According to the eForms team's response to this issue OP-TED/eForms-SDK#1187, the properties epo:isWithdrawn, epo:hasWithdrawalDate, epo:hasWithdrawalReason and epo:hasReviewRequestFee should be on the class epo:ReviewDecision , NOT on the epo:ReviewRequest class.
If the ePO WG agrees with this modelling, we can update the mappings of fields BT-795-Review , BT-796-Review, BT-797-Review and BT-798-Review, which will resolve the SPARQL validation problems described in this issue OP-TED/ted-rdf-mapping-eforms#99, but will not resolve the related SHACL violations (in ePO 4.0.0).
If this is fixed in ePO 5.0.0, then in the future, when we will map to ePO 5.0.0, the SHACL violations will disappear as well.