In a prospective study of 100 subjects, an ELISA assay was used to measure platelet-rich plasma (PRP) serotonin levels in patients with polycythemia vera (PV; n=26), essential thrombocythemia (ET; n=14), myelofibrosis with myeloid metaplasia (MMM; n=26), secondary or spurious polycythemia (SP; n=20) and controls (n=14). Neutrophil PRV-1 expression was concurrently assayed by real-time PCR in 69 patients (23 PV, 17 SP, 12 ET, 7 MMM, 10 controls). Table 1 outlines the median and range of measured values across the different study groups.

Table 1

ControlsSPPVETMMM
Age in years N/A 49 (18–77) 64 (32–83) 52 (18–81) 60 (22–80) 
Plt count x 109/L N/A 249 (157–371) 509 (191–940) 613 (284–1100) 259 (64–955) 
PRP serotonin; ng/109 plts 567 (360–1071) 609 (369–1780) 205 (0–496) 385 (137–1026) 90 (0–278) 
PRV-1/GAPDH ratio 1.28 (1.20–1.43) 1.23 (1.13–1.38) 1.03 (0.84–1.28) 1.24 (0.99–1.44) 1.25 (1.11–1.40) 
ControlsSPPVETMMM
Age in years N/A 49 (18–77) 64 (32–83) 52 (18–81) 60 (22–80) 
Plt count x 109/L N/A 249 (157–371) 509 (191–940) 613 (284–1100) 259 (64–955) 
PRP serotonin; ng/109 plts 567 (360–1071) 609 (369–1780) 205 (0–496) 385 (137–1026) 90 (0–278) 
PRV-1/GAPDH ratio 1.28 (1.20–1.43) 1.23 (1.13–1.38) 1.03 (0.84–1.28) 1.24 (0.99–1.44) 1.25 (1.11–1.40) 

A highly significant difference (p < 0.0001), in PRP serotonin levels, was seen among the different disease categories and control group (table 1). The lowest levels were registered for MMM and PV with a statistically significant difference between the two groups (p = 0.015) but a large degree of overlap. The results in both were significantly as well as markedly lower than those of controls as well as SP with only slight overlap in the PV group and no overlap in the MMM group (p < 0.0001; table 1). The PRP serotonin levels in patients with ET were significantly reduced compared to both controls (p = 0.019) and SP (p = 0.002) and significantly increased compared to both PV (p = 0.0008) and MMM (p < 0.0001). However, there was a substantial degree of overlap in values between ET and other disease categories and controls. Neither aspirin nor cytoreductive therapy correlated with PRP serotonin levels (p = 0.116 and p = 0.148, respectively). Similarly, within a specific disease category, PRP serotonin level did not significantly correlate with platelet count, hemoglobin level, leukocyte count, spleen size, disease duration, or gender. However, an inverse correlation with age was noted only in MMM. In general, there was not a significant correlation between PRP serotonin level and neutrophil PRV-1 expression in each of the three myeloproliferative disease variants. Tests of sensitivity, specificity, and accuracy in distinguishing PV from SP were 92, 95, and 93% for the PRP serotonin assay and 78, 95, 86% for neutrophil PRV-1 assay, respectively. The current study suggests that PRP serotonin concentration might be considered as one of several biological markers that complement each other for the diagnosis of PV.

Author notes

Corresponding author

Sign in via your Institution