Key words
breast cancer - radiotherapy - regional lymphatic radiotherapy
Schlüsselwörter
Brustkrebs - Radiotherapie - Regionäre Lymphabflussbestrahlung
Introduction
Radiotherapy of the regional lymphatic drainage pathways (regional nodal irradiation,
RNI) in patients with breast carcinoma has been a subject of controversy in recent
decades. In nearly all studies on post-mastectomy radiotherapy (PMRT), the entire
regional lymphatic drainage routes – i. e., the supraclavicular nodes (SCNs), axillary
lymph nodes (ALNs), and internal mammary nodes (IMNs) – have been irradiated. By contrast,
adjuvant radiotherapy after breast-conserving surgery (BCS) has often been restricted
to whole-breast irradiation (WBI) and RNI has not been routinely included. Most studies
have carried out randomized comparisons of adjuvant radiotherapy versus refraining
from any radiotherapy, but not using or refraining from additional RNI [1]
[2]. It has therefore not been possible to quantitatively analyze the effect of RNI
on survival. This gap has been closed to some extent by a few recent randomized studies,
and the published results may have led to a change of direction.
The introduction of sentinel node biopsy (SNB) created a paradigm shift in the radicality
of axillary surgery. For radio-oncology, it also raised fresh challenges – e. g.,
the question of whether and with what risk patterns axillary node radiotherapy (ALN-RT)
might represent an alternative to axillary lymph-node dissection (ALND). Newly published
randomized data [3] are suggesting a potential for future individualization of the treatment approaches
here.
The aim of the present study is to provide an overview of the current data and their
implications for the indication for regional nodal irradiation.
RNI with one to three positive axillary lymph nodes
RNI with one to three positive axillary lymph nodes
While the indication for RNI when there are four or more positive axillary lymph nodes
remained undisputed, it became a matter of controversy in patients with one to three
positive lymph nodes (pN1) or with pT3 pN0 tumors [4]
[5]
[6]. This question was first explicitly investigated in the Canadian study (NCIC-CTG
MA20, recruiting period 2000 – 2007) [7]. A total of 1832 patients were included, most of whom had one to three positive
axillary lymph nodes (85 %), while 10 % had negative lymph nodes and a high-risk pattern.
After breast-conserving surgery and axillary lymph-node dissection, the patients were
randomly assigned either to WBI alone or WBI with additional RNI. The RNI in the study
consisted of irradiation of the supraclavicular nodes and internal mammary nodes.
Around 8 % of the women in both arms of the study also received axillary radiotherapy.
After 10 years, locoregional control was 95.2 % with RNI and 92.2 % without it (P = 0.009),
and the rate of in-breast recurrences was comparable in the two groups, as expected
(4.3 % vs. 3.6 %). By contrast, lymph-node recurrences were five times less frequent
after RNI (2.5 % vs. 0.5 %). Overall, the recurrence-free survival after RNI was 82 %,
in comparison with only 77.4 % in the control arm (P = 0.01), and similarly the metastasis-free
survival (MFS) was 86.3 % versus 82.4 % (P = 0.03). These differences go beyond the
improvement in locoregional tumor control. Avoiding lymph-node recurrences thus appears
to prevent distant metastases. Improvement in the probability of survival with additional
RNI was only observed in estrogen receptor–negative patients (n = 465; 10-year overall
survival 81.3 % vs. 73.9 %), while in the group as a whole the overall survival rates
with and without RNI were approximately 81.8 % versus 82.8 %.
The European study (EORTC 22 922 – 10 925, recruiting period 1996 – 2004) [8] included 4004 patients mainly with pT1 – 2 tumors (95 %) and either positive axillary
lymph nodes (55.6 %) and/or a medial tumor location (44.5 %). After BCS (76 %) or
mastectomy (23 %), they were randomly assigned either to WBI/PMRT alone or additional
RNI. Nearly all of the patients received adjuvant systemic treatment. After 10 years,
a significant improvement in the disease-free survival (DFS) was noted in the RNI
group (72 % vs. 69 %; P = 0.04) and also in the MFS (78 % vs. 75 %; P = 0.02). The
overall survival also showed a trend in favor of RNI (82.3 % vs. 80.7 %; P = 0.06).
