MIRACLE: A Time-oriented Data Validation and Analysis System for Clinical Research

MIRACLE is a medical information system designed to validate and analyse time-oriented data. The chronological patient database is composed of user-defined repeating groups stored in numerically coded character records of variable length. A flexible query language gives access to semantic validation, cross-sectional and mixed-longitudinal analysis. The main application is at present a large hospital-based tumor register.

Hippocampal size and dementia in stroke patients: the Sydney stroke study

BACKGROUND

Hippocampal atrophy is an early feature of Alzheimer's disease (AD) but it has also been reported in vascular dementia (VaD). It is uncertain whether hippocampal size can help differentiate the two disorders.

METHODS

We assessed 90 stroke/TIA patients 3-6 months after the event, and 75 control subjects, with neuropsychological tests, medical and psychiatric examination and brain MRI scans. A diagnosis of VaD, vascular mild cognitive impairment (VaMCI) or no cognitive impairment (NCI) was reached by consensus on agreed criteria. T1-weighted MRI was used to obtain total intracranial volume (TICV), gray and white matter volume, CSF volume, hippocampus and amygdala volumes, and T2-weighted scans for white matter hyperintensity (WMH) ratings.

RESULTS

Stroke/TIA patients had more white matter hyperintensities (WMHs), larger ventricle-to-brain ratios and smaller amygdalae than controls, but hippocampus size and gray and white matter volumes were not different. WMHs and amygdala but not hippocampal volume distinguished stroke/TIA patients with VaD and VaMCI and without NCI and amygdala volumes. Right hippocampus volume significantly correlated with new visual learning.

CONCLUSIONS

Stroke/TIA patients and patients with post-stroke VaMCI or mild VaD do not have hippocampal atrophy. The amygdala is smaller in stroke/TIA patients, especially in those with cognitive impairment, and this may be accounted for by white matter lesions. The hippocampus volume relates to episodic memory, especially right hippocampus and new visual learning. A longitudinal study of these subjects will determine whether hippocampal atrophy is a late development in VaD.

Multivisceral resections in colorectal cancer

From 1969 to 1983 a total of 1918 patients with colorectal cancer were treated by curative resection. One hundred twenty one patients in this group had multivisceral organ involvement, necessitating extended multivisceral radical resection. Tumor infiltration was proven histologically in 55 percent, while 45 percent had inflammatory adherence to the attached organ only. Postoperative mortality was 12 percent. Dukes' A and B stages were present in 57 percent. The five-year survival rate (postoperative mortality included) was 54 percent for patients with inflammatory adherence, 49 percent for patients with tumor infiltration resected en bloc without tumor tears of rupture, and 17 percent when the surgeon inadvertently had torn or cut into tumor tissue during resection.

[Operative tactics and results in extensive operations for colorectal cancer]

Between 1969 and 1983 a total of 1918 patients with colorectal carcinoma were curatively operated. In 121 of these patients a curative procedure could be achieved despite tumor invasion of adjacent organs by extending the operation to include these involved structures. In 45% of these patients the histological examination revealed that there were only peritumorous inflammatory adhesions while in 55% of the cases a direct tumor invasion of the neighbouring organ was present. The 5-year-survival-rate in the 121 patients was 44%. The operative mortality of 12% (14 out of 121) is included. The final results were considerably worse if the tumor could not be removed en bloc.

[Retained cholelithiasis--a risk factor after endoscopic papillotomy?]

Endoscopic papillotomy (EPT) today is in well defined indications an accepted procedure for the therapy of common bile duct stones. In patients with stone gallbladder in situ and bile duct stones it has been recommended to remove the gallbladder after papillotomy in order to prevent late complications. The present study shows that the gallbladder in situ is not necessarily a significant risk factor. On an average of 34,9 months (16-84) after EPT of 223 patients with a follow up, 83,7% were free of symptoms or significantly improved. Cholecystectomy was performed in 43 patients (18,9%): in 24 (11,6%) within an interval , in 13 (6,3%) as an emergency operation following a complication of EPT and in 2 (1%) months later because of recurrent cholecystitis. In patients over 70 years of age with increased risk for surgery therefore a gallbladder with stones after EPT must not absolutely be removed. A wait and see attitude may be justified.

