Computer-aided individual prognoses (CIP) is a software-package developed on the basis of an empirical study and can be installed on any IBM-compatible personal computer. The project which went into the making of CIP was called “Prospective DÖSAK-study on squamous cell carcinomas of the lips, oral cavity and oropharynx”. In the course of the study 1485 patients were treated between 1977 and 1982, and followed up through 1985. CIP facilitates individual prognoses and comparisons of independent patient groups with parallel groups from the data of the above-mentioned study. In practical clinical work individual prognoses allow exact and reliable judgements on individual patients. In clinical cancer research it provides the information about prognostic factors required for controlled clinical studies. The comparison of independent patient populations allows for an ongoing qualitative control of the patients in each clinical institution. In clinical cancer research such a comparison means that certain characteristics of the patient, the tumor and of the disease can be given higher prognostic value. The same is true of the clinical testing of therapeutic measures which is typically carried out during phase-II-studies. Due to its menu-based organization CIP does not presuppose any specialist knowledge on the part of its users and can be regarded as particularly user-friendly.

The "Treatment Dependent Prognostic Index" (TPI), which has been published by the authors in 1982 as a result of a retrospective observational study of the DOSAK (German-Austrian-Swiss Association for Head and Neck Tumors), is submitted to a critical review. For this purpose a sample of 1485 patients representing the "Prospective DOSAK-Study on Squamous Cell Carcinomas of the Lips, the Oral Cavity and the Oropharynx" were classified according to TPI. 13 TPI groups exhibited greater than or equal to 20 patients and were suitable for calculation of observed survival and comparison to their TPI prognoses. In 10 of 13 TPI groups observed and predicted survival are almost identical. 3 TPI groups show a considerable lack of correspondence between observation and prognosis. In general, the analyses revealed a high prognostic validity of TPI in its substantial parts. A further improvement of TPI by taking several tumor sites into consideration seems possible and will be a subject of future investigations.

Von 1972–1984 wurden an der Abteilung für Kiefer- und Gesichtschirurgie in Linz/ Österreich insgesamt 330 Nasenkorrekturen vorgenommen. Bei 52 Patienten bestand eine Einsattelung des Nasenrückens, welche die Implantation von autologem oder konserviertem Knorpel notwendig machte.

A short review of regression models for the analysis of censored survival data is given. These include multiplicative hazard rate models, log-linear models (accelerated failure time models), linear models and polynomial models. An application of some of these models to the analysis of a large retrospective study on carcinomas of the oral cavity is described. The results obtained by parametric and semiparametric analyses are compared.

In einer multizentrischen, retrospektiven Beobachtungsstudie über Karzinome der Mundhöhle einschließlich Lippen und Oropharynx wurden anhand eines Datenmaterials von 1021 Patienten 3 wesent-liche Ziele verfolgt: Überprüfung vorhandener Klassifizierungsvorschläge; Analyse prognostisch relevanter Faktoren der Tumorkrankheit; Erstellung eines Prognoseindex zur individuellen und kollektiven Prognosestellung. Die Untersuchungen erbrachten folgende Ergebnisse: ad 1: Die TNM-Klassifizierungen für Mundhöhlenkarzinome weisen schwere Mangel auf und sind deshalb nicht zielführend. ad 2: Die ohne und mit Berücksichtigung der Therapie erfolgten mehrdimensionalen Analysen prognostisch relevanter Faktoren zeigen eindeutig, daß eine verläßliche Prognosestellung nur unter verschiedenen therapeutischen Voraussetzungen erfolgen kann. ad 3: Letzteres führte zur Erstellung des therapieabhängigen Prognoseindex TPI, der den Anforderungen der klinisch-therapeu-tischen Krebsforschung und der klinischen Praxis gerecht wird.

In a retrospective study, on a sample of 1021 patients with carcinoma of the lips, oral cavity and oropharynx, the clinically available factors at the time of the patient's first admission are analysed with reference to their prognostic relevance. Prognostically relevant factors are determined by both univariate and multivariate analyses. Of the 18 factors analysed, the following seven finally proved to be prognostically relevant: tumour size, degree of infiltration, degree of histological differentiation and site by organ of the primary tumour, the combination of evidence + clinical appearance + degree of fixation of the regional lymph nodes, age of the patient, and evidence of distant metastases. On the basis of these results it will be possible to create a prognostic index. This prognostic index should be eligible for use in clinical practice, as opposed to usual classification models.

Nonspecific immune reactivity (lymphocyte stimulation with PHA, PWM; % T-cells; absolute T-cell levels; skin reactivity to DNCB) was determined in 30 patients with squamous cell carcinoma of the head and neck region and in 30 age- and sex-matched healthy controls. The tests were carried out in each patient at 4-week intervals for at least 1 year. The tumor patients were regularly controlled over a period of 5 years. A reduced nonspecific immune reactivity was detected for tumor patients as compared with healthy controls. However, there was no correlation between the follow-up tests on unspecific immune reactivity and the clinical course of the disease. Moreover it was not possible—on the basis of pretherapeutic unspecific immune reactivity and 5 years' clinical follow-up—to make any prognostic statement for the tumor patients tested.

In a multicentric, retrospective observational study on carcinomas of the oral cavity including the lips and the oropharynx a data material of 1021 patients has been analyzed. The specific goals of this study were: Review of existing proposals for classification; Analysis of prognostically relevant factors of the tumor disease; Construction of a prognostic index for the determination of individual and collective prognoses. The following results were gained: None of the TNM-classifications of oral cavity carcinomas does the necessary criteria justice not even approximately. The multivariate analyses of prognostically relevant factors were performed with and without taking therapeutical factors into account. The results unequivocally show that reliable prognoses are only possible under consideration of various treatment modalities. This led to the construction of the treatment-dependent prognostic index TPI, which will be eligible for the use within the clinical-therapeutical cancer research and within the clinical practice.

In a retrospective study of 802 patients with carcinomas of the oral cavity, the most frequently used classifications—AJC 1967, UICC 1973, UICC 1978, and AJC 1978—were analyzed. In this study, the unequivocally defined objective of a classification of malignant tumors, namely, the creation of prognostically homogeneous categories which differ mutually in their prognosis, is not attained by any of the 4 classifications. In several experiments, the possibility of improvements in the T and N categories is demonstrated. In order to achieve a more exact assessment of the prognosis of patients with carcinomas of the oral cavity, however, multidimensional analyses of prognostically relevant factors, with a prognosis index as a goal, are necessary.