoimmune disease of the specific organ, such as autoimmune hepatitis (AIH). However, detail of the pathogenesis of irAEs has not been well estimated. In this review, we specially focused on this important issue and discussed the liver toxicity of this type of agent in the context of comparison of clinical and pathological findings of liver damage related to irAEs and AIH.

Introduction

Immune checkpoint inhibitors exert anti-tumor effect by blocking the interaction between regulatory receptors and ligand of checkpoint molecules on T cells, antigen-resenting cells and tumor cells [1]. These are expected to restore the anti-cancer immunity and are becoming one of the key agents for the treatment of cancers [2, 3, 4]. Currently, anti-programmed cell death (PD)-1, anti-PD-ligand 1 (PD-L1), and anti-cytotoxic T-lymphocyte associated antigen (CTLA-4) antibodies are available for several types of malignancies, including melanoma, renal cell carcinoma, non-small cell lung cancer, Hodgkin lymphoma, cervical cancer, and gastric cancer. In addition, combination therapies of immune checkpoint inhibitors, such as anti-PD-1 and anti-CTLA-4 combination, are approved for the treatment of advanced melanoma, and under clinical trials for many types of cancers [5, 6].

On the other hand, owing to the unique action of immune checkpoint inhibitors on immune system, this type of agent causes immune-related adverse events (irAEs) that mainly involve digestive system, lung, skin, endocrine glands, liver but can potentially affect any tissue. So far, it has been reported that irAEs caused by immune checkpoint monotherapies were manageable under steroid therapy and/or discontinuation of the agent [7, 8, 9, 10]. However, it should be noted that irAEs induced by combination therapies, that aim to enhance the anti-tumor response, have not been well estimated [11, 12, 13].

It has been suspected that irAEs mimic the autoimmune disorders that target specific organs in terms of its pathogenesis. However, recent reports suggested that irAEs could represent a different feature compared to the autoimmune disorders [14]. In this review, we focused on the liver damage induced by immune checkpoint inhibitors and discussed the similarity and difference of clinical and pathological feature between irAEs of liver and autoimmune hepatitis (AIH).

Hepatotoxicity by immune checkpoint inhibitorsThe frequencies of liver damage related to immune checkpoint inhibitors have been reported in the clinical trial of several types of cancers [15, 16]. Generally, it is reported that onset of liver damage was observed mainly within 3 months after the treatment but can emerge anytime during the treatment [17]. On the other hand, incidence of increase of aspartate transaminase (AST) or alanine transaminase (ALT) was observed in 2–5% of the cases and grade 3/4 increase of transaminase was detected in 1–4% [16, 17]. We summarized the frequencies of elevation of AST or ALT for melanoma patients treated with immune checkpoint inhibitor in Table 1. Although the dose and agent differed among the studies, the total frequencies of AST or ALT abnormalities ranged from 1.8–6.2%, and frequencies of grade 3/4 abnormalities were reported as 1.1–1.8% for the treatment of anti-PD-1 monotherapies [18, 19, 20]. On the other hand, the frequencies of liver damage related to the CTLA-4 monotherapy ranged from 1.2–14.6% for all grade, and 0.4–5.7% for grade 3/4 toxicity, suggesting that liver damage related to immune checkpoint inhibitors is more common in anti-CTLA-4 than in anti-PD-1 therapies [19, 20, 21, 22, 23].

Table 1

Frequencies of liver damage in the treatment with immune checkpoint inhibitors

[td]

