Contributions
Abstract: PB2089
Type: Publication Only
Background
Patient clinical situation at diagnosis (Dx) of MDS is a function of general condition (GC) before Dx and the impact of the new disease. ECOG performance status probably depends on both of them, and consequently may be highly dependent on the very moment that it is measured. GC at baseline can be measured by Lee Index (LIx) for Older Adults (Lee SJ et al, 2006), a 12-item validated scale that has been shown to predict 4-year mortality among community-dwelling adults in USA and also to predict overall survival (OS) in a prospective and molecularly annotated cohort of Spanish MDS patients (Ramos F et al, 2017). To our knowledge, LIx has not been applied yet to MDS-related conditions (CMML and AML).
Aims
To analyze whether patient GC before Dx can be considered an independent determinant of OS in a heterogenous population of Pts diagnosed with MDS and related diseases.
Methods
We have reviewed the clinical charts of 132 pts (84 M, 48 F), median age 79y (IQR 73-82y), diagnosed in our center with MDS (n=88), CMML (17) or AML (27), whose GC had been evaluated by means of the LIx. The study was approved by our Institutional Review Board. All Pts were Dx and classified according to WHO 2008 and stratified prognostically (Px) according to disease-specific criteria (IPSS-R in MDS, CPSS in CMML and MRC/LRF scoring system in AML). ECOG was 0 in 23 Pts (17.4%), 1-2 in 93 (70.4%), 3-4 in 15 (11.4%) and NA in 1 (0.75%). Twenty-two Pts (16.7%) had received therapy known to prolong OS, while the rest had received only supportive care. After a median follow-up of 38.1 months (IQR 37.7-71.6), 70 Pts (53.0%) had died, 45 (34.1%) were alive, and 17 (12.9%) were lost to follow-up. The proportion of Pts that received disease-modifying therapy was similar across the Dx and Px groups. OS was evaluated by Log-rank tests (for trend, as appropriate) and Cox regression models.
Results
OS was longest for CMML Pts (median NR), intermediate for MDS (53.4 months, IQR 29.1-77.7) and shortest for AML (8.5, IQR 0-19.1), p<0.001 (Log-rank). As expected, it was also progressively shorter as the disease-specific Px categories were worse (p<0.001, Log-rank for trend). ECOG categories were strongly associated with both LIx and Px categories (Chi-Square for trend 10.84, p=0.001 and 7.94, p=0.006, respectively), but not to the Dx category (p=0.55).
LIx score ranged 0-19 points (median 8, IQR 6-10). Fifty-three Pts (40.2%) were included in the Q1 (score 0-5), 45 Pts (34.1%) in Q2 (6-9), 25 (18.9%) in Q3 (10-13) and 9 (6.8%) in Q4 (score 14+). As expected, LIx categories were associated neither with the Dx nor with the Px categories (p=0.15 and p=0.36, respectively). At first sight, we observed a non-significant univariate trend towards a shorter OS as the LIx was higher: Q1 68.2 months (IQR 25.6-110.9), Q2+Q3 45.2 (IQR 14.9-75.4) and Q4 16.2 (IQR 0-41.8); p=0.211, Log-rank for trend. Interestingly, multivariate models disclosed that LIx score was an independent determinant of OS in this heterogeneous population, whether the type of treatment is included in the model (p= 0.019) or not (p=0.034).
Conclusion
Patient general condition, as evaluated multidimensionally by the Lee Index for Older Adults, is an independent determinant of OS in patients with MDS, CMML and AML. This statement is now being evaluated in a larger and prospective patient cohort.
Session topic: 10. Myelodysplastic syndromes – Clinical
Keyword(s): Acute Myeloid Leukemia, Chronic myelomonocytic leukemia, Myelodysplasia, Survival
Abstract: PB2089
Type: Publication Only
Background
Patient clinical situation at diagnosis (Dx) of MDS is a function of general condition (GC) before Dx and the impact of the new disease. ECOG performance status probably depends on both of them, and consequently may be highly dependent on the very moment that it is measured. GC at baseline can be measured by Lee Index (LIx) for Older Adults (Lee SJ et al, 2006), a 12-item validated scale that has been shown to predict 4-year mortality among community-dwelling adults in USA and also to predict overall survival (OS) in a prospective and molecularly annotated cohort of Spanish MDS patients (Ramos F et al, 2017). To our knowledge, LIx has not been applied yet to MDS-related conditions (CMML and AML).
Aims
To analyze whether patient GC before Dx can be considered an independent determinant of OS in a heterogenous population of Pts diagnosed with MDS and related diseases.
Methods
We have reviewed the clinical charts of 132 pts (84 M, 48 F), median age 79y (IQR 73-82y), diagnosed in our center with MDS (n=88), CMML (17) or AML (27), whose GC had been evaluated by means of the LIx. The study was approved by our Institutional Review Board. All Pts were Dx and classified according to WHO 2008 and stratified prognostically (Px) according to disease-specific criteria (IPSS-R in MDS, CPSS in CMML and MRC/LRF scoring system in AML). ECOG was 0 in 23 Pts (17.4%), 1-2 in 93 (70.4%), 3-4 in 15 (11.4%) and NA in 1 (0.75%). Twenty-two Pts (16.7%) had received therapy known to prolong OS, while the rest had received only supportive care. After a median follow-up of 38.1 months (IQR 37.7-71.6), 70 Pts (53.0%) had died, 45 (34.1%) were alive, and 17 (12.9%) were lost to follow-up. The proportion of Pts that received disease-modifying therapy was similar across the Dx and Px groups. OS was evaluated by Log-rank tests (for trend, as appropriate) and Cox regression models.
Results
OS was longest for CMML Pts (median NR), intermediate for MDS (53.4 months, IQR 29.1-77.7) and shortest for AML (8.5, IQR 0-19.1), p<0.001 (Log-rank). As expected, it was also progressively shorter as the disease-specific Px categories were worse (p<0.001, Log-rank for trend). ECOG categories were strongly associated with both LIx and Px categories (Chi-Square for trend 10.84, p=0.001 and 7.94, p=0.006, respectively), but not to the Dx category (p=0.55).
LIx score ranged 0-19 points (median 8, IQR 6-10). Fifty-three Pts (40.2%) were included in the Q1 (score 0-5), 45 Pts (34.1%) in Q2 (6-9), 25 (18.9%) in Q3 (10-13) and 9 (6.8%) in Q4 (score 14+). As expected, LIx categories were associated neither with the Dx nor with the Px categories (p=0.15 and p=0.36, respectively). At first sight, we observed a non-significant univariate trend towards a shorter OS as the LIx was higher: Q1 68.2 months (IQR 25.6-110.9), Q2+Q3 45.2 (IQR 14.9-75.4) and Q4 16.2 (IQR 0-41.8); p=0.211, Log-rank for trend. Interestingly, multivariate models disclosed that LIx score was an independent determinant of OS in this heterogeneous population, whether the type of treatment is included in the model (p= 0.019) or not (p=0.034).
Conclusion
Patient general condition, as evaluated multidimensionally by the Lee Index for Older Adults, is an independent determinant of OS in patients with MDS, CMML and AML. This statement is now being evaluated in a larger and prospective patient cohort.
Session topic: 10. Myelodysplastic syndromes – Clinical
Keyword(s): Acute Myeloid Leukemia, Chronic myelomonocytic leukemia, Myelodysplasia, Survival