Combining the Number of Infectious Episodes and Interval of Chemotherapy to Transplantation May Predict a Transplant Outcome in Patients with Acute Myeloid Leukemia: A Retrospective Analysis of JALSG AML201 Study

Background

Allogeneic hematopoietic stem cell transplantation (allo-HSCT) is a potentially curative therapy for acute myeloid leukemia (AML), but non-relapse mortality (NRM) remains obstacles to this treatments. Therefore, precise comorbidity evaluation before transplantation is crucial for better outcomes. One of the most frequently used tools for assessing the risk profile of transplant recipients is the hematopoietic cell transplantation comorbidity index (HCT-CI), which only incorporates information of the antibiotics use on day 0 as an infection-related factor, however, some important components, such as the number and severity of infectious episodes, are not included.

Methods

From December 2001 to December 2005, 1,057 newly diagnosed adult AML patients were registered to JALSG AML201 (Blood 2011; 117: 2358-65, 2366-72). Of them, the 121 evaluable patients who underwent allo-HSCT in the first hematological complete remission were analyzed. We have checked the above mentioned components, in addition to the interval from the last chemotherapy to transplantation, in these 121 patients and analyzed whether these factors might affect the subsequent transplant outcome.

Results

The median age was 37 years (range, 15-61). Regarding cytogenetic risk, 7 (6%), 92 (76%), and 15 (12%) patients were categorized as favorable, intermediate and high risk, respectively. The median follow-up period after transplantation was 939 days (range 19-2204). At 2 years after allo-HSCT, overall survival, leukemia free survival, cumulative incidence of relapse (CIR), and NRM for all patients were 73.2%, 65.4%, 19.4%, and 15.8%, respectively. A total of 106 (88%) patients developed at least one infectious episode as 102 (84%) patients faced to febrile neutropenia (FN), 45 (37%) patients experienced severe infection (≥ grade 3), and 27 (22%) patients eventually developed sepsis. Regarding the episode of FN, 52 (43%) patients experienced more than 2 times during chemotherapy. The median cycles of chemotherapy before transplantation was 4 (range, 1-6) and the median interval from the first day of last chemotherapy to transplantation was 77 days. The ROC curve indicated mild correlations between NRM and the interval with an area under the curve of 52% and the cutoff value at 113 days.

The number of infectious episodes or severity of infection did not carry a significant impact on NRM at 2 years as 19.2% in FN with ≥3 episodes vs. 13.2% in FN with 0-2 episodes; p=0.18, 18.0% in patients with severe infection vs. 14.5% in patients without severe infection; p=0.49. However, in view of bone marrow transplant (BMT) recipients (n=76), multiple FN episodes was significantly associated with higher NRM (23.5% vs. 7.2% at 2 years, p=0.029) (Fig 1A). While, the relative shorter interval (<113 days) from the first day of last chemotherapy to transplantation might have a negative impact on NRM as 19.4% vs. 6.2% at 2 years, p=0.054 (Fig 1B), but not affect CIR rate (18.4% vs. 22.3% at 2 years, p=0.67). Furthermore, the relative shorter interval in BMT recipients with multiple FN episodes showed a significant adverse impact on NRM (32.0% vs. 0% at 2 years, p=0.038) (Fig 1C).

Conclusion

Combining the number of infectious episodes and interval of chemotherapy to transplantation may predict a transplant outcome in AML patients. Providing an adequate interval from the last chemotherapy to transplantation may reduce NRM especially in BMT recipients who developed multiple FN episodes during induction or consolidation chemotherapy.

Figure 1.

View largeDownload slide

Figure 1.

View largeDownload slide

Close modal