The rate of lymph-node recurrences was in favor of RNI, at 2.7 % versus 4.2 %. Lymph-node
tumor control thus appears to have an influence on metastasization here as well. The
results are summed up in [Table 1].
Table 1
Summary of the results of studies on regional radiotherapy.
|
without RT of regional LNs
|
with RT of regional LNs
|
P
|
|
Canadian study, NCIC-CTG MA20; n = 1832, 10-y data [7]
|
|
|
92.2 %
|
95.2 %
|
0.009
|
|
|
82.4 %
|
86.3 %
|
0.03
|
|
|
77.4 %
|
82.0 %
|
0.01
|
|
|
81.8 %
|
82.8 %
|
n.s.
|
|
European study, EORTIC 22 922 – 10 925; n = 4004, 10-y data [8]
|
|
|
4.2 %
|
2.7 %
|
–
|
|
|
75 %
|
78 %
|
0.02
|
|
|
69 %
|
72 %
|
0.04
|
|
|
80.7 %
|
82.3 %
|
0.06
|
|
French study, n = 1334, 10-y data [18]
|
|
|
59.3 %
|
62.5 %
|
n.s.
|
|
Danish cohort study; n = 3089, 8-y data [19]
|
|
|
72.2 %
|
75.9 %
|
0.005
|
A meta-analysis of these data [9] found that all of the survival advantages were significant ([Table 2]). The hazard ratio (HR) for the overall survival (OS) was 0.88 (95 % CI, 0.78 to
0.99; P = 0.034) and the HR for the DFS was 0.86 (95 % CI, 0.78 to 0.95). The greatest
effect was seen in the MFS (HR 0.84; 95 % CI, 0.75 to 0.94; P = 0.002), and this may
support Hellman’s hypothesis that radiotherapy “stops metastases at their source”
[10].
Table 2
Meta-analysis of randomized studies on prophylactic regional radiotherapy [9].
|
end point
|
hazard ratio
|
P
|
|
metastasis-free survival
|
0.84 (95 % CI, 0.75 to 0.95)
|
0.002
|
|
disease-free survival
|
0.86 (95 % CI, 0.78 to 0.95)
|
0.002
|
|
overall survival
|
0.88 (95 % CI, 0.78 to 0.99)
|
0.034
|
Radiotherapy of the internal mammary nodes (IMNs)
Radiotherapy of the internal mammary nodes (IMNs)
Data from extensive surgical groups showed in the 1960 s that there is a high rate
of IMN metastases when the tumor is in a medial location and there is axillary involvement
[11]. A Chinese study published in 2008 confirmed similar findings in a group of 1679
patients in whom dissection of the IMNs was carried out in addition to extended mastectomy,
and IMN metastases were confirmed at histopathology in more than 20 % of the cases
when risk factors were present (medial location, four or more positive axillary lymph
nodes, T3, age < 35y) [12]. In contrast to what might be expected following these data, however, clinically
manifest IMN recurrences are actually only observed rarely in patients with breast
carcinomas (only in approx. 1 %) [13]. A hypothetical explanation for this discrepancy might be unintentional simultaneous
irradiation of the IMNs when the intention is to carry out radiotherapy of the breast
or chest wall “alone,” using the “tangential fields” that were formerly customary.
Since in addition systematic imaging of the IMNs does not form part of routine follow-up,
metastases in this area sometimes remain clinically unrecognized, particularly when
a distant metastases has in the meantime become the focus of clinical symptoms [14]. It is therefore not surprising that the indication for IMN radiotherapy has been
severely restricted during the last decade, particularly since increased cardiovascular
toxicity was reported in older studies [15]
[16]. This can be largely avoided with modern techniques, however, as the dosage to the
heart can be minimized [17].
Only two prospective studies have explicitly investigated the effect of IMN radiotherapy.
