What is the lifetime risk of developing cancer?: the effect of adjusting for multiple primaries.

BACKGROUND: The 'lifetime risk' of cancer is generally estimated by combining current incidence rates with current all-cause mortality ('current probability' method) rather than by describing the experience of a birth cohort. As individuals may get more than one type of cancer, what is generally estimated is the average (mean) number of cancers over a lifetime. This is not the same as the probability of getting cancer. METHODS: We describe a method for estimating lifetime risk that corrects for the inclusion of multiple primary cancers in the incidence rates routinely published by cancer registries. The new method applies cancer incidence rates to the estimated probability of being alive without a previous cancer. The new method is illustrated using data from the Scottish Cancer Registry and is compared with 'gold-standard' estimates that use (unpublished) data on first primaries. RESULTS: The effect of this correction is to make the estimated 'lifetime risk' smaller. The new estimates are extremely similar to those obtained using incidence based on first primaries. The usual 'current probability' method considerably overestimates the lifetime risk of all cancers combined, although the correction for any single cancer site is minimal. CONCLUSION: Estimation of the lifetime risk of cancer should either be based on first primaries or should use the new method.

Nm23 protein expression in ductal in situ and invasive human breast carcinoma.

BACKGROUND: Mortality associated with human breast carcinoma is almost entirely due to subsequent metastatic disease, but the molecular basis of this metastasis is not understood. Elucidation of the genetic control of metastatic propensity of a tumor is important in determining prognosis and choice of therapy. Expression of nm23, a putative metastasis suppressor gene, has been detected in human breast cancers, but studies have not consistently shown high levels of the Nm23 messenger RNA or protein to be associated with better histological differentiation. This inconsistency suggests that Nm23 protein may act independently as a metastasis suppressor. PURPOSE: The purpose of this retrospective study was to investigate the relationship of Nm23 protein expression with 1) histology in ductal breast carcinoma in situ and 2) the variables considered to be the major prognostic indicators in invasive breast carcinoma. METHODS: We obtained formalin-fixed biopsy specimens of breast tissue excised from 128 patients with breast lesions detected by mammography. Of these patients, 35 had been diagnosed with benign breast disease, 26 with ductal carcinoma in situ (DCIS), and 67 with invasive carcinoma. Tissue sections were embedded in paraffin blocks, and immunohistochemical staining was used to determine Nm23 expression. Specimens were rated positive if all lesional epithelium was stained and negative if any lesional epithelium was unstained. Statistical analysis was performed by multiple regression analysis because of nonorthogonality of the data. RESULTS: All 35 examples of benign breast disease showed uniform epithelial cell staining. The seven cases of comedo DCIS were negative for Nm23 protein; all 18 noncomedo types were positive. Nm23 negativity was significantly associated with worsening invasive ductal carcinoma grade and advancing lymph node stage but not with tumor diameter or vascular invasion. Despite the putative antimetastatic role of the nm23 gene, no statistically significant association was found between Nm23 protein expression and vascular invasion. CONCLUSIONS: The precise role of the nm23 gene remains to be established, but our simplified immunohistochemical rating system shows an association between Nm23 protein expression and the two most significant prognostic factors relating to histologic grade and stage. Nm23 negativity distinguished comedo ductal carcinoma in situ from the other histological types, a finding consistent with the fact that comedo histology is known to have a higher likelihood of becoming invasive and of having higher cell proliferation rates and higher expression of growth factor (c-erb B2) receptor.

Measuring repeatability and validity of histological diagnosis--a brief review with some practical examples.

Evaluation of histological diagnosis requires an index of agreement (to measure repeatability and validity) together with a method of assessing bias. Cohen's kappa statistic appears to be the most suitable tool for measuring levels of agreement, which if unsatisfactory may be caused by bias. Further study of bias is possible by examining levels of agreement for each diagnostic category or by searching for categories of disagreement in which more observations occur than would be expected by chance alone. This article gives reasons for choosing the kappa statistic, with examples illustrating its calculation and the investigation of bias.

Prognostic implications of qualitative assessment of left ventricular function compared to simple routine quantitative echocardiography.

OBJECTIVE: To compare the prognostic value of qualitative estimates of left ventricular function with that of routine simple quantitative indices used in echocardiography. DESIGN: Retrospective follow up study. SETTING: University hospital. PATIENTS: The records of 2,964 patients who had undergone echocardiography and who could be traced on the family health services register were examined; 919 cases were included in the study, and a further 458 were used to validate the statistical models for prognostic assessment. There were 928 exclusions on the basis of referral for or diagnosis of alternative conditions, and 659 because of incomplete collection of the qualitative and quantitative data used in the study. MAIN OUTCOME MEASURE: Survival over the study period. RESULTS: A qualitative "eyeball" estimate of left ventricular function was of prognostic significance (relative risk of poor v good, 2.248; P << 0.001; 95% confidence interval 1.620 to 3.119). None of the quantitative echocardiographic indices was of independent prognostic significance when all variables were tested simultaneously in the regression model. CONCLUSIONS: A qualitative echocardiographic estimate of left ventricular dysfunction is of prognostic value, supporting the view of many cardiologists who use their overall impression of left ventricular function at echocardiography as the basis for treatment decisions.