Agent (dose per infusion)	Tumor type	Number of the patients	Incidence of liver damage (%)		References
			Total	Grade ¾
Anti-PD-1 antibody
Pembrolizumab (10 mg/kg)	Melanoma	277	1.8	1.8	Robert et al. [19]
Nivolumab (3 mg/kg)	Melanoma	313	3.8	1.3	Larken et al. [18]
Nivolumab (3 mg/kg)	Melanoma	313	4	1	Wolchok et al. [23]
Nivolumab (3 mg/kg)	Melanoma	452	6.2	1.1	Weber et al. [20]
Nivolumaba	HCC	48	15	6	El-Khoueiry et al. [28]
Pembrolizumabb	HCC	104	9	4	Zhu et al. [29]
Anti-CTLA-4 antibody
Ipilimumab (3 mg/kg)	Melanoma	256	1.2	0.4	Robert et al. [19]
Ipilimumab (3 mg/kg)	Melanoma	311	3.9	1.6	Larken et al. [18]
Ipilimumab (3 mg/kg)	Melanoma	311	4	2	Wolchok et al. [23]
Ipilimumab (10 mg/kg)	Melanoma	453	14.6	5.7	Weber et al. [20]
Tremelimumab (15 mg/kg)c	HCC	20	55	25	Sangro et al. [30]
Anti-PD-1 + anti-CTLA-4 antibodies
Nivolumab + ipilimumabd	Melanoma	53	21	11	Wolchok et al. [22]
Nivolumab + ipilimumabe	Melanoma	313	17.6	8.3	Larken et al [18]
Nivolumab + ipilimumabe	Melanoma	313	19	9	Wolchok et al. [23]

For the trial on melanoma, cases positive for hepatitis B (HBV) and hepatitis C virus, (HCV) were excluded

aNivolumab 0.1–10 mg/kg. Among 48 patients, 15 and 10 were positive for HBV and HCV, respectively

bPembrolizumab 200 mg

cNivolumab 0.1–10 mg/kg. All patients showed chronic HCV infection

dNivolumab 0.3–3 mg/kg + ipilimumab 1–3 mg/kg

eNivolumab 1 mg/kg + ipilimumab 3 mg/kg

On the other hand, several systemic therapies are found to be effective in advanced HCC cases [24, 25, 26, 27] and several critical trials of immune checkpoint inhibitors are ongoing [28, 29]. Although number of the patients treated with this type of agents is much smaller in HCC than in melanoma cases, liver damage is more frequently observed in HCC cases, where 9–15% of the cases showed increase of ALT, and grade 3/4 toxicity was observed in 4–6% for anti-PD-1 treatment [28, 29]. Furthermore, anti-CTLA-4 antibody, tremelimumab, induced liver damage in 55% for all grades and 25% for grade 3/4 toxicity in the HCC patients, respectively [30]. Therefore, increase of ALT should be carefully monitored during the treatment of immune checkpoint inhibitors in HCC cases, although all liver damages were manageable by discontinuation of the agent and corticosteroid. On the other hand, grade 3/4 liver damage was much more severe for the combination of anti-PD-1 and anti-CTLA4 antibodies: 17.6–21% for all grades and 8.3–11% for grade 3/4 toxicity in melanoma cases [18, 21, 22, 23].

Difference of clinical and pathological feature between autoimmune hepatitis and liver damage induced by immune checkpoint inhibitorsDe Martin et al. reported 16 cases with liver injury related to immune checkpoint inhibitors; 7 cases were treated with anti-CTLA-4 antibody or anti-PD-1 and anti-CTLA-4 combination, and 9 cases underwent anti-PD-1 monotherapy. Among them, 6 cases (37.5%) were accompanied by high fever (mainly observed in the cases with anti-CTLA-4 or combination therapy); five cases (31.3%) showed skin rash [31]. In contrast to the cases of autoimmune hepatitis (AIH), no female predominance was reported for the incidence of liver damage of irAEs (Table 2). In addition, the characteristic findings of serological examination of AIH, such as increase of γ-globulin and appearance of anti-nuclear antibody (ANA) and anti-smooth muscle antibody (SMA), were not always observed in the case with liver damage related to immune checkpoint inhibitors [31].

Differences of histological features between AIH and liver damage of irAEs have also been described. Histology of the liver related to irAEs is heterogeneous that should be attributed to the complexity of pathogenesis, including lobular hepatitis, steatosis and steatohepatitis, as well as bile duct injury [32, 33]. In addition to the autoimmune-like hepatitis in liver parenchyma [34], damage in ductal and endothelial cells is observed, suggesting that cellular rejection may be induced by this type of agent; the cases with liver allograft failure have been reported in association with immune checkpoint inhibitors [35, 36, 37].

