Local Recurrence After Breast-Conserving

Surgery and Radiotherapy

Frequency, Time Course, and Prognosis

JOHN M. KURTZ, MD,' ROBERT AMALRIC, MD,t HENRl BRANDONE, MD,* YVES AYME, MD,*
JOCELYNE JACQUEMIER, MD,§ JEAN-CLAUDE PIETRA, MD,t DANIEL HANS, MD,*
JEAN-FRANCOIS POLLET, MD,* CLAUDE BRESSAC, MD,$ AND JEAN-MAURICE SPITALIER, MD*

Mammary recurrences were studied in 1593 patients with Stage I and I1 breast cancer treated by mac-
roscopically complete tumor excision followed by megavoltage radiotherapy, including a boost to the
tumor bed (mean dose, 78 Gy). The actuarial freedom from mammary recurrence was 93% at 5,86% at
10, 82% at 15, and 80% at 20 years. Seventy-nine percent of the recurrences were in the vicinity of the
tumor bed, but with increasing time interval, an increasing percentage of recurrences was located elsewhere
in the breast. A majority of recurrences after 10 years could be considered new tumors. Only ten of 181
patients with recurrence had prior or concomitant distant metastases, and 159 of 171 isolated mammary
recurrences (93%) were operable. Uncorrected overall survival after operable recurrence was 69% at 5
and 57% at 10 years. Prognosis after late recurrence (after 5 years) was favorable (84% 5-year survival).
Operable early recurrences retained a favorable prognosis if smaller than 2 cm and confined to the breast
(74% 5-year survival). Disease-free interval and histologic grade also appeared to be important prognostic
factors after early recurrence. Survival after recurrence did not depend upon the type of salvage operation.
Locoregional control was 88% at 5 years after salvage mastectomy and 64% after breast-conserving
salvage procedures. The role of adjuvant systemic therapy at time of local recurrence requires additional
study. This experience illustrates the important differences between mammary failure and chest wall
recurrence after mastectomy, in particular the protracted time course and more favorable prognosis as-
sociated with the former.
Cancer 63:1912-1917,1989.

B REAST-CONSERVING THERAPY differs from primary

radical surgical treatment principally in the pres-
ervation of the affected breast and the requirement for
breast irradiation. Although local control rates appear to
be similar for both forms of treatment,'-3 concern about
the continued risk of recurrence in the retained breast
remains the principal argument against the use of con-
servative surgery.
Because this less mutilating treatment approach is being
implemented with increasing frequency both in Europe
and in North America, it is of considerable interest that
the problem of recurrent cancer in the treated breast be
studied in-depth. Because mammary failure after adequate
treatment is an uncommon event, and many years of fol-
low-up must be accrued to evalute subsequent prognosis,
few treatment centers are in a position to carry out such
an analysis. The combination of breast-conserving surgery
and megavoltage radiotherapy has been practiced at the
Marseille Cancer Institute since the early 1 9 6 0 ~conse-
~;
quently, we have had the opportunity to treat many local
recurrences and to observe their subsequent c o u r ~ eThis
.~
report analyzes the development of breast recurrence as
a function of time and studies the prognosis after treat-
ment of local failure.

From the Departments of ?Radiotherapy, $Surgery, and $Pathology,
Cancer Institute, Marseille, France, and the *Radiation Oncology De-
partment, University Hospital, Basel, Switzerland.
The authors thank Jakob Roth, PhD, for his contribution of the sta-
tistical analysis.
Address for reprints: John M. Kurtz, MD, Radiation Oncology, Uni-
versity Hospital, CH-403 1 Basel, Switzerland.
Accepted for publication December 13, 1988.
Materials and Methods
The study population consisted of 1593 patients with
clinical Stages I and I1 breast cancers (American Joint
Committee') treated between November 1963 and De-
cember 1982 at the Cancer Institute and associated clinics
in Marseille, employing primary limited surgery followed