The only randomized study, conducted in France [18], included 1334 patients who after mastectomy were found to have either positive
axillary LNs (75 %) or a central/medial tumor location (25 %). The PMRT always included
the supraclavicular nodes. The target of randomization was additional IMN radiotherapy.
After 10 years, no survival advantage was seen (62.5 % vs. 59.3 %) with IMN radiotherapy,
and in lymph node–negative patients there was even a trend towards a reduced probability
of survival. By contrast, lymph node–positive patients had a nonsignificant survival
advantage. The authors admit that the study may not have been sufficiently powered
to demonstrate significant effects. No increased cardiotoxicity was observed.
In contrast to those findings, a more recent prospective study in Denmark reported
a significant benefit with additional IMN radiotherapy [19]. This population-based cohort study included 3089 patients with positive lymph nodes
(2003 – 2007), among whom 1492 women with right-sided tumors received simultaneous
irradiation of the IMNs while 1597 women with left-sided tumors underwent radiotherapy
without the IMNs (to avoid potential cardiotoxicity). ALND and adjuvant systemic therapy
were carried out in all cases. The tumor characteristics and risk factors were similar
in the two groups. The 8-year OS was 75.9 % with IMN radiotherapy versus 72.2 % without
it, and thus improved by 3.7 % (P = 0.005). Mortality rates due to ischemic cardiac
diseases were similar in the two groups. The positive effect of IMN radiotherapy was
clearest when there was a high risk of IMN metastases. A subgroup analysis of patients
with a medial/central location and/or four or more involved lymph nodes thus showed
an 8-year OS that was 7.4 % better (72.2 % vs. 64.8 %; P = 0.001). Although the study
was not randomized, it has a prospective design, homogeneous patient groups, and a
short period of 4 years in which all patients with breast carcinoma were treated in
a standardized way in the national tumor centers.
According to current studies, in addition to modern techniques the extent of the target
volume should be limited craniocaudally to the first to third intercostal spaces,
in order to avoid the increased cardiac toxicity that used to be seen in earlier studies
when the IMNs were irradiated simultaneously [7]
[8].
Assessment of the state of research
Assessment of the state of research
It should be pointed out when discussing these studies that systemic therapy was the
standard practice at the time; adjuvant chemotherapy followed in all studies, with
anthracycline- and taxane-containing regimens, and adjuvant endocrine therapy was
mainly with tamoxifen. Aromatase inhibitors and trastuzumab had not yet been approved
during the recruitment period in most cases; aromatase inhibitors were used in the
Danish cohort study starting in 2004. The possibility can therefore not be excluded
that the results might have been fundamentally different with intensified “modern”
systemic therapy; but on the other hand this is unlikely. In general, the relative
effect of radiotherapy is greater the more effective the systemic therapy is.
It is particularly noteworthy that significant cardiac toxicity was excluded by regional
irradiation in the two current studies, with median follow-up periods of 10.9 years
in the EORTC study and 9.5 years in the MA20 study. As a disadvantageous effect of
radiotherapy was already evident even after a relatively short follow-up period in
the last meta-analysis of the earlier studies, relevant cardiac toxicity during the
subsequent course is very unlikely [15].
The most important finding in these research studies is surely the fact that the effect
of regional irradiation on the metastasis-free survival was at least as large or even
larger than the effect on the locoregional recurrence rate. This type of “systemic
effect of radiotherapy” has not previously been observed with breast carcinoma, but
it has recently come to be regarded as quite plausible. Radiotherapy can produce immunogenic
cell death, so that it can trigger “off-scope” effects (remissions outside of the
target volume of the radiotherapy). This is currently being investigated in clinical
studies. If such an immunological effect does indeed exist, it will probably mainly
be found in patients with a low “residual risk” – i. e., after (optimal) systemic
therapy.
Procedure with positive sentinel lymph nodes: nothing, or X-rays instead of scalpels?
Procedure with positive sentinel lymph nodes: nothing, or X-rays instead of scalpels?
