Clinical and Radiologic Responses to Cladribine for the Treatment of Erdheim-Chester Disease

Key Points

Question  What is the efficacy of cladribine in the management of Erdheim-Chester disease?

Findings  In this medical record review, 21 patients with Erdheim-Chester disease receiving cladribine therapy demonstrated overall clinical and radiological response rates of 52% and 54%, respectively. The response with cladribine was durable in some cases, and the drug was generally well tolerated in most patients.

Meaning  Our study findings suggest that cladribine has moderate clinical efficacy in the treatment of Erdheim-Chester disease and can be considered a treatment option in cases without the BRAF V600E mutation.

Importance  While cladribine is best known for the treatment of hairy cell leukemia and other lymphoid cancers, it also has activity against myeloid neoplasms, such as Erdheim-Chester disease (ECD).

Objective  To assess the efficacy of cladribine (2-chloro-2′-deoxyadenosine) in the treatment of ECD.

Design, Setting, and Participants  This study was a single-institution retrospective medical record review from January 1, 1998, to April 6, 2016, at a tertiary academic medical center. In all eligible cases, the diagnosis of ECD was made using clinical criteria in conjunction with histopathologic findings.

Exposure  Cladribine therapy in first-line treatment or later.

Main Outcomes and Measures  Two response criteria were used, clinical and radiological. For clinical response, the following criteria were used: complete response (complete resolution of symptoms attributed to ECD), partial response (partial resolution of symptoms attributed to ECD), stable disease (no change in symptoms attributed to ECD), and progressive disease (worsening of symptoms attributed to ECD). For radiological response, the following categories were used: complete response (complete resolution of proven or suspected lesion due to ECD), partial response (partial resolution of proven or suspected lesion due to ECD), stable disease (no significant change in proven or suspected lesion due to ECD for ≥3 months), and progressive disease (progression or worsening of proven or suspected lesion due to ECD).

Results  A total of 63 adult patients with confirmed ECD were identified. Their median age at diagnosis of ECD was 54 years (age range, 18-80 years), and 67% (42 of 63) were male. Cladribine was the most commonly used chemotherapeutic agent and was administered in 21 of 63 patients (33%). Their median age at the time of cladribine therapy was 62 years (age range, 40-78 years). Cladribine was used as the first-line treatment in 9 patients and as later-line treatment in the remaining 12 patients. The median number of cycles of cladribine administered was 2.5 (range, 1-6). The overall clinical response rate was 52% (9 of 17) (6% [1 of 17] complete response and 46% [8 of 17] partial response), with 18% (3 of 17) stable disease and 30% (5 of 17) progressive disease. Among patients who responded to cladribine therapy, the median duration of clinical response was 9 months (range, 6-129 months), with ongoing response in 2 patients. The overall radiological response rate was 54% (8 of 15) (all partial response), with 26% (4 of 15) stable disease and 20% (2 of 15) progressive disease. Treatment-related adverse effects included 2 infectious complications (pneumonia and central line infection, both requiring hospitalization) and 2 hematologic adverse effects (grade 4 neutropenia and thrombocytopenia, and grade 3 neutropenia, both requiring therapy discontinuation).

Conclusions and Relevance  Cladribine has moderate clinical efficacy in the treatment of ECD and can be considered a treatment option in cases without the BRAF V600E mutation. It is generally well tolerated and may result in a durable response.

Erdheim-Chester disease (ECD) is an uncommon histiocytic disorder that is considered a hematologic cancer.¹ Because of the rarity of the disease, there are limited data on treatment options and their efficacy. Historically, interferon alfa has been used with variable efficacy, albeit at the cost of adverse effects.² This limitation has led to the use of other agents in an attempt to find more efficacious and less toxic therapies. Cladribine (2-chloro-2′-deoxyadenosine) is an antimetabolite approved by the US Food and Drug Administration for the treatment of hairy cell leukemia. It has also been shown to have activity in other lymphoid cancers (chronic lymphocytic leukemia, lymphoplasmacytic lymphoma, and certain low-grade lymphomas), as well as Langerhans cell histiocytosis. Increasing numbers of case reports demonstrate activity of cladribine in the treatment of ECD.^3-10 However, contemporary trial or prospective data are lacking to help discern the response rate and outcomes with this agent. In this study, our objective was to assess the efficacy of cladribine in the treatment of ECD.