1912
No. 10 MAMMARY
RECURRENCE
by megavoltage radiotherapy. Patient selection criteria, a
description of treatment techniques, and the results of
therapy have appeared in previous publication^.^.^.^ Seven
hundred twelve patients were in clinical Stage I and 88 1
in Stage 11. Mean age was 51 years.
Briefly, the surgical treatment for all patients involved
a macroscopically complete tumor removal by simple tu-
morectomy or wide excision. In addition, 68 1 more recent
patients were subjected to an axillary dissection, which
was usually limited to the lower two levels. Radiotherapy
involved treatment to the entire breast and the draining
lymph node areas, generally applying 50 to 60 Gy in 5 to
6 weeks, employing a telecesium or telecobalt machine.’
Additional boost radiation was given to the tumor bed,
most commonly with electron beam therapy (mean total
prescribed dose 78 Gy). Interstitial implant therapy was
not employed.
Although 580 selected premenopausal patients were
surgically castrated, adjuvant medical therapy was not
commonly employed prior to 1980. Two hundred eigh-
teen patients received some form of hormone therapy,
and 145 patients adjuvant chemotherapy, most com-
monly as a single alkylating agent in the perioperative
period.
After completion of treatment, patients were followed
closely by their family physicians and at least once yearly
by the responsible surgeon or radiation oncologist. In ad-
dition to the clinical examination, the follow-up program
for detection of local recurrence included yearly mam-
mography and infrared therm~graphy,~ as well as ultra-
sound examination in more recent years.
Upon documentation of local recurrence, patients hav-
ing no evidence of distant metastases were treated with a
secondary operation (salvage surgery). This most com-
monly involved a total mastectomy, with or without ax-
illary dissection. Small, well demarcated recurrences were
treated with increasing frequency using wide excision,
without mastectomy.” Adjuvant therapy at time of re-
currence was individualized. Hormone therapy was lib-
erally employed, most commonly with tamoxifen. Com-
bination chemotherapy was reserved for recurrences that
were thought to be unfavorable, especially in premeno-
pausal patients.
Follow-up from time of primary therapy ranged from
5 to 24 years, with a median of 1 1 years. Actuarial cal-
culations were performed by the life table method, as rec-
ommended by the American Joint Committee. I I Actuarial
freedom from mammary recurrence was calculated using
the effective number of breasts at risk during the interval
in question, with documented failure in the parenchyma
or skin of the treated breast as endpoint. Patients devel-
oping distant metastases, dying, or lost to follow-up were
treated as censored observations. In the calculation of
- Kurtz et al. 1913
overall survival rates, all causes of death were included,
and patients lost to follow-up while clinically free of cancer
were censored at the time of last consultation. Patients
dying of unknown cause were considered as cancer deaths.
Differences between survival curves were tested for sig-
nificance using the log-rank test.12
Results
Time Course of Recurrence
As of December 1987, recurrent cancer in the treated
breast has been observed in 178 of the 1593patients (1 1%),
without previously documented distant metastases. An
additional three patients developed recurrence in the
breast after the appearance of metastatic disease. The ac-
tuarial freedom from mammary recurrence was 93% at 5
years, 86% at 10 years, 82% at 15 years, and 80% at 20
years (Fig. 1). Isolated axillary recurrences were recorded
in an additional 32 patients; these are not given additional
consideration here.
The yearly actuarial risk of breast recurrence, as cal-
culated from life tables, averaged l .5% during both of the
first two 5-year periods. A decreasing risk ensued, aver-
aging 1.1 % per year during the third 5 years. The risk
thereafter was small, with three additional recurrences
among 120 preserved breasts at the beginning of the 16th
year.
Location of Recurrences Within the Breast
To quantify the contribution of multicentricity, an ef-
fort was made to distinguish between regrowth of cancer
within the vicinity of the original primary tumor and new
tumor formation elsewhere within the breast. Recurrent
tumors were considered true recurrences if they occurred
within the tumor bed or in its vicinity, not more than a
few centimeters from the edge of the electron boost field.
In contrast, a recurrence was scored as clearly distant when
it could be judged to be at least 5 cm away from the orig-
inal primary. In case of doubt, recurrences were prefer-
entially placed into the true recurrence category, as were
seven cases for which the available descriptions were in-
adequate. The resulting analysis will thus tend somewhat
to underestimate the importance of new tumor formation.
Table 1 characterizes the locations of recurrences within
the breast. Only 38 of the recurrences (21%) arose clearly
at a distance from the original tumor bed. With increasing
interval from primary treatment, however, an increasing
percentage of the recurrences occurred elsewhere within
the breast. Whereas only 10% of recurrences observed
prior to 24 months were new tumors, nine of 14 recur-
rences (64%) occurring after 10 years were located else-
where in the breast, including all three recurrences diag-
nosed after 15 years.
1914 CANCERMay IS 1989 Vol. 63

