Safety and diagnostic yield of splenic core biopsy; a methodical approach using combined Haematology/Radiology assessment in a tertiary referral centre

Abstract

Core biopsy of the spleen is performed less often than in other solid organs due to a perceived increased risk of post-procedure haemorrhage. In the majority of cases of splenomegaly or abnormalities of splenic tissue, a diagnosis can be reached using blood, bone marrow and/or lymph node analysis without the need for splenic tissue sampling. However, a subset of patients remains in whom prior investigations have not yielded a diagnosis and have persistent symptoms or cytopenias that might relate to a disease involving the spleen. Recently, there has been interest in the use of radiologically-guided core biopsies of the spleen, which have the advantage over traditional fine needle aspiration (FNA) of preserving tissue architecture, allowing more detailed analysis and accurate grading of malignancies. The procedure has been reported to be sensitive and safe with high diagnostic yield (Cavanna et al, 1992; Lucey et al, 2002; Liang et al, 2007). We report a prospective series of 30 consecutive patients who underwent splenic core biopsy as part of the investigation of hitherto unexplained splenic abnormalities. All patients were referred and reviewed through a single pathway from Physicians, Surgeons and Haematologists in the West of Scotland. To be considered for biopsy, patients had to show on-going symptoms that might be related to spleen size or an associated disease or show persistent or progressive cytopenias. All patients underwent an imaging review by a dedicated radiologist to identify other features that might explain splenomegaly (e.g. portal hypertension) and to identify possible alternative biopsy sites. In cases with indeterminate splenic lesions we utilized positron emission tomography/computed tomography (CT) imaging, with patients showing focal or diffuse splenic fluorodeoxyglucoseuptake proceeding to biopsy. A dedicated haematologist reviewed clinical and laboratory databases to ensure all other appropriate investigations for potential haemolytic, lymphoproliferative and myeloproliferative disorders (blood film, blood flow cytometry, bone marrow aspirate and trephine biopsy as a minimum) were complete. All biopsies were performed by a single dedicated radiologist at Gartnavel General Hospital between 2008 and 2018. Patients had baseline full blood count, coagulation screen, biochemistry and group-and-save specimens taken. A normal coagulation profile and platelet count >50 9 10/l were deemed essential. Procedures were performed from an inpatient bed with facilities for overnight observation. Twenty biopsies were performed under ultrasound and 10 under CT guidance. The core biopsies were all 18-gauge and obtained with a BioPince device (Argon Medical Devices, Frisco, TX, USA) under local anaesthesia; no coaxial device was utilised. Between 2–5 cores were taken from each patient. All specimens were reported by a dedicated haematopathologist. The biopsies were fixed in 10% buffered formalin and embedded in paraffin wax. An early 4-lm section was stained with haematoxylin and eosin and reviewed, with appropriate immunohistochemistry then performed. In some cases, molecular testing (polymerase chain reaction for clonality studies, fluorescence in situ hybridisation for detection of translocations) was performed. Patient demographics and selected laboratory parameters in relation to the type of splenic abnormality (focal or diffuse) are summarized in Table I. Ten patients had a history of treated malignancy, most typically lymphoma. All biopsies showed viable cores of splenic tissue origin. None were considered insufficient for interpretation by the reporting haematopathologist. A definitive diagnosis was reached in 24 patients. The most frequent disease entity was lymphoma (Table II). Of the 5 patients who had a “reactive” spleen, 2 went on to have splenectomy for persistent symptoms, with tissue revealing a diagnosis of littoral cell angioma in one patient and sclerosing angiomatoid nodular transformation of the spleen (SANT) in the other. The other 3 patients with a “reactive” spleen were managed with observation with no subsequent evidence of malignancy (follow-up 18–64 months). One patient found to have fibrotic change only has been managed with observation (follow-up 8 months). One patient had a bleeding episode from a biopsy taken from a lesion close to the splenic hilum on the evening of the procedure. This was confirmed on CT imaging and effectively treated by radiology-guided embolisation. The patient received supportive treatment including blood transfusion and made a full recovery. No other post-procedure complications were identified and all other patients were discharged the following day. We have demonstrated that radiologically-guided splenic core biopsy has a high diagnostic yield in patients with unexplained splenic abnormalities. Twenty of the 30 patients (67%) had treatment initiated following the biopsy for diseases (malignancy, infection, sarcoidosis) that would not have been possible to diagnose without splenectomy. We feel that the careful patient selection process and consistent Correspondence