Preview Mode Links will not work in preview mode

Journal of Clinical Oncology recognizes that readers do not always have time to review an article in depth, and yet they still wish to understand how the results will influence their clinical practice or research. To address this need, we offer podcasts that will enhance the readership experience by presenting the key results of high-profile publications in a convenient audio format. Our podcasts are designed to place selected articles into a clinically useful perspective that is easy to listen to in the office or while on the road.

May 21, 2019

This JCO Podcast provides observations and commentary on the JCO article “PET Score Has Greater Prognostic Significance Than Pre-Treatment Risk Stratification in Early-Stage Hodgkin Lymphoma in the UK NCRI RAPID Study” by Barrington et al. My name is Brue Cheson, and I am at Georgetown University Hospital, Lombardi Comprehensive Cancer Center. My Hematologic-oncologic specialty is Lymphoma.

 

Hodgkin lymphoma is clearly one of the most dramatic success stories in modern oncology. More than 90% of patients with limited disease and about 85% with advanced disease are cured using conventional chemotherapy regimens.  As a consequence, current clinical trials are focusing on augmenting or modifying treatment for those at higher risk and decreasing the intensity or duration of therapy for those at a lower risk of treatment failure.

 

One important question has been: how best to distinguish those disparate groups?  Over the years, various prognostic scoring systems have been devised.  The International Prognostic Scoring System (IPSS) differentiated patients into 6 groups using 7 clinical and laboratory factors.  However, only 7% of patients are in both the most and least favorable groups.  The German Hodgkin study Group (GHSG) and the EORTC each published criteria slightly different from each other for treatment selection.  Nevertheless, it is not clear that any of these schemas remains relevant in the context of current Hodgkin regimens.  More importantly, they do not reliably dictate how to treat patients, nor do they offer therapeutic targets.

 

FDG-PET scanning has revolutionized our management of patients with lymphoma.  In 2005 we first demonstrated that integration of PET into standard response assessment improved the ability to distinguish between residual tumor and fibrosis in patients with diffuse large B-cell lymphoma, leading to a revision of standardized response criteria. More recent studies have confirmed this observation in Hodgkin lymphoma and other histologies. Patients with advanced Hodgkin lymphoma can be distinguished into high and low risk groups based on PET scan results after 2 cycles of standard ABVD chemotherapy, regardless of their pretreatment IPSS score.  In a number of studies, reacting to the positive interim scan by intensifying therapy achieved outcomes markedly improved over expected.

 

In the paper that accompanies this podcast, Sally Barrington and her colleagues performed a secondary analysis of the RAPID trial to evaluate the role of pre-treatment risk factors and PET results in predicting outcome of patients with early stage Hodgkin lymphoma.  This study accrued 602 patients who were treated with standard ABVD and underwent PET scanning after the third cycle.  Those with a negative scan (a Deauville score of 1-2) were randomized to no further treatment vs involved field radiotherapy.  Despite a failure to demonstrate non-inferiority of progression-free survival in this cohort, the overall survival was the same, thus sparing 90% of patients unnecessary radiotherapy.  Those with a positive scan (defined as a Deauville score of 3-5), the primary focus of the current manuscript, received an additional cycle of ABVD plus radiotherapy.  Only the 21 patients with a Deauville score of 5, defined as an SUV at least 3 times greater than that of the liver, had an inferior time to progression or greater risk of Hodgkin-related death.  Importantly, this finding was independent of pretreatment prognostic factors using either the GHSG or EORTC scores.  Whether this observation can be extrapolated to patients with features not eligible for the RAPID study, such as those with bulky mediastinal disease or B-symptoms, is presently unknown.  Nonetheless, these data support the role of metabolic imaging over standard clinical and laboratory risk factors.

 

But we are clearly doing this all wrong.  Why do we treat all patients the same and then wait until the disease has demonstrated resistance before we alter therapy?  Several recent papers support the notion that anticipatory, biologically-based, risk-adapted approaches may be feasible. Pre-treatment total metabolic tumor volume (defined as the sum total of all metabolically active lesions) can predict outcome in Hodgkin lymphoma as well as follicular, diffuse large B-cell and primary mediastinal large B-cell lymphoma.  High heterogeneity of intratumoral FDG uptake distribution on PET-CT may be a marker of chemoresistance in solid tumors as well as various lymphoma histologies.  Unfortunately, those tests do not provide a target against which to direct a specific agent.  In contrast, a number of investigators have demonstrated a correlation between bcl-2, p53, FOXP3, CD68, STAT1, pattern of PD-1 expression, mutational patterns derived from next generation sequencing, and other factors in pre-treatment Hodgkin node biopsies and outcome.  Thus, if we are able to predict outcome prior to treatment, why do we expose patients to drugs to which we know they will not benefit? The goal of treatment should be anticipatory, risk-adapted strategies whereby patients with a high likelihood of benefit may receive standard of care, unless there is another clinical question being addressed.  On the other hand, those unlikely to benefit as determined prior to therapy should be spared the waste of time and toxicity and treated with novel regimens directed at specific targets.  Both groups should be monitored during treatment for the emergence of mutations, with treatment altered accordingly.  Yes, we may be a long way from having the appropriate tools for such an approach.  But, to quote the geneticist, molecular engineer and chemist George M. Church, “The best way to predict the future is to change it”.  Anticipatory, risk-adapted strategies could do just that.

 

This concludes this JCO Podcast. Thank you for listening.