Imaging of Head and Neck Cancer Patients Before, During and After Radiotherapy

Abstract

This thesis shows that before, during and after radiotherapy medical imaging can aid the treatment of HNSCC patients. This appears to be particularly true for MR imaging. It also shows that imaging has to be applied correctly in order to be effective. In part one, the role of imaging prior to the start of treatment is discussed. The role of FDG PET-CT for the detection of metastatic lymph nodes is examined in chapter 2. Metastatic lymph nodes are routinely diagnosed using cytology acquired by USgFNAC. However, in patients that already received a PET-CT, for example to detect distant metastases, some information on the status of the lymph nodes is already available. By using the information freely available from the PET-CT around half of all lymph nodes could be spared an unnecessary USgFNAC examination. In chapter 3, the power of DW-MRI in the delineation process is examined. Although the conformity of the delineations was good, indicating that DW-MRI could be helpful during the pretreatment delineation phase, we also saw clear instances with a different interpretation of the images. Chapter 4 introduces ADC as a quantitative measurement. We show that the TNM stage, is significantly more powerful as a prognostic factor than pretreatment ADC. Part two starts by looking into ADC and in particular the change in ADC as it is measured during treatment. A prospective trial (the PREDICT trial) is described in chapter 5. PREDICT has the primary aim of predicting treatment outcome of HNSCC patients using the change in ADC induced by the radiation therapy. Chapter 6 illustrates a potential issue with MRI tumor identification and delineation during treatment. The change in the appearance of tumors on T2 weighted imaging hinders delineation and consequently reduces interobserver agreement. The most effective way to address this issue is probably to create clear and easy to follow guidelines on the delineation of tumors during treatment. In the last part of this thesis, part 3, imaging in the posttreatment phase is presented. Chapter 7 discusses a prospective comparative study of MRI with diffusion weighted sequences and PET-CT, the RETURNED study. Both are used to detect local recurrence. The diagnosis of recurrent HNSCC is significantly more challenging than the diagnosis of primary disease. This is due to the difficulty of differentiating between post treatment effects and recurrent tumor. In the study, MRI and PET-CT were both able to detect recurrences with an accuracy of around 70%. However, MRI showed a higher specificity but a lower sensitivity than PET-CT. Unfortunately, by adding the two modalities together the accuracy of detection did not improve much. Therefore we advocate that PET-CT, already routinely used for recurrence detection, should remain the modality of choice if there is clinical suspicion of recurrent tumor. A reason for the lack of accuracy of the MRI with diffusion weighted images is sought in chapter 8. Here, ten radiologist tried to determine if the MRI images of ten RETURNED patients showed a local recurrence or local control. The interobserver agreement was moderate. Other studies reported also a similar interobserver agreement for PET-CT and also the presence of easy and difficult cases. Perhaps that MRI examinations performed at regular intervals could improve the detection of recurrent MRI, however this was not examined in this thesis. Finally, chapter 9 describes a framework for implementing MRI in the radiotherapy process. This is relevant as innovations in the field resulted in hybrid radiotherapy and MRI devices that are currently being introduced to the clinic. It provides answers to questions a radiation oncologist or clinical physicist might have when introducing MRI for HNSCC radiotherapy treatment in their institutions.