After obtaining waiver of informed consent and approval of the study from the Mayo Clinic, Rochester, Minnesota, Institutional Review Board, we retrospectively reviewed the medical records of patients with ECD evaluated at Mayo Clinic, Rochester (a tertiary academic medical center) from January 1, 1998, to April 6, 2016, who received cladribine as any line of therapy. In all eligible cases, the diagnosis of ECD was made using clinical criteria in conjunction with histopathological findings.¹¹ All biopsy specimens were independently reviewed at our institution by 3 of us (G.G., M.V.S., and R.S.G.). All eligible patients received cladribine using 1 of 2 intravenous dosing regimens (0.14 mg/kg for days 1-5 every 28 days or 5 mg/m² for days 1-5 every 28 days).

Because posttherapy response assessment was not uniformly performed, we used 2 response criteria, clinical and radiological. Clinical response to therapy was assessed using the clinical documentation. Clinical response was categorized as (1) complete response (CR) (complete resolution of symptoms attributed to ECD), (2) partial response (PR) (partial resolution of symptoms attributed to ECD), (3) stable disease (SD) (no change in symptoms attributed to ECD), or (4) progressive disease (PD) (worsening of symptoms attributed to ECD). The minimum duration for symptom improvement to be considered a response was arbitrarily set at 3 months.

Radiological response to therapy was assessed using the imaging studies available. In our series, most patients did not undergo positron emission tomography–computed tomography. Response was assessed based on alternative imaging studies available, including radiographs, magnetic resonance imaging, computed tomography, and radionuclide bone scans. Radiological response was categorized as follows based on change in the size or appearance of lesions suspected or proven to be ECD: (1) CR (complete resolution of proven or suspected lesion due to ECD), (2) PR (partial resolution of proven or suspected lesion due to ECD), (3) SD (no significant change in proven or suspected lesion due to ECD for ≥3 months), or (4) PD (progression or worsening of proven or suspected lesion due to ECD).

A total of 63 adult patients with confirmed ECD were identified. Their median age at diagnosis of ECD was 54 years (age range, 18-80 years), and 67% (42 of 63) were male. Cladribine was the most commonly used chemotherapeutic agent and was administered in 21 of 63 patients (33%). Their median age at the time of cladribine therapy was 62 years (age range, 40-78 years). Nine patients received cladribine as their first-line treatment, and the remaining 12 patients received it as second-line treatment or later. The distribution of men to women in the cohort was 14:7. Results of BRAF V600E (NCBI Reference Sequence NM_004333.4) mutation testing were available in 5 of 21 patients (24%), most of whom did not express the mutation by immunohistochemistry, although one was positive by circulating tumor DNA polymerase chain reaction testing. The median number of cycles of cladribine administered was 2.5 (range, 1-6). Among the 21 patients treated with cladribine, 8 (38%) received it between 1998 and 2006, and 13 (62%) received it between 2007 and 2016.