Liver parenchymal damage induced by anti-PD-1 antibody is mainly represented lobular hepatitis with mild lobular infiltration [38]. It may be accompanied by cholangiolitis, bile duct injury and endothelialitis [14, 38, 39]; although the degree of liver injury was milder compared to that induced by anti-CTLA-4 antibody. Portal fibrosis was also reported for liver injury related to anti-PD-1 antibody. However, infiltration of plasma cell and lymphoid follicles, which was one of the characteristic findings of AIH, was rarely observed [31, 38, 40]. On the other hand, bile duct injury during the treatment using anti-PD-1 antibody may lead to vanishing bile duct syndrome that is attributed to lymphocytic cholangitis [32, 41, 42]. Reportedly, extrahepatic bile duct could also be involved, which represented dilatation of extrahepatic biliary [32]. Elevation of alkaline phosphatase, γ-glutamyl transpeptidase and bilirubin are predominant laboratory findings in the cases with cholangiopathy [43].

Liver injury related to anti-CTLA-4 antibody and its combination with anti-PD-1 antibody could be more severe and, reportedly, represented unique pathological features. The liver injury induced by anti-CTLA-4 may occur earlier than that induced by anti-PD-1 antibody [31]. Most patients showed pan-lobular hepatitis with infiltration of lymphocyte and histiocytes. The infiltrated T lymphocyte mainly represented CD8+ in anti-CTLA-4 antibody-related liver injury, although both CD4+ and CD8+ T lymphocytes were observed in AIH as well as liver injury related to anti-PD-1 antibody [31, 44]. The centrilobular hepatitis, which had been described in AIH, could also be observed in anti-CTLA-4 related liver injury. Bile duct injury and steatohepatitis might take place, but not characteristic findings [31, 44]. In many cases, central vein endothelialitis and sinusoidal distribution of macrophage were reportedly detected. Everett et al. reported two cases of the fibrin ring granulomas in patients with liver damage related to ipilimumab and nivolumab combination therapy, which could be observed in a variety of liver injuries, such as induced by infectious and toxic agents [45]. It generally consists of several layers with a central lipid vacuole surrounded by macrophages, histiocytes and fibrin ring. De Martin et al. also reported that granulomatous hepatitis with fibrin deposit accompanied by severe lobular necrotic and inflammatory activity was frequently detected in the patients treated with anti-PD-1 and anti-CTLA-4 combination [31]. Endothelialitis in central and portal vein also occurred in patients treated with combination therapy [31, 44].

So far, there are no specific predictors that predict emergence of liver damage after the start of immune checkpoint inhibitors. Based on the previous analysis that analyzed 16 cases of liver damage associated with immune checkpoint inhibitors, 9 cases showed positive for ANA or SMA, although their titers are low [31]. Although previous report shows high flare rate of autoimmune disorders in patients with inactive status, risk of liver toxicity in patients with autoimmune liver disease has not been evaluated [33]. In addition, association between the presence of ANA and SMA and risk of emergence of liver damage on the treatment using immune checkpoint inhibitors has not been clarified yet. On the other hand, skin and gastrointestinal complication could be accompanied by liver damage; emergence of such symptom could be a precursor of liver complication [33].

Conclusion

The treatment using immune checkpoint inhibitors is expanding; its liver toxicity is poorly understood [46]. It has been considered that immune checkpoint inhibitors arouse autoimmune reaction, but clinical and pathological feature of liver injury related to this type of agent is, in some respects, different from that of AIH [14]. The typical serological finding of AIH was not observed in liver injury related to irAEs, where major infiltrating cells are histiocyte and macrophage but not plasma cells. Severe interface hepatitis and rosette formation are rare, and granulomatous lobular hepatitis, fibrin ring granulomas, and endothelialitis and bile duct injury are more frequently observed compared to AIH.

