Although the prognostic impact is still controversial, the clinical outcome of children with AML associated with OOGS is generally poor. There are only a few reports that describe the biological and clinical features of these children.

Figure 5. Comparison of survival curves of OOGS (+) and negative AML patients (8).

In 1989 Cavdar et al reported ⁽²⁾ a retrospective analysis of clinical, hematological, ultrastructural and therapeutical results in 33 Turkish children presenting with OOGS and acute myelomonocytic leukemia (AMML) between the years 1963-1983. OOGS, characterized by exophthalmos, chemosis and orbital masses, was observed in 33 (27.2%) of 121 patients with AMML. Light and electron microscopy of the eye tumors demonstrated primitive myeloid and monocytic cells similar to those seen in the bone marrow. The patients with OOGS were compared with 41 children with AMML without ophthalmic tumors seen during the same period. Hematological results in the two groups were not significantly different. Identical chemotherapy regimens were administered to all patients, and although this was not a randomized trial, the mean survival time of 8.7 months in the OOGS group was significantly shorter than that of patients without OOGS (28.6 months) (p<0.01) (Figure 5) ^(3,8), consistent with other reports of a poor prognosis ⁽²⁾. In this series the OOGS patients had low socioeconomic status. They also had diminished delayed hypersensitivity reaction, and low T-cell counts prior to the commencement of chemotherapy protocol ⁽⁵⁾, although whether this is relevant to the presence of GS is not known. Indeed, the etiology of OOGS remains an enigma. Characteristically, OOGS occurs in patients with FAB M4 (myelomonocytic) or M5 (monocytic) subtypes, both of which include monoblasts and monocytes in the leukemic infiltrate. Thus, local transformation of monocytes or histiocytes to malignant blasts ⁽⁶⁾ or the “homing” of such cells to the tumor sites ⁽⁷⁾ have been considered as possible mechanisms of OOGS development. However, GS has been described in other FAB types, and associated with a number of the chromosomal translocations that occur in AML, including t8;21, t15;17, inv16, trisomies 9 and 16 and translocations involving 11q23. Our recent studies have been focused on detecting the expression of tissue adhesion molecules (CD44 and CD56) on the blast cells of the patients with OOGS, which could account for homing to certain tissue sites. We have also studied the expression of the P-glycoprotein, associated with pleotropic multidrug resistance ⁽⁹⁾, which could account for the worse prognosis of these patients. The strikingly high rate of relapses in GS, however, might also result from initial misdiagnosis, or to less intensive or delayed therapy. Regardless of the reason for a poor prognosis, Cavdar et al. have suggested that patients with OOGS should be classified as a 'high risk' subgroup of childhood AML, and that such children should receive more intensive chemotherapy with hematopoetic stem cell rescue and local irradiation to the orbits in an attempt to improve the overall survival.

In conclusion, we believe that patients presenting with OOGS and AML comprise a distinct subgroup arising in several AML subtypes. Turkish children, who have a high incidence of OOGS associated with AML, appear to have a poorer prognosis compared to patients without OOGS. Further studies are necessary to identify the reasons for tissue involvement, and to improve the clinical outcome of these patients in our setting.

G. Yavuz, E. Unal, N. Tacyildiz, H. Ugur, A. Ikinciogullari, S. Gozdasoglu, E. Babacan, A. O. Cavdar, Ankara University School of Medicine, Ankara, TURKEY

References

1. Golub TR, Arceci RJ. Acute myelogenous leukemia 545-589. Principles and Practice of Pediatric Oncology. Pizzo and Poplack,2002.
2. Reinhardt D, Pekrun A, Lakomek M, Zimmermann M, Ritter J, Creutzig U.Primary myelosarcomas are associated with a high rate of relapse: report on 34 children from acute myeloid leukemia-Berlin-Frankfurt - Münster studies. Br J Hematol. 2000; 110:863-866.
3. Cavdar AO, Babacan E, Gözdasoglu S, Kilicturgay K, Arcasoy A, Cin S, Ertem U, Erten J. High risk subgroup of acute myelomonocytic leukemia in with orbito-ocular granulocytic sarcoma (OOGS) in Turkish children. (retrospective analysis of clinical hematological ultrastructural and therapeutical findings) Acta Haematol 1989; 81: 80-85.
4. Meiss JM, Butler J. Granulocytic sarcoma in nonleukemic patients. Cancer 1986;58:2697.
5. Cavdar AO, Babacan E, Gözdasoglu S, Yavuz G, Unal E. Pamir A. T cell subsets in OOGS and AML in Turkish Children. Med Oncol Pharmacother 1993; 10: 113.
6. Byrd JC, Edenfield WJ, Shields DJ, Dawson NA.Extramedullary myeloid cell tumors in acute nonlymphocytic leukemia; a clinical review. J Clin Oncol .1995;13:1800-1816.
7. Schiffer CA, Wiernik PH.Functional evaluation of circulating leukemic cells in acute nonlymphocytic leukemia. Leuk Res 1985;10:271-277.
8. Gozdasoglu S, Yavuz G, Unal E, Tacyildiz N, Cavdar AO. Orbital granulocytic sarcoma and AML with poor prognosis in Turkish children. Leukemia 2002; 16: 962
9. Cavdar AO, Tacyildiz N, Aksoylar S, Unal E, Gozdasoglu S, Yavuz G. Orbital granulocytic sarcoma (OGS) and acute myelocytic leukemia (AML) in Turkish children; new findings on multidrug resistance, adhesion molecules, cytokines, cytogenetic and immunophenotypic analyses, XXXth SIOP Congress, Yokohama, Japan, 4-8 October 1998.