ESTIMATION OF THE CURRENT SURVIVAL MEASURES IN CHRONIC MYELOID LEUKEMIA: METHODOLOGY AND NEW SOFTWARE TOOLS

Konference: 2013 18th Congress of the European Hematology Association - účast ČR

Kategorie: Maligní lymfomy a leukémie

Téma: Chronic myeloid leukemia - Biology

Číslo abstraktu: P130

Autoři: prof. MUDr. Jiří Mayer, CSc.; RNDr. Tomáš Pavlík, Ph.D.; RNDr. Eva Janoušová; MUDr. Hana Klamová, CSc.; MUDr. Daniela Žáčková, Ph.D.; Prof. MUDr. Zdeněk Ráčil, Ph.D.; Mgr. Kateřina Machová (Poláková), Ph.D.; Mgr. Veronika Némethová; RNDr. Jana Březinová, Ph.D.; prof. MUDr. Petr Cetkovský, Ph.D.; Prof. RNDr. Ladislav Dušek, Ph.D.

Background:

Common ways of survival assessment, the leukemia-free survival (LFS) and the cumulative incidence (CI), are not fully comprehensive for the outcome assessment in chronic myeloid leukemia (CML) because these measures do not account for multiple relapses and leukemia-free periods during the treatment course. Therefore, the concept of the so-called current survival measures, which accounts for the proportion of patients who have lost the first disease remission as well as the proportion of leukemia-free patients who achieved the subsequent remissions, has recently become discussed in the literature. Proper estimation methods and publicly available software tools are needed to make the current survival measures widely usable.

Aims:

To demonstrate the validity of the current survival measures. To introduce the methodology, software, and web-based calculator for the CCI and CLFS estimation.

Methods:

The current survival measures, the CCI and CLFS, were estimated using nonparametric statistical methods, which are commonly used in survival analysis (Pavlik et al, BMC Med Res Methodol. 2011; 11:140). R software package currentSurvival (http://cran.r-project.org/web/packages/currentSurvival/) was also made publicly available for the CCI and CLFS calculation. Moreover, a web-based calculator at http://www.iba.muni.cz/data-analysis-tools/currentSurvival/ will be launched on May 1st 2013. In total, 233 Czech CML patients in chronic phase received the first-line imatinib between July 2003 and December 2011; records were registered in the Czech database INFINITY (http://www.leukemia-cell.org/en/databaze/).

Results:

Regarding all 233 patients, the estimated CCI at 3 and 5 years after starting imatinib therapy was 75.4% and 73.4%, respectively. On the other hand, the common CI at 3 and 5 years after starting imatinib was estimated as 85.2% and 87.1%, respectively. Thus, the estimated difference between the CCI and CI curves reached 9.8% and 13.7% at 3 and 5 years after starting imatinib, respectively. Only 185 patients (79.4%) who achieved at least one CCgR were available for the CLFS calculation. The estimated CLFS at 3 and 5 years after achieving the first CCgR was 90.9% and 92.8%, respectively. The LFS was estimated as 74.2% and 64.0% at 3 and 5 years after achieving the first CCgR, respectively. Therefore, at 3 and 5 years after the achievement of the first CCgR, the difference between the CLFS and LFS estimates reached 16.7% and 28.8%, respectively.

Summary / Conclusion:

The common CI overestimates the probability of being alive and in CCgR after initiating imatinib therapy, whereas the common LFS underestimates the probability of being alive and in CCgR after the achievement of first CCgR on the imatinib therapy. Thus, both current survival measures, the CCI and CLFS, more reliably illustrate a CML patient’s disease status in time because they account for multiple leukemia-free periods during the treatment course. Moreover, the methodology for CCI and CLFS estimation is now available for public use in the form of either R software package currentSurvival or web-based calculator.

Abstrakta v časopise Haematologica 2013, Suppl1

Online Program

Datum přednesení příspěvku: 14. 6. 2013