Skip to main content
Premium Trial:

Request an Annual Quote

A Biomarker for Survival


Some Hodgkin's lymphoma patients respond well to treatment while others relapse. Randy Gascoyne, a hematopathologist and researcher at the British Columbia Cancer Agency, and his colleagues wanted to know why that is. Genome Technology's Christie Rizk spoke to Gascoyne to discuss his recent study, published in the New England Journal of Medicine. What follows is an excerpt of their conversation, edited for space and clarity.

Genome Technology: Why did you take this approach to studying Hodgkin's disease?

Randy Gascoyne: We were trying to determine molecular markers that distinguish treatment outcome. We believed that the answer might lay in the microenvironment. Hodgkin's lymphoma is very unusual because essentially 99 percent of the biopsy is benign and the malignant cells constitute only 1 percent of the tumor. The hypothesis is that the malignant cells somehow cultivate the micro-environment in favor of the tumor, so-called protumoral immunity. So we built a study around the concept that the answer to distinguishing people who do well after primary therapy, in contrast to the 15 or 20 percent who do not, might indeed be found in the microenvironment. We reasoned that by doing gene expression profiling on diagnostic biopsies, we might see a signal from the microenvironment that might give us clues.

GT: What was your approach?

RG: We identified diagnostic biopsies in our database that were in the freezer and were able to identify quite a large number of cases. We pooled 130 cases of diagnostic biopsies of classic Hodgkin's lymphoma for which we had clinical information, and they were dichotomized into 92 treatment successes and 38 treatment failures. We extracted RNA from those cases and did Affy one-cycles [a standard protocol] using U133 arrays, then analyzed the data in a supervised manner. From this analysis, we discovered a number of signatures that traveled with treatment outcome, but one that jumped out at several levels was the tumor-infiltrating macrophage-type signature. We then took 166 independent cases and we built a tissue microarray. We enriched it for failures and decided the easiest thing to do was straightforward immunohistochemistry for CD68. We were able to count the macrophages, which showed a dramatic difference between people who were successful in their initial therapy versus those who were not.

GT: Can you quantify the effect of macrophages on successful treatment versus failure?

RG: When we analyzed the data, there was quite a striking difference between the two clinical extremes. The thing that was particularly exciting was that when we focused on the patients with limited-stage disease, the distinction between having a few macrophages versus having a lot was 100 percent survival versus about 60 percent. So in particular for limited-stage disease, it looked like a very good predictor of people who will do very well after their primary therapy.

GT: Using these findings, do you think clinicians could accurately predict a patient's chances of survival or recurrence?

RG: I think the answer is yes. It needs, though, to be validated in another cohort. However, if it does validate, then I think it offers the possibility of a useful biomarker to identify patients at risk. The other interesting thing is that it identifies those patients who are likely to do extremely well, and so the study also offers information on those patients who may be able to have their therapy dialed back with the goal of reducing long-term sequelae such as secondary tumors, which can be the consequences of those treatments. Also, by testing the original diagnostic biopsy we were not only able to predict the results of success versus failure after primary therapy, but also very successful at predicting the response to salvage therapy like bone-marrow transplantation.

The Scan

Study Links Genetic Risk for ADHD With Alzheimer's Disease

A higher polygenic risk score for attention-deficit/hyperactivity disorder is also linked to cognitive decline and Alzheimer's disease, a new study in Molecular Psychiatry finds.

Study Offers Insights Into Role of Structural Variants in Cancer

A new study in Nature using cell lines shows that structural variants can enable oncogene activation.

Computer Model Uses Genetics, Health Data to Predict Mental Disorders

A new model in JAMA Psychiatry finds combining genetic and health record data can predict a mental disorder diagnosis before one is made clinically.

Study Tracks Off-Target Gene Edits Linked to Epigenetic Features

Using machine learning, researchers characterize in BMC Genomics the potential off-target effects of 19 computed or experimentally determined epigenetic features during CRISPR-Cas9 editing.