Temporal changes in survival among adult patients with acute myeloid leukaemia in the period 2000–2016: a Danish population-based study

Data S1. Supplementary methods.

Table S1. Results from two Cox models (i) adjusting for age and (ii) adjusting for age, Eastern Cooperative Oncology Group (ECOG) Performance Status and AML subtype (de novo, secondary AML, therapy-related AML). Age was treated as a continuous variable and modelled using natural cubic splines with three degrees of freedom. The likelihood ratio test indicates whether a temporal change in overall survival (OS) was observed. The null model in the likelihood ratio test included the variables in the reported model except the diagnostic time period.

Table S2. Results from three Cox models investigating (i) sex difference in overall survival (OS), (ii) sex differences in OS adjusted for age, and (iii) sex differences in the temporal improvement of OS. The likelihood ratio test was conducted by comparing to a model without an interaction between sex and calendar period.

Table S3. Results from Cox proportional hazards models stratified on cytogenetic risk group. For each risk group, a univariable (including only diagnostic time period) and adjusted (also including age at diagnosis modelled by a natural cubic spline) Cox model is fitted and a likelihood ratio test is performed. The null model in the likelihood ratio test included the variables in the reported model except time period of diagnosis. P values for each likelihood ratio test are reported in the table. LRT, likelihood ratio test.

Table S4. Hazard ratios comparing mortality in different time periods for adult patients with AML treated with intensive chemotherapy. Age was treated as a continuous variable and modelled using natural cubic splines with three degrees of freedom. The likelihood ratio test indicates whether a temporal change in overall survival (OS) was observed. The null model in the likelihood ratio test included the variables in the reported model except time period of diagnosis.

Table S5. Results from Cox proportional hazards models stratified on treatment group. For each treatment group, a univariable (including only diagnostic time period) and adjusted (also including age at diagnosis modelled by a natural cubic spline) Cox model is fitted and a likelihood ratio test is performed. The null model in the likelihood ratio test included the variables in the reported model except time period of diagnosis. P values for each likelihood ratio test are reported in the table. LRT, likelihood ratio test.

Table S6. Degrees of freedom for all natural cubic splines used to model 2-year overall survival (OS).

Fig S1. (A) The 2-year crude and age-standardised relative survival against year of diagnosis for adult Danish patients with AML. (B) Age-stratified 2-year relative survival as function of diagnosis year. The shaded areas indicate pointwise 95% confidence intervals.

Fig S2. The 2-year age-standardised overall survival against year of diagnosis for adult Danish patients with AML stratified according to cytogenetic risk profile. Modelling the 2-year age-standardised overall survival for patients with a favourable cytogenetic risk profile was not feasible due to the limited sample size.

Fig S3. Temporal overall survival for Danish patients with AML diagnosed in 2000–2016 and treated with intensive chemotherapy within 30 days of diagnosis.

Fig S4. The 2-year overall survival for patients with AML treated with intensive chemotherapy stratified by age at diagnosis. Results for patients aged >75 years were excluded due to the small number of patients treated with intensive chemotherapy in this age group. The shaded areas indicate pointwise 95% confidence intervals.

Fig S5. The 2-year crude and age-standardised survival for patients receiving ASCT after first CR following intensive chemotherapy. The age-standardised survival is based on the age distribution of all transplanted patients who were diagnosed in the period 2007–2016.

Fig S6. (A) The 2-year crude and age-standardised overall survival (OS) against year of diagnosis for adult Danish patients with AML with ≥30% blasts in blood or bone marrow at diagnosis. (B) Age-stratified 2-year OS as function of diagnosis year. The shaded areas indicate pointwise 95% confidence intervals.

Fig S7. Temporal overall survival estimates for patients with AML with ≥30% blasts in blood or bone marrow at diagnosis. P values were obtained from log-rank tests.