Liver involvement with Langerhans cell histiocytosis in adults

Abstract

Background and Aims

Liver involvement portends poor prognosis in adults. We aimed to characterize the clinical features, liver function tests, radiologic findings, molecular profiles, therapeutic approaches and outcomes of adults patients with Langerhans cell histiocytosis (LCH) with liver involvement.

Methods

We conducted a retrospective analysis of all adults with LCH (≥ 18 years) seen at Peking Union Medical College Hospital (Beijing, China) between January 2001 and December 2022.

Results

Among the 445 newly diagnosed adults with LCH, 90 patients had liver involvement at diagnosis and 22 patients at relapse. The median age was 32 years (range, 18–66 years). Of 112 evaluable patients, 108 had full liver function testing, including alanine aminotransferase, aspartate aminotransferase, alkaline phosphatase (ALP), γ-glutamyl transpeptidase (GGT), and total bilirubin and albumin. Elevated ALP was seen in 63.0% and GGT in 86.1%; 14.8% had elevated bilirubin. Next-generation sequencing of 54 patients revealed frequent BRAF^N486_P490 (29.6%), BRAF^V600E (18.5%), and MAP2K1 (14.8%).

Outcomes

After a median 40 months’ follow-up (range 1-168 months), 3-year progression-free survival (PFS) and overall survival were 49.7% and 86.6% respectively. In multivariable analyses, ≥3 abnormal liver function tests (HR 3.384, 95% CI 1.550–7.388, P = .002) associated with inferior PFS; immunomodulatory drug therapy (HR 0.073, 95% CI, 0.010-0.541, P = .010) correlated with superior PFS versus chemotherapy.

Conclusions

In summary, elevated GGT and ALP were common in adults with LCH liver involvement. Greater than equal to 3 abnormal liver function tests predicted poor outcomes. Immunomodulatory drug therapy was associated with favorable progression-free survival compared to chemotherapy.

Introduction

Langerhans cell histiocytosis (LCH) is a rare, heterogeneous histiocytic neoplasm originating from CD1a-positive and CD207-positive immature myeloid dendritic cells.^1,2 LCH exhibits a wide spectrum of clinical manifestations and outcomes, spanning from benign single-system (SS), disease to aggressive multisystem (MS) disease affecting bone (70%), pituitary (62%), and lungs (61%).^3,4 Liver, spleen, and hematopoietic system are considered “risk organs” in pediatric LCH.⁵ Our previous research has demonstrated that liver involvement at diagnosis was associated with poor survival in adults with LCH.⁴ LCH with liver involvement may manifest as tumorous lesions or progress to sclerosing cholangitis, the latter of which often requires liver transplantation.⁶ Despite a case series reporting 23 patients with liver LCH by Abdallah et al,⁷ only 39% of them underwent liver imaging and none received molecular testing. The patterns of liver involvement in LCH, abnormalities in liver function, radiologic findings, treatment, and correlations with outcome remain poorly defined.

The discovery of recurrent BRAF^V600E mutations were identified in 57% of LCH samples in 2010,⁸ stimulating further exploration of other activating MAPK pathway mutations in LCH.^9,10 In pediatric LCH, BRAF^V600E mutation was associated with high-risk organ involvement, including liver disease.¹¹ However, little is known about the molecular landscape of liver LCH in adults.

To address these questions, we conducted a retrospective study investigating the clinical features, mutational characteristics, treatments, and outcomes of this subgroup of patients at our institution over the past 2 decades.

Methods

Patients

We conducted a retrospective analysis of all adults with LCH (≥ 18 years) seen at Peking Union Medical College Hospital (Beijing, China) between January 2001 and December 2022. Histological findings were consistent with LCH on the basis of the World Health Organization classification of hematopoietic neoplasms.¹² Liver involvement indicated by one or more of the following: either hepatomegaly, defined as a liver edge >3 cm below the costal margin at the midclavicular line, or liver dysfunction defined either by abnormal serum biochemical tests including bilirubin >1.5 times the upper limit of normal, alkaline phosphatase (ALP) greater than 1.5 times the upper limit of normal, γ-glutamyl transpeptidase (GGT) >1.5 times the upper limit of normal; or histopathological findings of active disease.¹³ The study was approved by the Peking Union Medical College Hospital Ethics Committee. The study was performed in accordance with the ethical standards of the 1964 Declaration of Helsinki and its later amendments.

