Systematic review

Meta-analysis of aberrant lymphatic drainage in recurrent

breast cancer
M. Ahmed1 , R. Baker2 and I. T. Rubio3
1 Research Oncology, Division of Cancer Studies, King’s College London, London, and 2 Department of Statistics, School of Business 612, University of
Salford, Salford, UK, and 3 Breast Surgical Unit, Breast Cancer Centre, Hospital Universitario Vall d’Hebron, Barcelona, Spain
Correspondence to: Mr M. Ahmed, Research Oncology, Division of Cancer Studies, King’s College London, Guy’s Hospital Campus, Great Maze Pond,
London SE1 9RT, UK (e-mail: muneer.ahmed@kcl.ac.uk)

Background: Sentinel node biopsy (SNB) in recurrent breast cancer offers targeted axillary staging
compared with axillary lymph node dissection (ALND) or no treatment. The evidence for lymphatic
mapping in recurrent breast cancer is reviewed, focusing on aberrant drainage and its implications for
patient management.
Methods: A meta-analysis of studies evaluating lymphatic mapping in recurrent breast cancer was
performed. Outcomes included sentinel node identification, aberrant lymphatic pathways and metastatic
node rates in aberrant drainage and ipsilateral axilla. Pooled odds ratios (ORs) and 95 per cent confidence
intervals (c.i.) were estimated using fixed-effect analyses, or random-effects analyses in the event of
statistically significant heterogeneity.
Results: Seven studies reported data on lymphatic mapping in 1053 patients with recurrent breast cancer.
The intraoperative sentinel node identification rate was 59⋅6 (95 per cent c.i. 56⋅7 to 62⋅6) per cent, and
significantly greater when the original axillary surgery was SNB compared with ALND (OR 2⋅97, 95 per
cent c.i. 1⋅66 to 5⋅32). The rate of aberrant lymphatic drainage identification was 25⋅7 (23⋅0 to 28⋅3) per
cent, and significantly greater when the original axillary surgery was ALND (OR 0⋅27, 0⋅19 to 0⋅38). The
metastatic sentinel node rate was 10⋅4 (8⋅6 to 12⋅3) per cent, and a significantly greater metastatic nodal
burden was identified in the ipsilateral axilla (OR 6⋅31, 1⋅03 to 38⋅79).
Conclusion: Lymphatic mapping is feasible in recurrent breast cancer. It avoids ALND in over 50 per cent
of patients who have undergone SNB, and allows the 4 per cent of patients with metastatically involved
aberrant nodes to receive targeted surgical and adjuvant therapies.

Paper accepted 3 July 2016

Published online 6 September 2016 in Wiley Online Library (www.bjs.co.uk). DOI: 10.1002/bjs.10289

Introduction feasible in patients with recurrent breast cancer10 and this

Sentinel node biopsy (SNB) is the standard of care for
axillary staging in patients with a clinically and radiolog-
ically negative axilla1 – 5 . However, between 5 and 10 per
cent of patients with breast cancer will experience a local
recurrence in the breast within 10 years6,7 . The standard
management for patients who have previously undergone
SNB has been to proceed to an axillary lymph node dissec-
tion (ALND), with its associated risk of morbidity8 . This
has been based on the perceived disruption of lymphatic
drainage pathways to the axilla by surgical and adjuvant
therapy9 , which may interfere with the efficacy of the SNB
technique, through non-sentinel node uptake of radio-
active tracer and blue dye, resulting in a high false-negative
rate. However, there is increasing evidence that SNB is
has resulted in the increased use of repeat SNB for axillary
staging. This approach raises the possibility of more aber-
rant drainage pathways (lymphatic pathways outside the
ipsilateral axilla) being identified during surgery than with
current management. The implications of these aberrant
pathways for further management and prognostic outcome
have not been assessed previously, and form the basis of this
review.
Methods
Study selection
A systematic review of the literature was performed using
PubMed, Embase and the Cochrane Library databases to

