The value of bone marrow, liver, and spleen imaging in diagnosis, prognostication, and follow-up monitoring of myeloproliferative neoplasms: a systematic review

Abstract Background Diagnostic and treatment response criteria for the JAK2/CALR/MPL mutation-related myeloproliferative neoplasms (MPNs) are largely based on bone marrow (BM) biopsy results. However, these biopsies have several limitations, such as the risk of sampling error. Also, the prognostic impact of BM abnormalities is largely unclear. Although not currently used in clinical practice, imaging techniques might offer additional information. In this review, we investigated the value of BM, liver, and spleen imaging for diagnosis, prognostication, and response monitoring of the JAK2/CALR/MPL mutation-related MPNs (i.e. essential thrombocythemia (ET), polycythemia vera (PV), and myelofibrosis (MF)). Methods A systematic literature search was performed via PubMed, Embase and the Cochrane Library up to 2020 March 26th. Of 5505 identified records, 55 publications met the eligibility criteria (i.e. containing original data on the imaging appearance of BM, spleen, or liver in adult ET, PV, or MF patients, published in a peer-reviewed journal, written in English). Results Many explorative studies described imaging features, sometimes with comparisons to clinical characteristics. Studies reporting measures of diagnostic accuracy included 1) splenic transient elastography to predict BM fibrosis grade in MF, 2) dynamic contrast-enhanced MRI to discern MF patients from ET patients and healthy controls, and 3) 18-fluorodeoxyglucose PET to detect residual disease after stem cell transplantation in MF. The diagnostic accuracies of radiography and 99mTc-colloid scintigraphy were derived from several other articles. Except for the study on 18-fluorodeoxyglucose PET, we established substantial concerns regarding risk of bias and applicability across these studies, using the QUADAS-2 tool. Three publications described a correlation between imaging results and prognosis, of which one quantified the effect. Conclusions Based on current data, MRI (T1-weighted/STIR, Dixon) seems especially promising for the evaluation of BM fat content - and indirectly cellularity/fibrosis - in MF, and possibly for estimating BM cellularity in ET/PV. 18-fluorodeoxyglucose and 18-fluorothymidine PET/CT might be useful for evaluating BM fibrosis, with good reported accuracy of the former for the diagnosis of residual disease. Further research on these and other techniques is warranted to determine their exact value. Future researchers should improve methodology and focus on evaluation of diagnostic accuracy and prognostic implications of results.