Data collection and molecular analysis

We collected data on baseline clinical features, medical history, physical exam, complete blood count; liver function test (alanine aminotransferase [ALT], aspartate aminotransferase [AST], ALP, GGT, total bilirubin, conjugated bilirubin, and albumin), inflammatory markers (high-sensitive C reactive protein [hsCRP], erythrocyte sedimentation rate [ESR]), imaging studies (abdominal ultrasound, computed tomography [CT], magnetic resonance imaging [MRI], 18F-fluorodeoxyglucose positron emission tomography [FDG-PET]), treatment and survival. Patients with available samples underwent next-generation sequencing of 183 genes as described.¹⁴ Patients were classified as having unifocal disease (solitary lesion involving any single organ) or multisystem (MS) disease (involvement of multiple systems) based on organ involvement when hepatic involvement was established through diagnosis.⁵

Treatment

Systemic treatments of LCH were categorized into 3 broad groups: (i) conventional chemotherapies, including methotrexate/cytarabine (MA) combination,¹⁵ single-agent cytarabine, vindesine, and prednisone (VP)-based regimens,¹⁵ and single-agent cladribine and corticosteroids; (ii) immunomodulatory drug-based (IMIDs) therapies, including TCD regimen (thalidomide, cyclophosphamide, and dexamethasone),¹⁶ RD regimen (lenalidomide and dexamethasone)¹⁷; and (iii) target therapies, including BRAF inhibitors and MEK inhibitors.

Outcomes

Overall survival (OS) was calculated from the date of diagnosis of liver involvement by LCH to the date of death or the last contact. Progression-free survival (PFS) was calculated from the initiation of treatment for LCH (after diagnosis of liver involvement) to disease progression or death and patients without events were censored at last contact. PFS2 was calculated from the initiation of salvage therapy after first progression of LCH (after diagnosis of liver involvement) to the second progression or death.

Statistical analysis

All analyses were performed using SPSS statistics version 24.0 (SPSS Inc., Chicago, IL, USA). Fisher’s exact test was used to compare categorical variables, whereas the Mann-Whitney test was used to compare continuous variables between groups. OS and PFS were estimated according to Kaplan-Meier survival analysis and compared with the log-rank test. Comparisons between the variables of interest were performed using univariate and multivariate Cox regression models. Only variables with a certain level of significance (P < .10) in the univariate analysis were included in the multivariate models. A Cox regression model was used to estimate hazard ratios (HRs) and their 95% CIs for factors associated with OS and PFS. P < .05 was considered statistically significant. The final follow-up date was December 31, 2022.

Results

Demographic data

Among 445 newly diagnosed adults with LCH between January 2001 and December 2022, 90 patients had liver involvement at diagnosis and 22 patients had liver involvement at relapse. Overall, 112 patients were included in this study including 78 males (69.6%) and 34 females (30.4%), and the median age at diagnosis of liver involvement was 32 years (range, 18-66 years). Fifteen patients (13.4%) were diagnosed by liver biopsy. Among patients who had liver involvement at relapse, the median time from disease diagnosis to diagnosis of liver involvement was 32 months (range 3-108 months), and the median number of prior lines of treatment was 1 (range 1-3). Three patients (2.7%) had a solitary liver tumor at diagnosis and were classified as unifocal, 109 patients (97.3%) had MS LCH. The median number of involved organs was 4 (range 1-7). The most common organ involved besides liver was bone (65.2%), followed by the pituitary (61.6%), lung (55.4%), lymph nodes (34.8%), skin (21.4%), thyroid (20.5%), and spleen (19.6%).

