B-C List friends,
This message deals with the 1994 study by Cuzick, et al., on some of the effects of radiotherapy. I welcome any comments, as I try to sort through why radiotherapy went into decline, and why it is coming back amongst some doctors after mastectomies (or after that plus chemotherapy). The issues include changes in techniques and a now-emerging survival benefit for a subset of women (basically, node-positive, or many nodes positive, plus some other categories). Ê
My Part One looked at the November 1995 "Oxford Overview" and reactions to it. Those reactions criticized it for mixing older and newer studies together into single survival figures (numerous studies that used quite old radiotherapy techniques, which were both more heart-harmful and less cancer-beneficial, and more recent studies that used modern radiotherapy techniques that were better targeted at sites likely to cause recurrences and less targeted at the heart where survival could be compromised by heart damage).
The important 1994 Cuzick study separated old trials from new ones in terms of the potential side-effects from radiotherapy, presaging the criticisms that would later be aimed at the Oxford Overview. The Cuzick analysis also pointed the way toward newer studies that show patients gaining survival benefits from radiotherapy while minimizing any risks. Indeed, an editorial accompanying the 1994 Cuzick study raised exactly that possibility.
J. Cuzick, H. Stewart, R. Peto, et al, "Overview of randomized trials of postoperative adjuvant radiotherapy in breast cancer," Cancer Treat Rep 71:15-29 (1987)
In 1987, an international team of researchers reported that there were more deaths after the 10-year mark for post-mastectomy in patients in 8 studies around the world when radiotherapy had been administered (prior to 1975) than when it was not. This study may, according to some medical observers, have led to a further decline in the trend that was already occurring toward omitting radiation treatment after mastectomy.
As Drs. Lori Pierce and Allen Lichter of the University of Michigan
pointed out in an editorial (which I will examine below) accompanying Cuzick's
1994 re examination of the 1987 conclusions, "The first Cuzick meta-analysis
further reduced support for the use of chest wall radiation by suggesting
that
it not only failed to increase survival, but actually reduced survival."
Probably aware of how their 1987 study was influencing therapy, Drs. Cuzick, et al., went back in the years following the 1987 study, collected more data on the causes of mortality, and made some important discoveries. This 1994 re-examination paints quite a different picture. It is discussed in the following.
(D-2) CUZICK, STEWART, RUTQVIST, PETO 1994 RE-ANALYSIS OF OLD AND NEW STUDIES
J. Cuzick, et al., "Cause-Specific Mortality in Long-Term Survivors of Breast Cancer Who Participated in Trials of Radiotherapy," J. Clin. Onc. 12:447-453 (1994)
Dr. Cuzick and other researchers from England, Scotland, Sweden, the US, Norway, and Germany determined that the reason for the 1987 finding of survival disadvantages to radiotherapy was that certain of the studies used had involved heart-damaging techniques that are now understood to be inappropriate. In short, his group performed an analysis in 1994 that had not ben performed in 1987 and that the Oxford Group (discussed in Part One) failed to perform subsequently in 1995 (as pointed out by Dr. Recht in 1996 in his strong criticisms against the Oxford Overview).
The Cuzick 1994 study originated when the international group decided to investigate the question why, in the overview published in 1987 by the same principle author, J. Cuzick, radiotherapy appeared to lead to lower overall survival, not to increased overall survival.
The earlier (1987) Cuzick study looked at one set of trials using older radiotherapy doses and techniques (prior to 1975). In 1994 he and others reviewed those trials again and,in addition, looked at another set of trials using newer radiotherapy doses and techniques (the Stockholm, Manchester Regional, Cancer Research Campaign, and NSABP trials, none of which had much 10-year data earlier).
COMPARISON OF OLDER AND NEWER TRIALS
Cuzick and the others concluded that there are several reasons for the older trials showing no benefit and even a detriment from radiotherapy techniques after mastectomy. And there are reasons, they said, that the newer trials are starting to show a net benefit, namely the use of newer techniques.
Of the older trials, they wrote: "The relevance of these results to current practice is difficult to assess, since the techniques and fields used have changed markedly over the course of these trials." This statement parallels the criticisms made later by Dr. Recht about the Oxford Overview.
