Adenocarcinoma of the prostate in Iceland: a population-based study of stage, Gleason grade, treatment and long-term survival in males diagnosed between 1983 and 1987.

OBJECTIVE: To investigate adenocarcinoma of the prostate in a single population with an extended follow-up period. MATERIAL AND METHODS: Using the Icelandic Cancer Registry, we identified all Icelandic men diagnosed with prostate cancer between 1983 and 1987. Disease stage, initial treatment and follow-up information were obtained from hospital records and death certificates. A critical evaluation was made of the accuracy of the death certificates regarding prostate cancer. All available histology information was reviewed and graded according to the Gleason grading system. RESULTS: A total of 414 men were diagnosed with adenocarcinoma of the prostate. Of these, 370 were alive at the time of diagnosis and stage could be determined. Four stage groups were defined: focal incidental (n=50); localized (n=164); local advanced (n=32); and metastatic disease (n=124). The mean age at diagnosis was 74.4 years (range 53-94 years). The combined Gleason score was 2-5 in 89, 6-7 in 117, 8-10 in 117 and unknown in 47 cases. The median follow-up period for the group was 6.15 years (range 0.3-19.8 years). Thirty men received treatment with curative intent: radiation therapy, n=20; and radical prostatectomy, n=10. A total of 334 patients died during the follow-up period, of whom 168 (50%) died of prostate cancer. Prostate cancer-specific survival at 10 and 15 years was 100% and 90.6%, respectively for focal incidental cancer; 73.1% and 60.8% for men with localized disease; 23.4% and 11.7% for local advanced disease; and 6.81% and 5.45% for metastatic disease. A Cox multivariate analysis showed age, stage and Gleason score to be independent predictors of prostate cancer death. A total of 104 patients with localized disease and a Gleason score of

Breast and ovarian cancer risks to carriers of the BRCA1 5382insC and 185delAG and BRCA2 6174delT mutations: a combined analysis of 22 population based studies.

A recent report estimated the breast cancer risks in carriers of the three Ashkenazi founder mutations to be higher than previously published estimates derived from population based studies. In an attempt to confirm this, the breast and ovarian cancer risks associated with the three Ashkenazi founder mutations were estimated using families included in a previous meta-analysis of populatrion based studies. The estimated breast cancer risks for each of the founder BRCA1 and BRCA2 mutations were similar to the corresponding estimates based on all BRCA1 or BRCA2 mutations in the meta-analysis. These estimates appear to be consistent with the observed prevalence of the mutations in the Ashkenazi Jewish population.

Cosmic radiation and cancer mortality among airline pilots: results from a European cohort study (ESCAPE).

Cosmic radiation is an occupational risk factor for commercial aircrews. In this large European cohort study (ESCAPE) its association with cancer mortality was investigated on the basis of individual effective dose estimates for 19,184 male pilots. Mean annual doses were in the range of 2-5 mSv and cumulative lifetime doses did not exceed 80 mSv. All-cause and all-cancer mortality was low for all exposure categories. A significant negative risk trend for all-cause mortality was seen with increasing dose. Neither external and internal comparisons nor nested case-control analyses showed any substantially increased risks for cancer mortality due to ionizing radiation. However, the number of deaths for specific types of cancer was low and the confidence intervals of the risk estimates were rather wide. Difficulties in interpreting mortality risk estimates for time-dependent exposures are discussed.

Familial risk of prostate cancer in Iceland.

OBJECTIVE: To estimate the risk of prostate and other types of cancer among relatives of Icelandic men diagnosed with prostate cancer over a 5-year period. PATIENTS AND METHODS: The risk ratio (RR) was used to estimate the risk among relatives of 371 patients with prostate cancer, all of whom lived in Iceland and were diagnosed when alive over a 5-year interval (1983-7). Information on cancer incidence was obtained from the population-based Icelandic Cancer Registry, and information on families from a comprehensive genealogical database covering the population of Iceland. RESULTS: First-degree male relatives were at a 1.7-fold greater age-adjusted risk of prostate cancer (1832 men; 95% confidence interval, CI, 1.28-2.34). The risk was independent of proband's age at diagnosis. First-degree male relatives of patients who died from prostate cancer were at a statistically significantly greater risk of the disease (784 men; RR 2.17; 95% CI 1.34-3.53) and relatives of patients with incidental disease (T1a) were at a greater risk but not statistically significant so (261; RR 1.86; 95% CI 0.75-4.58). Female first-degree relatives were not at greater risk of breast cancer. The risk of kidney cancer was higher in first- and second-degree female relatives, with an RR (n, CI) of 2.50 (1780, 1.10-5.66) and 2.67 (5534, 1.04-6.81), respectively. The risk of kidney cancer was not statistically significantly greater in male relatives. CONCLUSION: Family history is a risk factor for prostate cancer in Icelandic men. The risk is potentially higher for relatives of patients who die from the disease. Female relatives are not at greater risk of breast cancer but they may be at greater risk of kidney cancer.

