Among the piles of journals that invade our offices weekly, advancing on the sanctity of clean desk space like kudzu, occasionally—very occasionally—an article pops out that upends the conclusions we have always drawn from clinical trials.
Maybe you saw it. Not in the blue journal with the white writing, or the black journal with the enlarged cell, or the oversized white journal you're reading now, but in the other white journal from the Massachusetts Medical Society (you know, the one that senselessly keeps rejecting every manuscript I send them), there was a study with long-term follow-up in patients with Hodgkin Lymphoma, from the National Cancer Institute of Canada (NCIC) and the Eastern Cooperative Oncology Group (ECOG).
Basically, from 1994 to 2002, Hodgkin Lymphoma patients who had localized disease (non-bulky, Stage I-IIa) were randomized to receive chemotherapy with adriamycin, bleomycin, vinblastine, or dacarbazine (ABVD) for four to six cycles, or to receive radiation therapy with or without two cycles of ABVD chemotherapy. Whether or not these patients achieved a complete response was assessed, as was long-term freedom from progressive disease and overall survival.
Now, when I was in medical school, and then in residency and fellowship, I was taught that five-year outcome was the Holy Grail. In other words, if you have cancer, and you're treated for that cancer with surgery or chemotherapy or radiation therapy or a combination, and you're alive five years after your diagnosis without that cancer returning, you've won the game: that's the first time we get to use the “C” word (Cure, not Cancer—we use Cancer all the time).
Then, we realized that certain cancers can recur late, and five years probably isn't the best marker. This is true of slower-growing tumors, like breast or prostate cancer. But within the hematologic malignancy world, we still cling to the five-year mark, and often discharge patients from our practice at that timepoint, with non-evidence-based instructions on future cancer screening with their primary care physicians, and the hope that they will follow-up in our late effects clinic, which often remains an unfulfilled wish. I can't say that I blame them. If I had cancer, I'm sure I would want to be done with my connection to a cancer center the moment my doctor uses the “C” word.
So, what happens after I give them that hug goodbye?
In the study, 203 patients received subtotal nodal radiation therapy-based treatment, while 196 patients received only chemotherapy. The five-year freedom from disease progression significantly favored the arm of patients who received radiation therapy. In the world of Hodgkin Lymphoma, this difference was not a minor one, and you would think that everyone should be treated with radiation therapy. These results were published in 2005.
But then a funny thing happened on the way to the Forum.
The investigators continued to follow these patients for 12 years, looking at both disease-free and overall survival. And the survival curves crossed. Now, despite a freedom from progression that continued to favor the arm of patients who received radiation therapy, at 92% vs. 87% (p=0.05), those who received radiation therapy had a worse overall survival, at 87%, compared with those who just received chemotherapy, at 94% (p=0.04).
What went wrong?
The therapies we give are not benign. They cause immediate adverse effects (the ones you see in the movies), like vomiting and hair loss, but also delayed sequelae, like cardiovascular morbidity and secondary cancers. I'll never forget the 40-year-old woman I took care of as a resident in the cardiac intensive care unit who had a cardiomyopathy, with an ejection fraction of 15%, due to coronary artery stenosis that developed from mantle radiation she received in the late 1960s for her Hodgkin Lymphoma. She was married with two young children. I must have intimated something about how terrible it was that this happened to her more than 25 years later when she had a family, and I'll never forget her response.
“I'm not sorry at all. I never would have had a family if not for those radiation treatments, and it was the only therapy available at the time. I'm the only one still alive from my oncologist's waiting room.”
In the NCIC/ECOG Hodgkin's study, 23 patients who received radiation therapy developed secondary cancers compared with 10 in the chemotherapy-only arm. Additionally, 26 have cardiac events, compared with 16 in the chemotherapy arm. These late effects likely contributed to the worse survival in the group of ABVD patients who received radiation therapy. And it took 12 years to see this occur.
Now, this is not to invite everyone to gang-up on radiation oncologists. It's clear that this study was not conducted in the modern era of radiation therapy, and likely these long-term effects have been reduced with smaller radiation fields and different techniques. I use this trial to illustrate two key points that we should consider in designing studies in hematologic malignancies, moving forward:
1. Freedom from disease progression is not an acceptable proxy for overall survival.
2. The outcome of patients enrolled in hematologic malignancy studies should be followed for 10-15 years.
Yes, that will take a long time. Yes, it will be expensive. But we don't want to commit our patients to therapies that give false promise of early benefit, when equally beneficial alternatives are available, only to create deadly diseases later on. This study should be a model of how we move the field forward in hematologic malignancies.
More ‘Second Thoughts’!
Check out all the previous articles in Mikkael Sekeres' award-winning column in this collection on the OT website: bit.ly/OT-SekeresCollection