Laboratory findings

At diagnosis of liver involvement, 108 of 112 patients underwent comprehensive liver function testing. Baseline results are summarized in Table 1. Elevated GGT (93 patients, 86.1%) and ALP (68 patients, 63.0%) were most frequent. ALT and AST were elevated in 44.4% (48 of 108) of patients each. Hyperbilirubinemia was detected in 14.8% (16 patients). Only 9.3% (10 patients) had hypoalbuminemia, including 3 with albumin < 3 g/dL. Nearly, all patients (98 of 108, 90.7%) had at least one abnormal liver function test at baseline. Non-specific markers of inflammation were often elevated, including hsCRP in 61.1% (55 of 90) and ESR in 69.1% (65 of 94).

Radiology findings

Ultrasonography, CT, MRI, and FDG-PET were variably performed at diagnosis of liver LCH for initial staging and evaluation of liver involvement in 52, 83, 19, and 76 patients, respectively. The patterns observed are summarized in Supplementary Table S1.

On ultrasonography, diffuse liver hyperechogenicity with or without hepatomegaly/splenomegaly (51.9%) and heterogeneous liver with or without nodules (19.2%) were most frequent. CT typically showed heterogenous hypoattenuation with or without hepatomegaly/splenomegaly (44.6%) or hypodense nodules (12.0%). MRI findings were available for a minority of patients. Examples of MRI and FDG-PET images are shown in Figure 1 and Supplementary Figure S1. Nearly, 36.8% of patients (28 of 76) had no hypermetabolic liver lesions on FDG-PET at diagnosis, while 25.0% (19 of 76) had hypermetabolic liver lesions. Diffuse hypermetabolism was seen in 13.2% (10 of 76). These findings, together with the predominance of heterogenous and nodular patterns on CT and ultrasonography, suggest a spectrum of tumorous and mixed tumorous-sclerosing cholangitis liver diseases in this cohort.

Molecular analysis

Only 54 patients (48.2%) had sufficient DNA for next-generation sequencing. At least one nonsynonymous gene mutation was detected in 51 patients (94.4%). The median number of gene mutations was 2 (range 1-7). MAPK/ERK-related pathway mutations were detected in 39 patients (72.2%). BRAF^V600E was detected in 10 patients (18.5%), BRAF^N486_P490 was detected in 16 patients (29.6%). MAP2K1 was detected in 8 patients (14.8%). BRAF non-V600E alternations (p.A366P and p.E586G) were detected in 2 patients (3.7%). KRAS and NRAS were detected in 3 patients (5.6%) each. MAP3K1 and MAPK1 was detected in 3 patients (5.6%) and 2 patients (3.7%), respectively. The three patients with isolated liver involvement did not undergo NGS. Detailed genomic profiles of the entire cohort are displayed in Figure 2.

Treatment and outcomes

First-line treatment was administered to 96 patients after diagnosis of liver LCH. The most common regimens were MA (n = 37, 38.5%), cytarabine monotherapy (n = 28, 29.1%), IMIDs-based therapies (n = 15, 15.6%), and target therapies (8.3%, including 4 patients with BRAF inhibitors and 4 patients with MEK inhibitors). Two patients died before receiving any treatment and 7 patients went back to local hospitals for further treatment. The initial treatment of the whole cohort is illustrated in a flow diagram in Figure 3. All of the 3 patients with isolated liver involvement received cytarabine monotherapy, all were progression free without cirrhosis after a 4-month follow-up (range 1-6 months).

After a median 40-month follow-up (range 1-168 months), 42 patients had disease progression. The median PFS duration was 31.4 months, and the 3-year PFS rate was 49.7% (Figure 4A). Univariate analysis was performed to evaluate the prognostic factors of PFS. Patients with elevated hsCRP at baseline had significantly shorter PFS than patients with normal hsCRP (25.9 months vs not reached, P = .039) (Figure 4B). Patients who had 3 or more abnormal liver function test (including ALT, AST, ALP, GGT, total bilirubin, and ALB) had significantly shorter PFS than patients with 2 or less abnormality (22.5 months vs not reached, P = .007; Figure 4C). Patients who received targeted therapy or IMIDs-based therapy as first-line treatment had significantly longer PFS than patients who received other systemic treatment (not reached vs not reached vs 25.9 months, P = .023; Figure 4D). In multivariable analyses, patients who had 3 or more abnormal serum liver function tests (HR 3.384, 95% CI 1.550-7.388, P = .002), and treatment with IMIDs (HR 0.073, 95% CI 0.010-0.541, P = .010) compared to other systemic chemotherapy remained predictive factors for PFS (Supplementary Table S2).