Breast cancer risk in airline cabin attendants: a nested case-control study in Iceland.

AIMS: To investigate whether length of employment as a cabin attendant was related to breast cancer risk, when adjusted for reproductive factors. METHODS: Age matched case-control study nested in a cohort of cabin attendants. The cases were found from a nationwide cancer registry (followed up to end of year 2000) and the reproductive factors (age at first childbirth and number of children) from a registry of childbirth, in both instances by record linkage with the cabin attendants' identification numbers. The employment time of the cabin attendants at the airline companies and the reproductive factors had been systematically recorded prior to the diagnosis of breast cancer in the cohort. A total of 35 breast cancer cases and 140 age matched controls selected from a cohort of 1532 female cabin attendants were included in the study. RESULTS: The matched odds ratio from conditional logistic regression of breast cancer risk among cases and controls of cabin attendants was 5.24 (95% CI 1.58 to 17.38) for those who had five or more years of employment before 1971 compared with those with less than five years of employment before 1971, adjusted for age at first childbirth and length of employment from 1971 or later. CONCLUSIONS: The association between length of employment and risk of breast cancer, adjusted for reproductive factors, indicates that occupational factors may be an important cause of breast cancer among cabin attendants; the association is compatible with a long induction period.

Risk factors for cutaneous malignant melanoma among aircrews and a random sample of the population.

AIMS: To evaluate whether a difference in the prevalence of risk factors for malignant melanoma in a random sample of the population and among pilots and cabin attendants could explain the increased incidence of malignant melanoma which had been found in previous studies of aircrews. METHODS: A questionnaire was used to collect information on hair colour, eye colour, freckles, number of naevi, family history of skin cancer and naevi, skin type, history of sunburn, sunbed, all sunscreen use, and number of sunny vacations. RESULTS: The 239 pilots were all males and there were 856 female cabin attendants, which were compared with 454 males and 1464 females of the same age drawn randomly from the general population. The difference in constitutional and behavioural risk factors for malignant melanoma between the aircrews and the population sample was not substantial. The aircrews had more often used sunscreen and had taken more sunny vacations than the other men and women. The predictive values for use of sunscreen were 0.88 for pilots and 0.85 for cabin attendants and the predictive values for sunny vacation were 1.36 and 1.34 respectively. CONCLUSION: There was no substantial difference between the aircrew and the random sample of the population with respect to prevalence of risk factors for malignant melanoma. Thus it is unlikely that the increased incidence of malignant melanoma found in previous studies of pilots and cabin attendants can be solely explained by excessive sun exposure.

Average risks of breast and ovarian cancer associated with BRCA1 or BRCA2 mutations detected in case Series unselected for family history: a combined analysis of 22 studies.

Germline mutations in BRCA1 and BRCA2 confer high risks of breast and ovarian cancer, but the average magnitude of these risks is uncertain and may depend on the context. Estimates based on multiple-case families may be enriched for mutations of higher risk and/or other familial risk factors, whereas risk estimates from studies based on cases unselected for family history have been imprecise. We pooled pedigree data from 22 studies involving 8,139 index case patients unselected for family history with female (86%) or male (2%) breast cancer or epithelial ovarian cancer (12%), 500 of whom had been found to carry a germline mutation in BRCA1 or BRCA2. Breast and ovarian cancer incidence rates for mutation carriers were estimated using a modified segregation analysis, based on the occurrence of these cancers in the relatives of mutation-carrying index case patients. The average cumulative risks in BRCA1-mutation carriers by age 70 years were 65% (95% confidence interval 44%-78%) for breast cancer and 39% (18%-54%) for ovarian cancer. The corresponding estimates for BRCA2 were 45% (31%-56%) and 11% (2.4%-19%). Relative risks of breast cancer declined significantly with age for BRCA1-mutation carriers (P trend.0012) but not for BRCA2-mutation carriers. Risks in carriers were higher when based on index breast cancer cases diagnosed at <35 years of age. We found some evidence for a reduction in risk in women from earlier birth cohorts and for variation in risk by mutation position for both genes. The pattern of cancer risks was similar to those found in multiple-case families, but their absolute magnitudes were lower, particularly for BRCA2. The variation in risk by age at diagnosis of index case is consistent with the effects of other genes modifying cancer risk in carriers.