w
E
60 -
t-
ri,
<
W
-
E
a3
z
40 -
0
a
L
-
W
w 20-
a
LL

I 1 I I I I I I I I I I I I I I I I I 1
0 2. 4. b 8 1.0 12 I4 16 18 20
Y E A R S S I N C E THERAPY
1593 1295 442 120 17
FIG.1 . Actuarial freedom from recurrence in the breast for 1593 Stage 1-11 patients treated between 1963 and 1982. The numbers below the time
axis indicate the number of preserved breasts at risk in disease-free survivors.

Diagnosis of Mammary Failure
The way in which the recurrence was diagnosed was
adequately described for 161 patients. One hundred thirty-
nine recurrences (86%) were diagnosed primarily by the
clinical examination. In 20 instances (12%)mammogra-
phy had been indispensable in leading to the diagnosis.
For 113 recurrences for which mammography reports
were available, 75 (66%)were positive, 14 (12%)negative,
and 24 (21%)equivocal. The mammographic films them-
selves were not reviewed for this study. One recurrence
was diagnosed primarily because of reappearance of a hot
spot on infrared thermography and one recurrence was
diagnosed by ultrasound examination.
Prognosis of Mammary Failure
In contrast to chest wall failure afler mastectomy, failure
in the breast almost always presented as an isolated event.
Only three patients with recurrence had previously di-
TABLE1 . Location of Recurrences Within the Breast,
With ResDect to Orieinal Tumor
Interval
(Yd In vicinity of primary*
0-2 36 (90%)
3-7 80 (83%)
8-10 19 (68%)
11-15 5 (45%)
16+ 0 3 ( I 00%)
* Includes inoperable and skin recurrences.
agnosed distant metastases, and in seven additional cases
metastases were observed concomitantly with local re-
currence. Twelve patients had recurrences that were so
advanced as to preclude surgical treatment, usually with
extensive skin involvement (inflammatory recurrence). All
but one of these latter patients have died of breast cancer,
with a median survival of 9 months.
The remaining patients had recurrences that were an-
atomically operable (1 59 of 178, 89%). All but three of
the operable patients were treated by salvage surgery with
or without adjuvant chemotherapy and/or hormonal g a -
nipulation. One patient who refused operation was treated
with tamoxifen, and a second received an interstitial irid-
ium- 192 implant; both remain alive and well at 1 1 years
and 8 years after recurrence. The third patient had a lo-
calized recurrence in the skin of the breast, which was
treated with electron beam therapy. An extensive paren-
chymal recurrence necessitated mastectomy 2 years later,
and the patient died of metastatic disease 8 years afier
first recurrence. Of 88 patients whose salvage operations
included an axillary dissection, 34 (39%) were found to
have histologically positive lymph nodes. For all 159 pa-
tients with operable recurrences, overall survival was 69%
Clearly distant at 5 years and 57% at 10 years, as calculated from time
of salvage treatment (Fig. 2).
4 (10%)
16 (17%)
9 (32%) Prognostic Factors for Operable Mammary Recurrence
6 (55%)
Prognostic factors were investigated for the 159 patients
with operable mammary recurrences, using overall sur-
No. 10 MAMMARY
RECURRENCE - Kurtz et al. 1915

I00

80

c3
-z>
,- 60
>
of
3
m
+
W
40
0
of
W
a
20

0 I I I I I I 8 1
0 2 4 6. 8 I0 12 14
YEARS AFTER SALVAGE TREATMENT
I59 134 115 97 78 5.5 39 33 27 19 IG 14 9 7 6 4

FIG.2. Actuarial overall survival following treatment of I59 operable mammary recurrences. The numbers below the time axis indicate the number
of patients surviving.