StephenW

与免疫检查点抑制剂有关的肝损伤
抽象

最近，免疫检查点抑制剂正成为全身治疗癌症的关键因素之一。这类药物的抗癌机制与传统疗法完全不同;阻断调节受体和免疫检查点分子的配体产生具有持久反应的抗肿瘤免疫。然而，由于其对宿主免疫系统的独特作用，免疫检查点抑制剂有时会引起免疫相关的不良事件（irAE），这在传统的化学疗法中未被观察到。据报道，通过停药可以控制irAEs可以想象，在免疫检查点阻断治疗期间，irAE会搅动特定器官的自身免疫疾病，例如自身免疫性肝炎（AIH）。然而，irAE的发病机制的细节尚未得到很好的估计。在本综述中，我们特别关注这一重要问题，并在比较与irAE和AIH相关的肝损伤的临床和病理结果的背景下讨论了这类药物的肝毒性。
介绍

免疫检查点抑制剂通过阻断T细胞，抗原复发细胞和肿瘤细胞上调节受体和检查点分子配体之间的相互作用而发挥抗肿瘤作用[1]。预计这些将恢复抗癌免疫力，并成为治疗癌症的关键因素之一[2,3,4]。目前，抗程序性细胞死亡（PD）-1，抗PD配体1（PD-L1）和抗细胞毒性T淋巴细胞相关抗原（CTLA-4）抗体可用于几种类型的恶性肿瘤，包括黑素瘤，肾细胞癌，非小细胞肺癌，霍奇金淋巴瘤，宫颈癌和胃癌。此外，免疫检查点抑制剂的联合治疗，如抗PD-1和抗-CTLA-4组合，被批准用于治疗晚期黑色素瘤，并在许多类型的癌症的临床试验中[5,6]。

另一方面，由于免疫检查点抑制剂对免疫系统的独特作用，这种类型的药物引起免疫相关的不良事件（irAEs），主要涉及消化系统，肺，皮肤，内分泌腺，肝脏但可能影响任何组织。到目前为止，据报道，由免疫检查点单一疗法引起的irAE在类固醇治疗和/或停用该药物后是可控的[7,8,9,10]。然而，应该注意的是，旨在增强抗肿瘤反应的联合疗法诱导的irAE尚未得到很好的估计[11,12,13]。

已经怀疑irAE模仿针对特定器官的自身免疫性疾病的发病机理。然而，最近的报告表明，与自身免疫性疾病相比，irAE可能代表不同的特征[14]。在这篇综述中，我们关注免疫检查点抑制剂诱导的肝损伤，并讨论了肝脏和自身免疫性肝炎（AIH）的临床和病理特征的相似性和差异性。
免疫检查点抑制剂的肝毒性
在几种类型的癌症的临床试验中已经报道了与免疫检查点抑制剂相关的肝损伤的频率[15,16]。一般来说，据报道，肝损伤的发生主要在治疗后3个月内观察到，但可能出现另一方面，天冬氨酸转氨酶（AST）或丙氨酸转氨酶（ALT）的增加发生率在2-5％病例和转氨酶的3/4级增加检出率为1-4％[16,17]。我们总结了表1中用免疫检查点抑制剂治疗的黑素瘤患者的AST或ALT升高频率。虽然研究中剂量和药剂不同，但AST或ALT异常的总频率范围为1.8-6.2％，频率为据报道，治疗抗PD-1单一疗法的3/4级异常为1.1-1.8％[18,19,20]。另一方面，与CTLA-4单药治疗相关的肝损伤频率在所有等级中为1.2-14.6％，对于3/4级毒性为0.4-5.7％，表明与免疫检查点抑制剂相关的肝损伤是在抗-CTLA-4中比在抗-PD-1疗法中更常见[19,20,21,22,23
表格1

免疫检查点抑制剂治疗中肝损伤的频率

药剂（每次输液剂量）


肿瘤类型


患者人数


肝损害发生率（％）


参考

总


等级¾

抗PD-1抗体

Pembrolizumab（10 mg / kg）


黑色素瘤


277


1.8


1.8


罗伯特等人。 [19]

