Re: Coronary Artery Findings After Left-Sided Compared With Right-Sided Radiation Treatment for Early-Stage Breast Cancer

Journal of Clinical Oncology, Vol 23, No 30 (October 20), 2005: pp.
7391-7394

EDITORIAL

Radiation-Induced Heart Disease: Vigilance Is Still Required

Robert G. Prosnitz, Lawrence B. Marks
Department of Radiation Oncology, Duke University Medical Center,
Durham, NC

Radiation therapy (RT) plays an important role in the multimodality
management of patients with breast cancer. Trials conducted in the
1970s established that survival is equivalent after either breast
conservation (local excision and RT) or mastectomy. Studies performed
in the 1980s showed that the addition of RT to lumpectomy
significantly reduces the risk of local recurrence and, in a meta-
analysis, enhanced survival as well.1 Finally, trials conducted in the
1980s also demonstrated that RT after mastectomy dramatically
decreases locoregional recurrence and significantly improves overall
survival in patients with involved axillary nodes or large primary
tumors.2

This silver cloud unfortunately has a black lining. In the late 1980s,
data emerged that older radiotherapy (RT) techniques used in the
treatment of breast cancer, particularly after mastectomy, resulted in
increased rates of both cardiac morbidity and mortality. In a meta-
analysis involving 19,582 women with breast cancer enrolled onto 40
randomized trials begun before 1990, the Early Breast Cancer Trialists
Collaborative Group found that RT reduced the annual mortality from
breast cancer by 13% but increased the annual mortality rate from
other causes by 21% and that this increase was due primarily to an
excess number of deaths from vascular causes (death rate ratio, 1.3
[SE 0.09]).3 In a similar meta-analysis, Cuzick et al4 reviewed
individual patient-level data from 7,941 women enrolled onto 10
randomized trials of mastectomy, with or without RT, initiated before
1975 and found that the standardized mortality ratio was significantly
higher for patients treated with RT compared with controls (1.11 v
0.69; P < .001).

Critical factors in the genesis of radiation-induced heart disease are
the volume of heart exposed and the radiation dose deposited in that
volume. Older methods of delivering postmastectomy RT resulted in
relatively large volumes of heart being incidentally exposed to
moderate to high doses of radiation. The main culprit was thought to
be the use of anterior photons to treat the ipsilateral internal
mammary nodes, often as part of a larger L-shaped so-called hockey-
stick field that also covered the supraclavicular and sometimes the
axillary nodes.

Once the adverse effects of RT on the heart were recognized,
techniques evolved to irradiate the chest wall and regional nodes
while reducing exposure of the heart. The hockey-stick method was
abandoned by most radiation oncologists in favor of methods that
included the internal mammary nodes (IMNs) within the same tangential
RT fields used to irradiate the chest wall (so-called deep or
partially wide tangents) or treated the IMNs mostly or entirely with
superficially penetrating electrons, rather than deeply penetrating
photons. Given the controversy about the necessity of treating the
IMNs and the possible increased cardiac risk that came from electively
doing so, other radiation oncologists elected to abandon IMN treatment
altogether. The safety of irradiating the chest wall was also improved
by the development of treatment planning based on computed tomography
that enabled better visualization of the heart, leading to the
selection of RT fields that minimized cardiac exposure.

These heart-sparing RT methods for irradiating the chest wall (with or
without the regional nodes), were also applied to the treatment of
patients with an intact breast, which rapidly gained in popularity
following the publication of the randomized trials comparing
mastectomy with breast conservation. For patients with an intact
breast, but uninvolved axillary lymph nodes, tangential fields that
minimized cardiac exposure were used to treat only the breast itself.

Have these modern methods of delivering adjuvant RT following
mastectomy or lumpectomy eliminated the risk of cardiac injury? In
this issue of the Journal of Clinical Oncology, Patt et al5 from the
M.D. Anderson Cancer Center (Houston, TX) attempt to answer this
question. Using data from the Surveillance, Epidemiology, and End
Results (SEER) -Medicare database, they identified more than 16,000
women diagnosed with nonmetastatic breast cancer from 1986 to 1993 who
received adjuvant RT following breast surgery. They then compared the
rate of hospitalization for ischemic heart disease, valvular heart
disease, congestive heart failure, and conduction abnormalities in
patients with left- and right-sided breast cancer. With a mean follow-
up of 9.5 years, there were no significant differences in the rate of
hospitalization for any of the cardiac conditions examined in patients
with left- compared to right-sided tumors.