Relative risk estimation from vital statistical data: validation, a pitfall and an alternative method.

A previously described method of obtaining an estimate of relative risk from routinely available data was applied to data on cigarette consumption and lung cancer mortality to test its validity. Some shortcomings of the method were noted and an alternative approach using weighted logistic regression gave results closer to those predicted on the basis of other studies, without the disadvantages of the original technique.

Life expectancy as a summary of mortality in a population: statistical considerations and suitability for use by health authorities.

OBJECTIVE: To investigate the sampling distribution and usefulness of expectation of life in comparisons of mortality at health district level or below. DESIGN: Derivation of a formula for the variance of the expectation of life, confirmation of the result and generation of the sampling distribution by Monte Carlo simulation; comparison of expectation of life with standardised mortality ratio (SMR) and other summary indices of mortality. SETTING: A health district in Trent Region, England. SUBJECTS: Routinely available mortality statistics at electoral ward level and above. MAIN RESULTS: Given reasonable and simple assumptions the sampling distribution of the expectation of life is approximately normal. Expectation of life shows a high negative correlation with SMR even if the oldest age band for the SMR is open ended. CONCLUSIONS: Where sampling error is an issue, inference concerning differences in mortality rates between populations can be based on expectation of life, which is better for illustrative purposes than SMR. The formula for the variance of the expectation of life is more complex however. If the final age band is open ended, its lower bound should be as high as possible to avoid misleading results caused by hidden differences in age structure.

Effects of hormone replacement therapy on fine needle aspiration cytology of the breast.

To investigate the possibility that hormone replacement therapy (HRT) might induce morphologic changes in breast cytology simulating malignancy, we determined the subjective interpretation and objective morphometric features of 39 fine needle aspiration cytology samples from women who were using a variety of HRT preparations. Cases were matched for age and final diagnosis (assessed by a triple approach) with women who had never used HRT. Cytocentrifuge preparations stained with Papanicolaou stain were assessed by two observers for cellularity and diagnosis. Although diagnostic accuracy was slightly higher in HRT than non-HRT samples, there was no significant difference in the assessments of cellularity and diagnosis between the two observers or between HRT and non-HRT samples. Nuclear areas in HRT samples were larger than in non-HRT samples, measured using computerized image morphometry; this difference did not induce the observer to diagnose malignancy more frequently. HRT had no adverse effect upon the accuracy of interpreting breast fine needle aspiration cytology samples.

Clinicians and the coronial system: ability of clinicians to recognise reportable deaths.

OBJECTIVE: To assess the ability of clinicians to recognise deaths which require referral to the coroner. DESIGN: Postal questionnaire consisting of 16 fictitious case histories, 14 of which contained a clear indication for referral to the coroner. SETTING: Large teaching hospital. Coroner's office. SUBJECTS: 200 clinicians from general medical and surgical firms and senior staff of the local coroner's office (two coroner's officers and the two deputy coroners). MAIN OUTCOME MEASURES: Number of correct assessments on questionnaire. RESULTS: The mean recognition score for the clinicians was 9.11 (range 3-14) with no difference between the clinical grades. All of the coroner's senior staff recorded maximum recognition scores of 16. CONCLUSIONS: The study highlights several features of the coronial system which are poorly understood by clinicians and provides the basis for an initiative to improve the medicolegal education of all clinicians.

Space-time clustering and bone tumours: application of Knox's method to data from a population-based cancer registry.

The possibility of space-time clustering in osteosarcoma and Ewing's tumour was studied in a sample of 182 patients aged under 25 years, drawn from a population-based cancer registry. No evidence of space-time clustering was found for either disease and this finding supports previous work in this field.

Rapidly progressive cervical cancer: is it a real problem?

The appearance, in younger women, of a rapidly progressive form of cervical carcinoma has been suggested. This, together with evident deficiencies in the provision and technique of screening services, has been taken to indicate a need for more frequent screening. We show that a substantial increase in the observed number of such rapidly progressive cancers would be expected simply on the basis of changes in population structure, a general increase in cervical cancer rates and increases in the proportion of women screened.

Squamous and adenocarcinoma of the uterine cervix: a comparison using routine data.

We studied the clinical, demographic and survival characteristics of more than 5,000 women registered with either squamous or adenocarcinoma of the uterine cervix in South Thames Cancer Registry over the period 1968-81. There were similarities with respect to social class, smoking habit, oestrogen/oral contraceptive use and time trends in incidence but differences between the two cancers were found with respect to age distribution, parity, method of detection and survival. Some of the data are of limited value, having been obtained only from case notes, so the results need some caution in their interpretation. However the results are broadly consistent with those of studies performed in other countries on smaller samples. A methodological issue is also raised, viz. the appropriateness of a disease with well-known characteristics as a comparison group. From our results the likely size of various associations can be judged and used in the design of future studies to clarify the epidemiology of cervical adenocarcinoma.

