Potential biomarkers in Hodgkin lymphoma

Biomarker identification to predict prognosis in patients with Hodgkin lymphoma (HL) is challenging. Currently, positron emission tomography (PET) is the most useful predictive biomarker but better biomarkers are needed to prognosticate outcomes.

Characteristics of an ideal biomarker include being minimally invasive, sensitive, specific to a clinical condition, cost-effective, and providing readily available information to the clinician in order to facilitate decision making.

Prognostic factors, currently used at first diagnosis of HL are represented by: levels of serum albumin and hemoglobin, sex, clinical stage, age, white cell count, and lymphocyte count. However, these parameters have limited prognostic value compared to PET.

In a review, recently published in Expert Review of Hematology, Melita Cirillo and Sven Borchmann gave an overview of potential biomarkers for HL and their potential integration into clinical practice.

Evolving biomarkers

• PET-adapted therapy: Interim PET (iPET) is the most sensitive predictive biomarker that could be used to guide treatment decisions. Patients with advanced stage disease who are PET positive after two cycles of chemotherapy (PET-2) with doxorubicin, bleomycin, vinblastine and dacarbazine (ABVD), have a poor outcome. In these patients an initial treatment with bleomycin, etoposide, doxorubicin, cyclophosphamide, vincristine, procarbazine and prednisone (BEACOPP)-escalated (eBEACOPP) therapy and the use of iPET to guide the de-escalation can represent an alternative approach but further data are needed
• Metabolic tumor volume (MTV) as assessed by PET: MTV seems to correlate with baseline clinical risk factors in HL. Recent findings show that MTV could be useful in predicting response to autologous stem cell transplantation (auto-SCT) but additional studies are required. A combination of MTV with soluble biomarkers may improve specificity compared to single biomarkers

Serum biomarkers

Thymus and activation regulated chemokine (TARC) is the most validated blood-based biomarker in HL:

• At diagnosis, baseline levels of TARC correlated with known risk factors and very high levels seemed to be predictive of HL in young patients
• During first-line therapy, normalized TARC correlated with a negative PET-2 and 5-year PFS
• At the end of treatment, higher levels of TARC associated with poorer PFS in PET-2-negative patients
• Post-transplant, higher levels of TARC predicted relapse

Combinations of TARC with other soluble biomarkers, such as sCD163 or IL-10, have also been evaluated and some of these are better predictor of outcomes than TARC alone. TARC represents a good biomarker thanks to its rapid response-kinetics that allow an early identification of treatment responders before the standard PET-2 timepoint. In addition, it is minimally invasive and patients can avoid PET radiation exposure. However, patients with co-existing inflammatory disorders can have high levels of TARC and this represents a limitation of TARC specificity.

Tumor microenvironment (TME)-related biomarkers

• Immunohistochemistry (IHC) and molecular profiling of tumor biopsies are useful to identify prognostic component of the tumor microenvironment. IHC panels detecting Hodgkin Reed Sternberg cells (HRSC) include CD15, CD30, CD45, PAX5, and EBER plus Bob1 and Oct2. Other proposed prognostic IHC markers are cyclooxygenase 2 (expression associated with shortened PFS and OS) and galectin-1 (high expression predicted poorer event free-survival in young patients)
• The most validated prognostic component of the TME is represented by tumor-associated macrophages (TAM). CD68 is the marker used to quantify TAM, elevated TAM populations (> 5%) are associated with poorer PFS and OS. At relapse, elevated CD68+ TAM predict auto-SCT treatment failure. In addition, elevated macrophage content seems to be a predictor of treatment resistance
• Other TME components that correlate with outcome are
  • B cells expressing CD20
  • T regulatory cells expressing FOXP3

• CD20, CD68 and FOXP3 were validated as predictors of survival in newly diagnosed HL. A validation of the methods for routine classification and the optimal method to assess TME are required in order to incorporate TME into routine practice
• Molecular prognostication has been difficult but, thanks to the advances in the field of next-generation sequencing, key pathways involved in the disease pathogenesis, such as NFκB and JAK/STAT, are now recognized
• Recently, a 23-gene signature in patients with advanced stage HL and a 30-gene signature in relapsed HL have been identified as predictors of outcome

Circulating nucleic acids

• In patients with HL, at diagnosis, the viral DNA from Epstein-Barr virus (EBV) associated HL is present in whole blood or plasma and its presence up to 6 months after therapy has been correlated to poorer outcome and relapse
• Circulating tumor DNA (ctDNA) can be used to follow mutations detected at baseline to assess disease response
• Small micro-RNA (miR) fragments contained in extracellular vesicles (EV) detected in plasma of patients with HL are predictors of patient response to treatment
• The combination of ctDNA and EV-small micro-RNA (EV-miR) can represent a good prognostic tool for response and relapse monitoring

Non-tumor cell population

• In a study on a small group of pre-treated patients with HL, an increase in monocyte populations persisting at least 6 months after therapy has been observed. Elevated monocyte populations may be a source of elevated sCD163, a new biomarker (detectable in serum) that can be used at diagnosis and for monitoring patients during therapy
• Dendritic cell subsets, plasmacytoid and monocytic, are reduced in advanced stage, disease bulk, B symptoms and extra-nodal disease, but at the end of the treatment return to normal, suggesting a potential use to monitor therapy response

Conclusions

• PET and TARC are the most validated biomarkers in HL
• PET remains the most useful predictive biomarker in HL but it is a test with limitations. Limitations include low predictive value in early stage disease, and false-positive results due to inflammatory states
• The combination of PET with other biomarkers could be a better prognostic tool than PET alone
• Emerging biomarkers such as MTV, TAM, circulating nucleic acids, and non-tumor cells require additional evaluation
• Genotyping of tumor biopsies would be useful to identify high-risk patients and guide treatment choice
• Circulating nucleic acids in liquid biopsy are promising prognostic biomarkers for response monitoring and detection of measurable residual disease