INTRODUCTION: Identification of predictive biomarkers of response to new agents represents a challenging issue. Thymus and Activation-Regulated Chemokine (TARC) is a protein highly expressed by Reed-Sternberg cells in Hodgkin lymphoma (HL) and has been demonstrated to be a prognostic biomarker in HL patients treated with chemotherapy. We investigated the correlation between clinical response and early variation of serum TARC levels in patients with refractory/relapsed HL treated with kinase inhibitors. Patients were enrolled in two consecutive phase II trials evaluating activity and safety of the multikinase inhibitor Sorafenib (Nexavar®, Bayer) used as single agent, or in combination with the AKT inhibitor Perifosine (Æterna Zentaris GmBH, Germany).

METHODS: Between July 2008 and November 2011, 30 patients (19 males and 11 females; median age 30 years, range, 17-73 years) with relapsed (n = 3) or refractory (n = 27) HL were analyzed. In one trial, patients received Sorafenib alone (400 mg BID, per os), and in the second trial Perifosine (50 mg BID, per os) for 4 weeks followed by combination therapy of Perifosine plus Sorafenib (400 mg BID, per os) until disease progression or clinically significant toxicity. Tumor responses were assessed according to the revised response criteria for malignant lymphoma of the International Working Group. Serum samples collected prior to therapy initiation and monthly thereafter were analyzed for TARC by ELISA. Early variation in TARC levels was calculated as percentage difference between values at baseline and at day 60 after starting therapy. Comparison of variation of TARC between different groups of patients was calculated using the Mann-Whitney Test, correlation between clinical response and TARC reduction was calculated using Fisher’s exact text and presented also as relative risk (RR). The study was approved by the Institutional Ethical Committee

RESULTS: Nine out of 30 patients (30%) achieved clinical responses, including complete response (CR, n = 1) and partial response (PR, n = 8), 12/30 patients achieved stable disease (SD), and 9/30 patients experienced progressive disease (PD). Median overall survival (OS), progression-free survival (PFS) and duration of response (DOR) were 18, 4.5 and 3 months, respectively. Data of TARC at baseline and during therapy were available for all 30 patients but only 29 patients were evaluated. The patient achieving CR who showed baseline and 60-day values below 500 pg/ml (212 pg/ml and 368 pg/ml, respectively) was in fact excluded from analysis. At baseline, the median TARC level for the whole population was 13451 pg/ml (range, 1010 - 89877). No differences in baseline TARC levels were detected between patients achieving PR, SD and PD as best response. The median percentage variation of TARC levels after 60 days of therapy in PR, SD and PD patients was -57% (IQ range, -79% to -28%), -15% (IQ range, -78% to +8%) and +6% (IQ range, -49% to >100%). Variation of TARC observed among PR patients was significantly different from variation detected in PD group (p=0.005), whereas no differences were observed between PR and SD patients. A decrease of TARC levels at day 60 was observed in 17 patients, median percentage reduction was 49% (range, 8% – 94%). TARC reduction discriminated patients sensitive to treatment, in fact 15/17 patients (88%) with evidence of a reduction achieved PR (n=7) or SD (n=8), whereas 7/12 patients showing no change or increasing values of TARC experienced PD (p=0.014). The chance of obtaining PR and SD was greater in patients with evidence of TARC reduction than in patients with stable or increasing values (RR 3.37, 95%CI, 0.97 – 11.7). When we considered the median percentage reduction of 49% as cut-off, this value didn’t discriminate PR from SD patients. Similarly, deeper reduction in TARC levels was not correlated to longer PFS or DOR in responding patients.

CONCLUSIONS: Analysis of TARC in HL patients treated with perifosine and sorafenib allowed to identify a biomarker predicting response to treatment with kinase inhibitors. Early reduction of TARC seems to be a useful tool for selecting patients who will receive a clinical benefit from therapy. However, correlation observed between early reduction and clinical response need to be validated in larger studies.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.

Sign in via your Institution