Abstract P033

Final analysis of correlation between serum TARC concentration and MTV,TLG in a cohort of Classical Hodgkin lymphoma during first line treatment.

TARC (Thymus and activation-regulated chemokine) is produced by Reed-Sternberg cells in classical Hodgkin's lymphoma (cHL). Correlation between treatment response and serum TARC (sTARC) concentration has been described in several studies. The aim of this one is to evaluate correlation between sTARC and PET variables [metabolic tumour volume (MTV) and total lesion glycolysis (TLG) during first line treatment]. Plasma samples were collected, at baseline, after 2 cycles (corresponding with interim PET, iPET), and at the end of treatment (EOT). Thresholds used for measuring MTV and TLG were SUVmax>2.5 and 41% of the SUVmax. To assess iPET and EOT response, variables were evaluated as logarithmic reduction (LogRED) of baseline vs iPET, and as logarithmic variation (LogΔ) of iPET and EOT. Logβ, logarithmic reduction of baseline vs EOT was added to evaluate pts receiving BV-AVD, being the role of iPET unknown.

We enrolled a total of 74 cHL pts: 6 (8%) and 12 pts (16%) were excluded due to missing samples and unavailability of PET images respectively. Total evaluable pts were 56 mostly advanced stage disease (92%). 6 patients (11%) and 9 (16%) patients were iPET and EOT-PET positive. 70 pts (95%) received ABVD regimen, 4 (5%) received BV-AVD which were not evaluated for logRED and LogΔ. A total of 52 pts was evaluable for logRED, 50 and 56 for LogΔ and Logβ. LogΔ and Logβ of sTARC were significantly different in EOT+ versus EOT- pts (p=0.0174 and p=0.0092), but not for LogRED (p=0.239). LogRED, LogΔ and Logβ of PET variables were significantly lower in iPET+ and EOT+ pts compared to iPET- and EOT- pts (LogRED p<0.001; LogΔ p=0.0001 and p=0.0003 for MTV 2.5; Logβ p<0.0001). The correlation between PET variables and sTARC showed a significant trend for LogRED using both thresholds for MTV and TLG, Fig1. Likewise, LogΔ (r=0.5328, p<0.0001 TLG 2.5; r= 0.5012, P=0.0002 for TLG 41%, r=0.5159, p<0.0001 for MTV 2.5 and r=0.4929, p <0.0003 for MTV 41%) and Logβ (r=0.3857, p=0.0040 TLG 2.5; r=0.3697, p=0.0059 for TLG 41%; 0.3783, p= 0.0048 for MTV 2.5 and r=0.3592, p =0.0076 for MTV 41%) were significantly correlated with sTARC.

The current study shows the deep interconnection between PET variables and sTARCwhich adds prognostic relevance in identifying iPET-/EOT-PET+ pts. As far as EOT PET, sTARC can be used as a useful biomarker also with BV-AVD regimen. The prognostic role of TARC should be evaluated in larger studies.