Sentinel node biopsies (SNBs) have become an established standard and have reduced
the role of ALND when there are clinically negative findings. On the basis of data
from a randomized study by the American College of Surgeons Oncology Group (ACOSOG),
various guidelines also allow avoidance of ALND and also ALN radiotherapy in patients
with a favorable risk profile, even when there are one or two involved lymph nodes
[4]
[5]
[6].
In the ACOSOG’s Z0011 study [20], 891 patients with T1 (70 %) to T2 (30 %) tumors and positive hormone receptor status
underwent randomization after SNBs had histologically demonstrated involved axillary
LNs (one or two involved LNs in approx. 90 % of cases, micrometastases in 45 %). No
ALND was carried out in the experimental arm, but it was done in the traditional way
in the control arm. In accordance with the protocol, all of the women were to undergo
whole-breast irradiation. After 5 years, no differences were seen in the DFS, OS,
or local recurrence rate. The authors concluded that in patients with “limited SLN
metastatic breast cancer with breast conservation and systemic therapy,” ALND does
not improve the prognosis – i. e., has no therapeutic effect. However, the study has
substantial methodological weaknesses: recruitment was interrupted prematurely, and
a total of 103 patients could not be analyzed. In violation of the explicit protocol
requirements, additional RNI was carried out in around 15 % of them [21]
[22],
With regard to the prognosis when there is sentinel node involvement, a countertrend
emerged in a subgroup analysis in the NSABP-B-32 study [23]. Fifteen percent of the lymph nodes initially evaluated as histopathologically negative
were found to have occult LN metastases at an additional histological examination.
After 5 years, these patients had slightly but significantly reduced DFS (2.8 %) and
OS (1.2 %) rates.
The effectiveness of ALN radiotherapy in comparison with ALND was investigated in
the randomized and prospective AMAROS study (EORTC 10 981 – 22 023) [3]. A total of 1425 patients with positive SNs underwent randomization (ALND, n = 744;
ALN radiotherapy, n = 681). The axillary 5-year recurrence rate was unexpectedly low
in both arms, at 0.43 % after ALND and 1.19 % after ALN radiotherapy (n.s.). The planned
noninferiority test was therefore not sufficiently powered. With comparable axillary
tumor spread in the two groups, it can be assumed that additional metastases were
present in the axilla after the SNB in around one-third of the patients who underwent
axillary radiotherapy. However, as the axillary recurrence rate after 5 years was
only 1.19 %, it can be concluded that these were effectively eliminated by the RT.
Nor were any differences seen in the DFS and OS. By contrast, lymphedema was reported
significantly more often after ALND, at 28 % after 5 years in comparison with 14 %
after ALN radiotherapy (P< 0.0001).
The question of whether patients with a clinically unremarkable axilla (cN0) require
axillary therapy, and if so which, can therefore hardly be answered at present. The
studies cited above show that supplementary ALND with one or two positive SNs does
not offer any therapeutic advantage. This has been confirmed by Li et al. recently
in a meta-analysis of 12 studies including a total of 130 575 patients [24]. In comparison with SNB alone, supplementary ALND showed no therapeutic benefit.
However, the authors limit their conclusions to patients with one to three positive
SNs, due to insufficient data for patients with more positive SNs and patients who
did not meet the ACOSOG Z0011 criteria.
Procedure with at least three positive SNs
Bonneau et al. investigated the question of whether supplementary ALND can also be
dispensed with when there are three or more positive LNs. Data were extracted from
the SEER database for a total of 9521 patients who were treated between 2003 and 2008
and whose prognostic characteristics matched the inclusion criteria for the ACOSOG-Z0011
study. The analysis showed that no prognostic benefit is achieved with supplementary
ALND even with at least three involved SNs [25].
The number of involved lymph nodes was for decades the most important parameter for
deciding on whether to administer adjuvant systemic therapy. In recent years, this
parameter has increasingly been replaced by assessment of intrinsic molecular biomarkers
and the use of multigene assays, however, so that in this respect as well, complete
ALND for diagnostic purposes has become dispensable [26]
[27]
[28].
