Background and objective. The aim of this study was to provide up-to-date cancer survival estimates using the Estonian Cancer Registry (ECR) database that was recently supplemented with previously missing information from death certifi cates.
Methods. Data on all cases of invasive cancerdiagnosed in adult patients in Estonia during 2000–2008 were obtained from the ECR. Death certifi cates only and autopsy cases were excluded. Five-year relative survival ratios (RSR) were calculated as the ratio of the observed survival of the cancer patients to the expected survival of the underlying general population; the latter was calculated using national life tables stratifi ed by sex, age and calendar year. Period estimates of relative survival for 2005–2009 are based on 44996 incident cases and presented for 26 cancer sites.
Results. THe highest 5-year RSRs were seen for non-melanoma skin cancer (around 100%), Hodgkin’s lymphoma (88%), thyroid cancer (85%), testicular cancer (78%), prostate cancer (77%), breast cancer (76%) and corpus uteri cancer (73%). The lowest 5-year RSRs were observed for pancreatic, liver and oesophageal cancer (3%, 4% and 5%, respectively).
The RSRs were signifi cantly higher among females compared with males for thyroid cancer (90% vs 64%), skin melanoma (74% vs 58%), kidney cancer (67 vs 57), and oral cancer (54% vs 32%).
Conclusions. The results showed a märkedvariability of the survival estimates across the cancer sites. In comparison with the EUNICE survival study, the relative survival for breast and prostate cancer and non-Hodgkin lymphoma exceeded predictions for 2005–2009; for many sites, however, the estimates remained lower than expected. For some sites, the 5-year RSRs in Estonia have reached the European average reported by the EUROCARE-4 study (1995–1999) or exceeded these; however, the Estonian estimates lag markedly behind for skin melanoma and cancers of urinary bladder, breast, corpusuteri, colon and rectum. Further studies should focus on examining survival by age and cancer stage, and continue to clarify the underlying mechanisms of survival variation using the high-resolution approach. Late detection is still an important problem for many sites and should be addressed by health care organizers.