viva1 from time of recurrence as the end point. The most defined as single recurrences 5 2 cm in size confined to
consistently important prognostic factor was disease-free the breast, were associated with a 74%overall survival 5
interval from time of primary therapy (Table 2), indicating years after recurrence. The prognosis for recurrences that
increasingly favorable 5-year survival with increasing in- were larger, multiple, or associated with axillary recurrence
terval. We have chosen to exclude the 70 patients recumng was significantly less favorable.
after 60 months from additional analysis because their Neither adjuvant hormone therapy nor chemotherapy
prognosis has been almost uniformly excellent and to re- administered at time of recurrence was associated with a
strict additional analysis to patients with early recurrence significantly superior survival compared with patients re-
diagnosed prior to 5 years. ceiving no adjuvant therapy. However, patients receiving
Table 3 presents the results of the analysis of potential adjuvant hormone treatments, including oophorectomy,
prognostic factors for the 89 patients with operable early fared better than those treated with chemotherapy. The
recurrence. The following parameters had no statistically latter patients were most likely selected for this treatment
significant influence on overall survival following recur- because of a presumed unfavorable prognosis.
rence: age at time of initial therapy (younger than 50 years
versus 50 years or older); initial clinical T-stage (T1 versus Local Control After Salvage Treatment
T2); initial pathologic nodal status (N+ versus N- versus
With a median follow-up of 53 months after mammary
unknown); estrogen receptor status (ER+ versus ER- ver-
failure, 3 1 of the 159 patients have developed additional
sus unknown); location of recurrent tumor (near tumor
bed versus elsewhere); and type of salvage therapy (mas-
tectomy versus breast-conserving). TABLE2. Influence of Disease-Free Interval on Overall Survival
After Treatment of Ooerable Recurrence
Statistically significant correlation with survival after
early recurrence could be demonstrated for disease-free Interval Five-year overall survival
interval, histologic grade (using a modified Bloom-Rich- (mo) N ("/.)
ardson ~ystem'~), and the anatomic extent of the recur- 524 31 48
rence. A relatively poor prognosis was associated with re- 25-36 24 61
currences diagnosed prior to 24 months and for recur- 37-60 34 75
2 61 70 84
rences of Grade 3 tumors. Recurrences of limited extent,
1916 CANCERMay 15 1989 Vol. 63

TABLE3. Study of Potential Prognostic Factors for Overall Survival the current experience serves to highlight some
After Treatment of Operable Early Mammary Recurrence
important differences between the local recurrences as-
5-year sociated with the two treatment approaches.
survival A unique feature of breast-conserving therapy is the
Factor N (9%) P value
continued risk of tumor formation in the treated breast.
Interval (yr) Although chest wall recurrence after 5 years is not rare in
<2 31 48 early stage disease,I4late recurrences are decidedly more
2-5 58 68 0.027
common with breast preservation. The yearly recurrence
Age* (yr) risk continues to approximate 1% per year through the
<50 55 59
250 34 67 0.8
first half of the second decade; after 15 years the risk ap-
pears to be small. Most recurrences occurring after 10
T-Stage*
T1 41 63
years are at some distance from the primary and could
T2 48 59 0.2 be viewed as new tumors.

-
Nodal status* Overall, new tumors accounted for only 38 of the 178
N- 22 ipsilateral recurrences in this series. During the same time
N+ 19 57 \ 0.8
56
period 98 of the 1593 patients developed cancer in the
Undetermined 48 66
contralateral breast. This suggests that whole breast irra-
Estrogen receptor diation was effective in the reduction of new tumor for-
ER+ 22
ER- 20 57 \
70 c--- 0.1
mation. As a result, the great majority of recurrences dur-
Undetermined 47 61 ing the first decade occurred in the vicinity of the tumor
Location of recurrence
bed. Similar observations have been made by other au-
Same 75 63 thors. I5-I8 Any improvement in recurrence rates, there-
Elsewhere 14 55 0.6 fore, is likely to come only through refinements in treat-
Extent of recurrence? ment of the primary tumor area, or possibly through the
Limited 37 74 additive effect of sytemic therapy in high-risk patients. *
Extensive 36 42 0.018 The intimate association between chest wall recurrence
Histologic grade
G 1-2
G3
Unknown
24
18
47
72
40
61
-
\ 0.003
and distant metastases has been well documented. Distant
metastases are clinically manifest in 25% to 50% of pa-
tients by the time local recurrence is diagnosed, and 5-
year survival after treatment of isolated chest wall recur-
Type of salvage operation
Mastectomy 43 53 rence in unselected surgical series ranges from 18% to
Conservative 46 70 0.321 32%.19-**In contrast, mammary recurrence is almost al-
Adjuvant therapy ways an isolated event, with only ten of 18l patients in
None
Hormones
Chemotherapy
44
28
17
59
74 \
47 -
0.03