The effect of a single BRCA2 mutation on cancer in Iceland.

OBJECTIVE: To estimate the risk of malignant diseases in families of probands with the same mutation in the BRCA2 gene. DESIGN: A cohort study using record linkage of a breast cancer family resource and the Icelandic Cancer Registry. SETTING: Iceland. SUBJECTS: Families of 995 breast cancer patients, from which 887 were tested for a single founder 999del5 mutation; 90 had the mutation and 797 did not. RESULTS: Relatives of probands with the mutation had significantly increased relative risk (RR) of breast cancer. For first degree relatives, the RR was 7.55 (95% CI 6.04 to 9.03) but was 1.72 (95% CI 1.49 to 1.96) in first degree relatives of probands without the mutation. For prostate and ovarian cancer, the first and second degree relatives of probands with the mutation had a significantly increased RR, but in families of probands without the mutation no significant familial risk was found. CONCLUSIONS: The 999del5 mutation in the BRCA2 gene explains a substantial proportion of familial risk of breast cancer in Iceland, but significant familial risk remains in relatives of probands without the mutation. For prostate and ovarian cancer, the mutation accounts for most of the familiality observed in families of breast cancer patients.

Cancer in siblings of children with cancer in the Nordic countries: a population-based cohort study.

BACKGROUND: In some rare inherited disorders such as Li-Fraumeni syndrome, relatives of children with cancer are at increased risk of cancer. We aimed to assess relations between childhood cancer and sibling risk, and evaluate the influence of recessive conditions in cancer causation. METHODS: We did a population-based cohort study in the Nordic countries of 42277 siblings of 25605 children with cancer. Children with cancer were identified from records in the five Nordic cancer registries, and their siblings from nationwide population registries. Cancers in siblings were documented through record linkage with cancer registries and compared with national incidence rates. We also assessed cancer incidence in parents to identify familial cancer syndromes. FINDINGS: 284.2 cancers were expected in siblings, whereas 353 were diagnosed (standardised incidence ratio 1.24 95% CI 1.12-1.38). Risk ratios for siblings were highest in the first decade of life (2.59, 1.89-3.46). We excluded 56 families with genetic syndromes linked to cancer, which reduced this ratio from 1.7 to 1.0 (0.7-1.3) for siblings younger than 20 years, and from 1.3 to 1.0 (0.8-1.3) for those aged 20-29 years. We found no new patterns of familial cancer that indicated inherited susceptibility, or evidence that recessive conditions might contribute to cancers not explained by syndromes. 40% of cancers in siblings that occurred before age 20 years could be attributed to known genetic factors, whereas 60% remained unexplained. INTERPRETATION: Apart from rare cancer syndromes, paediatric cancer is not an indicator of increased cancer risk in siblings.

Decreasing late mortality among five-year survivors of cancer in childhood and adolescence: a population-based study in the Nordic countries.

PURPOSE: To assess the risk of death in patients who survive more than 5 years after diagnosis of childhood cancer and to evaluate causes of death in fatal cases. PATIENTS AND METHODS: This was a population-based study in the five Nordic countries (Denmark, Finland, Iceland, Norway, and Sweden) using data of the nationwide cancer registries and the cause-of-death registries. The study cohort included 13,711 patients who were diagnosed with cancer before the age of 20 years between 1960 and 1989 and who survived at least 5 years from diagnosis. By December 31, 1995, 1,422 patients had died, and death certificates were assessed in 1,402. Standardized mortality ratios (SMRs) for validated causes of death were calculated based on 156,046 patient-years at risk. RESULTS: The overall SMR was 10.8 (95% confidence interval [CI], 10.3 to 11.5), mainly due to high excess mortality from the primary cancer. SMR for second cancer was 4.9 (95% CI, 3.9 to 5.9) and was 3.1 (95% CI, 2.8 to 3.5) for noncancer death. The pattern of causes of death varied markedly between different groups of primary cancer diagnoses and was highly dependent on time passed since diagnosis. Overall late mortality was significantly lower in patients treated during the most recent period of time, 1980 to 1989, compared with those treated from 1960 to 1979 (hazard ratio, 0.61; 95% CI, 0.54 to 0.70), and there was no increase in rates of death due to cancer treatment. CONCLUSION: Long-term survivors of childhood cancer had an increased mortality rate, mainly dying from primary cancers. However, modern treatments have reduced late cancer mortality without increasing the rate of therapy-related deaths.

