Low dose gemtuzumab ozogamicin for relapsed acute myeloid leukaemia in elderly.

Introduction. Acute myeloid leukemia (AML) in the elderly presents with peculiar characteristics, and current conventional approaches achieve only poor results, especially in relapsed patients. In particular, intensified regimens do not offer advantages in terms of clinical response, and even cause significant toxicity. Thus, alternative strategies able to combine efficacy and a low profile of toxicity, and in particular monoclonal antibodies, are under investigation. Gemtuzumab ozogamicin (formerly CMA-676; Wyeth Laboratories, Philadelphia, PA, USA), a humanized anti-CD33 monoclonal antibody conjugated to calicheamicin, is effective in relapsed AML patients, also in the elderly. Nevertheless, conventional doses are characterized by not irrelevant toxicity, especially in heavily pre-treated patients. We treated an AML patient not table to receive standard doses with low dose GO, in order to limit the possible occurrence of adverse events. Case Report. A 56-years old man was diagnosed with acute myeloid leukemia (M2 FAB-subtype; normal karyotype) in 1996. He received 2 courses of mini-ICE chemotherapy (idarubicin 10 mg/m/day for 3 days, etoposide 100 mg/m/day for 5 days, and cytarabine 100 mg/m/day in continuous i.v. perfusion for 5 days), achieving a complete remission. He was then submitted to autologous bone marrow transplantation. Subsequently, he received low-dose sub-cutaneous cytarabine (20 mg for 7 days, every 45 days), as maintenance therapy. This therapy was retained until the relapse occurred, in October 2002, because, despite a bone marrow CR, as assessed at morphologic examination, a minimal residual amount of blasts, around 3-5% was always present. At the time of relapse, the bone marrow aspirate was normo-hypocellular with 65% of blasts. We did not administer any cytotoxic treatment, considering the poor clinical condition and the good peripheral blood counts (WBC=4.7x10/L, 56% ANC and 0% blasts, Hb=12.9 g/dL, PLT=266x10/L). Five months later, due to the worsening blood counts, especially concerning the ANC number (WBC=2.1x10/L with 12% ANC, and 0% blasts Hb=10.6 g/dL, and PLT=198x10/L), we decided to treat the patient with low dose gemtuzumab ozogamicin (GO) (1.5 mg/m). At this time the bone marrow aspirate was normo-hypocellular with 90% of blasts. The peripheral blood cell count nadir was reached 5 days after therapy (ANC=0.2x10/L, PLT=100x10/L, Hb=10.5 g/dL). Thirteen days after the GO administration, ANC had recovered up to 1.0x10/L. The PLT count was never lower than 100x10/L. No transfusions were required. During neutropenia, the patient developed a pulmonary aspergillosis, treated with fungizone, 50 mg/m for 14 days. No laboratory abnormalities were observed. The post-treatment bone marrow aspirate was normo-hypocellular, with no blasts. Peripheral blood counts were: WBC=10.7x10/L, 72% ANC and 0% blasts, Hb=10.4 g/dL, PLT=255x10/L. The CR was maintained for 6 months, and then the patient relapsed. At present, the bone marrow aspirate is normocellular, with 70% of myeloid blasts, and the peripheral blood counts are: Hb 11 g/dL, WBC 3.9x10/L (50% ANC, and 0% blasts), PLT 162 x109/L, without transfusions. Due to the high risk of fungal infection, no further chemotherapy has been proposed. Discussion and Conclusion. GO is currently administered at a dose of 9 mg/m as a single agent, and 4-6 mg/m when given in combination with other chemotherapy. The pharmacokinetics of GO is not erratic, and doses lower than 4 mg/m seem not able to induce complete saturation of CD33 in bone marrow and peripheral blood. Thus, low doses are not used in the treatment of active leukemia, but have been proposed only as maintenance therapy. In the present case we documented the ability of very low doses of GO (1.5 mg/m) to induce CR in a relapsed elderly AML patient, a phenomenon not previously reported. Notably, the treatment-related toxicity was very limited. In fact, the only adverse event was a fungal infection, most likely related to neutropenia, a condition already present before GO administration. It should be considered that although an elderly relapsed AML patient always has very little chance of achieving a new remission, in this case the quite long first CR duration (approximately 6 years) can be considered as a relatively good prognostic feature. Nevertheless, in our opinion, our observation is noteworthy. In fact, this schedule could be proposed to patients not suitable for standard doses. Secondly, since GO in association with conventional chemotherapy seems quite toxic, lower doses of GO may be suggested for combination regimens. Nevertheless, larger studies are warranted in order to confirm our findings, and eventually assess the role of low-dose GO in AML therapy.