© 2016 BJS Society Ltd BJS 2016; 103: 1579–1588

Published by John Wiley & Sons Ltd
1580
identify all original articles published up to November
2015 that evaluated the role of SNB in recurrent breast
cancer. The search terms used were: sentinel node biopsy
AND lymphatic mapping AND recurrent breast cancer.
Studies were restricted to those published in the English
language and performed in humans. The related articles
function was used to broaden the search, and all abstracts,
studies and citations obtained were reviewed. References of
the acquired articles were also searched by hand. The last
search was conducted on 16 November 2015.
Inclusion criteria
Studies were included if they satisfied the following
eligibility criteria: performance of SNB or lymphatic
mapping in recurrent breast cancer; documented the
presence or absence of aberrant lymphatic drainage,
defined as lymphatic pathways outside the ipsilateral
axilla, either on lymphoscintigraphy or during SNB;
assessed more than 50 patients undergoing SNB or lym-
phatic mapping; attained a satisfactory quality assessment
score (at least 4 of 6); and were written in the English
language.
Exclusion criteria
Studies that failed to fulfil the inclusion criteria and those
in which the outcomes of interest were not reported
were excluded. Other exclusion criteria were: full text
not available; review article; letter to the editor; edi-
torial report; case report; duplicate publication; and
abstract.
Titles identified from database searches n = 111
PubMed n = 70
Embase n = 41
Cochrane Library n = 0
Title and abstract screened n = 93
Full-text articles assessed for eligibility n = 24
Included in review n = 7
Fig. 1 Selection of articles for review
© 2016 BJS Society Ltd
Published by John Wiley & Sons Ltd
M. Ahmed, R. Baker and I. T. Rubio
Data extraction
Data were extracted from the selected studies using a
data extraction form, which included information on:
publication details; study design; number of patients; orig-
inal breast and axillary surgery; adjuvant therapies; SNB
technique and timing of lymphoscintigraphy; follow-up
period; number of patients identified with sentinel nodes
on SNB and lymphoscintigraphy; number of patients
with aberrant lymphatic drainage identified using these
techniques; number of patients with involved sentinel
nodes (excluding isolated tumour cells where possible);
false-negative rate of SNB; and axillary, local and sys-
temic recurrence rates. The quality of cohort studies was
assessed according to the STROBE recommendations11
and six items of the STROBE statement were con-
sidered relevant for quality evaluation. Two reviewers
extracted data from included studies independently.
Comparison of the data extraction and quality score
was undertaken, and discrepancies were resolved by
consensus.
Statistical analysis
All extracted data were tabulated and presented in terms
of patient numbers. Ratios were calculated according
to the original type of axillary surgery performed;
involved sentinel nodes were also tabulated according
to the ratio of ipsilateral axilla to aberrant drainage
pathways to address the relevant outcomes of included
studies.
RevMan 5.0 software was used to perform the
meta-analysis12 . Pooled means were calculated with
Duplicates excluded
n = 18
Excluded based on title and abstract
n = 69
Excluded according to inclusion
criteria (< 50 patients) n = 17
www.bjs.co.uk BJS 2016; 103: 1579–1588
Aberrant lymphatic drainage in recurrent breast cancer 1581

Table 1 Study characteristics

No. of No. of
patients patients undergoing
Total Original Original receiving Radioisotope lymphoscintigraphy
Study no. of breast axillary adjuvant (± blue dye) and timing after Follow-up
Reference type patients surgery surgery therapy injection technique injection of radioisotope (months)*

Uth et al.21 R 147 BCS 133 SNB 74 RT 124 99m Tc-labelled nanocolloid 121 38(11⋅5)
Mx 14 ALND 73 (40 MBq); PT and PA for 2 h after injection
palpable cancer and PA alone
for non-palpable
Patent Blue dye; PA or at site of
recurrence, or both
Vugts et al.23 P 502‡ BCS 441 SNB 214 RT 441§ 99m Tc-labelled nanocolloid; PT, IT, 502 n.s.
Mx 61 ALND 288 PA or SA Timing n.s.
Patent Blue dye; site not specified
Cordoba et al.17 P 53 BCS 47 SNB 10 RT 46 99m Tc-labelled nanocolloid; SA or 53 16(9)
Mx 6 ALND 43 ST 29 on mastectomy scar Planar and SPECT–CT images
1 h after injection
van der Ploeg et al.22 R 115¶ BCS 115# SNB 103# RT 86 99m Tc-labelled nanocolloid 115 34(4)
ALND 12 (120 MBq); IT Static images 10 min, 2 and
1 ml Patent Blue dye; site not 4 h after injection;
specified SPECT–CT at 4 h
Cox et al.18 R 56 n.s. n.s. RT 41 99m Tc-labelled nanocolloid; PT 13 26
or PA Timing n.s.
5 ml isosulphan blue; PT or PA
Intra et al.19 R 65 BCS 65 SNB 65 RT 57 99m Tc-labelled nanocolloid 65 45⋅9†
(10–12 MBq); SD Planar images at 15, 30 and
120 min after injection
Port et al.20 P 115 BCS 115 SNB 54 RT 92 99m Tc-labelled
nanocolloid 112 26⋅4
ALND 63** ST 51 (0⋅1–0⋅5 mCi); ID Planar images at 30 and
Isosulphan blue; SD 120 min after injection

*Values are mean(s.d.), except †median. ‡Only patients who had undergone axillary interventions previously were included (34 patients without previous
axillary intervention excluded). §Based on patients receiving breast-conserving surgery (BCS). ¶Ten patients with no previous axillary surgery. #Includes
13 excisional biopsies that were benign on histopathology. **Represents number of procedures as two patients underwent bilateral breast and axillary
surgery. R, retrospective cohort; Mx, mastectomy; SNB, sentinel node biopsy; ALND, axillary lymph node dissection; PT, peritumoral; PA, periareolar;
P, prospective cohort; IT intratumoral; SA subareolar; n.s., not stated; ST, systemic therapy; SPECT, single-photon emission CT; SD, subdermal; ID,
intradermal.

95 per cent confidence intervals (c.i.). The odds ratio Study characteristics
(OR) with 95 per cent confidence interval was calcu-
The studies were published between 2007 and 2015.
lated for binary data variables. The Mantel–Haenszel
They comprised seven cohort studies17 – 23 (Table 1) of
method was used to combine the ORs for the outcomes
which three17,20,23 were prospective. Original breast
of interest. The fixed-effect model was used to calculate
surgery was predominantly breast conservation; axillary
the pooled outcome for binary variables. If heterogeneity
surgery demonstrated greater variation between studies
was present, the random-effects model was used13 – 15 .
(Table 1). Adjuvant radiotherapy was performed in all stud-
Potential heterogeneity of the study results was assessed
ies, whereas use of chemotherapy was reported in only
with a forest plot, displaying the OR estimates of dif-
two17,20 . Only two studies17,19 failed to use the dual tech-
ferent outcomes for SNB in recurrent breast cancer for
nique of radioisotope and blue dye for SNB; none relied
each study. Potential heterogeneity of the study results
on blue dye alone. Lymphoscintigraphy was performed
was also evaluated statistically using the χ2 test, with
a maximum I 2 value of 30 per cent identifying low in all studies, and in all patients in four of these17,19,22,23 .
heterogeneity16 . Follow-up ranged from 16 to 45⋅9 months17,19 .