The same group of investigators also used the SEER-Medicare database
to examine cardiac mortality after adjuvant RT for breast cancer.
Giordano et al6 identified more than 27,000 patients diagnosed with
early-stage breast cancer from 1973 to 1989 who received adjuvant RT.
The 15-year rate of mortality from ischemic heart disease in patients
with left- and right-sided tumors was compared for patients diagnosed
in three different time periods: 1973 to 1979, 1980 to 1984, and 1985
to 1989. All patients were censored at 12 years of follow-up to ensure
equal follow-up. Although ischemic heart disease mortality was higher
for patients with left-sided tumors who were diagnosed in the earliest
time period, no significant differences were seen for patients
diagnosed in either of the two later time periods.

The studies by Patt et al5 and the similar study by Giordano et al6
have a number of significant strengths. First, the sample size in both
of these of these studies is large. The study by Giordano et al
included 27,283 patients, and the study by Patt et al included 16,270
patients. These large sample sizes reduce the probability of type II
error. Second, the mean follow-up in both studies was relatively long:
9.5 and 9.3 years for Patt et al and Giordano et al, respectively.
Because radiation-induced heart disease is a late effect of
radiotherapy, studies with extended follow-up are most useful.
Finally, both studies address clinically meaningful end points.
Giordano et al6 examined overall survival, perhaps the most clinically
meaningful end point. The main drawback of this end point is that the
latency between the onset of radiation-induced heart disease and
secondary mortality may be many years. To overcome this obstacle, Patt
et al examined the surrogate end point of cardiac morbidity. As the
authors correctly point out, cardiac morbidity may be a more sensitive
and earlier predictor of cardiac damage than mortality.

Do the results of these studies provide definitive evidence that
current techniques for irradiating patients with breast cancer do not
endanger the heart? We think not. Although their analysis is
methodologically sound, there are several limitations, most of which
are acknowledged by the authors.

First, the median follow-up of 9.5 years, while relatively long, may
not be adequate to detect the long-term consequences of cardiac
irradiation. Studies of the long-term survivors of breast cancer and
Hodgkin's disease have demonstrated that the risk for radiation-
associated heart disease increases for at least 20 years after RT.7-9
The current report's finding of no difference in cardiac morbidity in
the first 10 years after RT is reassuring, but clinically meaningful
differences may emerge with longer follow-up.

Second, the SEER-Medicare data set does not contain information
regarding RT technique. RT technique and patient anatomy are the
primary factors that determine the volume and dose of incidental
cardiac irradiation. With different techniques, the amount of heart in
the RT fields may vary considerably. The study by Patt et al,5
therefore, does not ensure that all current RT methods are safe.

Third, the underlying assumption on which the study is based, that
patients with right-sided tumors did not receive any incidental
cardiac irradiation and could therefore serve as the control group,
may not be entirely true. Approximately 25% of the patients with both
left- and right-sided tumors in the study had involvement of the
regional lymph nodes and would therefore have been candidates for IMN
treatment. If a significant fraction of patients with both right- and
left-sided tumors had anterior IMN fields, long-term cardiac toxicity
may have occurred in both groups.

Fourth, these studies do not directly address the potentially
synergistic cardiotoxic effects of systemic therapies and RT.
Anthracyclines, currently a mainstay breast chemotherapy, and newer
agents such as trastuzumab, are potentially cardiotoxic. The use of
these agents likely increases the potential cardiac risks associated
with radiation. Although neither the Patt et al5 nor the Giordano et
al6 study observed differences in cardiac morbidity and mortality
among patients with regional-stage disease receiving left-sided RT
compared with right-sided RT (most of whom presumably received
adjuvant chemotherapy), the percentage of patients in either study who
received anthracyclines is unknown, due to the limitations of the SEER-
Medicare database.

Epidemiologic studies are useful in describing the scope of the
problem of radiation-induced heart disease from the perspective of the
larger population. However, they lack detailed individual dose and
volume information needed to establish a pathophysiologic link between
RT exposure and cardiac injury. We believe that this link is the key
to identifying the subsets of patients who may be at increased risk
for radiation-induced heart disease.