Clinical responses were available for 17 patients, and radiological responses were available for 15 patients. Clinical response rates with cladribine were 6% (1 of 17) CR, 46% (8 of 17) PR, 18% (3 of 17) SD, and 30% (5 of 17) PD, while radiological response rates were 0% (0 of 15) CR, 54% (8 of 15) PR, 26% (4 of 15) SD, and 20% (2 of 15) PD (eTable in the Supplement). The median duration of clinical response was 9 months (range, 6-129 months), with ongoing response in 2 patients (>6 months in both). All patients who did not express BRAF V600E mutation by immunohistochemistry demonstrated clinical response. The disease sites in the patients who responded included central nervous system (n = 5), bone (n = 2), pleura (n = 2), pericardium (n = 2), aorta (n = 2), and paraspinal (n = 1) (Figure and eFigure in the Supplement). Cladribine was generally well tolerated, and results of adverse effect assessment demonstrated 2 infectious complications (pneumonia and central line infection, both requiring hospitalization) and 2 hematologic adverse effects (grade 4 neutropenia and thrombocytopenia, and grade 3 neutropenia, both requiring therapy discontinuation). None of the patients died as a result of cladribine therapy.

Our study demonstrates that cladribine has moderate clinical efficacy in patients with ECD, with an overall clinical response rate of 52% (9 of 17) and an overall radiological response rate of 54% (8 of 15). To date, 17 case reports of ECD have been documented, with a collective response rate to cladribine approaching 60%.^3-10 Similar to our series, these case reports demonstrated a response with cladribine at various ECD sites, such as central nervous system, bone, pleura, and pericardium. In our study, there were no specific discernible characteristics of patients that could be used for predicting a positive response to cladribine therapy. Overall, cladribine has a favorable adverse effect profile; the major adverse effects were related to myelosuppression, leading to fatigue or infections in 20% (4 of 21) of our patients. This finding is comparable to the adverse effect rates seen in prior experience with cladribine in hairy cell leukemia, for which it is considered the standard of care.¹²

There is increasing evidence that almost all patients with BRAF V600E mutant ECD have an excellent response to the BRAF inhibitor vemurafenib.^13,14 This finding will likely lead to the emergence of vemurafenib as the first-line treatment in BRAF V600E mutant ECD, which constitutes approximately two-thirds of cases.¹⁵ However, there is still a lack of promising treatment options for patients with ECD who lack the mutation. Our study findings suggest that cladribine may be considered an alternative to interferon alfa as the first-line treatment in patients who do not harbor the BRAF V600E mutation. The choice of agent could be further individualized based on adverse effect profile, patient preference, and ease of administration. To our knowledge, the present study is the largest series reviewing the efficacy of cladribine in the treatment of ECD, which underscores the challenges in conducting large studies owing to the rarity of this condition.

The limitations of our study are primarily related to the fact that it is a single-institution retrospective medical record review, so our results may not be broadly generalizable. The study spanned 18 years, and it would be difficult to ascertain an effect of the role of advances in overall management and imaging modalities during this period. In addition, posttherapy response assessment was not uniformly performed after treatment in all patients, thereby leading us to devise the clinical and radiological response criteria. Nevertheless, the medical records were independently reviewed by 3 of us (G.G., M.V.S., and R.S.G.) to minimize any chance of error or bias.

Cladribine has moderate clinical efficacy in patients with ECD, with a favorable adverse effect profile in therapy-naive and previously treated patients. Our study findings suggest that cladribine can be considered a treatment option in patients with ECD without the BRAF V600E mutation. Further studies are needed to assess predictive factors for a favorable response to cladribine therapy.

Accepted for Publication: January 4, 2017.

Corresponding Author: Ronald S. Go, MD, Division of Hematology, Mayo Clinic, 200 First St SW, Rochester, MN 55905 (go.ronald@mayo.edu).

Published Online: March 2, 2017. doi:10.1001/jamaoncol.2017.0041

Author Contributions: Drs Goyal and Go had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Goyal, Shah, Call, Go.

Acquisition, analysis, or interpretation of data: All authors.

Drafting of the manuscript: Goyal, Litzow, Go.

Critical revision of the manuscript for important intellectual content: Goyal, Shah, Call, Hogan, Go.

Statistical analysis: Goyal, Go.

Administrative, technical, or material support: Goyal, Go.

Study supervision: Call, Hogan, Go.

Conflict of Interest Disclosures: None reported.