Results Study quality

The detailed literature search resulted in seven studies Relevant items of the STROBE statement that were used
being critically appraised in this review (Fig. 1)17 – 23 . for quality assessment of included cohort studies are shown

© 2016 BJS Society Ltd www.bjs.co.uk BJS 2016; 103: 1579–1588

Published by John Wiley & Sons Ltd
1582 M. Ahmed, R. Baker and I. T. Rubio

Table 2 Study quality assessment of cohort studies

Only patients with proven Patient Withdrawals

Study Clear inclusion Standardized malignancy follow-up from study
Reference objectives criteria technique (in situ or invasive) reported reported

Uth et al.21 Yes Yes Yes Yes Yes Yes

Vugts et al.23 Yes Yes Yes Yes No Yes
Cordoba et al.17 Yes Yes Yes Yes Yes Yes
van der Ploeg et al.22 Yes Yes Yes No Yes Yes
Cox et al.18 Yes Yes Yes Yes Yes Yes
Intra et al.19 Yes Yes Yes Yes Yes Yes
Port et al.20 Yes Yes Yes Yes Yes Yes

Study quality was assessed according to the STROBE statement11 .

Sentinel nodes identified

Reference SNB ALND Weight (~) Odds ratio Odds ratio
Cordoba et al.17 4 of 10 34 of 43 17·8 0·18 (0·04, 0·76)
Uth et al.21 32 of 74 32 of 73 29·2 0·98 (0·51, 1·87)
van der Ploeg et al.22 90 of 103 6 of 12 20·2 6·92 (1·94, 24·71)
Vugts et al.23 141 of 214 166 of 288 32·8 1·42 (0·98, 2·05)

Total 267 of 401 238 of 416 100·0 1·21 (0·50, 2·93)

Heterogeneity: 2 = 0·59; 2 = 14·77, 3 d.f., P = 0·002; I2 = 80~

0·01 0·1 1 10 100
Test for overall effect: Z = 0·42, P = 0·67
Favours ALND Favours SNB

a Lymphoscintigraphy

Sentinel nodes identified

Reference SNB ALND Weight (~) Odds ratio Odds ratio
Cordoba et al.17 4 of 10 24 of 43 11·6 0·53 (0·13, 2·14)
Port et al.20 40 of 54 24 of 63 21·2 4·64 (2·10, 10·26)
Uth et al.21 48 of 74 24 of 73 23·6 3·77 (1·90, 7·46)
van der Ploeg et al.22 91 of 103 6 of 12 13·0 7·58 (2·10, 27·32)
Vugts et al.23 136 of 214 123 of 288 30·7 2·34 (1·63, 3·36)

Total 319 of 455 201 of 479 100·0 2·97 (1·66, 5·32)

Heterogeneity: 2 = 0·25; 2 = 11·01, 4 d.f., P = 0·03; I2 = 64~

0·01 0·1 1 10 100
Test for overall effect: Z = 3·66, P < 0·001
Favours ALND Favours SNB

b SNB (second axillary procedure)

Fig. 2Forest plot comparing sentinel node identification rates determined by a lymphoscintigraphy and b sentinel node biopsy (SNB)
(second axillary procedure) following original surgery by SNB versus axillary lymph node dissection (ALND). A Mantel–Haenszel
random-effects model was used for meta-analysis. Odds ratios are shown with 95 per cent confidence intervals

in Table 2. The overall STROBE score ranged from 5 to 6.
The methodology and reported data of all included cohort
studies were considered adequate.
Study outcomes
Sentinel node identification rates
Six studies17,19 – 23 , comprising 997 patients, provided data
on the sentinel node identification rate on preoperative
lymphoscintigraphy. A sentinel node was identified in a
total of 631 patients, providing a pooled mean sentinel
node identification rate on preoperative lymphoscinti-
graphy of 63⋅3 (95 per cent c.i. 60⋅3 to 66⋅3) per cent. Four
studies17,21 – 23 provided data on the impact of previous
surgery (SNB versus ALND) on sentinel node identifica-
tion (Fig. 2a). There was significant heterogeneity (I 2 = 80
per cent, P = 0⋅002). In the random-effects model, there
was no statistically significant difference between previous

© 2016 BJS Society Ltd www.bjs.co.uk BJS 2016; 103: 1579–1588

Published by John Wiley & Sons Ltd
Aberrant lymphatic drainage in recurrent breast cancer 1583

Aberrant sentinel nodes identified

Reference SNB ALND Weight (~) Odds ratio Odds ratio
Cordoba et al.17 1 of 10 20 of 43 5·1 0·13 (0·01, 1·10)
Port et al.20 3 of 54 16 of 63 10·5 0·17 (0·05, 0·63)
Uth et al.21 9 of 74 15 of 73 10·0 0·54 (0·22, 1·32)
van der Ploeg et al.22 8 of 103 3 of 12 3·7 0·25 (0·06, 1·12)
Vugts et al.23 40 of 214 136 of 288 70·8 0·26 (0·17, 0·39)

Total 61 of 455 190 of 479 100·0 0·27 (0·19, 0·38)

