Aims: Nationwide, computer-based survey of all total joint arthroplasties performed in Finland has been carried out since January 1980. From these records a cohort of 9,443 patients, with 80,734 person-years, after primary operation with a total polyethylene-on-metal knee arthroplasty (TKA) has been followed up for cancer. In 1999 we published the material until December 31, 1996. The standardized incidence ratio (SIR) for all cancers was 0.98. The SIRs for non-Hodgkin lymphoma (1.40), Hodgkinñs disease (1.24), multiple myeloma (1.54) and prostate cancer (1.49) were increased, but only that of non-Hodgkin lymhoma was statistically signiþcant 3 to 10 years after the operation. The overall cancer risk after TKA done for primary osteoarthrosis appeared to be negligible. Material and methods: Follow-up for cancer was undertaken using the þles in the population-based, nationwide Finnish Cancer Registry, employing personal identiþcation numbers. Follow-up for cancer started at the date of þrst knee replacement and ended on emigration, death, or 31 December 2000, whichever occurred þrst. No subject was lost to follow-up. Standardized incidence ratios (SIRs) were calculated by dividing observed numbers of cases by expected numbers. Results: There were 2,001 men and 7,442 women left in the cohort to be followed-up. Numbers of person-years are now 15,679 and 650,552, respectively. The mean duration of follow-up per person was therefore 8,5 years. During the 21-year follow-up period, 358 cases of cancer occurred in the male subjects. The expected number was 349. In the female subjects 801 cases of cancer were observed vs. 804 expected. Combining data for men and women, the SIR for cancer at all sites was 1.00. SIRs were signiþ-cantly lower in relation to cancers of the lung (0.69), and colon (0.77). The incidence of risk of colon cancer was lower than would have been expected only in women (SIR 0.70). As far as cancers of the urinary organs is concerned, the SIR was within unity (SIR 1,03). The overall incidence of non-Hodgkin lymphoma (extranodal sites included) was elevated, but without any statistical signiþ-cance (SIR 1.22). The SIR values for multiple myeloma and leucemia have reduced to 0.80 and 0.53. Conclusions: We conclude, that this large nation-wide material could not establish any association between TKA and the all-over incidence of remote or adjacent cancers; on the contrary incidences of certain cancers were in fact lower than those reported in the formed material with four years less follow-up. All the alarming SIRs in the former material, especially those with non-Hodgin lym-phoma, urinary tract cancers and leucemia, have now reduced and are less than unity.

Aims: In 1999 we published a cohort of 24,638 polyethylene-on-metal total hip arthroplasty (THA) patients followed up for cancer, using Finnish Cancer Registry data, from 1980 to 1995. The number of person-years was then 173,022 (until 31st Dec, 1996). During follow-up, there were statistically significantly fewer cancers among the THA patients (standardized incidence ratio [SIR], 0.91; 95% confidence interval [Cl], 0.87–0.94). There was no significantly increased risk at any site, and for certain cancers that was even below the unity (lung and stomach).

On the longer run, however, certain tendency for increased risk for cancer of the urinary bladder, myeloma, and leukemia could be observed; SIRs were greater than unity with the THA patients followed up 3 to 9 years. Further follow-up of the cohort is therefore needed. Methods: The follow-up of the same cohort, originally identified in the National Register of Arthroplasties, maintained by the National Agency for Medicines (primary THA with primary arthritis as the indication) was expanded with a four year period (from 1st of Jan 1997 to 31st of Dec 2000). Follow-up for cancer was undertaken using the files in the population-based, nationwide Finnish Cancer Registry, employing personal identification numbers.

Follow-up for cancer started at the date of first hip replacement and ended on emigration, death, or December 31, 2001, whichever occurred first. Multiple cancers were taken account of in similar ways in relation to observed and expected numbers of cases. Results: After excluding revision and infective or systemic disease as indications for operation, there were 9,479 men and 15,157 women in the cohort followed. The updated numbers for person-years were 89,295 for males and 153,759 for females. The mean duration of follow-up was now 9,9 years. The total risk for cancer was now 0.93 (95%Cl 0.90–0.96) and that for stomach 0.89, for colon 0.90, lung 0.64, for urinary organs 1.01, and for connective tissues 0.88. The SIRs for non-Hogkin lymphoma, myeloma and leucemia were all under the untity, 0.88, 1.09 and 0.73, respectively, without any tendency for increased risk in the follow-up of plus ten years.

Conclusions: These findings indicate that the risk of hematopoietic cancers is not increased after THA using polyethylene-on-metal prostheses. Expanding of the follow-up with four years did not markedly change the profile of the SIRs at any specific site, and contrary to our earlier report there were increased risk left of myeloma and leucemia with increased follow-up time. The SIRs associated with soft tissue cancers and bone sarcomas were not significantly different from unity. No sarcomas developed at the site of a prosthesis.