Orbital extramedullary leukaemia is not a rare entity - response to Paydas.

Abstract

We previously reported the clinical, pathological and molecular genetic findings of a series of adult acute myeloid leukaemia (AML) patients with ocular extramedullary myeloid leukaemia (EML) (Ohanian et al, 2018). In response, Dr. Paydas took exception to our assertion that ocular EML is uncommon. The differences between our study and the studies cited by Dr. Paydas appear to reflect differences in patient age and geography. While the overall incidence of EML in AML patients of all ages is about 9%, its incidence in children may reach as high as 40% (Ohanian et al, 2013). At our institution, over a period of 12 years, 7% (252/3764) of adult AML patients demonstrated EML; 10 patients with EML (4%) had ocular involvement. Thus, at our institution, the overall incidence of ocular EML in adult AML patients is 10 out of 3,724 (0·3%) (Ohanian et al, 2018). Of the 10 cases of ocular AML that we reported, 4 were therapy-related, and 2 de novo AML showed the t(8;21)(q22;q22) (Ohanian et al, 2018). In contrast, most ocular EML cases reported in children represent extramedullary involvement by de novo AML, and the most common cytogenetic abnormality is t(8;21)(q22;q22), which has been identified in up to 100% (9/9) of cases (Schwyzer et al, 1998). In the Western world, ocular EML is uncommon in AML patients of all ages (Bisschop et al, 2001). In a recent retrospective analysis of 3240 AML patients in the United States ranging in age from 15 (Ganzel et al, 2016) to 86 years (Rowe et al, 2004), 769 (23·7%) evaluable patients had clinical evidence of EML at diagnosis, none of which were ocular (Ganzel et al, 2016). Compared to those without EML, patients with EML were generally younger and predominantly male, consistent with the literature (Cavdar et al, 1989; Bisschop et al, 2001). In a Dutch study of 477 children with AML, 25% had EML, none of which were ocular (Bisschop et al, 2001). The incidence of ocular EML appears much higher in children in Africa (Davies & Owor, 1965; Schwyzer et al, 1998) and the Middle-East than in the Western hemisphere (Cavdar et al, 1989). In a study from Uganda, among 8 children with granulocytic sarcoma, 6 demonstrated orbital involvement, 4 of which were confirmed by biopsy (Davies & Owor, 1965). In a South African study of 88 children of African descent with AML, 15 (17%) had EML, 9 with ocular involvement, although biopsy results were not provided (Schwyzer et al, 1998). Similarly, a 20-year retrospective study reported that 33/133 Turkish children with AML presented with ocular symptoms (exophthalmos, chemosis, orbital masses), 12 (36%) with biopsy-proven EML (Cavdar et al, 1989). These authors reported an association between paediatric ocular EML, lower socioeconomic status and defects in delayed hypersensitivity testing, suggesting T-cell deficiency (Cavdar et al, 1989). It is difficult to determine the true incidence of ocular EML in particular, and of EML in general. Many of the reported cases of ocular EML were based solely on physical examination and were not biopsy-proven, which may overestimate its incidence. On the other hand, because baseline and follow-up imaging studies are not currently standard practice in AML management, it is likely that EML is under-reported. Several small prospective studies using positron emission tomography (PET) or PET/computed tomography (CT) imaging in AML patients either before induction chemotherapy or at relapse have demonstrated that many EML cases are undetectable by physical examination alone. A pilot study of 10 AML patients who underwent PET or PET/CT imaging before induction chemotherapy identified EML in 40%, involving meninges, pericardium, abdomen, testes and lymph nodes (Buck et al, 2008). A larger study of 26 AML patients who underwent PET imaging before induction chemotherapy demonstrated that PET imaging detected EML in twice as many patients as physical examination alone (65% vs. 31%) (Cribe et al, 2013). In 10 patients with newly diagnosed or relapsed AML with known EML, PET/CT imaging detected new sites of EML in 60% of patients (Ohanian et al, 2013). The results of these studies provide a rationale for baseline imaging in AML to help establish the true incidence of EML, and possibly change the current practice of monitoring only bone marrow responses to AML therapy. MO and LA wrote the manuscript. CT reviewed and approved the manuscript.