Splenomegaly is a common clinical feature at diagnosis or during clinical course of chronic myeloproliferative neoplasms (MPN). It has been observed in approximately 30% of patients with Polycythemia Vera (PV), from 5% to about 20% in Essential Thrombocythemia (ET) patients, and in more than 50% in Primary Myelofibrosis (PMF) patients. Spleen enlargement can be the ultimate cause of cytoreductive treatment start or failure. Splenomegaly evaluation by physical examination lacks both accuracy and reliability. IWGMRT and ELN experts recommended objective spleen response assessments in MPN patients by magnetic resonance (MR).
This study evaluated whether spleen volume (SV) measured by ultrasonography (US) is comparable to that measured by MR in MPN patients.
Adult patients with MPNs diagnosis who underwent to ruxolitinib treatment were recruited in this prospective study. All subjects received both US and MR of the spleen before ruxolitinib starting (baseline) and at 32 weeks of treatment.
Spleen US was performed as described by Picardi et al. (Blood. 2002;99(11):4228–30.), using an EPIQ 5 Philips instrument with a 1–5 MHz broadband curvilinear probe. Perimeter, longitudinal diameter, and area, defined as the maximum measurements with splenic borders and angles clearly defined, were measured, and volume (in mL) was calculated automatically. MR calculated SV using a computer-assisted method based on Sorensen et al. (J Clin Oncol. 2001;19(2):551–7), as previously described in RESPONSE study. SV was obtained by outlining the circumference of the organ and determining the volume using the validated technique of last squares. We administered a patient satisfaction questionnaire regarding both exams as previously described (Clin Radiol. 1995;50(3):137–43).
US volume estimations were compared with MR volume estimations at the baseline and during follow-up using Sperman's rank correlation.
Patients distribution among MPN diagnosis (WHO 2016) was as follow: 14 PV, 9 pre-MF, 9 overt MF, 4 post-PV MF and 4 post-ET MF. All patients received scheduled treatment with ruxolitinib. At baseline, the median SV assessed by US was 600 ml (range 200–5000 ml) while median SV evaluated by MR was 553.1 ml (range 172–5140 ml). There was a good correlation between US and MR volume measurements, with spearman r coefficient of 0.957 (CI 95%: 0.918–0.978), p value <0.0001. On the other side, the median spleen longitudinal diameter (LD) evaluated by US was 14 cm (range 8–30 cm) whereas median spleen LD measured by MR was 15 cm (range 7–31 cm). Splenic length also showed a strong correlation between US and MR measurements, with spearman r coefficient of 0.906 (CI 95%: 0.824–0.951), p value <0.0001.
At week 32, the median SV assessed by US was 415 ml (range 130–4400 ml) while median SV evaluated by MR was 380.2 ml (range 111–4790 ml). There was a good correlation between US and MR SV measurements, with spearman r coefficient of 0.943, p value <0.0001. The questionnaires showed a greater appreciation for US in terms of exam duration and anxiety.
Baseline SV evaluation is important for MPN diagnosis. Significant spleen reduction responses of ruxolitinib in MF and PV patients have been previously assessed by MR. This technique appeared not applicable in routine clinical practice due to cost and exam duration. SV assessment by US is a cheap, rapid and a positively accepted substitute for initial diagnosis and SV response assessment in patients treated with ruxolitinib.