Enrico Martin

293 Summary consensus among all surgeons. Although some differences may in part be explained by specialty bias, significant differences in work-up may occur based on initial presentation. A multidisciplinary approach combining knowledge from both surgical oncologists and peripheral nerve surgeons may be beneficial. Chapter 6 Diagnostic Accuracy of Non-Invasive Tests MPNSTs are impossible to differentiate from benign nerve sheath tumors (BPNSTs) based on clinical presentation. Moreover, sampling errors are common and repetitive biopsies are cumbersome and possibly damaging. It is therefore crucial to find non- invasive ways to detect tumors with the highest probability of malignant transformation. MRI’s are widely used, but PET scans possibly offer additional value in NF1 patients. For this reason we systematically reviewed current literature of non-invasive tests that can be used to distinguish MPNSTs. This included conventional MRI, functional MRI, PET scans, and liquid biopsies. Diagnostic accuracy was quantified for several MRI and PET characteristics with Bayesian bivariate meta-analyses. These analyses allow for heterogeneity in threshold values even when combining a smaller sample of studies with few patients. The absence of a target sign was highly sensitive with a pooled negative likelihood ratio of 0.04 indicating high certainty of a BPNST when the target sign is present. However, the absence of a target sign is highly unspecific and additional characteristics such as ill-defined margins and perilesional edema should be taken into account to distinguish benign and malignant lesions. In NF1 populations PET scans offer higher accuracy. SUVmax and tumor-to-liver ratios are equally accurate. Ideal thresholds are lacking, but based on individual level data of 246 patients, meta-analyses suggested an ideal threshold for SUVmax at ≥3.5. Functional MRI’s may provide equal accuracy as PET scans and are therefore of special interest in the sporadic patient population, but their implementation requires more research. Liquid biopsies currently hold no role in diagnostic work-up of MPNSTs but may potentially gain interest in the future. Chapter 7 Emerging Therapeutic Targets Current cytotoxic systemic treatment options yield limited responses and new therapeutic targets are desperately warranted. By means of a systematic review we summarized and investigated current literature of all non–cytotoxic treatment possibilities in MPNST. We included 60 in vivo studies and found that targeting the PI3K/ Akt/mTOR pathway or vascularization may be promising, as well as the use of oncolytic viruses. Of 6 published trials, none has yet shown effective in MPNST. Currently 13 trials are still ongoing, recruiting MPNSTs in various degrees. Hopefully, some will provide further insights. A combination of drugs will most likely be pivotal to maximize treatment effect, because of the complex and heterogeneous biology of MPNSTs. A