Heterogeneity: 2 = 3·21, 4 d.f., P = 0·52; I2 = 0~

0·01 0·1 1 10 100
Test for overall effect: Z = 7·46, P < 0·001
Favours ALND Favours SNB

a Lymphoscintigraphy

Aberrant sentinel nodes identified

Reference SNB ALND Weight (~) Odds ratio Odds ratio
Uth et al.21 6 of 74 6 of 73 52·8 0·99 (0·30, 3·21)
van der Ploeg et al.22 8 of 103 3 of 12 47·2 0·25 (0·06, 1·12)

Total 14 of 177 9 of 85 100·0 0·64 (0·25, 1·64)

Heterogeneity: 2 = 2·00, 1 d.f., P = 0·16; I2 = 50~

0·01 0·1 1 10 100
Test for overall effect: Z = 0·93, P = 0·35
Favours ALND Favours SNB

b SNB (second axillary procedure)

Forest plot comparing aberrant sentinel node identification rates determined by a lymphoscintigraphy and b sentinel node biopsy
Fig. 3
(SNB) (second axillary procedure) following original surgery by SNB versus axillary lymph node dissection (ALND). A
Mantel–Haenszel fixed-effect model was used for meta-analysis. Odds ratios are shown with 95 per cent confidence intervals

surgery type and sentinel node identification rates (OR
1⋅21, 95 per cent c.i. 0⋅50 to 2⋅93).
A total of seven studies17 – 23 , comprising 1053 patients,
contributed to the intraoperative sentinel node identi-
fication rate. A sentinel node was identified in a total
of 628 patients, providing a pooled mean intraoperative
identification rate of 59⋅6 (56⋅7 to 62⋅6) per cent. Five
studies17,20 – 23 provided data on the impact of previous
surgery on intraoperative sentinel node identification rates
(Fig. 2b). There was significant heterogeneity (I 2 = 64 per
cent, P = 0⋅03). In the random-effects model, there was
a significantly greater sentinel node identification rate
when the original axillary surgery was SNB compared with
ALND (OR 2⋅97, 1⋅66 to 5⋅32).
Aberrant sentinel node identification
Six studies17,19 – 23 , including 997 patients, provided data
on aberrant lymphatic drainage pathways on preoperative
lymphoscintigraphy. This was standardized in all included
studies as referring to lymphatic drainage outside the
ipsilateral axilla. Aberrant lymphatic drainage was iden-
tified in a total of 256 patients, providing a pooled mean
aberrant sentinel node identification rate on preoperative
lymphoscintigraphy of 25⋅7 (23⋅0 to 28⋅3) per cent. Five
studies17,20 – 23 provided data on the impact of previous
surgery on aberrant sentinel node identification rates
(Fig. 3a). There was no significant heterogeneity (I 2 = 0
per cent, P = 0⋅52). In the fixed-effect model, there was a
significant difference favouring ALND for greater aberrant
sentinel node identification (OR 0⋅27, 0⋅19 to 0⋅38).
Five studies18 – 22 , comprising 498 patients, provided data
on aberrant lymphatic drainage pathways identified during
surgery. Once again, this was standardized in all included
studies as referring to lymphatic drainage outside the ipsi-
lateral axilla. Aberrant lymphatic drainage pathways were
identified in a total of 38 patients, providing a pooled mean
aberrant sentinel node identification rate at SNB of 7⋅6 (5⋅3
to 10⋅0) per cent. Only two studies21,22 provided adequate
data on the impact of previous surgery on aberrant sentinel
node identification rates at sentinel node biopsy (Fig. 3b).
There was no significant heterogeneity. In the fixed-effect
model, there was no statistically significant difference
between the type of previous surgery and aberrant sentinel
node identification at SNB (OR 0⋅64; 0⋅25 to 1⋅64).

© 2016 BJS Society Ltd www.bjs.co.uk BJS 2016; 103: 1579–1588

Published by John Wiley & Sons Ltd
1584 M. Ahmed, R. Baker and I. T. Rubio

Metastatic sentinel nodes

Reference Ipsilateral axilla Aberrant Weight (~) Odds ratio Odds ratio
Cordoba et al.17 4 of 8 4 of 8 16·8 1·00 (0·14, 7·10)
Intra et al.19 6 of 7 1 of 7 13·3 36·00 (1·80, 718·68)
Port et al.20 8 of 10 2 of 10 16·1 16·00 (1·79, 143·15)
Uth et al.21 14 of 17 3 of 17 17·5 21·78 (3·73, 127·02)
van der Ploeg et al.22 10 of 12 2 of 12 16·2 25·00 (2·92, 213·99)
Vugts et al.23 19 of 46 27 of 46 20·1 0·50 (0·22, 1·14)

Total 61 of 100 39 of 100 100·0 6·31 (1·03, 38·79)

Heterogeneity: = 4·10; = 30·86, 5 d.f., P < 0·001;

2 2 I2 = 84~
Test for overall effect: Z = 1·99, P = 0·05 0·001 0·1 1 10 1000
Favours aberrant Favours ipsilateral axilla

Fig. 4 Forest plot comparing identification of metastatically involved sentinel nodes in aberrant drainage pathways versus the ipsilateral
axilla. A Mantel–Haenszel random-effects model was used for meta-analysis. Odds ratios are shown with 95 per cent confidence
intervals