Among 42 patients who had disease progression after first-line therapy, 32 patients received second-line therapy, including IMIDs-based therapies (n = 18, 56.3%), VP-based regimens (n = 5, 15.6%), cytarabine/cladribine/MA (n = 5, 15.6%), and MEK inhibitor (n = 3, 9.4%; Figure 4). After a median 35-month follow-up (range 3-73 months), the estimated median PFS2 duration was 43.7 months (Supplementary Figure S2).

Thirteen patients died during follow-up period. The 3-year OS rate was 86.6%. Concerning the prognostic factors of OS according to the univariate analysis, elevated total bilirubin at baseline was associated with significantly shorter OS than patients with normal total bilirubin (P = .046; Supplementary Figure S3A). Patients who had 3 or more abnormal serum liver function test had significantly shorter OS than patients with 2 or less abnormality (P = .090; Supplementary Figure S3B).

Nine patients developed cirrhosis and 2 developed sclerosing cholangitis during follow-up. Two patients underwent liver transplantation, one required re-transplantation for progressive sclerosing cholangitis and liver failure from LCH.

Discussion

The features of liver involvement in children have been extensively described,² but the characteristics of this condition in adults remain poorly defined. To date, this is one of the largest studies of adults with liver involvement in LCH. In this study, we reported the clinical features, liver biochemistry, radiology findings, treatments, and outcomes of adults with patients with LCH with liver involvement. The vast majority (97.4%) of the liver involvement was observed in patients with multi-system LCH. Interestingly, this study first reported 3 patients with unifocal liver involvement, a rare finding in LCH. None of them had disease progression due to the follow-up time was quite short (median 4 months), necessitating longer follow-up of unifocal liver LCH to assess the risk of progression compared to MS LCH.

In LCH with liver involvement, patients typically present with hepatomegaly, elevated liver enzymes or jaundice. GGT has been identified as a sensitive indicator of liver infiltration in pediatric LCH.² In this study, we found that over 80% of patients with elevated GGT levels and more than 60% of patients with elevated ALP levels. Transaminases were less commonly elevated and tended to be mild. Less than 10% of patients showed decreased albumin levels, while elevated bilirubin levels, indicating sclerosing cholangitis or end-stage liver disease, was observed in less than 15% of patients and always in conjunction with other liver test abnormalities. It is noteworthy that a comprehensive liver function test should be conducted for the diagnosis of liver involvement as over 90% of the patients in this study exhibited at least one abnormal liver function.

This study reported a wide range of imaging patterns to date, which were heterogeneous in nature. Previous studies described two distinct forms of LCH with liver involvement²: early infiltration characterized by hepatomegaly, liver nodules, mild cholestasis, and elevated transaminases; and late stage involvement characterized by fibrosis, severe cholestasis, and liver failure.¹⁸ As liver biopsy is not always feasible for multi-system patients in clinical settings, imaging studies play a crucial role in assessing liver injury by revealing parenchymal infiltration or sclerosing cholangitis. Ultrasound commonly showed diffuse hyperechogenicity and heterogeneity, with or without hepatomegaly, accounting for over 70% of the patients. CT scans revealed heterogeneous hypoattenuation while MRI displayed varied T1/T2 changes with only 2 patients had normal MRI, lower than other modalities, implying MRI may have higher sensitivity for early recognition of liver involvement. FDG-PET was able to globally depict both the extent and the activity of other hematological tumors, aiding in the assessment of bone, lymph node, and pituitary gland involvement of patients with LCH. But more than two-thirds of patients exhibited a hypoattenuation or normal pattern FDG-PET scans.