Breastfeeding and reduced risk of breast cancer in an Icelandic cohort study.

Case-control studies on the association between breastfeeding and the subsequent risk of breast cancer have given inconsistent results. To date, only two cohort studies have been reported on this subject. The present nested case-control study uses data from an Icelandic cohort of 80,219 women visiting a Cancer Detection Clinic that offers population-based cervical and breast cancer screening, in the years 1979-1995. The 993 parous cases were aged 26-90 years at diagnosis, with 9,729 parous controls individually matched on birth year, vital status at case diagnosis, and age when giving information on several potential risk factors for breast cancer. Using conditional logistic regression and confining the analysis to the 84 cases who were <40 years at diagnosis, an inverse association was evident between total duration of breastfeeding and breast cancer, with the adjusted odds ratio = 0.77 per 6 months' increase in duration of breastfeeding (95% confidence interval: 0.59, 1.00), whereas for the remainder of the women, a much weaker trend was observed. Ever lactating was associated with decreased risk, with the adjusted odds ratio = 0.33 (95% confidence interval: 0.19, 0.56) for women diagnosed at all ages. This is the first cohort study to indicate a negative association between breastfeeding and breast cancer.

Risk of breast cancer in female flight attendants: a population-based study (Iceland).

OBJECTIVES: To study whether increased cancer risk, particularly of cancer types previously related to radiation, was found among cabin attendants, using employment time as a surrogate of exposure to cosmic radiation. METHODS: A cohort of 1690 cabin attendants, 158 men and 1532 women from the Icelandic Cabin Crew Association and two airline companies in Iceland, was established. Cancer sites were ascertained between 1955 and 1997 by follow-up in a cancer registry. The personal identification number of each subject was used in record linkage to population-based registers containing vital and emigration status, reproductive factors and histologically verified cancer diagnosis. Standardized incidence rates (SIR) of different cancer sites in relation to employment time and year of hiring were calculated, as well as predictive values of breast cancer risk for evaluating possible confounding due to reproductive factors. RESULTS: The total number of person-years was 27,148. Among the women, 64 cancers were observed whereas 51.63 were expected (SIR 1.2, 95% CI 1.0-1.6), and significantly increased risk for malignant melanoma (SIR 3.0, 95% CI 1.2-6.2) was found. Significantly increased risks of overall cancers (SIR 1.3, 95% CI 1.0-1.8) and breast cancer (SIR 1.6, 95% CI 1.0-2.4) were observed among the female cabin attendants when 15 years lag time was applied. Those hired in 1971 or later had the heaviest exposure to cosmic radiation at a young age and had significantly increased risk of overall cancer (SIR 2.8, 95% CI 1.4-4.9) and breast cancer (SIR 4.1, 95% CI 1.7-8.5). Predictive values calculated on the basis of reproductive factors among the cabin attendants and the population, and risk of breast cancer were 1.0 for parous vs. nulliparous, 1.0 for number of children, and 1.1 for age at birth of first child. CONCLUSION: The increased risk of breast cancer and malignant melanoma among cabin attendants seems to be occupationally related. The part played by occupational exposures, i.e. cosmic radiation, disturbance of the circadian rhythm, and electromagnetic fields or combination of these factors in the etiology of breast cancer among the cabin crew, is still a puzzle as confounding due to parity appears to be ruled out. The relationship between the sunbathing habits of the cabin crew and the increased risk of malignant melanoma needs to be clarified. There is also an urgent need to elucidate the importance of these findings for today's aviation.

[Hepatocellular carcinoma in Iceland.].