Ipsilateral axilla versus aberrant sentinel node metastatic
rates
Seven studies17 – 23 , comprising 1053 patients, provided
data on identification of metastatic sentinel nodes. These
were identified in a total of 109 patients, providing a
metastatic sentinel node rate of 10⋅4 (8⋅6 to 12⋅3) per
cent. Six studies17,19 – 23 provided data on the metastatic
node rates between the ipsilateral axilla and aberrant
drainage locations (Fig. 4). There was significant hetero-
geneity (I 2 = 84 per cent, P < 0⋅001). In the random-effects
model, there was a significant difference favouring greater
metastatic nodal involvement in the ipsilateral axilla
compared with aberrant lymphatic pathways (OR 6⋅31,
1⋅03 to 38⋅79).
False-negative rates
Only Cordoba et al.17 performed completion ALND in all
patients, irrespective of findings on SNB. In their cohort of
53 patients, there were no false-negatives among patients
who previously underwent SNB (only 10 patients), but six
false-negatives in the remainder, who originally had an
ALND. Vugts and colleagues23 assessed a small subcohort
of 31 patients who, despite a normal SLN at repeat SNB,
underwent completion ALND in the early part of their
study, and identified two involved nodes (macrometas-
tases). Port and co-workers20 similarly performed ‘back-up
ALND’ in a subcohort of 23 patients with a negative SNB,
and identified two involved nodes. They also identified a
single involved node at completion ALND among 53 failed
SNBs. The remaining studies were reliant on identification
of false-negatives from completion ALND in patients with
failed SNB. Uth et al.21 identified five involved nodes at
ALND in 72 failed SNBs; Cox and colleagues18 identified
one involved node at ALND among 11 failed SNBs; Intra
and co-workers19 identified one involved node at ALND
of only two failed SNBs; and van der Ploeg et al.22 did not
identify any involved nodes in the 15 patients who under-
went completion ALND after failed SNB.
Discussion
Traditional management of the axilla in recurrent breast
cancer has consisted of either leaving the axilla alone, in
the case of previous ALND, or proceeding to ALND in the
event of previous SNB. The former is based on the premise
that the axillary nodal content has been removed in its
entirety, and lymphatic drainage of the breast from the
areola and parenchyma converge on anteropectoral nodes
in the axilla24,25 , making further axillary staging futile. The
latter is based on the perceived disruption of lymphatic
drainage at initial SNB, meaning that the SLN identifica-
tion rate at repeat SNB will be low and the false-negative
rate high. However, the present results question the
validity of these traditional concepts by reporting a SLN
identification rate of 63⋅3 (95 per cent c.i. 60⋅3 to 66⋅3) per
cent at lymphoscintigraphy and 59⋅6 (56⋅7 to 62⋅6) per cent
at SNB. A previous meta-analysis10 of recurrent breast can-
cer, which had a cohort half the size of that in the present
study including case reports, reported values of 70⋅8 (66⋅9
to 74⋅5) and 81 (76 to 85⋅2) per cent for lymphoscintigraphy
and SNB respectively. Interestingly, that meta-analysis10
found higher identification rates at SNB. This may be arte-
factual owing to the smaller size of the study and particu-
larly the pooled number of patients in the SNB group being
half of that in the lymphoscintigraphy group. Nevertheless,
this clearly demonstrates that lymphatic mapping and SNB
are feasible in recurrent breast cancer despite lymphatic