Nivolumab（3 mg / kg）


黑色素瘤


313


3.8


1.3


拉肯等人。 [18]

Nivolumab（3 mg / kg）


黑色素瘤


313


4


1


Wolchok等。 [23]

Nivolumab（3 mg / kg）


黑色素瘤


452


6.2


1.1


韦伯等人。 [20]

Nivolumaba


HCC


48


15


6


El-Khoueiry等人。 [28]

Pembrolizumabb


HCC


104


9


4


朱等人。 [29]

抗CTLA-4抗体

Ipilimumab（3 mg / kg）


黑色素瘤


256


1.2


0.4


罗伯特等人。 [19]

Ipilimumab（3 mg / kg）


黑色素瘤


311


3.9


1.6


拉肯等人。 [18]

Ipilimumab（3 mg / kg）


黑色素瘤


311


4


2


Wolchok等。 [23]

Ipilimumab（10 mg / kg）


黑色素瘤


453


14.6


5.7


韦伯等人。 [20]

Tremelimumab（15 mg / kg）c


HCC


20


55


25


Sangro等人。 [30]

抗PD-1 +抗CTLA-4抗体

Nivolumab + ipilimumabd


黑色素瘤


53


21


11


Wolchok等。 [22]

Nivolumab + ipilimumabe


黑色素瘤


313


17.6


8.3


Larken等[18]

Nivolumab + ipilimumabe

免疫检查点抑制剂诱导自身免疫性肝炎与肝损害的临床和病理特征的差异
德马丁等人。报告16例与免疫检查点抑制剂有关的肝损伤; 7例采用抗CTLA-4抗体或抗PD-1和抗-CTLA-4联合治疗，9例采用抗PD-1单药治疗。其中6例（37.5％）伴有高热（主要见于抗CTLA-4或联合治疗的病例）; 5例（31.3％）出现皮疹[31]。与自身免疫性肝炎（AIH）的病例相反，IRAE的肝损伤发生率没有女性占优势（表2）。此外，在肝损害相关的病例中，并未总能观察到AIH血清学检查的特征性发现，如γ-球蛋白的增加和抗核抗体（ANA）和抗平滑肌抗体（SMA）的出现。免疫检查点抑制剂[31]。

还描述了AIH与irAE的肝损伤之间的组织学特征的差异。与irAE相关的肝脏组织学是异质的，应该归因于发病机制的复杂性，包括小叶肝炎，脂肪变性和脂肪性肝炎，以及胆管损伤[32,33]。除了肝实质中的自身免疫性肝炎[34]，还观察到导管和内皮细胞的损伤，表明这种类型的药物可能诱导细胞排斥;已报道肝移植失败的病例与免疫检查点抑制剂有关[35,36,37]。

抗PD-1抗体诱导的肝实质损伤主要表现为小叶性肝炎伴轻度小叶浸润[38]。它可能伴有胆管炎，胆管损伤和内皮炎[14,18,39];尽管与抗-CTLA-4抗体诱导的相比，肝损伤的程度更轻微。还报道了与抗PD-1抗体相关的肝损伤的门静脉纤维化。然而，很少观察到浆细胞和淋巴滤泡的浸润，这是AIH的特征性发现之一[31,38,40]。另一方面，使用抗PD-1抗体治疗期间的胆管损伤可能导致胆管综合征消失，这归因于淋巴细胞性胆管炎[32,41,42]。据报道，还可能涉及肝外胆管，这表示肝外胆管扩张[32]。在胆管病变的病例中，碱性磷酸酶，γ-谷氨酰转肽酶和胆红素的升高是主要的实验室检查结果[43]。

