Marieke van Son

98 CHAPTER 5 on MRI (50% versus 25% ³T3) than non-failing patients. MRI stage ³T3 and higher PSA were significantly associated with poorer biochemical control. Larger CTV was also a significant risk factor, but this was related to T-stage (p=0.003 by ANOVA test). For the “low risk of failure” group within our cohort, BDFS was very good compared with other focal salvage cohorts, where a vast majority of patients were “low risk”. 14,15,18,19 Another tumor characteristic potentially associated with treatment failure is radio- resistance. To evaluate this, we compared within-patient primary and recurrent tumor localizations. Out of 23 primary MRI scans available, 16 patients (70%) had an in-field recurrence with respect to the primary tumor. Stratifying for in- and out-of-field lesions, there was no clear relation between treating an in-field recurrence and salvage treatment failure: 9/16 (56%) “in-field patients” failed and 4/7 (57%) “out-of-field patients” failed. For local staging, additional functional imaging sequences (most importantly dif- fusion weighted imaging [DWI] and dynamic contrast enhanced [DCE]) have improved accuracy. 20 Although correct interpretation is challenging and should be performed by expert uro-radiologists, mp-MRI can aid in focal salvage treatment planning, intra-op- erative guidance and follow-up. 21 Studies comparing mp-MRI assessment to prostate biopsies (n=52) or full histologic prostate mapping (n=13) in patients with a suspected local recurrence report area under the receiver-operator curve (AUC)-values of 0.82-1 and 0.8-0.9, respectively. 22,23 For the detection of extra prostatic extension (EPE) and SVI the sensitivity and specificity are less optimistic, ranging between 50-75% and 70- 100%, respectively. 24 PSMA-PET/CT is the indicated modality for metastatic disease assessment, albeit with correct timing: a study on 248 patients with rising PSA reported detection rates of 58% at PSA-levels of <0.5 ng/mL versus 97% at PSA-levels of ≥2 ng/mL. 25 Two studies evaluating 30 and 65 patients with biochemical recurrence and PSMA-suspected lymph node metastases who underwent (extended) salvage lymph node dissection, reported positive and negative predictive values of 100% and 89-100% respectively. 26,27 For ultrafocal salvage HDR-BT, we delineate the recurrent lesion using both mp-MRI (in particular T2-hypointense and restricted diffusion areas) and PSMA-PET/CT. When these images do not overlap, the GTV is extended to encompass suspected areas from all sequences. This approach is supported by recent reports showing mp-MRI signifi- cantly underestimates the tumor volume, where PSMA-PET/CT has better agreement with histopathology. 28,29 All in all, it seems we can adequately exclude metastatic disease but tend to under- estimate the extent of local recurrence. Although we add a 5 mm margin to the GTV to account for this, going from ultrafocal to quadrant treatment could potentially improve oncologic control even further. Dose escalation or dose fractionation could also improve oncologic outcomes, especially considering our relatively high number (73%) of in-field recurrences, of which 13/16 had been underdosed (D95% <19 Gy) and 4/16 did not reach the lower threshold of D90% >17 Gy. By fractionating, tumor cells may become more susceptible to the next radiation course because of enhanced tumor cell division during