This multicentre retrospective study is characterised by a long period of accrual
(median follow-up: 136 months,; range: 16–292 months), which offers the
advantage of providing interesting information about the trend of the
disease. A certain number of patients (12 patients) enrolled in the first
years have died during this long period of observation; other patients (47 patients)
didn’t return for the planned examination to the Centre of Radiotherapy where
they had been treated, and were lost to follow-up. In addition only approximately
20% of patients were treated in the first period (from 1985 to 1990) and the
number of cases increased progressively during the following years (Figure 1). Most of the cases belong to the two decades covered by the
screening activity. Nevertheless, a large number of cases were also found in the
younger age group, between 41 and 50. Increased debate and publicity about breast
cancer screening after the introduction of screening programmes could have
determined an improvement in women’s awareness, with a consequent increase of
spontaneous mammography among age groups not invited for the screening, resulting in
an increse of DCIS radiographycally detected .
In the entire case series, 66% of cases were identified by mammography, since the
clinical pattern was negative, while in the series of patients treated between 1985
and 1990 the diagnosis was exclusively mammographic in only 30% of cases. These
data show that, since the diffusion of screening, mammographic examinations have
Quadrantectomy (74%) was the most frequent conservative surgery. In many cases,
even among the oldest ones, x-ray examination of the surgical specimen or
postoperative mammography was carried out to check completeness of the excision, as
stated by the Consensus Conference .
A considerable number of cases (41%) were also subjected to axillary lymph node
dissection. In the series of patients treated from 1985 to 1990, axillary
lymphadenectomy was carried out in 69% of the cases. In the 1970s and 1980s most
patients with DCIS underwent axillary lymph node dissection. During the 1990s the
number of women undergoing axillary lymph node dissection for DCIS declined, in
particular after the publication of the Consensus Conference . Neither the indication to remove the axillary lymph nodes nor that to
remove the sentinel lymph node is currently to be considered, except in the event of
extensive, high-grade, palpable lesions, with a greater risk of occult invasion [1, 27, 28].
The total median dose of RT to the residual breast was 50 Gy with a 2 Gy/fr.
conventional fractionation. Considering the entire case series, a boost to the
tumour bed was also administered in 50% of the cases. In the analysis of the
series treated in the 1985–1990 period, it was found that over 70% of the
patients had received a boost. Our study showed no correlation between the
prescription of the boost and the patients’ age and no significant difference
was found in the distribution of local recurrences in relation to it (p= 0.45).
In the literature, the role of the radiation boost to the tumour bed has not been
fully defined as yet. Some Authors have used a total dose of 10–20 Gy after
standard whole breast RT [29–31]; or a total dose of 7.5 Gy in three fractions after standard or
hypofractionated whole breast RT .
In the four randomised studies [8, 11] that evaluated the role of postoperative RT after conservative surgery
for DCIS , the boost was not envisaged.
The role of a higher dose on the initial site of the disease was evaluated in a
retrospective analysis  on 373 patients ≤ 45 years old; it significantly reduced the risk
of local recurrence (p < 0.0001). Besides the radiation boost, predictive
factors for local recurrence included patients’ age (≤ 39 years) and the
margin status. However, this study has some limits concerning its retrospective
nature, the absence of a centralised review of histological samples and data on
margin status, tumour size and grade. Randomised studies are underway and, while
waiting for their results, it is advisable to administer the boost to young
patients, in whom negative margins should always be obtained [1, 7].
Considering that our study is retrospective, a comparative analysis of the results of
the main retrospective studies [12–16] was performed. In our series, the incidence of local recurrences is
10%, and 63% of them are invasive. In the other series the incidence of
local failures ranges between 8% and 20%, and around 50% (from 31%
to 76%) are invasive. If one examines OS (98.5% at 5 years and 95.5% at
10 years) and DSS rates (100% at 5 years and 99% at 10 years), our data do
not differ significantly from those reported in the literature [12–16].
The long period of accrual of this study has also caused some disadvantages.
The database sheets provided for the collection of the following pathological
parameters: nuclear grade, margin status, estrogen and progesterone receptors,
necrosis and multifocality, in addition to histological type and tumour size. The
pathological records, especially those regarding the initial years of the study, did
not always include all the above-mentioned parameters and, as it was not possible to
review the specimens, they are unknown in a certain group of patients. We were able
to carry out a detailed analysis of the risk for local recurrence only with regard
to the following parameters: age of patients, tumour size and nuclear grade. Only
age turned out to be a statistically significant prognostic factor (p= 0.0009).
Several prognostic factors correlated to the risk of local recurrence have been
identified in the literature, such as clinical presentation, tumour size,
histological type, nuclear grade and the presence of central necrosis, state and
width of the margins and age of patients at diagnosis [12, 34–39].
Young age (≤ 40 years) is considered to be one of the most important prognostic
factors, in both the clinically palpable DCIS lesions [9, 30, 37–39] and the occult ones, diagnosed by mammography [14, 40]. A number of clinical studies [9, 14, 41, 42] report a local recurrence rate very similar to ours in the ≤ 40 age
In our study, most of the patients with local recurrence (78%) underwent
mastectomy with axillary lymph-node dissection. The reasons for choosing
breast-conserving surgery in 10 cases (16.9%) are unknown.
The therapeutic approach to local failure plays a major role, especially in the
presence of an invasive recurrence. Mastectomy is the elective treatment for
patients initially treated with conservative surgery and postoperative RT. In the
case of small-sized and unifocal recurrences, local excision followed by RT is also
suggested, if patients have not previously received RT [1, 37]. The prognosis for in-situ recurrences is excellent, as less than 1%
of patients develop further relapse, while it is much less favourable for invasive
recurrences, which can develop metastases in 15-20% of the cases [31, 43].
The use of accelerated partial breast irradiation (APBI) has been recently introduced
not only in the treatment of invasive carcinoma at an early stage, but also of DCIS [44–46]. In 2005, a phase-III clinical trial comparing whole breast RT with APBI
was conducted by means of one of the three following treatment techniques:
multi-catheter interstitial brachytherapy, brachytherapy with MammoSite and 3D
external conformal radiotherapy [44–47]. APBI experiences in the treatment of local recurrences are also reported
in patients previously treated with conservative therapy (surgery followed by whole
breast RT) for invasive breast carcinoma and DCIS [48, 49]. In highly selected cases of local relapse, this therapeutic approach may
find an important role in the future.
The incidence of distant metastases, after an invasive local recurrence, was 0.9%
and the incidence of a second contralateral breast tumour was 6.8% in our
series, in line with the results reported in the literature [17, 50].