Verification and comparison of two different predictive equations in Hodgkin's disease

Abstract

BACKGROUND AND OBJECTIVE: In recent years, two predictive equations to estimate median expected survival at diagnosis for patients affected with Hodgkin's disease have been developed at the University of Pavia Medical School. The present retrospective work was aimed at testing correlation between mean survival estimated using the two equations and observed survival, and at comparing the results of the two different equations. METHODS: Fifty-three deceased patients were considered from a series of 114 consecutive ones. All these patients had been treated in a conventional way according to therapeutic modalities similar to those used in the series from which the two equations were derived. Expected median survival values calculated with the older, linear equation and with the newer exponential one were compared with observed survival. RESULTS: Mean survival of the whole series was over 24 years, with survival probabilities of 85% after 5 years and 74% after 10 years. Using the first predictive equation on the 53 deceased patients resulted in a satisfactory correlation between estimated median survival and real survival: Pearson's R correlation coefficient value is 0.5996, with a t value of 5.35 and p < 0.001. The more recent exponential predictive equation showed a better correlation between estimated median survival and observed survival: R = 0.7338, t = 7.71, p < 0.001. INTERPRETATION AND CONCLUSIONS: The new exponential equation, while apparently complex, is superior to the older one, and is a very reliable and straightforward tool for estimating median expected survival: its forecast proves to be an important pretreatment parameter in every HD patient. These observations support widespread use of this tool in clinical practice to evaluate the prognosis of Hodgkin patients in a more accurate and flexible way.