In individual cases, and also only indirectly, nomograms may be helpful for assessing
the risk of additional LN metastases following positive sentinel nodes. The model
presented by Katz et al. [29], which calculates the probability of finding four or more positive LNs after supplementary
axillary dissection in patients with positive SNs, takes numerous established risk
factors into account (including the number of positive SNs, tumor size, extranodal
involvement, size of the positive SN, and histology).
The clinically negative axilla
As the data presented show, refraining from axillary dissection presupposes that the
axilla is clinically negative. According to the criteria in the ACOSOG Z0011 study,
negative palpation findings were sufficient to establish a clinically negative axilla.
Ultrasonography of the axilla has in the meantime become established, and in a meta-analysis
of 16 studies it was found to have moderate and widely varying sensitivity (49 – 87 %),
but higher specificity (56 – 97 %) [30]. It is expected that the INSEMA study that is currently in progress will clarify
the diagnostic and therapeutic procedure in patients with cN0. Two additional prospective
studies (NCT 02 466 737 and NCT 02 167 490) are also investigating whether it might
even be possible to dispense with SNB when the axilla is negative on ultrasound [31]
[32].
Complete inclusion of the axilla with extensive LN involvement
According to the present state of the data, extending RNI to the complete axilla does
not appear to be justified after complete axillary dissection, even when there is
evidence of more extensive involvement (pN3), except in cases of residual tumor [5]
[6]
[33].
Planning and technique of radiotherapy in RNI
Planning and technique of radiotherapy in RNI
CT-based three-dimensional planning of radiotherapy is the standard [33]
[34]
[35]. Contouring guides for the breast or breast wall and for the individual LN stations
[36]
[37]
[38] are helpful for defining the irradiation volume. Contouring of the at-risk organs
(particularly the heart, and ideally with partial volumes such as the left coronary
artery, lungs, and plexus) and documentation of the radiation dosage to each of these
structures are important. More recent techniques such as intensity-modulated radiotherapy
(IMRT) and volume-modulated radiotherapy (VMRT) can be used to optimize the dose distribution
and can markedly reduce the radiation burden on healthy organs [17]
[39]. When the tumor is left-sided, respiration-triggered radiotherapy or breath gating
techniques [40] and special positioning aids [41] can increase the distance between the target volume and the heart and thus reduce
the dosage.
The radiation dosage is usually 50.0 – 50.4 Gy in individual doses of 1.8 – 2.0 Gy.
Accelerated hypofractionation (HF; 40.0 – 42.5 Gy in individual doses of 2.60 – 2.66 Gy
over a treatment period of around 3 weeks) is not currently recommended as the fractionation
method of choice for patients receiving RNI, since higher individual doses may increase
the risk of late sequelae such as cardiac toxicity or plexopathy [42]
[43] and there are as yet insufficient data on this from randomized HF studies [44]
[45]. In particular, the risk of lymphedema after hypofractionation is unclear and is
currently being investigated in a study in Denmark.
Implications for everyday practice
Implications for everyday practice
The studies published in recent years have for the first time investigated the value
of lymph-node irradiation as an individual measure; previously there had only been
indirect comparisons between studies with local radiotherapy (breast, breast wall)
and studies with local plus regional radiotherapy. The most important finding of the
current research on lymph-node irradiation is: “more radiotherapy may be beneficial
in some circumstances.” There is thus a clear discrepancy from surgical treatment
of the axillary lymph nodes, for which a survival advantage resulting from “more surgery”
has never been demonstrated. An immunological effect (via immunogenic cell death after
radiotherapy) is under discussion as a possible explanation for this. However, it
is difficult to transfer these results into everyday practice. It is very likely that
regional radiotherapy is still advantageous for specific groups of patients even today.
The issue of which patients who are receiving modern systemic treatments may be able
to benefit from regional radiotherapy still requires better research. By contrast,
it is very unlikely that lymph-node irradiation is disadvantageous.