* Age, stage, and nodal status at the time of primary treatment.

t The extent of recurrence is unknown for 16 patients.
recurrence in the preserved breast, on the chest wall, or
in the regional nodal areas. The actuarial freedom from
second locoregional recurrence for patients treated by sal-
vage mastectomy was 88% at 5 years. For patients treated
with breast-conserving salvage surgery, the corresponding
figure was 64%. Many of the recurrences in the latter pa-
tients could be treated by additional surgery.
Discussion
The most commonly employed methods for treatment
of the breast in early mammary carcinoma are total mas-
tectomy alone and tumor excision in conjunction with
breast irradiation. Although it is well established that both
methods are associated with similar 5-year local failure
this series having the prior or concomitant diagnosis of
metastatic disease. In addition, 85% to 95% of isolated
breast recurrences are operable, and patients can be ren-
dered disease-free by potentially curative salvage surgery.
In contrast to chest wall recurrence after mastectomy,
prognosis following mammary failure is comparatively
favorable, especially for late recurrences. Essentially all
inoperable local failures occur early, and for operable re-
currence diagnosed prior to 5 years the majority of patients
continue to have a relatively good outlook. A poor prog-
nosis is associated with extensive local or locoregional
recurrence, or recurrences showing other signs of aggres-
sive biology (disease-freeinterval less than 2 years, or his-
tologic Grade 3). Patients with small recurrences confined
to the breast continue to have a good prognosis, especially
if the recurrence is diagnosed after 24 months (Table 3).
Locoregional control after treatment of recurrence has
been satisfactory, with a 12% probability of subsequent
recurrence in the chest wall or nodal areas 5 years after
salvage mastectomy. Because many patients continue to
desire breast preservation even in the face of local recur-
No. 10 MAMMARY
RECURRENCE
rence, we have turned increasingly to wide excision as
surgical therapy of small, well-localized recurrences. l o
Most patients suffering subsequent local recurrence can
then be treated by additional surgery. The actuarial risk
of subsequent recurrence in the breast after conservative
salvage surgery (36% at 5 years) is similar to local recur-
rence rates reported after segmental mastectomy alone as
primary treatment.'
In summary, mammary failure differs from chest wall
recurrence both in its more protracted time course and
in its more favorable prognosis. The continued risk of
breast recurrence with time would be a source of great
concern were it not for the excellent outlook of such late
failures. Painstaking life-long follow-up by experienced
physicians is clearly an important element in the breast-
conserving treatment strategy. Salvage mastectomy rep-
resents effective local therapy for breast failure, and wide
excision is an alternative surgical treatment for selected,
limited recurrences in patients continuing to desire breast
preservation. The indications for adjuvant hormonal or
chemotherapy in this setting remain to be defined. The
use of such therapies for unfavorable recurrences appears
justified. It is unclear, however, whether patients with late
recurrence or patients with small recurrences confined to
the breast would benefit from adjuvant medical treatment.
Such questions require study within the framework of a
prospective clinical trial.
REFERENCES
I . Fisher B, Bauer M, Margolese R et al. Five year results of a ran-
domized clinical trial comparing total mastectomy and segmental mas-
tectomy with or without radiation in the treatment of breast cancer. N
Engl JMed 1985; 312:665-673.
2. Veronesi U, Banfi A, Del Vecchio M et al. Comparison of Halsted
