Haematologica 2002; 87:(04)ECR13
[Medline] [prev] [index] [next]Leukaemia cutis: clinical features and treatment strategies
Sonja Zweegman,1 Maarten H. Vermeer,2 Marcel W. Bekkink,2 Paul van der Valk,3 Prabath Nanayakkara,1 Gert J. Ossenkoppele1
1Departments of Haematology, 2Dermatology, 3Pathology, Vrije Universiteit Medical Centre, Amsterdam, the Netherlands.
Correspondence: S.Zweegman, MD, Department of Haematology, BR 240, Vrije Universiteit Medical Centre, de Boelelaan 1117, 1081 HV Amsterdam, the Netherlands. E-mail: s.zweegman@vumc.nl
Addition of radiotherapy to chemotherapeutic regimens in patients with acute myeloid leukemia (AML) with skin localization might be indicated, as a high incidence of skin relapses suggest that the skin could provide a sanctuary site for leukaemic cells. We describe a patient with leukaemia cutis and reviewed the literature to evaluate whether these patients indeed require different treatment strategies than patients with AML without skin involvement.
A 66-year old man presented with erythematous nodules (Figure 1). Laboratory studies showed normal blood counts and leukocyte differentiation. A skin biopsy revealed proliferation of blasts (Figure 2a and b). Bone marrow examination enabled the diagnosis of AML, FAB classification M5b. Blasts were present in the spinal fluid. Treatment with daunorubicin and cytarabine, combined with intrathecal methotrexate twice weekly was started. After two identical courses of chemotherapy, bone marrow examination showed a complete remission of the AML, however skin localization persisted. The performance status did not allow further treatment.
Revision of English Medical literature (Pubmed) of 1965-2001 identified 92 adult patients with leukaemia cutis.1,2 Characteristics were compared with two series of patients with AML without skin involvement (Table 1).3,4 There is a predominance of (myelo)monocytic leukaemia in patients with leukaemia cutis. The high percentage of hepatosplenomegaly and lymphadenopathy is probably accounted for by the predominance of (myelo)monocytic leukemia, since these percentages are similar among patients with monocytic leukemia.5,6 Central nervous localization occurs more frequently than would be expected from the higher incidence of the (myelo)monocytic subtype in patients with leukemia cutis; 17% versus 5-7% in patients with acute monocytic leukemia.5,6 Therefore, these patients require a diagnostic lumbar punction and prophylactic intrathecal chemotherapy.
A subset of patients shows leukemia cutis prior to bone marrow involvement.7 Therapy should not be postponed in these patients, as bone marrow involvement occurred in all patients within one year. Second, analysis of morphologically normal bone marrow showed abnormal karyotypes and fusion transcripts, identical to the ones found in tissue samples of these patients.8
The overall survival of patients with leukemia cutis is worse than that of to patients without skin involvement. In approximately 55% of the patients developing a relapse it appears at an extramedullary site, mostly in the skin and central nervous system. Also in the setting of allogeneic bone marrow transplantation a propensity of extramedullary relapse has been described (38.5%, of which 23.1% in the skin, versus 4.5% in the non-leukaemia cutis group).9 Although there are no randomized trials comparing combination therapy with single entity therapy, in view of the concept of the skin as a sanctuary site, combination therapy consisting of chemo- and radiotherapy seems preferable. The use of electron beam radiation (EBR) has been described as a valuable alternative for conventional irradiation.10 Electrons are delivered to a limited depth of the entire skin surface, thereby preventing systemic toxicity. Since skin toxicity has been described after EBR in conjunction with anthracyclines, but not with cytarabine, combination of EBR with cytarabine is preferable.1,10 Furthermore, it might be advocated not only to radiate the involved skin, but the whole body in case leukaemia cutis is widespread, as recurrences frequently occur outside of local irradiation fields.1
Acknowledgments
We thank Dr. W.L.J van Putten, (statistician), for reviewing the incidence of extramedullary localization in the HOVON AML studies and Peter C. Huijgens, MD PhD, for critically reading the manuscript.
References