INTRODUCTION: The incidence of hepatocellular carcinoma (HCC) varies throughout the world, being relatively low in Northern Europe (less than five per 100,000 population) where the majority of the patients have cirrhosis. In Iceland the prevalence of viral hepatitis and cirrhosis, the main risk factors for HCC, is lower than reported in many other countries. The aim of our study was to investigate the incidence and etiology of HCC in Iceland. MATERIAL AND METHODS: All patients diagnosed with HCC in Iceland in 1984-1998 were included in the study. Histologic diagnosis was required for inclusion. Patients were identified from the Icelandic Cancer Registry and by reviewing autopsy and histopathology reports. Further information was obtained from medical records. RESULTS: A total of 71 cases of HCC were identified, 51 males and 20 females. The mean age for males was 69.3 years (18-95) and 73 years (52-89) for females. The age-standardized annual incidence rate of HCC was 1.08/100,000 (males 2.10, females 0.67). The incidence did not increase significantly during the study period. Alcohol abuse (15.5%) and hemochromatosis (11%) were the most common risk factors. Twenty-three (32%) had cirrhosis but 39 (55%) had no known risk factors. Of 55 cases where non-neoplastic tissue was available for examination, 27 had liver disease. CONCLUSIONS: 1) The incidence of HCC in Iceland is lower than reported in other countries. 2) Alcohol abuse and hemochromatosis are the most common risk factors. 3) The ratio of patients with cirrhosis is low.

Helicobacter pylori antibodies and gastric cancer in Iceland - The decline in IgG antibody level is a risk factor.

H. pylori infection is considered a causal agent of duodenal ulcer and a significant risk factor for gastric cancer. Retrospective cohort studies have demonstrated a significant association between presence of antibody to H. pylori and gastric cancer when using samples obtained years before the diagnosis but not at the time of diagnosis. The present study investigates, in a population-based cohort, whether a decline occurs in H. pylori antibody levels before the diagnosis of stomach cancer. Repeat samples (2 to 5) were available from 23 persons with gastric cancer taken up to 20 years before the diagnosis and 128 control subjects matched for gender, age, time and number of repeat samples. The odds ratio of developing stomach cancer was 1.16 (95% CI 1.05-1.28) for those showing decline in antibody levels of 1 relative antibody activity unit per year versus those with constant or rising levels. We conclude that this decline in antibody levels in cases, and not in controls, supports an active role of H. pylori in the pathogenesis of gastric cancer by causing atrophic gastritis, and provides a better risk assessment for gastric cancer compared to single measurements.

Second malignant neoplasms after cancer in childhood and adolescence: a population-based case-control study in the 5 Nordic countries. The Nordic Society for Pediatric Hematology and Oncology. The Association of the Nordic Cancer Registries.

Our purpose was to assess the risk of developing a second malignant neoplasm (SMN) after cancer in childhood and adolescence associated with different treatment modalities. Our investigation was performed as a nested case-control study within a Nordic cohort of 25,120 patients younger than 20 years old at first malignant neoplasm (FMN) diagnosed in 1960 through 1987. SMNs were diagnosed in 1960 through 1991. For each case of SMN, 3 controls were sampled, matched by sex, age, calendar year of diagnosis and length of follow-up. For the final analysis, there were 234 cases and 678 controls. Relative risks (RRs) of various exposures were estimated by means of conditional logistic regression, with non-exposed as the reference. The RR of developing SMN in the radiated volume was 4.3 (95% confidence interval 3.0-6.2). The risk was highest in children diagnosed before the age of 5 years; it increased with the dose of radiation and with increasing follow-up time after FMN. Chemotherapy alone was not associated with an increased RR, but it significantly potentiated the effect of radiotherapy. RRs were unchanged between the periods 1960-1973 and 1974-1987, and since the use of chemotherapy increased in the latter period, the number of SMNs may increase. Hereditary factors were important for the occurrence of SMN independently of therapy. We conclude that radiation was the most important treatment-related risk factor for the development of SMN. Chemotherapy appeared to play only an accessory role during the study period, potentiating the carcinogenic effect of radiotherapy.

Incidence of cancer among commercial airline pilots.