PIP: Compared were the clinical, demographic, and survival characteristics of over 5000 women registered in the South Thames Cancer Registry in 1968-81 with either squamous carcinoma (4599 women) or adenocarcinoma (704 women) of the uterine cervix. There were no significant differences between these 2 groups in terms of social class, cigarette smoking, use of estrogen therapy or oral contraception, and time trends in cancer incidence. However, there were differences between these 2 cancers with respect to age distribution, parity, methods of detection, and survival. Women with adenocarcinoma were 2.7 years older on average than those with squamous carcinoma, were more likely to be single, and were more often nulliparous. Patients with adenocarcinoma were also more likely to present with symptoms of advanced disease rather than through screening, and their median survival time was 2 years less than that for women with squamous carcinoma. The methodology of this study can be criticized for its use of a disease with well-known characteristics as a comparison group. Moreover, it is recommended that future studies specify histologic type of cancer wherever possible: the quality of histologic diagnosis and its changes with time should also be studied to validate conclusions based on routine data. In general, the findings of this comparative study are consistent with those performed in other countries on smaller samples. The results of the present study can be used to judge the likely size of various associations and to design future studies aimed at clarifying the epidemiology of cervical adenocarcinoma.

Making cancer statistics more informative: measures of the quality of recorded diagnosis in a population-based registry.

The quality of the recorded diagnosis is a major limit to the usefulness of Cancer Registry statistics that is easily overlooked by users of the data. With data from a large population-based cancer registry as an example, we demonstrate how Registry statistics could be improved by wider use of three simple indices, namely (1) the proportion histologically verified (adjusted for age), (2) the proportion of verified cases with an uninformative diagnosis, and (3) the proportion of cases that are staged. We believe that greater awareness of the deficiencies of Cancer Registry statistics will lead to a more critical interpretation of them, and help stimulate efforts to rectify matters.

Evaluation of PAS-diastase and carcinoembryonic antigen staining in the differential diagnosis of malignant mesothelioma and papillary serous carcinoma of the ovary.

A series of malignant mesotheliomas and papillary serous carcinomas of the ovary were stained using the periodic acid Schiff reaction with diastase digestion (PAS-D) and an immunoperoxidase technique for carcinoembryonic antigen (CEA). The methods were evaluated in terms of repeatability and validity. Staining reactions were classed as positive or negative by three 'blinded' observers. Within-observer repeatability was greater for PAS-D than for CEA but there was no statistically significant difference for between-observer repeatability. The validity of the methods clearly differed, however, since the sensitivity and specificity of CEA were consistently less than that of PAS-D. Our most consistent estimates of sensitivity and specificity were compared with those of other studies and the reasons for this difference were analysed. Positive PAS-D and CEA staining were uncorrelated but a combination of the two did not appear significantly superior to PAS-D alone. We believe that PAS-D remains the method of choice for differentiating peritoneal mesotheliomas and disseminated papillary serous carcinoma of the ovary in the absence of reliable evidence of an ovarian primary because the significance of CEA positivity in occasional mesothelial tumours remains to be determined.

Can we achieve 100% ascertainment in cancer registration?

Completeness of ascertainment is a central issue for any disease register. We outline the process of cancer registration as carried out in a large population-based Registry and argue that none of the various methods proposed to estimate the completeness of ascertainment are satisfactory. We suggest an alternative way of looking at the issue and describe various constraints that hamper attempts to achieve complete registration.

General practitioners, skin lesions and the new contract.

We reviewed all skin lesions received in our laboratory from general practitioners (GPs) during a three-month period before the introduction of the new contract for GPs and during the same period a year later. For comparison we also reviewed skin lesions received from the general and plastic surgeons. Particular attention was paid to the completeness of excision. There was a significant increase in the number of skin lesions removed in general practice after the introduction of the new contract. Both benign and malignant lesions were more likely to be incompletely excised by GPs compared with surgeons. The GPs first noted to carry out minor surgery after the new contract came into force were less likely to have completely excised lesions than their more experienced colleagues. We suggest the future monitoring of lesions removed in general practice, possibly by the formation of a joint audit group.

Catecholamines and experimental stress ulcer: morphological and biochemical changes in the gastric mucosa.

A model for the production of gastric mucosal haemorrhages and ulcers, similar in morphology and distribution to those seen in stress states in man, has been produced by the intraperitoneal injection of adrenaline in single, double or treble doses of 0.6 mg/kg. Lesions were produced in 2 or 3 h, without ligation of the pylorus. Severe mucosal changes were associated with a fall in gastric mucosal histamine concentration. Mucosal serotonin did not change. Rises in mucosal adrenaline were accompanied by a fall noradrenaline, total catecholamine levels being unchanged.