BERKELEY, CA (UroToday.com) - Wilms Tumor (WT) is a very common renal tumor in childhood with more than 75% of cases diagnosed before the age of 5 years.
The progress in treatment of nephroblastoma was largely due to interdisciplinary approaches accomplished by cooperative study groups including the Children’s Oncology Group (COG) in North America and the Société Internationale d’Oncologie Pédiatrique (SIOP) in Europe. While the COG treatment protocol starts with primary nephrectomy, the SIOP treatment of children older than 6 months is initiated by preoperative chemotherapy. The primary aims of both protocols are risk-stratified therapies, improved patient outcome, and diminished toxicity. Until now, using the SIOP protocol, diagnosis of WTs is based only on clinical features, patient’s age at presentation, and imaging data. There is a lack of molecular markers that can be used for diagnosis or prognosis.
In this study, we set out to elucidate the autoantibody response in patients with WTs utilizing an array with 1,827 antigen-expressing clones. We screened this array with 54 sera of untreated WT patients collected by SIOP, 75 sera of untreated WT patients collected by COG, 35 sera of treated WT patients collected by SIOP and 50 sera of children without known WT disease. Analysis was automated by newly developed computer aided image analysis software that was tailored to the evaluation of protein macroarrays to identify seroreactivity patterns. Autoantibody–antigen complexes were detected by fluorescence signals. The reactivity of each of the clones against serum autoantibodies was measured by a customized automated image analysis system.
Interestingly, we found differences between the frequencies of antigen reactivity in WT sera and in control sera. To investigate whether the composite autoantibody signatures allow separating WT patients from controls, we applied linear-kernel SVM and performed evaluation with 20 independent replications. We first compared control sera with sera from SIOP patients that had not yet undergone chemotherapy. In this case, a separation with a mean accuracy of 0.82, a mean specificity of 0.83, and a mean sensitivity of 0.82 was achieved. The non-parametric permutation tests showed significantly decreased classification rates with an accuracy of 0.5, a specificity of 0.52 and a sensitivity of 0.48. These rates correspond to random guessing and demonstrate the significant classification performance of our approach. Likewise, comparison between COG sera and control sera also yielded an accuracy of 0.83 while permutation tests showed a classification accuracy of 0.52, essentially confirming the significant results obtained with sera from SIOP. To find the proteins that are the most informative for the separation, we computed the AUC value for each clone. We found 40 clones, including 16 in-frame clones that were informative for the classification of controls, versus SIOP patients that had not yet undergone chemotherapy. For the separation between controls and COG, we found 60 informative clones, including 24 in-frame clones. The best-suited in-frame clone for the classification of control versus SIOP patients showed sequence identity to the gene for ‘‘Amyloid beta A4 precursor protein-binding family B member 1’’ with an AUC value of 0.207. The most informative clone for the separation between controls and COG showed sequence identity to the gene for ‘‘coiled-coil domain containing 92’’ with an AUC value of 0.775. Moreover, we found two clones, namely fascin and zinc finger protein (ZFP) 346, that showed significantly increased reactivity – both with sera of patients recruited by COG and untreated patients recruited by SIOP compared with sera from controls. In addition, we found several in-frame clones for ribosomal proteins and for ZFP transcription factors and several out-of-frame clones for brain creatine kinase, thioredoxin peroxidase, and nuclease sensitive element binding protein 1 that were informative for both the separations between untreated SIOP patients and controls and between COG patients and controls.
As a first attempt to investigate the influence of chemotherapy on the autoantibody signature in WT patients, we studied a subgroup of 35 SIOP patients that underwent chemotherapy prior to serum collection. We compared the auto-antibody response of the treated SIOP patients with the response found in untreated SIOP patients and asked if, and to what extent, the auto-antibody signature allows separation between treated SIOP patients and non-WT controls. Here, we computed an accuracy of 0.68, a specificity of 0.56 and a sensitivity of 0.76, only. Permutation tests showed a still significantly decreased accuracy of 0.51. These results indicate that a separation between controls and WT patients is less accurate after chemotherapy. The less accurate separation between treated SIOP patients and controls is also reflected in the lower number of clones that were informative for this comparison based on their AUC values. Specifically, both fascin and ZFP 346, each of which was informative for the separation between untreated WT patients and controls, were no longer informative for the separation between treated WT patients and controls. Likewise, the out-of-frame clones thioredoxin peroxidase and nuclease sensitive element binding protein 1, both of which were informative for the separations between untreated WT patients and controls, were no longer informative for the separation between treated WT patients and controls. We found 20 clones that were informative for all classifications analyzed, e.g., for the classification between WT patients recruited by SIOP prior to chemotherapy versus controls, for WT patients recruited by COG prior to chemotherapy versus controls, and for WT patients recruited by SIOP after chemotherapy versus controls.
Our study provides first evidence for a WT-specific autoantibody signature and also for specific changes of the autoantibody signature in WT patients as result of chemotherapy.
Jana Schmitt as part of Beyond the Abstract on UroToday.com. This initiative offers a method of publishing for the professional urology community. Authors are given an opportunity to expand on the circumstances, limitations etc... of their research by referencing the published abstract.