OBJECTIVES: To describe the cancer pattern in a cohort of commercial pilots by follow up through the Icelandic Cancer Registry. METHODS: This is a retrospective cohort study of 458 pilots with emphasis on subcohort working for an airline operating on international routes. A computerised file of the cohort was record linked to the Cancer Registry by making use of personal identification numbers. Expected numbers of cancer cases were calculated on the basis of number of person-years and incidences of cancer at specific sites for men provided by the Cancer Registry. Numbers of separate analyses were made according to different exposure variables. RESULTS: The standardised incidence ratio (SIR) for all cancers was 0.97 (95% confidence interval (95% CI) 0.62 to 1.46) in the total cohort and 1.16 (95% CI 0.70 to 1.81) among those operating on international routes. The SIR for malignant melanoma of the skin was 10.20, 95% CI 3.29 to 23.81 in the total cohort and 15.63, 95% CI 5.04 to 36.46 in the restricted cohort. Analyses according to number of block-hours and radiation dose showed that malignant melanomas were found in the subgroups with highest exposure estimates, the SIRs were 13.04 and 28.57 respectively. The SIR was 25.00 for malignant melanoma among those who had been flying over five time zones. CONCLUSIONS: The study shows a high occurrence of malignant melanoma among pilots. It is open to discussion what role exposure of cosmic radiation, numbers of block-hours flown, or lifestyle factors--such as possible excessive sunbathing--play in the aetiology of cancer among pilots. This calls for further and more powerful studies. The excess of malignant melanoma among those flying over five time zones suggests that the importance of disturbance of the circadian rhythm should be taken into consideration in future studies.

Serum thyroglobulin as a risk factor for thyroid carcinoma.

Samples from a biological serum bank taken up to 23 years prior to diagnosis of thyroid carcinoma were analysed for human thyroglobulin, thyroid-stimulating hormone and thyroxin. After exclusions, the final study material consisted of 59 cases of papillary and follicular carcinomas. These cases were compared with 164 controls, matched for sex, age and time of sample taking. The most interesting finding was that concentrations of thyroglobulin in serum were abnormally elevated in cases compared with controls, equal to or above 30 microg/L, with odds ratio 7.0 (CI 3.1-15.7). This elevation of serum thyroglobulin occurred in 44% of the carcinoma cases. Sensitivity was around 50 for measurements taken up to 15 years prior to diagnosis, but 21 when the interval was over 15 years. Specificity was 89. No differences were found between cases and controls in values for thyroid-stimulating hormone and thyroxin.

[The effect of pregnancy on the survival of women diagnosed with breast cancer.].

OBJECTIVE: To answer the question whether the prognosis of women with breast cancer is affected by pregnancy after diagnosis. MATERIAL AND METHODS: We used information from the Icelandic Cancer Registry, the Birth Registry and the Genetic Committee. We identified all women who were diagnosed with breast cancer in the years 1927-1992 and who later became pregnant. Controls were women without a history of childbirth after diagnosis of breast cancer. They were matched on tumour size, axillary lymph node status and years of birth and diagnosis, with four year deviation. RESULTS: In the years 1927-1992, 838 women at ages below 50 were diagnosed with breast cancer in Iceland. Of those, 29 gave birth to a child after the diagnosis. Fourteen cases and 33 matched controls fulfilled the inclusion criteria of the study. Survival was better in the group of women who became pregnant after diagnosis, but the difference was not statistically significant (P=0.06). DISCUSSION: Our results do not indicate that the prognosis of women who become pregnant after the diagnosis of breast cancer is worse than of those who do not become pregnant. The group was too small to make definite conclusions. However, the results are in concordance with results from other studies.

[Liver metastases of unknown origin.].

OBJECTIVE: Approximately 5% of cancer patients are diagnosed with tumour of unknown origin (3-4% in Iceland). Of those 10-30% have liver metastases. Liver metastases of unknown origin is thus not an uncommon problem. In the present study information about the origin and histology of liver metastases of unknown origin was compiled. MATERIAL AND METHODS: Records of all biopsies from liver metastases performed in the years 1987-1996 were retrieved from the medical database of the Department of Pathology at the University of Iceland. The biopsies came from a group of 176 patients. Ninety-two cases, in which the origin of the primary tumour was suspected or known, were excluded from the study, leaving 84 cases where the primary was completely unknown. The database of the Icelandic Cancer Society was used to gather data about the final tissue diagnosis and the location of the primary tumour when known. RESULTS: The Cancer Society data revealed the location of the primary tumour in 55 of the 84 cases of liver metastases of unknown origin. The most prevalent (75%) primary tumours were cancers of the pancreas (15), lung (13) and colon/rectum (12). The tissue diagnosis was adenocarcinoma in 33 of the 55 cases. In the male patients 83% of the adenocarcinoma metastases came from the colon/rectum or pancreas. The corresponding figure for the female patients was 67%, while 20% of the tumours in females originated in the gallbladder and biliary tree. CONCLUSIONS: In two thirds of the cases of liver metastases of unknown origin the primary tumour was later discovered. The most prevalent tumours were cancers in the pancreas, lung and colon/rectum. Adenocarcinoma was the tissue diagnosis in 60% of cases.