To understand the pathophysiology of radiation-induced heart disease,
we and others have examined the surrogate end points of radiologic
changes in myocardial perfusion, wall motion, or ejection fraction
(EF) after RT. In 1998, we began enrolling patients with left-sided
breast cancer onto a prospective study to determine the potential
cardiotoxic effects of RT using modern techniques. Patients had modern
treatment planning based on computed tomography and pre-RT and serial
post-RT single-photon emission computed tomography-gated cardiac
myocardial perfusion scans to assess for changes in heart function.
New perfusion defects occurred in 50% to 63% of women 6 to 24 months
after RT.10,11 The incidence of perfusion defects was strongly
correlated with the volume of left ventricle (LV) in the RT field,
occurring in 25% of patients with 1% to 5% of the LV within the
tangent fields, and in 55% of patients with more than 5% of the LV
within the field. These perfusion defects generally persist at least 3
to 5 years after RT.12

The clinical significance of these perfusion defects is unknown.
However, they appear to be associated with abnormalities in wall
motion, declines in EF, and episodes of chest pain (likely
pericarditis).13,14 The data from our study and others are summarized
in the Table 1. Changes in EF have been apparent only in patients with
relatively large fractions of the LV affected by perfusion defects.
13,20




View this table:
[in this window]
[in a new window]
Table 1. Summary of Myocardial Perfusion Scintigraphy Studies
Assessing Cardiac Injury in Patients With Breast Cancer




Thus, these data suggest that RT-induced cardiac injury may still
occur with modern techniques. Furthermore, these cardiac abnormalities
can be seen in patients with extremely small fractions (eg, 5%) of
their LV included within the RT fields. Five percent of the LV
corresponds to approximately 2% to 3% of the heart; a small fraction
that is often difficult to discern on traditional dose volume
histograms, which would generally be considered "safe" by most
radiation oncologists. On the basis of these results, we frequently
use a heart block to markedly reduce, and sometimes eliminate,
incidental cardiac irradiation in our patients. In patients in whom
such a block may result in inadequate coverage of tissues at high risk
of harboring microscopic disease, compromises are made, or alternative
techniques such as electron-patches or respiratory gating are
considered.
Although modern RT techniques have reduced radiation exposure to the
heart, they may not have eliminated cardiotoxicity. It appears that
contemporary RT methods may still cause cardiovascular disease.
Changes in myocardial perfusion, wall motion, and EF have been
demonstrated in patients undergoing treatment with modern techniques.
Whether these radiographic changes will ultimately have clinical
significance is unclear. The reports by Patt et al5 and Giordano et
al6 provide some reassurance that the magnitude of the problem for the
entire population of patients undergoing RT for breast cancer may not
be large. However, they do not preclude the possibility that certain
subsets of patients may be at high risk of radiation-induced heart
disease. At present, we believe that radiation oncologists should use
contemporary RT planning and delivery methods that minimize cardiac
exposure, such as heart blocks21 and partially wide tangents.22
Treatment needs to be delivered accurately, given that small errors in
patient setup may increase a patient's risk of developing a perfusion
defect.21 We and others currently are investigating techniques to
further reduce cardiac exposure, such as intensity-modulated radiation
therapy, respiratory gating, mixed electron/photon beams and
tomotherapy.23-29 With treatment innovation, it is our goal to
eliminate incidental irradiation of the heart and to make radiation-
induced heart disease a historical footnote.


.

Relevant Pages

  • Re: ATTACKING THE HEART ATTACK
    ... I posted the following in 1999 about reversing heart disease: ... Cardiac patients surrounded by banks of equipment, ... intensive lifestyle changes reduced their LDL cholesterol ...
    (sci.med)
  • Re: ATTACKING THE HEART ATTACK
    ... I posted the following in 1999 about reversing heart disease: ... Cardiac patients surrounded by banks of equipment, ... intensive lifestyle changes reduced their LDL cholesterol ...
    (sci.med.cardiology)
  • Re: Hope and a Debate (Full Article in HTML)
    ... Heart Scanner Stirs New Hope and a Debate ... that doctors would not hesitate to use them. ... Diagnosis of heart disease, he argued, will be transformed, and lives saved. ... and patients could be harmed. ...
    (sci.med.cardiology)
  • REDUCING STRESS REDUCES HEART DISEASE
    ... Reducing Stress Reduces Heart Disease ... A new study of patients with heart disease found that relaxation, ...
    (sci.med.cardiology)
  • REDUCING STRESS REDUCES HEART DISEASE
    ... Reducing Stress Reduces Heart Disease ... A new study of patients with heart disease found that relaxation, ...
    (sci.med)