The team wrote about the newer trials (in contrast to the older trials), "The reduction in breast cancer deaths, especially in the more recent trials, is noteworthy and suggests that radiotherapy may have value beyond the clearly establishedÊ improvements obtainable for local control."
LIMITING IMPACTS ON THE HEART
What the team concluded was that the disadvantages from older radiotherapy came almost entirely from damage to the heart. The implication was, as they stated at the end of this 1994 study, that any benefit from radiotherapy in preventing distant metastases that lead to death should not be sought with radiotherapy approaches that themselves cause equal or greater amounts of death from heart problems. In their words:
"However, if this is not to be counterbalanced by increased cardiac mortality, it is essential to use procedures that minimize doses to cardiac tissue and the coronary arteries. This can be facilitated by individual treatment planning and avoidance of large fraction sizes and high total doses."
This strong warning leads us to the more recent studies. But before examining them, I will first include comments by another set of researchers who wrote an editorial in the same issue of the Journal of Clinical Oncology in 1994, commenting on Cuzick's new look at the old data.
(D-3) PIERCE &;LICHTER 1994 EDITORIAL ON CUZICK STUDY
Lori J. Pierce &;Allen S. Lichter, Editorial, "Postmastectomy Radiotherapy: More than Locoregional Control," J. of Clin. Onc. 12:444-446 (1994)
In the same issue of the Journal of Clinical Oncology, Drs. Pierce and Lichter of the University of Michigan wrote an editorial on the significance of the study by Cuzick, et al. Their goal was not only to deal with what Cuzick had found, but also to assess its implications for a potential survival benefit from post-mastectomy radiotherapy.
They pointed out in clear language that Cuzick's 1994 update of his 1987 study " suggest that the conclusions from the first meta-analysis are no longer valid." They cautioned that "every effort must be made to individualize treatment planning to reduce cardiac toxicity" and said that if this is done, then "survival benefit" can be realized.
OLDER RADIOTHERAPY TRIALS DID SHOW LOCO-REGIONAL CONTROL
Pierce & Lichter began by reminding readers that radiation reduces local recurrences. "Radiation of the postmastectomy chest wall was one of the first cancer treatments to be studied using the mechanism of the randomized clinical trial. Beginning in 1949, more than 20 trials have been performed testing the worth of this therapy. These studies have repeatedly shown a significant reduction in locoregional failures with the addition of postoperative radiotherapy."
So their first point was the well-known one that, quite apart from the issue of overall survival, radiotherapy can reduce the likelihood of having breast cancers show up again in the chest area after mastectomy.
DID OLD RADIOTHERAPY TRIALS DILUTE THE DATA?
Pierce & Lichter went on to write: "However, even with the substantial improvements in locoregional control, radiotherapy has been reported to have little or no impact on survival." Indeed, there was evidence from some studies that it may have actually worsened survival, including Cuzick's original 1987 study.
But Pierce & Lichter pointed out that the earlier studies didn't focus on the patients having the greatest risks of locoregional recurrences after mastectomy (and therefore at greater risk of distant metastases and earlier death caused by secondary dissemination from the local recurrences). Rather, those studies included many others as well. This is a familiar issue: who should be counted in the averages? Do some sub-groups benefit, while others do not? If so, then therapy should be tailored for those who can benefit from it.
If there was a beneficial effect for one group, it may have been lost or overwhelmed by the data from another. One might say that the data from those for whom radiotherapy never had much chance of making any difference (either because they were node-negative and thus at little risk of distant recurrences, or for other reasons) can "dilute" the effects shown in those for whom it can make a difference.
PAST FAILURES TO TREAT PROPER AREAS
A second concern is that the earlier Cuzick study mixed together earlier studies that we now know treated the wrong areas, making it unlikely they could show much, if any, benefit.
"To evaluate the earlier Cuzick conclusion, the technical aspects of the radiotherapy used in these older series must be examined. Some of the trials irradiated the regional nodes only and excluded the chest wall from treatment. Patterns of failure after surgery have identified the chest wall to be the most common site of locoregional failure, with up to 60% of locoregional recurrences including the chest wall as a component of failure.
