West Midlands Oncology Association trial of adjuvant chemotherapy in node-negative breast cancer.

Between 1976 and 1984, 574 patients with operable breast cancer and histologically negative axillary lymph nodes were randomly assigned after mastectomy to receive either no further treatment or chemotherapy with oral LMF (fluorouracil, 500 mg, methotrexate, 25 mg, and chlorambucil, 10 mg, on day 1; fluorouracil, 500 mg, and chlorambucil, 10 mg, on day 2). There is no overall survival or relapse-free survival benefit at a median follow-up of 10 years and 8 years, respectively. There are significantly more local relapses in the control group (P less than .01), but an excess of distant relapses in the treated group is not statistically significant (P = .24). A positive treatment effect in small tumors (relapse-free survival, odds ratio = 0.55, P = .01) and a negative effect in progesterone receptor-positive tumors (survival, odds ratios = 2.04, P = .04) is probably ascribable to chance. Analysis of various prognostic factors shows that tumor size and histological grade have a clear effect on both relapse-free interval and survival.

The Epping jaundice outbreak: a 24 year follow up.

STUDY OBJECTIVE: The aim was to trace 84 cases of jaundice that occurred following accidental ingestion of methylene dianiline (MDA) in Epping in 1965, and to look at long term health effects. DESIGN: The original case notes of the cases were used to identify the patients. Subsequent tracing procedures included local general practitioners, the Central NHS Registry, electoral rolls, and company records. SETTING: This was a community based survey. MAIN RESULTS: The health status of 68 (81%) of the group was established with 18 deaths. Of the 50 cases known to be alive, 58% completed a health questionnaire. The causes of death were unremarkable except for one case of carcinoma of the biliary tract. Two surviving cases had suffered retinal pathology. Four other surviving cases had had a further, perhaps unrelated, episode of jaundice. CONCLUSIONS: Although the dose and route of administration in the epidemic differed from occupational exposure, this follow up study a generation on provides little, if any, evidence of long term health sequelae. Nevertheless, in the absence of well documented exposure and health effects data, such accidental poisonings with proven animal carcinogens warrant long term follow up. The identified cohort will be the subject of further study.

Etiological factors in oral squamous cell carcinoma.

A retrospective survey of 75 patients with oral squamous cell carcinoma (ICD Nos. 143, 144, 145) resident in the County Borough of Stoke-on-Trent, England, and 150 controls has been carried out by interview. The controls were matched for age, sex, occupation and place of residence. There was no difference in the prevalence or duration of denture wearing, although male controls had their dentures remade more frequently. Male patients practised oral hygiene procedures less frequently when they had their own teeth than controls. Habitual beer-drinking was more common and greater quantities were drunk by male patients than controls. Habitual cigarette smoking was less and pipe smoking greater among male patients than controls. Tobacco chewing, which was restricted to miners, was equally common (45%) in the two groups. Among miners, the combined habit of tobacco chewing and pipe smoking was more common among patients (1005) than among controls (25%).

Selection, follow-up, and analysis in the Birmingham Study.

The Birmingham, England, Cancer Registry is so organized that every case of cancer in its territory of 5,200,000 persons is included. This coverage allows the staff to detail every epidemiologic aspect of the cancer experience of a whole population. For example, this registry system made it possible for us not only to demonstrate that the Birmingham region had four times the incidence of scrotal cancer as another region had but to identify the locations and the specific practices in the workplace responsible for the excess. The result was the successful adoption of protective measures. Other instances are presented of the inestimable value of a population-based registry to cancer epidemiology.

Evaluation of TNM classification of carcinoma of the breast.

A survey of the records of almost 2000 cases of carcinoma of the breast in the years 1960-67, since acceptance of the principles of the TNM system of classification, found only 9·2% lacking description. Although the TNM staging was not always allocated by the surgeon making the initial examination, the survey demonstrated that acceptance of the system has resulted in the inclusion of a much more detailed clinical description by the surgeon in the patient's record, from which it was generally possible to assess the TNM staging in the Radiotherapy Department. The more detailed survival rates available in the case of the TNM system, in comparison with the Manchester staging system, is cited as a further recommendation for its wider use. Although detailed localization of the tumour to subsites within the breast appears to have little influence on prognosis, the addition of a supplementary histological classification of node involvement is a valuable adjunct to the assessment of likely survival.

Second primary cancers in patients with tumours of the salivary glands.

In a series of patients drawn from the Birmingham Regional Cancer Registry (England) with tumours of the salivary glands, a significant excess of second primary tumours was observed. For females, the excess was found mainly in breast and bronchus and, for males, in prostate and skin. In a parallel series of female breast-cancer patients, the observed number of second primary tumours in salivary glands significantly exceeded expectation. These results support the reported association between salivary gland and breast cancer, and suggest that other hormone-dependent sites are also at risk.