© 2016 BJS Society Ltd www.bjs.co.uk BJS 2016; 103: 1579–1588

Published by John Wiley & Sons Ltd
Aberrant lymphatic drainage in recurrent breast cancer
disruption at previous surgery. Here, a significantly greater
sentinel node identification rate was identified at surgery
when the original axillary procedure was SNB rather than
ALND (OR 2⋅97, 95 per cent c.i. 1⋅66 to 5⋅32). This
finding is consistent with the targeted, minimally invasive
nature of SNB, which allows minimal lymphatic disrup-
tion and a larger number of axillary nodes to be retained
compared with ALND. This finding was not replicated
on lymphoscintigraphy, for which there was no significant
difference for sentinel node identification according to
type of previous axillary surgery (OR 1⋅21, 0⋅50 to 2⋅93).
This suggests that the technical performance of SNB is
adversely influenced far more by the presence of greater
axillary disruption, thus making targeted dissection diffi-
cult, than lymphatic mapping using lymphoscintigraphy,
and that remnant lymph nodes are present after ALND
making lymphatic mapping and SNB feasible in these
scenarios.
Only Port and colleagues20 made a brief assessment of
the number of nodes retrieved at original ALND, in five
patients who had a positive node identified at SNB in breast
cancer recurrence. In all these patients more than ten nodes
were excised at the original ALND. Unfortunately, owing
to limitations in the available data, it was not possible to
gain more information on the number of nodes retrieved
at original and recurrent axillary procedures. It may be the
case that the extent of nodal retrieval at primary ALND
influences the SLN identification rates in cancer recur-
rence and the establishment of aberrant drainage pathways.
However, this review is unable to provide data to support
or refute this, and it is recommended that future clinical
trials or registries record such data.
The metastatic sentinel node rate of 10⋅4 (8⋅6 to 12⋅3) per
cent and SLN identification rate of 59⋅6 per cent at SLNB
suggest that about 50 per cent of patients with recurrent
breast cancer could avoid ALND, with its established mor-
bidity, by undergoing SNB. However, the greatest poten-
tial benefit of SNB in the recurrent breast cancer setting is
the targeted visualization that it provides over ALND, with
particular reference to extra-axillary or aberrant lymphatic
drainage, and whether this finding may change manage-
ment and patient outcomes. The aberrant sentinel node
identification rate was 25⋅7 (23⋅0 to 28⋅3) and 7⋅6 (5⋅3 to
10⋅0) per cent on lymphoscintigraphy and SNB respec-
tively. The latter lower value is possibly due to confounding
owing the smaller number of studies contributing to it, but
could be explained by the fact that surgical exploration of
aberrant nodes, performed in all studies by open surgical
excision, may be limited. This lack of surgical exploration
is likely attributable to the potential additional morbidity
associated with invasive surgical excision (including risk of
© 2016 BJS Society Ltd
Published by John Wiley & Sons Ltd
1585
pneumothorax and bleeding) and current limited evidence
for recommendation26 .
Although the presence of metastatic sentinel nodes was
significantly greater within ipsilateral axillary nodes (OR
6⋅31, 1⋅03 to 38⋅79), aberrant nodes still contributed almost
40 per cent of all metastatic nodes (3⋅7 per cent of all 1053
patients). In this review, none of the 39 aberrant metastatic
nodes identified was associated with any concurrent, ipsi-
lateral axillary involvement, which potentially altered adju-
vant management. The impact of aberrant metastatic sen-
tinel node identification was specified in three of the
included studies17,19,22 . The four metastatic aberrant nodes
identified by Cordoba and colleages17 resulted in adjuvant
radiotherapy. The identification of two patients with aber-
rant sentinel nodes in the contralateral axilla by van der
Ploeg et al.22 led to completion ALND in the contralat-
eral axilla, and the single internal mammary node identi-
fied by Intra and co-workers19 allowed avoidance of ALND
and administration of adjuvant radiotherapy to the internal
mammary chain instead. Therefore, SNB allowed the adju-
vant treatment plan to be altered based on the outcome of
lymphatic mapping in these patients, which could include
adjuvant systemic therapy or regional radiotherapy to the
aberrant nodal site, or both, or contralateral ALND. The
identification of these aberrant nodes would not have been
possible through standard ALND of the ipsilateral axilla.
However, it is not exclusively patients with aberrant lym-
phatic drainage who benefit from alterations in adjuvant
management. In the study by Vugts and colleagues23 , 29 of
46 patients with positive sentinel nodes had the addition of
adjuvant radiotherapy or systemic therapy, or both, based
on the SNB findings, whereas Cordoba et al.17 reported
that all patients with a positive sentinel node received sys-
temic therapy. Intra and co-workers19 and van der Ploeg
et al.22 reported that all patients with positive ipsilateral
sentinel nodes, six and ten respectively, underwent comple-
tion ALND. Therefore, from the four trials17,19,22,23 that
providing data, identification of a positive sentinel node
resulted in a change to adjuvant management in 56 of a
total of 735 patients (7⋅6 per cent). However, with sen-
tinel node identification rate at SNB being about 60 per
cent and a positive node rate of 10 per cent, the greatest
benefit of SNB in recurrent breast cancer is the avoid-
ance of ALND in 50 per cent of patients. This ability to
alter adjuvant therapy applies particularly to patients who
have undergone ALND previously, as conventional man-
agement would involve no further staging of the axilla.
Therefore, the axilla could be understaged in the absence
of SNB, highlighting the benefits of this technique.
Regional nodal irradiation has been demonstrated to pro-
duce a marginal improvement in overall survival at median
www.bjs.co.uk BJS 2016; 103: 1579–1588
1586
follow-up of 10⋅9 years (82⋅3 versus 80⋅7 per cent; hazard