In short, if you aim the radiation beam at a place where recurrences are less likely, you are not going to show as much benefit from radiotherapy as if you aim the beam at the places where tiny clumps of cancer cells are more likely to be hiding, namely in the chest wall.
Pierce & Lichter also pointed out that the older, inadequate-voltage radiotherapy failed to treat target tissues evenly that way that newer, high-energy units do, and that in many of the older studies the inadequate total dosages simply were not enough to treat the cancer. Ê
PAST FAILURE TO LIMIT HEART EXPOSURE
The older studies also harmed the heart more. For example, they say, we now know that those that aimed radiation at the parasternal area delivered radiation to places of maximum danger for the patient.
"Such studies, with their parasternal treatment directly precisely at the heart while avoiding treatment to the high-risk chest wall itself, could be viewed as maximizing the long-term risk of treatment while minimizing the benefit," wrote Pierce & Lichter. Ê
That is quite an indictment of some of the older radiotherapy techniques. Ê
But there were even more shortcomings of the older radiotherapy, Pierce & Lichter pointed out, such as daily radiation doses so high that they caused more heart damage (something that can be limited by giving more days of treatments, but at lower daily radiation doses -- the 25+ day treatments seen now, as opposed to 15-day treatments of the past with higher daily doses). Ê
COMPARATIVE EFFECTS OF OLD AND NEW RADIOTHERAPY Ê
In words that I quoted partially earlier, Pierce & Lichter wrote, "In this issue of the Journal of Clinical Oncology, Cuzick et al, in an important update of the postmastectomy radiotherapy trials, suggest that the conclusions from the first meta analysis are no longer valid." Ê
They wrote that the data in his study "underscore the value that improved radiotherapy technique can have on patient outcome and illustrate the danger of placing too much weight on older studies in a discipline that has made considerable technical progress in the past three decades." Of course, that is the same type of criticism aimed at the Oxford Overview the following year.
Pierce & Lichter pointed out that where cardiac toxicity (heart damage) has been seen, it has been, even in the studies of the 1970s, attributable to such things as excessive daily doses ("thus treating the heart to a high dose in a relatively short time") or (in part of the Stockholm trial) "deeply directed tangential fields using cobalt 60 therapy" on a left breast, aiming to get deeply set internal mammary nodes on that side and skimming through the heart excessively in the process.
"In contrast, a subset of patients in the same trial who were treated with an electron beam technique that considerably reduced the dose of radiation to the heart did not show an increase in cardiac morbidity." If one does need to irradiate the internal mammary nodes of a left-sided breast cancer, one can now use CT-directed planning and varying of photon- and electron-field arrangements, among other things. Ê
CAN CHEMOTHERAPY ALONE CONTROL RECURRENCES AND INCREASE SURVIVAL? Ê
Why not just leave the local and regional recurrences to chemotherapy? Pierce & Lichter addressed this in their editorial: "It is clear from a number of series that chest wall failures persist despite the apparent success of adjuvant chemotherapy in reducing the risk of systemic recurrences." Ê
And can chemotherapy alone produce the maximum survival benefits? Pierce & Lichter noted that "two recent studies have demonstrated improved survival using combined chemotherapy and radiotherapy." Ê
These recent trials, from British Columbia and Denmark, showed improvements in overall survival. The wrote of a 10% benefit in 10-year overall survival in the B.C. trial for women with stage II disease, and an unspecified 5-year benefit in overall survival in the Danish study. Both involved use of CMF, by the way.
SUMMARY Ê
There was very little in the Pierce & Lichter editorial about the actual reasons that modern radiotherapy produces survival benefits. Ê
To obtain the benefits of radiotherapy, however, Pierce & Lichter wrote, "It is also clear from this report that every effort must be made to individualize treatment planning to reduce cardiac toxicity if a survival benefit is to be realized. This is particularly relevant because many patients requiring postmastectomy radiotherapy will have received a doxorubicin-containing regimen, further compromising cardiac tolerance." Ê
Clearer data showing a survival benefit would be forthcoming in the following two years, as I will discuss next.
Ê
John
----------------------------------------------------
John E. Bonine (jbonine@oregon.uoregon.edu)