The malignant potential of colorectal polyps--a new statistical approach

To investigate the factors influencing the malignant potential of adenomas, a logit analysis was carried out. The malignancy rate (frequency of malignant areas infiltrating into submucosa) in adenomas is influenced by 1. the size of the adenoma, 2. the interrelationship between size and histological type (tubular, tubulo-villous, villous) and 3. the macroscopic growth pattern (pedunculated, semipedunculated, sessile). No influence is exercised by the number of adenomas, their localization, or the sex of the patient.

[Surgical therapy of stomach cancer-stagnation or progress? Results of 2665 stomach cancer patients]

The results of surgical therapy in 1165 patients with gastric cancer who underwent treatment between 1949 and 1964 are compared with those of 1500 patients who were operated on between 1969 and 1981. Special emphasis is placed on curability, operative methods and postoperative mortality. The 5-year survival rates of patients from the time periods 1950-1959 and 1969-1976 are compared. 1. The percentage of curatively treated patients increased significantly from 46,7% (1949-1964) to 57,0% (1969-1981). 2. The postoperative mortality after total gastrectomy was further reduced significantly: it was 25% from 1957-1964, but only 13% from 1976-1981. 3. The 5-year survival rate of all patients with gastric cancer increased significantly from 10% (1950-1959) to 17% (1969-1976). 4. The 5-year survival rate after curative operations increased from 22% to 30%; after curative distal partial gastrectomy from 22% to 42% and after curative total gastrectomy from 6% to 20%. 5. The median survival time after palliative operations without resection of the tumor was 3 months. Since 1977 more than one third of the patients who could not be treated curatively underwent resection of the tumor. The median survival time in this group is 6 months with improved quality of life.

Classification of colorectal carcinomas with regional lymphatic metastases

According to the UICC classification of colorectal patients with histologically confirmed regional lymph node metastases (pN1) are not further subdivided. In these patients, however, subgroups with widely differing prognosis can be established. A number of classifications proposed to cover these patients are discussed. The analysis of the Erlangen case material shows that it is decisive for the prognosis to know whether involved lymph nodes are located only at the branches of the supplying vessels ("peripheral"), or whether the lymph nodes along the trunks of these vessels ("proximal") are also involved. Among the patients with proximal lymph node metastases, the question as to whether the lymph node at the margin of the resected lymph drainage area ("marginal node") is involved or not, has no essential influence on the prognosis. The counting of metastatic lymph nodes does not represent an essential criterion for prognosis.

[Prediction of local recurrences after surgery of carcinoma of the middle reticulum (author's transl)]

On the basis of completely and exactly documented follow-up data of 237 patients on whom carcinomas of the middle rectum were resected for cure, a multivariate statistical model was developed and the corresponding parameters were calculated. With this model the individual risk of local recurrences can be estimated after anterior resection with a small margin of clearance (11-30 mm on the fresh specimen without stretching), after anterior resection with a wide margin of clearance (more than 30 mm), and after excision of the rectum. Seven prognostic factors are taken into consideration that can be determined pre- and intraoperatively. Estimations of the risk of local recurrences on the basis of observed and documented follow-up data can be used as decisive factor for differential indication between restorative and excisional surgery on future patients. After anterior resection the risk of local recurrences is calculated according to macroscopic findings on the specimen and to frozen section histology. If the risks in anterior resection and excision are approximately equal, the resection is ended by anastomosis; otherwise the operation will be extended to excision of the rectum. This statistical model enables the surgeon to select the proper operative procedure for each individual situation (histology- and stage-adapted surgery).

Prognostic groups in colorectal carcinoma

A system of classifying colorectal carcinomas into five prognostic groups is described. This classification includes not only tumor spread, regional lymph node involvement, and distant metastases, but also resectability and R-classification (residual tumor or curability). All patients with colorectal tumors can be classified by means of this system. The five prognostic groups differ statistically significantly (p less than 0.01) from one another with respect to age-corrected 5-year survival rates.

[Developments in the diagnosis of colorectal carcinoma. Comparison of various diagnostic groups]

Epidemiological observations on colorectal cancer and the consequences of modern diagnostic methods are reported on patients treated in the Medical and Surgical Hospital of the University of Erlangen during the past 9 years. In this period the distribution of the rectal carcinomas according to the Dukes-classification was constant. Using coloscopy consequently as routine method and by proving occult intestinal bleeding as screening method, we succeeded in diagnosing colorectal carcinomas in an earlier stage of Dukes' classification.