Zelanna,

Thanks for summarizing and providing your interpretation of this important study. I had heard a bit about this trial, just based on news snippets, so I asked a few radiation oncologists about their impressions and the somewhat puzzling results. Much of what I heard reflected a similar skepticism, that this was a remarkably small trial to draw any large conclusions from, and that perhaps it wouldn’t have carried the weight it did, getting presented in the plenary session, had it not come from MD Anderson: essentially, that the strength of the actual data weren’t ideal, ad that it helped greatly that it came from a leading institution. Moreover, I believe I read that there was a biologically implausible difference in distant outcomes, such as that patients who received WBRT actually had more distant spread outside of the brain. If that’s correct, it seems far more likely to argue for imbalance in a very small trial than a truly meaningful treatment difference.

However, before I get carried away making any of my own conclusions, I need to recognize that I haven’t seen any details myself, and I may be misremembering aspects of the trial.

Could you please comment on these issues? Was this trial enough to change practice patterns? Regardless of what the data support, will it lead to so much concern for neurocognitive problems with WBRT, among both physicians and patients, that it makes it very hard to pursue this, even perhaps with a larger number of brain metastases?

Thanks,
Jack

Dr. West

Hi Jack:

Yes, yes and yes. The trial was too small and took too long to accrue to draw strong conclusions from. I would go one step further: it was given the prominence it was on the presentation stage because it confirms a pattern of practise already adopted and believed in by many. The survival data (stereotactic radiation therapy had prolonged survival over WBRT) doesn’t make any sense from what we know of radiation therapy. There was increased distant mets in the arm of the trial with WBRT, almost certainly the cause of the decreased survival, but likely just a statistical fluke (as you say, from the small size of the trial). Given that WBRT has been shown to improve survival in another, larger trial in NSCLC and is known to do so in SCLC these results of decreased survival with WBRT do not jibe. I do not think that this ends the discussion of whether there is a survival disadvantage to further brain failure, but I must say that there will be several high profile radiation oncologists that probably disagree with me. Given the biases already out there to treat with SRS, this study confirms the pattern of practise.

What I do think is a legitimate conclusion is the issue of cognitive effects at 4 months post 30 Gy to the whole brain. Those patients who have their brains irradiated have worse word list recall, and so we have to accept that there is a short term toxicity cost to WBRT that SRS patients avoid. The mean survival in this study was just over 7 months, so worse cognition for 4 months can be a significant quality of life issue.

Should this trial change pattern of practise? No, but given that for many the pattern of practise has already changed, this trial can be used to support it and at least make a claim that “survival data comparing the two types of brain treatment is mixed.” as opposed to having to suggest that there is evidence that survival may be decreased with WBRT.

cheers,
zg

Dr. Goldberg

One last point — I wish that I could say that doctors and patients will not over generalize this to treating 4 or more mets with SRS. Unfortunately, that seems unlikely, and the technical issues there, aside from lack of data and resource utilization issues, are significant. Probably worth a longer post if technical radiation issues are of interest to people.

cheers,
zg

Dr. Goldberg

Dr Goldberg -thanks for posting this. I agree with the observations that you and Dr. West have made, particularly that the longer survival for SRS patients over WBR patients is probably an anomaly related to the small sample size here. I have seen previous literature state that WBR did not extrend survival, but never that it actually worsened it.

If the results of this study were adopted into common practice, would by husband have been given WBR after his craniotomy? His situation is probably different from those in the study -
since his tumor was resected, there would be nothing to do SRS on, correct? My recall is that craniotomy was selected for him due to the largeish size of his met (about 3cm); his rapidly worsening vision with the tumor; the fact that there was only one met; and total lack of response to steroids in the pre-surgical period. He was also NED in all other ways.

Would all of that mean that, using the principles of this trial, his treatment would have been different?

Also - I’m not surprised to see that the trial found worsened neurocognitive decline in the WBR + SRS group at 4 months. My guess, and I certainly have a bias, is that this is mostly related to the WBR, and that the WBR group would continue to worsen in NC functioning for quite a while beyond that.

Thanks again for posting this.

mary

Hi Mary:

Looking backwards in time to decisions made is always done with 20-20 hindsight, so I tread cautiously.

This study presented at ASTRO would not have included your husband for he had only a single metastasis and had it resected, so the results couldn’t be generalized to him. What is likely true, though, is people are increasingly accepting in-brain failure (tumor recurrence) rather than giving WBRT. So, if a patient in a similar situation presented today, there is a good chance that his oncologists would suggest deferring WBRT and taking a ‘wait and see’ approach.

While that may have a bitter edge to it, given the neurocognitive problems you are now living with, the unknown is whether your husband would have had in-brain failure and then required either WBRT or some other treatment/lived with some other complication. This is the piece of the MDA trial that may be very misleading: there is little to hang our hats on regarding the price paid for not giving WBRT, except that there was no survival disadvantage and no one needed WBRT for salvage. Median survival was only just over 7 months, though.

I am not sure if there is a silver lining in here for you, or not, but I don’t put too much weight in this trial personally, for all the reasons we have discussed.

zg.

Dr. Goldberg

Thank you for your reply, Dr. Goldberg. I do understand the hindsight issues, and knew that I may have them when we opted for the WBR. When we were making the decision, I asked my husband what he wanted most out of treatment, and he said that he never, ever wanted another brain met - so, WBR was and still would be the best answer for that, and I told him that was the case. Anyway, I don’t ruminate on this, I just have a need to understand it as well as possible. Some decisions in life just don’t offer optimal choices, and we just do our best with what we have at the time. I also don’t ignore the fact that he is alive 16 months post treatment, not a small thing at all.