Carrere et al observed the presence of a BRAF^V600E mutation in 100% of pediatric patients with liver involvement.¹⁹ Our previous study demonstrated that BRAF deletion was more prevalent in patients with liver involvement as compared to those without liver involvement.⁴ However, in the present study, BRAF^V600E mutation was present in only approximately 20% of the patients, which was less common than BRAF deletion. The proportion of BRAF^V600E in our cohort was lower than previously reported. One of the reasons may be that BRAF^V600E mutation is associated with lower age at diagnosis in pediatric patients.²⁰ Another reason was that we include most patients with multi-system LCH. Our understanding of the genomic landscape and its association with organs involvement in adult patients remains limited. Further investigation is necessary to determine the clinical significance of these mutations in this subset of patients.

Previous study at our institution showed that MA regimen was 87% effective in the treatment of 83 adults with LCH.¹⁵ Patients with liver involvement at baseline had significantly shorter event-free survival compared with patients without liver involvement. A phase II study in our center showed that TCD regimen was 87.5% effective in recurrent/refractory patients with LCH, and patients with or without risk organ involvement had similar event free survival.¹⁶ In this study, we found that patients who received IMIDs-based therapy had better outcomes than chemotherapy, which suggests that IMIDs-based therapy could be considered as first-line therapy instead of chemotherapy in patients with liver involvement. Although patients who received targeted therapy exhibited significantly longer PFS than those who received chemotherapy in the univariate analysis, this was not observed in the multivariate analysis, possibly due to the limited number of patients. Notably, PFS2 was considerably longer than PFS with first-line therapy, and this could be attributed to the higher proportion of patients receiving IMIDs-based and targeted therapy in the second-line settings as compared to the first-line treatment.

Furthermore, our study observed that patients with more than 3 abnormal liver functions had poor outcomes in terms of PFS and OS. Patients with diffuse liver enzyme abnormalities may potentially benefit from upfront IMIDs-based or targeted therapy, and further investigation in this patient population is warranted.

This study has several limitations. Firstly, it is a retrospective study conducted at a single center, which may limit the generalizability of our findings. Additionally, the treatment selection was based on physician`s choice, introducing potential bias. Another limitation is the lack of uniform imaging studies at diagnosis and during follow-up for evaluation in all patients. In the future, prospective trials should be designed to help identify a more precise imaging study and better therapeutic options for LCH with liver involvement.

Conclusion

Elevated GGT and/or ALP were found in the majority of adults patients with LCH and liver involvement. Elevated hsCRP and 3 or more abnormal liver function tests exhibited poor outcomes. Targeted or IMIDs-based therapy was associated with superior outcomes compared with conventional treatments.

Supplementary material

Supplementary material is available at The Oncologist online.

Acknowledgments

The authors thank the patients and their families.

Author contributions

Long Chang (Conceptualization, Data curation, Formal Analysis, Writing—original draft, Writing—review & editing), Hua-Cong Cai (Conceptualization, Formal Analysis, Data curation, Writing—review & editing), Min Lang (Formal analysis, Patient follow-up, Writing—review & editing), He Lin (Formal analysis, Patient follow-up, Writing—review & editing), Ya-Ping Luo (Review of image studies, Data curation, Writing—review & editing), Ming-hui Duan (Resources, Writing—review & editing), Dao-bin Zhou (Data curation, Review of patient records, Writing—review & editing), Gaurav Goya (Data curation, Review of patient records, Writing—review & editing), and Xin-xin Cao (Conceptualization, Writing—original draft, Writing—review & editing)

Funding

This study was funded by the Beijing Natural Science Haidian frontier Foundation (grant no. L222081 to CXX), National Key R&D Program of China, Ministry of Science and Technology of the People’s Republic of China (2022YFC2304605), and the National High Level Hospital Clinical Research Funding (2022-PUMCH-A-080).

Ethical approval

All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. Informed consent was obtained from all patients included in the study.

Conflicts of interest

Gaurav Goyal served on the advisory board for Opna Bio LLC and Seagen, and received consulting fees from Recordati. The other authors indicated no financial relationships.

Data availability

The datasets during and/or analyzed during the current study are available from the corresponding author on reasonable request.

References

Author notes