与抗-CTLA-4抗体相关的肝损伤及其与抗PD-1抗体的组合可能更严重，据报道，其代表了独特的病理特征。抗-CTLA-4诱导的肝损伤可能早于抗PD-1抗体诱导的肝损伤[31]。大多数患者表现为泛小叶肝炎，淋巴细胞和组织细胞浸润。浸润的T淋巴细胞主要代表抗-CTLA-4抗体相关肝损伤中的CD8 +，尽管在AIH中观察到CD4 +和CD8 + T淋巴细胞以及与抗PD-1抗体相关的肝损伤[31,44]。已经在AIH中描述的小叶中心肝炎也可以在抗-CTLA-4相关的肝损伤中观察到。可能发生胆管损伤和脂肪性肝炎，但不是特征性发现[31,44]。在许多情况下，据报道检测到中心静脉内皮炎和巨噬细胞的正弦分布。埃弗雷特等人。报道了2例与ipilimumab和nivolumab联合治疗相关的肝损伤患者的纤维蛋白环肉芽肿，可观察到多种肝损伤，如感染性和毒性剂诱发[45]。它通常由几层组成，中央脂质空泡被巨噬细胞，组织细胞和纤维蛋白环包围。德马丁等人。还报道了在用抗PD-1和抗-CTLA-4组合治疗的患者中经常检测到纤维蛋白沉积伴有严重小叶坏死和炎症活动的肉芽肿性肝炎[31]。中心和门静脉的内皮炎也发生在联合治疗的患者[31,44]。
到目前为止，没有特定的预测因子预测免疫检查点抑制剂开始后肝脏损害的出现。根据之前的分析，分析了16例与免疫检查点抑制剂相关的肝损伤，9例ANA或SMA阳性，尽管它们的滴度较低[31]。虽然之前的报告显示无活动状态患者自身免疫性疾病的高发作率，但尚未评估自身免疫性肝病患者的肝毒性风险[33]。此外，尚未阐明ANA和SMA的存在与使用免疫检查点抑制剂的治疗中出现肝损伤的风险之间的关联。另一方面，皮肤和胃肠道并发症可伴有肝脏损害;这种症状的出现可能是肝脏并发症的前兆[33]。
结论

使用免疫检查点抑制剂的治疗正在扩大;其肝脏毒性尚不清楚[46]。已经认为免疫检查点抑制剂引起自身免疫反应，但是与这种类型的药剂相关的肝损伤的临床和病理特征在某些方面不同于AIH [14]。在与irAE相关的肝损伤中未观察到典型的AIH血清学发现，其中主要的浸润细胞是组织细胞和巨噬细胞，但不是浆细胞。与AIH相比，严重的界面性肝炎和玫瑰花结形成罕见，肉芽肿性小叶性肝炎，纤维蛋白环肉芽肿，内皮炎和胆管损伤更常见
黑色素瘤


313


19


9


Wolchok等。 [23]

对于黑色素瘤的试验，排除了乙型肝炎（HBV）和丙型肝炎病毒（HCV）阳性的病例

aNivolumab 0.1-10 mg / kg。在48名患者中，15名和10名分别为HBV和HCV阳性

bPembrolizumab 200毫克

cNivolumab 0.1-10 mg / kg。所有患者均表现出慢性HCV感染

dNivolumab 0.3-3 mg / kg + ipilimumab 1-3 mg / kg

eNivolumab 1 mg / kg + ipilimumab 3 mg / kg

另一方面，发现几种全身治疗在晚期HCC病例中有效[24,25,26,27]，并且正在进行免疫检查点抑制剂的几项关键试验[28,29]。尽管HCC患者使用这类药物治疗的患者数量远低于黑色素瘤患者，但在HCC患者中更常见肝损害，其中9-15％的病例显示ALT增加，并且3/4级毒性对于抗PD-1治疗，观察到4-6％[28,29]。此外，抗-CTLA-4抗体，tremelimumab，分别在HCC患者中诱导所有等级55％的肝损伤和25％的3/4级毒性[30]。因此，在HCC病例中免疫检查点抑制剂治疗期间应仔细监测ALT的增加，尽管所有肝损伤均可通过停用药物和皮质类固醇来控制。另一方面，抗-PD-1和抗-CTLA4抗体组合的3/4级肝损伤更为严重：所有级别为17.6-21％，黑素瘤为3/4级毒性为8.3-11％案件[18,21,22,23]