ratio (HR) 0⋅87, 95 per cent c.i. 0⋅76 to 1⋅00; P = 0⋅06) when
added to whole breast or thoracic irradiation in patients
with medial or centrally located primary tumours (irre-
spective of axillary involvement) or an externally located
tumour with axillary involvement27 . Until long-term
follow-up data from current RCTs assessing the benefit
of regional irradiation become available, lymphoscinti-
graphy and aberrant sentinel nodes may be used to guide
decisions regarding locoregional treatment. In terms of
systemic treatment, this may be not so clear in view of the
recent results of the CALOR (chemotherapy for isolated
loco-regional recurrence of breast cancer) trial28 , where
patients with isolated tumour recurrences who underwent
complete surgical excision with clear margins (lumpectomy
or mastectomy) were randomized to chemotherapy or not.
Five-year disease-free survival rates were 69 and 57 per
cent in the chemotherapy and no-chemotherapy groups
respectively (HR 0⋅59, 0⋅35 to 0⋅99; P = 0⋅05), and the
benefit of adding chemotherapy was significantly greater
in patients with hormone receptor-negative tumours
(P = 0⋅05)28 .
However, the recurrent cancer setting is unique com-
pared with primary breast cancer, where the identification
of extra-axillary nodes is significantly associated with
peritumoral injection of radioisotopes and the presence of
concurrent ipsilateral axillary involvement, which conse-
quently has a limited impact on adjuvant treatment plans26 .
In the recurrent breast cancer setting, potential lymphatic
(A)
Fig. 5Lymphoscintigraphy of aberrant lymphatic drainage of the
breast to the contralateral axilla, showing the radioisotope
injection site in the ipsilateral breast (A) and radioactive uptake
in a contralateral axillary sentinel node (arrow) (reproduced
courtesy of Vugts et al.23 )
© 2016 BJS Society Ltd
Published by John Wiley & Sons Ltd
M. Ahmed, R. Baker and I. T. Rubio
drainage outside the ipsilateral axilla should be anticipated
and the presence of contralateral axillary involvement con-
sidered as originating from aberrant lymphatic drainage of
the ipsilateral breast rather than systemic spread of circu-
lating tumour cells (Fig. 5). This has a positive implication
for prognostic outcome. No axillary recurrences were
reported in the included studies, one study19 identified
a local recurrence (2 per cent) and three studies17,19,20
reported systemic recurrence rates ranging between
1⋅5 and 5⋅2 per cent. Follow-up varied between studies
and benefit would be gained from longer-term figures.
However, the low event rate demonstrates the excellent
prognosis of these patients. This was also demonstrated in
a small retrospective review29 of 48 patients with contra-
lateral lymph node recurrence, which reported an overall
survival rate of 83 per cent (mean 50⋅3 months’ follow-up);
this is not comparable to the poor prognosis of metastatic
breast cancer, with nearly all patients being treated with
curative rather than palliative intent.
The significantly greater number of aberrant nodes iden-
tified on lymphoscintigraphy after previous ALND com-
pared with SNB (OR 0⋅27, 0⋅19 to 0⋅38) also supports the
established concept of greater lymphatic disruption and
aberrant drainage development from more extensive pre-
vious surgical manipulation, and the importance of lym-
phatic mapping in this group of patients rather than ignor-
ing the axilla completely. This finding was not replicated
during SNB (OR 0⋅64, 0⋅25 to 1⋅64), raising the possibility
of a false-positive result (type I error), although the likely
explanation is that only two studies contributed data to this
outcome (Fig. 3b); if a larger number of studies were avail-
able, the findings for lymphoscintigraphy might be repli-
cated for SNB. Unfortunately, it was also not possible from
the available data to determine whether aberrant nodes
identified after previous ALND had a greater incidence
of metastatic involvement; this could be a future area of
exploration.
This review included 1053 patients who underwent SNB
after recurrent breast cancer. Strict inclusion criteria were
applied, and case series and small cohort studies, which
have formed the basis for the previous reviews on the topic,
were excluded. The study suffers from heterogeneity in
the available data, but this was taken into account dur-
ing statistical analysis. Despite these limitations, the ben-
efit of SNB in recurrent breast cancer in identification of
patients who can receive targeted localized surgery and
adjuvant therapy, and avoid unnecessary ALND, has been
demonstrated. Further long-term evaluation of patient
outcomes is required, which would be best addressed
through RCTs or at least prospectively maintained data
registries. This would allow stratification of groups to
www.bjs.co.uk BJS 2016; 103: 1579–1588
Aberrant lymphatic drainage in recurrent breast cancer
reduce heterogeneity and strengthen the conclusions about
the ability of SNB to alter adjuvant patient management
and improve long-term outcomes.
Disclosure
The authors declare no conflict of interest.
References
1 Gill G; SNAC Trial Group of the Royal Australasian
College of Surgeons (RACS) and NHMRC Clinical Trials
Centre. Sentinel-lymph-node-based management or routine
axillary clearance? One-year outcomes of sentinel node
biopsy versus axillary clearance (SNAC): a randomized
controlled surgical trial. Ann Surg Oncol 2009; 16: 266–275.
2 Krag DN, Anderson SJ, Julian TB, Brown AM, Harlow SP,
Ashikaga T et al.; National Surgical Adjuvant Breast and
Bowel Project. Technical outcomes of sentinel-lymph-node
resection and conventional axillary-lymph-node dissection
in patients with clinically node-negative breast cancer:
results from the NSABP B-32 randomised phase III trial.
Lancet Oncol 2007; 8: 881–888.
3 Mansel RE, Fallowfield L, Kissin M, Goyal A, Newcombe
RG, Dixon JM et al. Randomized multicenter trial of
sentinel node biopsy versus standard axillary treatment in
operable breast cancer: the ALMANAC Trial. J Natl Cancer
Inst 2006; 98: 599–609.