mastectomy with quadrantectomy, axillary dissection, and radiotherapy
in early breast cancer: Long-term results. Eur J Cancer Clin Oncol1986;
22: 1085-1089.
3. Sarrazin D, Lt M, RouessC J et al. Conservative treatment versus
mastectomy in breast cancer tumors with macroscopic diameter of 20
millimeters or less. Cancer 1984; 53:1209-1213.
4. Spitalier JM, Gambarelli J, Brandone H et al. Breast-conserving
surgery with radiation therapy for operable mammary carcinoma: A 25-
year experience. World JSurg 1986; 10:1014-1020.
5. Kurtz JM, Amalric R, Brandone H, Ayme Y, Spitalier JM. Results
- Kurtz et al. 1917
of salvage surgery for mammary recurrence following breast-conserving
therapy. Ann Surg 1988; 207:347-351.
6. American Joint Committee on Cancer Manual for Staging of Can-
cer. Philadelphia: J. B. Lippincott, 1983; 127-133.
7. Amalric R, Santamaria F, Robert F et al. Conservation therapy of
operable breast cancer: Results at five, ten, and fifteen years in 2216
consecutive cases. In: Harris JR, Hellman S, Silen W, eds. Conservative
Management of Breast Cancer. Philadelphia: J. B. Lippincott, 1983;15-
21.
8. Amalric R, Santamaria F, Robert F et al. Radiation therapy with
or without primary limited surgery for operable breast cancer: A 20-year
experience at the Marseille Cancer Institute. Cancer 1982; 49:30-34.
9. Amalric R, Spitalier JM. Radiation as the sole method of treatment
in carcinoma of the breast. In: Zuppinger A, Hellriegel W, eds. Handbuch
der medizinischen Radiologie, vol. 19, Teil 2. Berlin: Springer-Verlag,
1982;301-346.
10. Kurtz JM, Spitalier JM, Amalric R, Brandone H, Ayme Y. Results
of wide excision for local recurrence after breast-conserving therapy.
Cancer 1988; 61:1969-1972.
1 I . American Joint Committee on Cancer. Manual for Staging of
Cancer. Philadelphia: J. B. Lippincott, 1983;11-21.
12. Pet0 R, Pike MC, Armitage P et al. Design and analysis of ran-
domized clinical trials which require prolonged observations of each
patient: 11. Analysis and examples. Br J Cancer 1977; 35:l-39.
13. Contesso G , Mouriesse H, Friedman S, Genin J, Sarrazin D,
Rouesst J. The importance of histologic grade in long-term prognosis
of breast cancer: A study of 1,010 patients, uniformly treated at the
Institute Gustave-Roussy. J Clin Oncol 1987; 9: 1378-1386.
14. Montague ED. Conservation surgery and radiation therapy in the
treatment of operable breast cancer. Cancer 1984; 53:700-704.
15. Clark RM, Wilkinson RH, Mahoney U,Reid JG, MacDonald
WD. Breast cancer: A 2 1 year experience with conservative surgery and
radiation. Int J Radiat Oncol Biol Phys 1982; 8:967-975.
16. Fisher ER, Sass R, Fisher B et al. Pathologic findings from the
National Surgical Adjuvant Breast Project (Protocol 6): 11. Relation of
local breast recurrence to multicentricity. Cancer 1986; 57: 1717-1724.
17. Leung S, Otmezguine Y, Calitchi E, Mazeron JJ, Le Bourgois JP,
Pierquin B. Locoregional recurrences following radical external beam
irradiation and interstitial implantation for operable breast cancer: A
twenty three year experience. Radiother Oncol 1986; 5: 1-10.
18. Recht A, Silen W, Schnitt SJ et al. Timecourse of local recurrence
following conservative surgery and radiotherapy for early stage breast
cancer. Int J Radiat Oncol Biol Phys 1988; 15:255-26 I .
19. Donegan WL, Perez-Mesa CM, Watson FR. A biostatistical study
of locally recurrent breast carcinoma. Surg Gynecol Obstet 1966; 122:
529-540.
20. Fentiman IS, Matthews PN, Davison OW, Millis RR, Hayward
JL. Survival following local skin recurrence after mastectomy. Br JSurg
1985: 72:14-16.
21. Gilliland MD, Barton RM, Copeland EM. The implications of
local recurrence of breast cancer as the first site of therapeutic failure.
Ann Surg 1983; 197:284-287.
22. Heitanen P, Meittinen M, Makinen J. Survival after first recurrence
in breast cancer. Eur J Cancer Clin Oncol 1986; 22:913-919.