Parenchymal patterns of the breast defined by real time ultrasound.

The association of mammographic parenchymal patterns of the breast with breast cancer risk has been studied extensively but there is little information about the distribution of different patterns in populations at different risks for breast cancer. Such information could be obtained if a risk-free method of breast examination were available that could be applied to the general population. We have evaluated real time ultrasound for this application by comparing the parenchymal pattern as assessed by mammography with the extent of echogenicity in the breast on ultrasound examination in 102 subjects. Subjects were examined by both methods, the mammographic and ultrasound images independently classified, and the proportion of the breast occupied by radiological density or ductal prominence compared with the extent of echogenic areas on ultrasound. These two methods of classifying mammographic parenchymal patterns were found to be strongly correlated. Real time ultrasound may therefore be useful in the epidemiological study of mammographic pattern and breast cancer risk.

Mediastinal lymph node metastases from bronchogenic carcinoma: detection with MR imaging and CT.

Magnetic resonance (MR) imaging and computed tomography (CT) were compared in a prospective study of 48 patients for the detection of metastatic mediastinal lymphadenopathy from bronchogenic carcinoma. The images were interpreted by three experienced radiologists using a five-point rating scale, enabling receiver operating characteristic (ROC) analysis. Imaging results were evaluated against "truth" data based on analysis of surgical specimens from mediastinoscopy and thoracotomy. All MR images were cardiac gated to reduce cardiac motion artifacts in the mediastinum. MR and CT both performed well, as indicated by similar areas under the ROC curves of 0.779 +/- 0.039 for MR imaging and 0.781 +/- 0.038 for CT scanning. No strong correlation between nodal size and metastatic involvement could be found for either MR or CT results. As long as nodal size remains the sole criterion in the detection of metastatic mediastinal lymphadenopathy, MR imaging is unlikely to enable better interpretations than CT scanning.

Computed tomography in advanced ovarian cancer: an evaluation of diagnostic accuracy.

The authors have investigated the diagnostic accuracy of computed tomography (CT) of the abdomen and pelvis in the assessment of patients prior to second-look laparotomy for advanced ovarian cancer. CT studies (read independently by three radiologists) and laparotomy findings were analyzed in 50 patients. Sensitivity varied from 0.30 to 0.65 among the radiologists, specificity from 0.44 to 0.89, positive predictive value from 0.50 to 0.73, and negative predictive value from 0.60 to 0.70. Receiver operator curve analysis of the data indicated poor performance of CT as a diagnostic test. The authors also examined some problems in interpreting their data and that of others in the literature, in particular, the influence of potential sources of bias.

Assessment of acute radiation-induced pulmonary changes using computed tomography.

In a prospective study of acute radiation-induced pulmonary changes, CT scans of 54 patients were performed before and at preselected times during the 6 months following fractionated radiation therapy of the thorax. The CT films were evaluated independently by three diagnostic radiologists and 36 patients were scored as having postirradiation pulmonary findings. The average interobserver agreement for this scoring was approximately 85%. The end point was observed as an increase in lung density within the irradiated volume on a follow-up CT examination. All 36 patients demonstrated lung opacities in an irregular, homogeneous, or nonhomogeneous pattern within the radiation beam boundaries. In addition, the following characteristics were observed at various frequencies in these 36 patients: extension of the changes across anatomic tissue boundaries (50%), air bronchograms (25%), loss of lung volume (15%), and pleural thickening (15%). Confinement of the findings within the irradiated volume was the only specific characteristic of postirradiation changes. In two patients the changes appeared as sharply defined, nodular opacities and were considered to be atypical of radiation damage. These were subsequently confirmed to be metastases. Prospective assessment of an adequate number of patients has helped to establish the CT appearance of acute radiation-induced pulmonary effects and, hence, to minimize its confusion with malignancies and other abnormalities.

Computed tomography in advanced ovarian cancer. Inter- and intraobserver reliability.

We have examined the reliability of computed tomography (CT) in the assessment of patients with advanced ovarian cancer. Three radiologists independently read the same set of 140 CT scans. Fifty percent of the films were reread by the same radiologist to assess intraobserver agreement. In addition we assessed variability among radiologists in reporting signs of disease such as a mass and ascites and the variability in reporting a change in signs in response to therapy. Intraobserver agreement on the identification of signs such as a mass and ascites and on overall response to therapy was moderate to excellent (Kappa 0.52-0.84). Agreement among observers on the same signs was not as good (Kappa 0.36-0.79).

Observer variation in abdominal CT.

Observer performance in radiologic interpretation has traditionally been based upon Receiver Operating Characteristic (ROC) analysis. A statistical method that evaluates the reproducibility or consistency of radiologic interpretation is presented. Since this method is not based upon pathologic verification, it cannot be used to assess accuracy of interpretation and is therefore not a substitute for ROC analysis. Rather, it can either supplement ROC analysis or be of use in situations in which the latter is not feasible or practical.