4 Veronesi U, Paganelli G, Viale G, Luini A, Zurrida S,
Galimberti V et al. A randomized comparison of
sentinel-node biopsy with routine axillary dissection in
breast cancer. N Engl J Med 2003; 349: 546–553.
5 Zavagno G, De Salvo GL, Scalco G, Bozza F, Barutta L,
Del Bianco P et al. A randomized clinical trial on sentinel
lymph node biopsy versus axillary lymph node dissection in
breast cancer: results of the Sentinella/GIVOM trial. Ann
Surg 2008; 247: 207–213.
6 Veronesi U, Cascinelli N, Mariani L, Greco M, Saccozzi R,
Luini A et al. Twenty-year follow-up of a randomized study
comparing breast-conserving surgery with radical
mastectomy for early breast cancer. N Engl J Med 2002; 347:
1227–1232.
7 Voogd AC, Nielsen M, Peterse JL, Blichert-Toft M,
Bartelink H, Overgaard M et al.; Danish Breast Cancer
Cooperative Group, Breast Cancer Cooperative Group of
the European Organization for Research and Treatment of
Cancer. Differences in risk factors for local and distant
recurrence after breast-conserving therapy or mastectomy
for stage I and II breast cancer: pooled results of two large
European randomized trials. J Clin Oncol 2001; 19:
1688–1697.
8 Lyman GH, Temin S, Edge SB, Newman LA, Turner RR,
Weaver DL et al.; American Society of Clinical Oncology
Clinical Practice. Sentinel lymph node biopsy for patients
with early-stage breast cancer: American Society of Clinical
© 2016 BJS Society Ltd
Published by John Wiley & Sons Ltd
1587
Oncology clinical practice guideline update. J Clin Oncol
2014; 32: 1365–1383.
9 Perre CI, Hoefnagel CA, Kroon BB, Zoetmulder FA,
Rutgers EJ. Altered lymphatic drainage after
lymphadenectomy or radiotherapy of the axilla in patients
with breast cancer. Br J Surg 1996; 83: 1258.
10 Maaskant-Braat AJ, Voogd AC, Roumen RM,
Nieuwenhuijzen GA. Repeat sentinel node biopsy in
patients with locally recurrent breast cancer: a systematic
review and meta-analysis of the literature. Breast Cancer Res
Treat 2013; 138: 13–20.
11 von Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC,
Vandenbroucke JP; STROBE, Initiative. The Strengthening
the Reporting of Observational Studies in Epidemiology
(STROBE) statement: guidelines for reporting
observational studies. Lancet 2007; 370: 1453–1457.
12 The Cochrane Collaboration. Review Manager (RevMan)
[Computer program], Version 5.0. The Nordic Cochrane
Centre, The Cochrane Collaboration: Copenhagen, 2008.
13 Demets DL. Methods for combining randomized clinical
trials: strengths and limitations. Stat Med 1987; 6: 341–350.
14 Higgins JPT, Green S (eds). Cochrane Handbook for
Systematic Reviews of Interventions, Version 5.0.0 (updated
February 2008). http://handbook.cochrane.org [accessed 1
August 2016].
15 DerSimonian R, Laird N. Meta-analysis in clinical trials.
Controlled Clin Trials 1986; 7: 177–188.
16 Higgins JP, Thompson SG. Quantifying heterogeneity in a
meta-analysis. Stat Med 2002; 21: 1539–1558.
17 Cordoba O, Perez-Ceresuela F, Espinosa-Bravo M,
Cortadellas T, Esgueva A, Rodriguez-Revuelto R et al.
Detection of sentinel lymph node in breast cancer
recurrence may change adjuvant treatment decision in
patients with breast cancer recurrence and previous axillary
surgery. Breast 2014; 23: 460–465.
18 Cox CE, Furman BT, Kiluk JV, Jara J, Koeppel W, Meade
T et al. Use of reoperative sentinel lymph node biopsy in
breast cancer patients. J Am Coll Surg 2008; 207: 57–61.
19 Intra M, Trifiro G, Galimberti V, Gentilini O, Rotmensz N,
Veronesi P. Second axillary sentinel node biopsy for
ipsilateral breast tumour recurrence. Br J Surg 2007; 94:
1216–1219.
20 Port ER, Garcia-Etienne CA, Park J, Fey J, Borgen PI,
Cody HS III. Reoperative sentinel lymph node biopsy: a
new frontier in the management of ipsilateral breast tumor
recurrence. Ann Surg Oncol 2007; 14: 2209–2214.
21 Uth CC, Christensen MH, Oldenbourg MH, Kjaer C,
Garne JP, Teilum D et al. Sentinel lymph node dissection in
locally recurrent breast cancer. Ann Surg Oncol 2015; 22:
2526–2531.
22 van der Ploeg IM, Oldenburg HS, Rutgers EJ,
Baas-Vrancken Peeters MJ, Kroon BB, Valdes Olmos RA
et al. Lymphatic drainage patterns from the treated breast.
Ann Surg Oncol 2010; 17: 1069–1075.
23 Vugts G, Maaskant-Braat AJ, Voogd AC, van Riet YE,
Luiten EJ, Rutgers EJ et al. Repeat sentinel node biopsy
www.bjs.co.uk BJS 2016; 103: 1579–1588
1588
should be considered in patients with locally recurrent
breast cancer. Breast Cancer Res Treat 2015; 153: 549–556.
24 Sappey MPC. Anatomie, Physiologie, Pathologie des Vaisseaux
Lymphatiques Consideres Chez l’Homme et les Vertebres. A.
Delahaye and E. Lecrosnier: Paris, 1874.
25 Fowler JC, Solanki CK, Guenther I, Barber R, Miller F,
Bobrow L et al. A pilot study of dual-isotope
lymphoscintigraphy for breast sentinel node biopsy
comparing intradermal and intraparenchymal injection. Eur
J Surg Oncol 2009; 35: 1041–1047.
26 Ahmed M, Purushotham AD, Horgan K, Klaase JM, Douek
M. Meta-analysis of superficial versus deep injection of
radioactive tracer and blue dye for lymphatic mapping and
detection of sentinel lymph nodes in breast cancer. Br J Surg
2015; 102: 169–181.
If you wish to comment on this, or any other article published in the BJS, please
visit the online Your Views section of the website (www.bjs.co.uk).
© 2016 BJS Society Ltd
Published by John Wiley & Sons Ltd
M. Ahmed, R. Baker and I. T. Rubio
27 Poortmans PM, Collette S, Kirkove C, Van Limbergen E,
Budach V, Struikmans H et al.; EORTC Radiation
Oncology and Breast Cancer Groups. Internal mammary
and medial supraclavicular irradiation in breast cancer.
N Engl J Med 2015; 373: 317–327.
28 Aebi S, Gelber S, Anderson SJ, Lang I, Robidoux A, Martin
M et al.; CALOR Investigators. Chemotherapy for isolated
locoregional recurrence of breast cancer (CALOR):
a randomised trial. Lancet Oncol 2014; 15:
156–163.
29 Moossdorff M, Vugts G, Maaskant-Braat AJ, Strobbe LJ,
Voogd AC, Smidt ML et al. Contralateral lymph node
recurrence in breast cancer: regional event rather than
distant metastatic disease. A systematic review of the
literature. Eur J Surg Oncol 2015; 41: 1128–1136.
www.bjs.co.uk BJS 2016; 103: 1579–1588