Summary

To the Editor

Extramedullary disease (EMD) is not an uncommon presentation at the time of acute myeloid leukemia (AML) diagnosis and/or relapse, and can present either as an isolated disease (myeloid sarcoma) or concurrently with a bone marrow involvement [1,2]. Extramedullary involvement in AML can manifest at any site; most frequently involving soft tissues, central nervous system (CNS), skin, lymphatic system, and bones [3]. Monocytic differentiation and CD56 expression, as well as cytogenetic findings of inversion (16)(p13.1q22) and translocation (8;21)(q22;q22.1) appear to be overrepresented in AML with EMD [3,4].

The treatment paradigm for AML with EMD follows the same standard treatment of AML in general [5]. However, in relapsed AML with rapidly progressive EMD, physicians usually favor administering high-intensity chemotherapy alone or with local radiation, especially if the involved site places adjacent organs at risk of compromise, in which case a rapid response is required. Chemotherapy is often administered even in truly chemotherapy-refractory cases, mainly due to lack of other effective salvage options.

Until recently, AML regimens were restricted to combination intensive chemotherapies in patients who can tolerate such therapy. However, the newly approved combination of venetoclax and hypomethylating agents (VEN-HMA) regimen has produced high response rates in elderly and unfit patients with newly diagnosed AML as well as patients with relapsed/refractory (r/r) AML [6,7]. The response data for VEN-HMA is mainly derived from patients with bone marrow involvement, and the impact of VEN-HMA on the subset of AML with EMD remains largely unknown. Therefore, we performed this analysis with the objective of understanding the activity of the VEN-HMA in AML with EMD.

We report a retrospective analysis in which we reviewed all myeloid sarcoma cases who were treated with VEN-HMA at City of Hope Medical Center between November 2016 and October 2019. We identified 18 patients with evidence of EMD at the time of initiating VEN-HMA. Response in EMD sites was evaluated using repeat imaging such as CT and/or PET scans. Our primary objective was evaluating response rate, defined as achievement of complete remission (CR)/CR with incomplete count recovery (CRi) [8].

The median age of patients was 57.5 (range 35–78). Seven (40%) patients had isolated myeloid sarcoma while 11 (60%) had EMD diagnosed concurrently with bone marrow involvement. The most common sites for EMD in our cohort was soft tissue (n= 6, 33%) followed by skin and CNS (n= 3 each; 16%). There were 10 (55%) men and 8 (45%) women. Five (28%) patients were treated with VENHMA for newly diagnosed AML selleck chemicals llc while 13 (72%) patients were treated for r/r disease, among which 4 patients (22%) had prior allogeneic hematopoietic cell transplantation (HCT). There were equal number of de novo and secondary AML cases (9 each). Leukemia risk was classified per European LeukemiaNet (ELN) risk stratification as high risk in 12 patients (67%), intermediate-risk in 5 (28%) and good-risk in one case (5%) [8]. The HMA used was decitabine in 14 patients (78%) and azacitidine in 4 patients (22%). Among patients who received decitabine, the first cycle was given as 5-day course in 5 patients and as 10day course in 9 patients. The median number of days that venetoclax was administered during the first cycle was 21days (range; 14–28). Six (33%) patients received additional concurrent therapies including radiation (n= 4) or intrathecal chemotherapy for CNS involvement (n= 3).

The overall response rate (CR/CRi) was 45%, medium-sized ring including 5 (28%) CR and 3 (17%) CRi. When we restricted response evaluation to the 12 patients who did not receive concurrent additional therapies (intrathecal chemotherapy or local radiation), the CR/CRi was 50% (n= 6). Among the four patients with isolated myeloid sarcoma treated without concurrent additional therapy, two achieved CR/CRi. The CR/CRi rate was 60% for newly diagnosed AML compared to 38% in r/r AML. The CR/CRi according to the type and schedule of HMA was 40% in decitabine 5days, 78% in decitabine 10-days, and 25% in azacitidine. The CR/CRi according to leukemia genetic risk was 42% (n= 5) for adverse-risk and was 50% (n= 3) for favorable/ intermediate-risk. Among the three patients with CSF involvement pretreatment, two cleared the CSF with VEN-HMA treatment in combination with intrathecal chemotherapy.

Venetoclax was safely administered with concurrent radiation and intrathecal chemotherapy, with no excessive encountered toxicity. The median neutrophils recovery for responders was 55days (range; 23–85). For patients with isolated EMD without BM involvement, the median duration for neutropenia (ANC<500) and thrombocytopenia (<50 k) during the first cycle were 22 (range; 6–31) days and 17.5 (range; 0–64) days, respectively. No patient died as the result of treatment in the first 60days in this cohort. The median number of venetoclax and HMA cycles in responders and non-responders were 4 and 3, respectively, and it was 4 and 2 for newly diagnosed and r/r AML, respectively. Four (22%) of responders underwent subsequent allogeneic HCT consolidation, following which only one patient relapsed at the time of data analysis. The median relapse-free survival (RFS) and overall survival (OS) were 141 and 175days, respectively. The median RFS and OS for newly diagnosed and r/r patients orthopedic medicine were (105 vs. 141days, p= 0.7) and (119 vs. 175days, 0.8), respectively. Refer to Table 1 for patients and leukemia characteristics, and treatment outcomes.

Here, for the first time, we described the activity of VEN and HMA exclusively in a cohort of AML patients with EMD. The response rate (CR+ CRi) was 45% in this setting, which includes half of patients who did not receive any additional concurrent therapy. Therefore, the activity of VEN-HMA for AML with EMD is promising and this combination represents a viable option for patients with AML with EMD involvement who are ineligible for or have failed prior chemotherapy. This is important since EMD occurs not infrequently in AML, especially in r/r setting as well as in those who underwent prior allogeneic HCT. In such patients, the use of intensive conventional chemotherapies may not only yield low response rate but also could subject them to substantial toxicity. Furthermore, a subset of responders (22%) were able to proceed to consolidation with allogeneic HCT which implies that the response was meaningful in a subset of physically fit patients.

The regimen was well-tolerated with no treatmentrelated death in the first 60-days of therapy. Furthermore, we have shown that VEN and HMA could be safely combined with local therapies such radiation for the EMD site as well as intrathecal chemotherapy for patients with CNS relapse. In a recently published case of chronic lymphocytic leukemia with CNS involvement, single agent venetoclax was given along with intrathecal chemotherapy and led to disease clearance. Authors have measured venetoclax concentration in the CSF and have shown that venetoclax can cross the blood brain barrier and able to achieve a potential therapeutic concentrations there [9]. Our analysis is limited by the retrospective nature of the study and the small number of patients. However, we conclude that VEN-HMA offers a viable option for patients with AML who are either unfit and present with EMD at the time of diagnosis or those who have relapsed with chemotherapy-refractory AML involving EM sites and are unlikely to benefit from standard high dose chemotherapies, especially if the patient fitness is a concern.