Although Hodgkin lymphoma (HL) is highly curable today, this comes at the expense of treatment-related toxicities, underscoring the need to identify patients who would be candidates for less intensive therapy regimens. There is also a fraction of 10 to 15% of patients who will not be cured by first-line therapy, who will be difficult to manage, and who are equally difficult to identify upfront. The development of effective therapies has taken place despite limited knowledge of only the biology of this disease: this is related to the low abundance of the Hodgkin/Reed-Sternberg (HRS) cell, the malignant cell in HL, which is present as only 0.1% to 2% of the total cells in HL biopsies and is outnumbered by inflammatory cells in the microenvironment. The rarity of these cells has been an obstacle to sequencing or genomic studies of HL. Purification of HRS cells by laser microdissection has been done by some for molecular studies or gene expression profiling. Although it is an elegant approach, it is labor-intensive, time-consuming, and not applicable on a large scale.

Peter Vandenberghe, MD, PhD, Center for Human Genetics, Leuven, Belgium, discussed the noninvasive detection of genomic imbalances in HRS cells in early and advanced-stage HL by sequencing of circulating cell-free DNA (ccfDNA) in plasma. His group applied massive parallel sequencing to ccfDNA in a prospective study of patients with biopsy-proven stage IIA to IVB nodular sclerosis classical HL (NSHL). The pipeline used was developed for noninvasive prenatal testing, allowing genome-wide detection of fetal aneuploidies and segmental imbalances. In one pregnant patient, a complex profile with several genomic imbalances was identified. After exclusion of fetal and maternal constitutional abnormalities, the possibility of a maternal or fetal tumor was considered, leading to a biopsy-proven diagnosis of early-stage (IIA) NSHL in the pregnant mother. To verify the origin of the genomic imbalances, HRS cells in formalin-fixed paraffin-embedded biopsy specimens were investigated by fluorescence in situ hybridization (FISH). HRS cells were identified by the size of their nucleus and CD30 immunostaining; gains of 8q24, 9p24, and 14q were found in these cells, which matched with imbalances in the ccfDNA profile; and this strongly suggested that DNA derived from them was causing the abnormal ccfDNA profile in this patient.

The group then conducted a prospective study in NSHL. In 9 additional patients examined (2 with stage IVB disease, and 7 with stage IIA disease), genomic imbalances were identified in 8 patients by massive parallel sequencing of ccfDNA. The profiles were most pronouncedly abnormal in patients with stage IV disease. The regions that were recurrently imbalanced in this series have all previously been described in the literature based on array comparative genomic hybridization on microdissected HRS cells. These imbalances were also validated by FISH analysis of HRS cells from biopsies from all patients. Although several imbalances were identified as recurrent, they did not occur uniformly: As such, profiling of ccfDNA promises to reveal patient heterogeneity. More patients need to be studied in order to identify patterns of imbalances and their frequency, as well as to tune the technology to the context of noninvasive cancer testing rather than noninvasive prenatal diagnosis.

All patients in this study were treated and reached complete remission. This was paralleled by rapid normalization of the ccfDNA profiles. Therefore, ccfDNA profiling also appears promising for noninvasive disease monitoring.