ISSN 1866-8836
Клеточная терапия и трансплантация

Toxicity and efficacy gemtuzumab ozogamicin with chemotherapy in patients with relapses or refractory acute myeloid leukemia

Bella I. Ayubova1, Sergey N. Bondarenko1, Olga S. Uspenskaya2, Elena V. Karyagina3, Elena I. Darskaya1, Irina A. Samorodova1, Anna G. Smirnova1, Olga V. Pirogova1, Elena V. Babenko1, Ildar M. Barkhatov1, Tatyana L. Gindina1, Ivan S. Moiseev1, Alexander D. Kulagin1, Boris V. Afanasyev1

1 Raisa Gorbacheva Memorial Research Institute of Pediatric Oncology, Hematology and Transplantation, Pavlov First Saint Petersburg State Medical University, St. Petersburg, Russia
2 Leningrad Regional Clinical Hospital, St. Petersburg
3 City Hospital No.15, St. Petersburg, Russia

Contact: Dr. Bella I. Ayubova


The remission rate in patients (pts) with acute myeloid leukemia (AML) is achieved approximately in 70-80% after induction chemotherapy (CT). In pts with relapsed AML the main goal of therapy is to achieve remission followed by allogeneic bone marrow transplantation (allo-HSCT). However, the frequency of second remission rate in the refractory/relapsed AML (RR AML) does not exceed 50%. Introduction of targeted drugs is the most promising strategy in modern therapy of hematological malignancies, in particular RR AML. Gemtuzumab ozogamicin (GO) is a recombinant, humanized anti-CD33 monoclonal antibody covalently attached to the cytotoxic antitumor antibiotic calicheamicin, which effectiveness depends on more than 75% expression of CD33-glycoprotein on leukemic blasts. The aim of this work was evaluation of GO effects in combination with chemotherapy in AML treatment.

Patients and methods

The study included 75 pts with RR AML. The median age was 36 (18-76) years. 30 (40%) pts were with primary refractory (Ref) AML, 45 (60%) pts were with relapsed AML. Pts with the first relapse (Rel1) comprised 73% (33), with the second or subsequent relapse (Rel≥2), 27% (12); the early relapse (eRel) was observed in 34 (76%) of cases, and late relapse (lRel) were registered in 11 (23%) of cases. The following distribution of ELN-2017 molecular genetic risk groups was observed: favorable, 15 (20%) pts; intermediate, 28 (37%) pts; unfavorable, 32 (43%) pts. All the pts received GO at the dose of 3 mg/m2 per administration (no more than 5 mg) from 1 to 3 times [one/two times – 26 (35%) pts, three times – 49 (65%) pts], in combination with high-dose CT (HDCT) (FLAG + Ida, HAM, HDAC, ICE, HAI) or standard and low doses of CT (SLCT) (7 + 3, LDAC, AzaIdaAraC, MetA) in 46 (61%), and 29 (39%) pts, respectively. Allo-HSCT was performed after GO in 21 pts (3 – related, 7 – unrelated, 11 – haplo), including 3 second allo-HSCTs from another donor. In 10 pts GO therapy was performed in a relapse of AML after allo-HSCT. The median timing of HSCT after GO therapy was 67 (17-157) days.


The 2-year OS was 34% (95%CI 17-51). The overall response (OR) was 52% (39/75): complete remission was achieved in 23 (31%) pts, complete remission with incomplete hematologic recovery – 13 (17%) pts, partial remission – 3 (4%) pts. The median duration of the OR was 81 (6-701) days. OR in the group which received one/two doses of GO was achieved in 31% (8/26), after three doses of GO, in 63% (31/49), p=0.007. In combination of GO and SLCT, the OR was 35% (10/29); with GO+HDCT, 63% (29/46), p=0.016. Dependence on the ELN 2017 risk group: in favorable group OR was obtained in 93% (14/15) pts, in the intermediate group, in 40% (14/28) pts; in unfavorable, in 34% (11/32) pts, p=0.006. OR was achieved in 30% (9/30) of pts with Ref AML, 64% (21/33) in Rel1 and 75% (9/12) in Rel>2, p=0.006. No statistically significant correlation was found in OR occurence between eRel or LRel (65% vs 73%, p=0.624), and for different CD33 expression on blast cells for the pts with CD33 levels over 60% or below 60% (40% vs 52%, 0.496). Depending on the timing of relapse: OR was 65% (22/34) in the pts with eRel, whereas it was 8/11 (73%) in pts with lRel. The pts with CD33 level on blasts of <60% OR was 40% (4/10), and with a level of >60%, 52% (32/62). The 2-year OS and 2-year RFS values of the pts with OR censored by allo-HSCT were 70% (95% CI 43-97) and 41% (95% CI 16-66), respectively. The relapse rate was 28% (95%CI 12-47). Hepatotoxicity manifested as a transient increase in transaminase levels (<10 ULN). Sinusoidal obstruction syndrome and long QT syndrome during induction did not occur in any of the patients. Neutropenia of 3-4 grade in the group with one/two doses of GO was observed in 88% (22/26), while being 98% (48/49) with three doses of GO. Neutropenia frequency of 3-4 grade and thrombocytopenia (grade 3-4 grade) in the pts who received GO +SLCT, occurred in 87% and 73%, respectively. In the GO + HDCT group, such complications were noted in all pts. Early mortality was 11% (95%CI 3-24). The causes of death were leukemia progression (4 pts), infectious complications (3 pts), intracerebral hemorrhage (1 pts). No direct association with the GO and death was observed.


Addition of GO to standard CT regimen for patients with RR AML demonstrated the efficacy and acceptable toxicity, which provides the basis for inclusion of GO in the combination salvage chemotherapy.


Acute myeloid leukemia, target therapy, gemtuzumab ozogamicin.

Volume 8, number 3

Download PDF version

Back to the list