Efficacy of computed tomography in the management of cancer patients.

An evaluation was made of 1030 computed tomography (CT) scans performed on cancer patients. Seventy percent of the scans taken were of the pelvis, abdomen, or thorax. Both referring clinicians and radiologists rated the diagnostic and management efficacy of each scan using a simple scale of values. Over 50% of the scans provided unique diagnostic information: 39% resulted in a change in diagnosis or in known extent of disease; and 14% led to a change in patient management. The fraction of abnormal results was not a reliable measure of efficacy. Specific groups of patients were identified, which demonstrated that both diagnostic and management efficacy should be measured in assessing the value of CT as a diagnostic modality.

Results of treating Hodgkin's disease without a policy of laparotomy staging.

Results of the treatment of 780 primary patients with Hodgkin's disease at the Princess Margaret Hospital (PMH) between 1968 and 1977 are analyzed. Treatment decisions were based on the evaluation of the extent of disease by clinical methods. A marked improvement in relapse-free survival and overall survival was observed for 1973-1977 as compared to 1968-1972. This improvement did not result from differences in the distribution of important prognostic attributes (clinical stage, pathology, and age) between the two periods, and there was no improvement in our ability to rescue relapsed patients. Improved relapse-free and overall survival during the second period was observed for all stages in patients less than 50 years of age, but not in the older group. The improved survival of patients treated between 1973 and 1977 is attributed to more effective initial therapy, which reduced the fraction of patients who relapsed. These observations provide indirect evidence that relapse has a negative effect on prognosis, and that the initial treatment of patients with Hodgkin's disease should be designed to reduce the risk of relapse to a minimum without causing an unacceptable increase in late complications. The observed/expected incidence of acute leukemia and non-Hodgkin's lymphoma in the PMH series was increased to 41.9 and 13.9 respectively. The question of whether a policy of doing routine staging laparotomies improves the results of treatment of patients with Hodgkin's disease is considered only in general terms by comparing the total PMH series with the total Stanford Medical Center series of patients treated between 1968 and 1977. Relapse-free survival at 10 years is 48.9% and 66.8% respectively, at the two institutions, while overall survival at 10 years is identified.

Discrepancies between radionuclide and computed tomographic scans in detecting secondary neoplastic involvement of the brain.

A radionuclide (RN) brain scan and a computed tomographic (CT) brain scan were performed on the same patient within one month on 207 occasions between February 1978 and October 1979 in order to investigate secondary neoplastic involvement of either metastatic or lymphomatous nature. The two scans were read independently of each other. There was a significant difference between the results of the two techniques in 26 of the 207 cases (13%). In 15 cases, the RN brain scan found clinically suspected secondary neoplastic involvement that was not detected by CT. In the other 11 cases, lesions detected by CT were not detected with the RN scan. Discrepancies were more frequent in patients with lymphoma than in patients with carcinoma.

Evaluation of computed tomography in oncology.

An evaluation was made of 1343 CT scans performed at The Princell Margaret Hospital using a 20-second body scanner. Six patient groups were defined according to the clinician's objective in ordering the scan, and to the relevance of symptoms to the region scanned. Computer-assisted analysis revealed that only by dividing patients into categories using three parameters (scan region, primary diagnosis and clinician's objective) did significant differences in the ratio of normal to abnormal CT scan results become apparent. Very specific groups of patients for whom CT is a low-yield procedure could be identified; nevertheless, further investigation showed that a normal CT result was of benefit in the management of these patients. The overall use of CT was efficacious and low diagnostic yield alone is not a sufficient measure of inappropriate use of CT.

Patterns of breast parenchyma on mammography.

The hypothesis that two of four mammographic patterns show a marked predisposition to develop breast cancer was examined in a retrospective study. Four hundred patients with breast carcinoma were randomly chosen. The xeromammograms of the opposite breasts in these patients had been done and were studied. The conclusion of Wolfe was not supported by our findings.

Radiologic studies of larynx after radiotherapy for carcinoma.

Antero-posterior linear tomography was used in conjunction with a xeroradiographic lateral view of the neck to study 35 cases of carcinoma of larynx previously treated by irradiation. Although false negatives and false positives did occur, this method is reasonably reliable in the diagnosis or exclusion of recurrent malignancy.

Appearances of the larynx after radiation therapy.

Laryngeal tomography and lateral Xerox radiographs from 21 cases of carcinoma of the larynx previously treated by irradiation were analyzed. Although false negatives and false positives did occur, radiology made a useful contribution to the diagnosis or exclusion of recurrence in a number of cases. The most relevant positive findings were: 1. Ulceration, 2. Asymmetrical swelling, 3. Failure of tumor mass to shrink.