December 10, 2008 (San Francisco, California) — A large meta-analysis of individual patient data from clinical trials of erythropoietin-stimulating agents (ESAs) in cancer patients has confirmed that there is a significantly increased risk for death associated with these agents. While the finding is not new — it has already led to restrictive labeling and black box warnings for these products — it will reignite concern over the use of ESAs in cancer patients, say experts.
These latest data come from a detailed analysis of 13,933 cancer patients from 53 clinical trials. They show cancer patients who were treated for anemia with ESAs and blood transfusions as needed had a 6% lower overall survival than those treated with blood transfusions alone.
"The increased risk of death must be balanced against the benefits of ESAs, taking into account each patient's clinical circumstances and preferences," commented lead researcher Julia Bohlius MD, MscPH, from the University of Bern. Her presentation at a late-breaking session here at the American Society of Hematology (ASH) 50th Annual Meeting and Exposition elicited many congratulatory comments from the audience, one of whom thanked the researchers for "undertaking this Herculean task."
Dr. Bohlius and colleagues collected individual data for every patient from individual trialists and from the makers (Amgen, Johnson & Johnson, and Roche). She emphasized that the makers had no involvement in the study design or analysis of the data, and that funding for the study came from an independent source (the German Ministry of Education and Research and OncoSuisse).
Impact on Both On-Study Mortality and Overall Survival
Overall, for cancer patients who received ESAs, there was a significant increase in on-study mortality (ie, deaths during the active study phase) and a significant decrease in overall survival (ie, deaths during the longest follow-up available). Both of these values are higher than have been reported previously from individual studies and literature-based meta-analyses, Dr. Bohlius commented.
Table 1. Results for All Cancer Patients Who Received ESAs (N = 13,933)
| Outcome | Hazard Ratio (95% Confidence Interval) | P value |
|---|---|---|
| On-study mortality | 1.17 (1.06 - 130) | .002 |
| Overall survival | 1.06 (1.00 - 1.12) | .005 |
From these results, Dr. Bohlius calculated the number-needed-to-harm (NNH) for cancer patients with various prognoses. For low-risk patients who have a 5% chance of dying within the next 4 months, the NNH was 121 (95% confidence interval [CI], 69 - 343). "You would have to treat 121 patients with ESAs for 1 additional patient to die," she explained. For medium-risk patients who have a 20% risk of dying within months, the NNH was 34 (95% CI, 19 - 94), whereas for high-risk patients with a 70% risk of dying within 4 months, the NNH was 24 (95% CI, 14 - 67).
No Strong Evidence That Any Factor Influences the Effect
The majority of the cancer patients in these clinical trials were receiving chemotherapy (10,441 [75%] of 13,933 patients), and this patient population was analyzed separately. (Chemotherapy-induced anemia is currently the only approved indication for the use of ESAs in cancer patients.) In this patient population, the increase in on-study mortality and decrease in overall survival was not significant. For patients undergoing chemotherapy, the increased risk for death "was less pronounced, but could not be excluded," Dr. Bohlius commented.
Table 2. Results for Cancer Patients Receiving Chemotherapy Who Received ESAs (n = 10,441)
| Outcome | Hazard Ratio (95% Confidence Interval) | P value |
|---|---|---|
| On-study mortality | 1.10 (0.98 - 1.24) | .12 |
| Overall survival | 1.04 (0.97 - 1.11) | .26 |
No significant differences were seen between any of the different cancer treatments that patients received: chemotherapy, radiotherapy, chemoradiotherapy, or no treatment. There were small differences between these groups, but "they can be explained by chance," Dr. Bohlius said.
There was also no evidence that the effect of ESAs was any different in patients with different tumor types, he said.
In total, the researchers investigated 20 factors and found no strong evidence that any factor modified the treatment effect, although there was some evidence for an effect modification by a history of thromboembolic disease, hematocrit at baseline, and the planned frequency of ESA administration.
"Not all patients are affected, but some are affected," Dr. Bohlius commented, adding that "at the moment, we cannot predict whether the drug will be harmful or not." She said further study at the molecular level is needed because so far analysis of clinical factors has not found any answers.
What Is the Mechanism Involved?
Why there is an increase in mortality in cancer patients receiving erythropoietin is not known, commented senior author Andreas Engert, MD, from the University of Cologne, Germany. There has been speculation, and some early data, that erythropoietin receptors are found on tumors, and that the ESA may act via these receptors to stimulate and promote tumor growth. However, there is no evidence to show this, Dr. Engert commented at an ASH press conference. "There is no convincing data to suggest that this is clinically relevant."
Another explanation that has been put forward is that the increase in mortality results from an increase in thromboembolic events. "I think that this may be the case," Dr. Engert commented. "Dialysis patients treated with ESAs who have too high a level of haemoglobin have a higher risk of dying from cardiovascular disease, and this may also be the case in cancer patients."
"I suspect that there has been an under-reporting of thromboembolic events," Dr. Engert commented. He pointed out that in one of the trials — the BEST study in breast cancer, one of the individual studies that found an increase in mortality with ESA use — some of the deaths did not have an autopsy report.
These latest results will reignite concern about the use of ESAs in cancer patients, predicted J. Evan Sadler, MD, PhD, Washington University, St. Louis, Missouri. "The increase in mortality is very worrisome," he told Medscape Oncology, and these products should be used with "appropriate caution."
"There are hazards associated with these drugs, and I think we have to watch the blood cell count very carefully," said Leonard Zon, MD, from the Children's Hospital, Boston, Massachusetts. "There is still a lot of discussion about these agents, and new clinical guidelines are being developed by a number of organizations, including the National Institutes of Health, about exactly how we should prescribe erythropoietin."
"There is no new cause for concern," commented Samuel Silver, MD, PhD, FACP, from the University of Michigan. The recent changes to these products' labeling (black box warnings, restrictions on use, etc) are in line with these findings, and he said that he did not anticipate any new moves soon.
But, the concern persists and "it makes sense to review and discuss the issue once again, and these data are going to be very important going forward," he said. "This reanalysis of the data at the patient level is critically important," he told the attendees at the press conference. "It was a huge amount of work, and the authors should be congratulated."
Δεν υπάρχουν σχόλια:
Δημοσίευση σχολίου