Multiple primary cancers of the breast and ovary.

Multiple primary cancer of the breast and ovary were investigated as part of a survey being undertaken at the Birmingham and West Midlands Regional Cancer Registry. Population-based data relating to 17,756 registrations for breast and 3470 for ovarian cancer between 1950 and 1964 were analysed. On the basis of person-years at risk and incidence rates for the Region (1960-1962), an increased risk of a second primary tumour in the ovary was observed in patients diagnosed with a first primary in the breast before 45 years of age (O = 8; E = 1.83; P less than 0.001). No excess was found in patients diagnosed after 45 years of age (O = 15; E = 17.06). In patients with an index tumour of the ovary, the observed number of second primary tumours of the breast was not significantly different from the expected number (O = 19; E = 12.95). Complementary analysis (a combined assessment for the 2 sites) showed that the development of a first primary at either site before 45 years of age carried a 2.8-fold risk of a second primary tumour at the other site (O = 9; E = 3.21; P less than 0.01). After 45 years of age no increased risk was found (O = 32; E = 28.63). Over all ages a 1.3-fold risk was observed (O = 42; E = 31.54; P less than 0.05). No evidence of subfertility was found in the 9 patients in the high-risk premenopausal group who developed the 2 tumours. The results are more consistent with an aetiology of early exposure to an external carcinogen than with one of abnormal hormone production.

Incidence of bilateral tumours in a population-based series of breast-cancer patients. I. Two approaches to an epidemiological analysis.

This paper gives the incidence in the Birmingham Regional Cancer Registry (England) of a second primary tumour in the contralateral breast among nearly 22,000 patients registered with a first primary in the breast between the years 1936 and 1964. The results, based on more than 90,000 women-years at risk and 399 second primary tumours, are presented with reference to 2 methods of analysis. In assessing risk, the principal factors investigated were age at first and second primary diagnoses and the interval between diagnoses. The results are discussed in terms of current aetiological hypotheses. On the basis of a method which included coincidental tumours, the overall risk of a tumour in the contraleteral breast was found to be 3.0 times that in the general population of a first primary. The corresponding risks for 3 main age-ranges (at the time of diagnosis of the first primary tumour) were 5.6 (ages 15-44 years), 3.7 (45--59 years) and 1.8 (60+ years). When coincidental tumours were excluded from the analysis, the relative risk was found to be 2.4 overall and 5.3, 3.0 and 1.0 for the 3 age-ranges, respectively. The level of risk was negatively correlated with age at first primary and the relative risk remained substantially constant over time.

The incidence of bilateral breast cancer: II. A proposed model for the analysis of coincidental tumours.

A statistical model has been proposed in an attempt to integrate coincidental (or synchronous) diagnoses of multiple primary cancers into a general method of analysis. In the context of population-based surveys, such diagnoses form an integral part of the pattern of incidence within the population. Because of clinical surveillance, the diagnosis of subsequent tumours may be advanced in time in comparison with a first primary diagnosis. The model has been used to predict the altered pattern of diagnosis in order to adjust the value of expected numbers. Data from a previously reported survey of bilateral breast cancer have been used to illustrate the model. Analysis in terms of the model showed a 2.6-fold increase in risk for a second primary tumour in the contralateral breast in a series of nearly 22,000 breast-cancer patients. The corresponding risks for 3 main age-ranges (at the time of diagnosis of the first primary) were 5.3 (age 15-44), 3.3 (45-49) and 1.5 (60+). In addition, a maximal risk of 5.0-fold was observed in the series as a whole during the third year after the diagnosis of the first primary.

Multiple primary cancers of breast and cervix uteri: an epidemiological approach to analysis.

Index sites of breast and cervix uteri were selected from populationbased data held at the West Midlands and Birmingham Regional Cancer Registry, and the expected numbers of second primary cancers in cervix and breast were computed (sequence analyses). In the breast series (17,756 patients) a small deficit of cervical tumours was observed (O = 16, E = 2·119, O/E = 0·76, P > 0·05), while in the cervix series (4817 patients) a small excess of breast tumours was found (O = 29, E = 23·38, O/E = 1·24, P > 0·05) over a period of 15 years.A theoretical statement of the combined risk of the 2 tumours occurring in the same individual of a general population was developed and was compared with the practical approach of summing the sequence analyses (complementary analysis). Complementary analysis indicated that there was no excess of women with the 2 primary tumours (O = 45, E = 44·57, O/E = 1·01) and that cancers of the breast and cervix uteri are not aetiologically related.

The generation of simulated data in the form of industrial cohorts.

A computer program INDCOGEN, which generates data in the form of industrial cohorts, is being made available to interested users. Values for variables which include selection effects, dose-response effects, length of study, staff turnover, plant exposure levels, and underlying rates of mortality are defined by the user. The program can be used to estimate the likelihood of detecting a cancer "hazard" given the parameters of a field study cohort, the likely size of selection effects, and an assumed form of dose-response.

