In the past few decades Hodgkin lymphoma (HL) has become a highly curable malignant disease, as a result of using modern polychemotherapy and irradiation. Differentiation of active tumor from fibrosis or necrosis within residual radiographic masses represents a problem of interpretation.
Aims: The aim of this retrospective study is to assess the value of FDG-PET for prediction of remission or relapse in HL.
Patients and methods: Data of 128 patients, who had residual masses on CT after completion of their planned treatment, have been analyzed. FDG-PET was performed between January 1995 and February 2005.
Results: The median duration of the follow-up from PET was 75.5 months (range: 20-180 months). 89 (70%) patients had negative and 39 (30%) patients had positive FDG-PET results. The numbers of true-positive, true-negative, false-positive and false-negative subjects were 29, 83, 10 and 6, respectively. Sensitivity of post-treatment FDG-PET was 83%, specificity 93%, positive predictive value 74%, negative predictive value 93%, and accuracy 88%. The difference between the event free survival of PET positive and negative cases is highly significant (p = 0.0000), according to the Mantel-Cox test.
Conclusion: Our results, in accordance with literature, clearly indicate that patients with negative FDG-PET results are unlikely to progress or relapse during a long follow-up. However, false positive uptake is a problem. We have investigated the effect of age, histological subtype, clinical stage and the type of treatment on the accuracy, but on the basis of these facts we could not find any significant difference. However, the date of the investigation influenced the results: before 2000 the number of false results was significantly higher than after that time, which shows the importance of investigators' experience.