Thanks again

mary

Drs. Douglas Kondziolka and L. Dade Lunsford of the University of Pittsburgh School of Medicine did a study, “Treating four or more brain tumors in a single radiosurgery session resulted in improved survival compared to whole brain radiation therapy alone,” presented at the 47th Annual Meeting of ASTRO.

In the study, patients with primary malignancies that had metastasized to the brain underwent Gamma-Knife radiosurgery and the results indicated that treating four or more brain tumors with radiosurgery is safe and effective and translates into a survival benefit for patients.

Two hundred and five patients with primary malignancies that had metastasized to the brain underwent gamma knife radiosurgery for four or more tumors during one session. The average number of brain tumors for patients in the study was five, with a range from four to 18. Radiosurgery was used alone, in combination with whole brain radiation or after failure of whole brain radiation.

Radiosurgery patients with the most prognostic factors associated with survival from brain metastases (defined as class 1 according to the Radiation Therapy Oncology Group classification system for patients with brain metastases) survived an average of 18 months, compared to a reported historical average of seven months for those who received whole brain radiation alone.

Patients defined as class 2 who received radiosurgery survived nine months compared to the historical average survival of four months for patients who received whole brain radiation.

Patients with the least prognostic factors associated with survival (class 3) who received radiosurgery survived an average of three months compared to the historical average survival of two months for patients who received whole brain radiation.

The average overall survival for patients who received radiosurgery was eight months and the average time to progression and new brain metastases was nine months.
The study also found that the sum of the volume of all treated brain tumors was a more significant predictor of length of survival than was the total number of brain metastases, indicating that tumor volume should be used as a criterion for radiosurgery rather than number of brain metastases.

Typically, only patients with one to three brain metastases are considered candidates for stereotactic radiosurgery. However, based on the results from this study, they concluded that the number of brain metastases should not necessarily preclude patients with multiple lesions from this potentially life-saving treatment option.

Primary lesions generally involve invasion into adjacent brain tissue and it could make sense to have postoperative whole brain radiation in an attempt to destroy any residual cancer cells. However, metastatic lesions have relative lack of invasion into adjacent brain tissue, making them ideal for radiosurgery or postoperative focal radiation.

Anonymous

Dear Anonymous:

Once again your post is a good encapsulation of a contentious area. The U of Pittsburg is a leading proponent of radiosurgery and will offer it to patients regardless of the number of mets. They also have substantial backing (which if I am recalling correctly includes for this study, but I am working from memory here so I could be mistaken)includes this study was funded by industry — that is, the producers of machines dedicated to SRS. Now, I do not personally think that the fact that this was (I think) an industry study negates its value, but it must be read with that in mind.

It is also a single institution experience with one of, if not, the most experienced SRS team in the USA.

The results are also compared to a historical control, which, again, must be kept in mind when considering the results. (I don’t recall if they specified how many received SRS in addition to WBRT or as a single brain treatment — ie, should SRS replace WBRT or suppliment it?)

Overall, this study can be said to ask the question of whether SRS should become a standard of care for patients with 4 or greater brain mets, but it certainly doesn’t answer it. A single institution study compared to historical controls cannot do that. It will be interesting to see if the large cooperative groups chose to take up this question and do a randomized trial.

As you can read in this post, I remain in the “to be convinced” group that this isn’t a lot of patient selection with limited true benefit to the therapy–unfortunately, the literature is littered with such single institution studies in brain tumors.

Thanks for the post. It is good to have someone else reading the abstracts.

zg.

Dr. Goldberg

Very good point Dr. Goldberg. I often thought about the problem of unintended self-interest bias which may be a factor in this problem.

Some years back, the government had asked Joseph P. Newhouse, a health policy professor at Harvard, and his colleagues to look into how the Medicare reimbursement system may affect how doctors prescribe chemotherapy.

His study “Does Reimbursement Influence Chemotherapy Treatment For Cancer Patients?” co-authored with Dr. Craig C. Earle, was finally published in Health Affairs in 2006. This joint Michigan/Harvard study added to the ’smoking gun’ survey by Dr. Neil Love, “Patterns of Care.”

I wrote to both of them to ask if their study methodology on reimbursements influencing chemotherapy treatments, could be applied to reimbursements influencing radiation treatment?

Before the days and widespread use of Stereotatic, Gamma-Knife, Cyber-Knife, and the like, the most expensive treatment for postoperative brain surgery for a solitary brain metastasis was whole brain radiation. With the newer treatments, whole brain radiation was abandoned because of the substantial neurological deficits that resulted with its use, sometimes appearing a considerable time after treatment. Today, cutting-edge clinical practices use a more “focused” radiation field.

During the last twenty years when the preponderance of cancer care shifted from the institution-based, inpatient setting to community-based, ambulatory sites for treating the majority of the nation’s cancer patients, many of these community-based settings did not have the cutting-edge high-tech toys.

Was there an incentive for radiation oncologists at community cancer centers to chose whole brain radiation treatments, as these were the most expensive, for them? Could Newhouse’s methodology collect data documenting a clear association between reimbursement to radiation oncologists for whole brain radiation treatment which is based on how much incentive occurs to the radiation oncologist?

They thought that there were similar issues, but their methodology would be different because radiation isn’t something that individual doctors buy, sometimes at a discount, and then profit from if they’re reimbursed more for it, as in the case with chemotherapy.

They relied upon price variation across regions in Medicare, which was pseudo-random and had been eliminated. To their knowledge, there was no comparable price variation in radiology that they could have used.

However, they did mention a radiation oncologist in Michigan, who had done some work looking at the number of palliative fractions of radiation given to patients with advanced lung cancer as being a situation in which there is a lot of discretion on the part of the physicians: one fraction is as good as 10, but 10 will reimburse more. I’m not sure if he ever published or presented his results? Interesting!

Anonymous