Mortality study of nickel-cadmium battery workers by the method of regression models in life tables.

The mortality experienced by a cohort of 3025 nickel-cadmium battery workers during the period 1946-81 has been investigated. Occupational histories were described in terms of some 75 jobs: eight with "high", 14 with "moderate" or slight, and 53 with minimal exposure to cadmium oxide (hydroxide). The method of regression models in life tables (RMLT) was used to compare the estimated cadmium exposures (durations of exposed employment) of those dying from causes of interest with those of matching survivors in the same year of follow up, while controlling for sex, for year and age of starting employment, and for duration of employment. No new evidence of any association between occupational exposure to cadmium oxide (hydroxide) and cancer of the prostate was found.

A mortality study of nickel/chromium platers.

The mortality experienced by a cohort of 2689 nickel/chromium platers between 1946 and 1983 has been investigated. All members of the study cohort had some period of chrome exposed employment. Overall, compared with the general population of England and Wales, statistically significant differences relating to cancer were found for cancer of the stomach (E = 16.2, O = 25), primary cancer of the liver (E = 0.8, O = 4), cancer of the nose and nasal cavities (E = 0.3, O = 3), cancers of the lung and bronchus (E = 48.1, O = 72), and all cancers (E = 164.2, O = 213). Chrome bath workers are the more heavily exposed workers, and a striking difference in SMRs was found for lung cancer among men first employed as chrome bath workers (SMR = 199) and men first employed as other chrome workers (SMR = 101). The method of regression models in life tables (RMLT) was used to compare the durations of chrome exposed employment of those dying from causes of interest with those of all matching survivors in the same year of follow up, while controlling for sex, and for year and age of starting employment. Significant positive associations were found only for cancers of the lung and bronchus and duration of chrome bath work. In this study exposure to nickel was shown not to be an important confounding exposure.

Analysis of mortality from nephritis and nephrosis among nickel-cadmium battery workers.

Mortality from nephritis and nephrosis (and other diseases of the genitourinary system) experienced by a study cohort of 3,025 nickel-cadmium battery workers was investigated. Observed (O) numbers of deaths from these diseases for the period 1946 through 1981 were compared with those that might be expected (E) to occur on the basis of rates of mortality for the general population of England and Wales (nephritis and nephrosis, International Classification of Diseases, eighth revision [ICD 8th] Code No. 580-584; E = 5.49, O = 10: all disease of the genitourinary system (noncancers), ICD 8th 580-629; E = 11.50, O = 17). These differences were not statistically significant at the 5% level. When the estimated cadmium exposures of those who died from these causes were compared with those of all matching survivors using the method of regression models in life-tables, no statistically significant association was found between occupational exposure to cadmium and mortality from these diseases. A separate analysis was made of the mortality experience of 39 workers with cadmium nephropathy.

Obstetric ultrasound and childhood malignancies.

The mothers of 1731 children who died of cancer in the United Kingdom between 1972 and 1981, together with the mothers of 1731 matched controls, were asked about exposure to diagnostic ultrasound during the relevant pregnancy. The claims of 103 case mothers and 103 control mothers were verified by hospital records. Exposure to ultrasound in utero did not differ between cancer and control children. It is concluded that ultrasound is safe as regards the risk of cancer and leukaemia. The observed difference between cases and controls exposed during the earlier years of ultrasound use may be due to the selective application of this technique to abnormal pregnancies at that time.

Scrotal cancer in the West Midlands 1936-76.

Between 1936 and 1976, 344 cases of scrotal cancer were registered in the West Midlands Region. There was a considerable increase in the number of registrations after 1955, the figures remaining high until the 1970s since when they appear to be falling towards the low level seen in the early years of registration. By the end of December 1981 250 deaths had occurred. The cause of death was known in 226 cases and of these, 97 (42.9%) were attributed to cancer of the scrotum, 40 (17.7%) to other malignancies, and 89 (39.4%) to other causes, scrotal cancer not being mentioned on the death certificate. The mean age at registration was 59.3 years and at death 66 years. Mean survival was 6.6 years with a five year survival rate of 51%. Survival was shorter in men who presented with palpable lymph nodes and greater in those treated with surgery alone. Occupations were known for all but 28 of the cases and of these, 61.9% had been exposed to mineral oil and a further 7.8% to pitch and tar. Among the occupations with exposure to mineral oil, tool setters and tool fitters accounted for the greatest proportion, 89 men having followed those trades. In 74 cases second primary tumours were registered after the scrotal primary, mostly in the skin and bronchus. There was no relation between the development of second primary tumours and exposure to mineral oil or to pitch and tar.