CLINICAL INVESTIGATION Breast
LOCOREGIONAL OUTCOMES OF INFLAMMATORY BREAST CANCER PATIENTS
TREATEDWITH STANDARD FRACTIONATION RADIATION AND DAILY SKIN BOLUS
IN THE TAXANE ERA
SHARI DAMAST, M.D.,* ALICE Y. HO, M.D.,* LESLIE MONTGOMERY, M.D.,yMONICA N. FORNIER, M.D.,z
NICOLE ISHILL, M.S.,xELENA ELKIN, PH.D.,xKATHRYN BEAL, M.D.,* AND BERYL MCCORMICK, M.D.*
Departments of *Radiation Oncology,ySurgery,zBreast Cancer Medicine, andxEpidemiology and Biostatistics, Memorial
Sloan-Kettering Cancer Center, New York, NY
Purpose: To assess locoregional outcomes of inflammatory breast cancer (IBC) patients who received standard
fractionation radiation with daily skin bolus and taxanes as part of combined-modality therapy (CMT).
Methods and Materials: We retrospectively reviewed the charts of 107 patients diagnosed with IBC between
January 1995 and March 2006 who presented to our department for adjuvant radiation therapy (RT).
Results: All patients received chemotherapy (95% anthracycline and 95% taxane), modified radical mastectomy,
and RT to the chest wall and regional lymphatics using standard fractionation to 50 Gy and daily skin bolus. The
RT to the chest wall was delivered via electrons(55%) or photons (45%) in daily fractions of 180 cGy (73%) or 200
cGy (27%). Scar boost was performed in 11%. A majority (84%) of patients completed the prescribed treatment.
Median follow-up was 47 months (range, 10–134 months). Locoregional control (LRC) at 3 years and 5 years was
90% and 87%, respectively. Distant metastases–free survival (DMFS) at 3 years and 5 years was 61% and 47%,
fractionation radiation with daily skin bolus and taxanes as part of combined-modality therapy. Distant metasta-
ses–free survival remains a significant therapeutic challenge.
? 2010 Elsevier Inc.
Radiation, Inflammatory breast cancer, Combined modality treatment, Standard fractionation, Taxanes.
Inflammatory breast cancer (IBC) is a unique entity com-
pared to other non-inflammatory locally advanced breast
cancers.Theterm‘‘inflammatory’’was firstcoined atMemo-
rial Hospital by Lee and Tannenbaum in 1924 (1). Clinically
IBC is diagnosed by erythema, edema, and induration devel-
oping in the breast over a rapid time course, and pathologi-
cally, widespread tumor emboli invading the dermal
lymphatic system may be observed (2). The condition of
IBC exhibits a tendency for early recurrence (3) and inferior
prognosis (4, 5). Historically, outcomes with radiation and
surgery alone were poor (6). Since the incorporation of neo-
adjuvant (NA) chemotherapy into treatment over the past
several decades, a significant proportion of patients have
demonstrated long-term survival (7).
Neoadjuvant chemotherapy in the management of IBC has
been refined over the past decade. Anthracyclines have been
shown to improve outcomes and are considered part of
standard induction treatment for IBC (8, 9). The National
Surgical Adjuvant Breast and Bowel Project (NSABP)
B-27 trial demonstrated that the addition of preoperative tax-
anes to anthracyclines improves response in operable breast
cancer (10), and the Cancer and Leukemia Group B
(CALGB) Trial 9741 showed that a dose-dense schedule of
chemotherapy administration improves survival in breast
have indirectly contributed to the incorporation of taxanes
and dose-dense chemotherapy schedules into combined
modality therapy (CMT) for IBC. Several studies have inves-
tigated the impact of these agents on survival outcomes in
specific IBC populations (12–15).
When integrated with NA chemotherapy, both mastec-
tomy and radiation therapy (RT) continue to have an impor-
tant role in the locoregional management of IBC (16, 17).
Given the rarity of the diagnosis, which represents fewer
than 5% of newly diagnosed breast cancers (18), there are
no randomized trials delineating optimal radiation regimens
for this disease. Various single institutions have published
Reprint requests to: Shari Damast, M.D., Department of Radia-
tion Oncology, Memorial Sloan-Kettering Cancer Center, New
York, NY 10065. Tel: (212) 639-6828; Fax: (212) 639-2417;
Presented in part as a poster at the 50th Annual Meeting of the
American Society of Therapeutic Radiology and Oncology, Boston,
MA, September 21–25, 2008.
Conflict of interest: none.
Received March 3, 2009, and in revised form June 12, 2009.
Accepted for publication June 12, 2009.
Int. J. Radiation Oncology Biol. Phys., Vol. 77, No. 4, pp. 1105–1112, 2010
Copyright ? 2010 Elsevier Inc.
Printed in the USA. All rights reserved
0360-3016/$–see front matter
their radiation techniques in patients receiving radiation as
part of CMT (19–22), and 5-year locoregional control rates
in these studies range from 78% to 92%. These studies, how-
ever, vary in sequencing of radiation (pre vs. post mastec-
tomy), total dose (50–66 Gy), and fractionation schedules
(once daily vs twice daily), making it challenging to draw
definitive conclusions regarding radiation techniques. In
addition, the use of taxanes, now common for IBC, was
incorporated in only small proportions of the study popula-
tions, ranging from 4% to 40% (20–22). There is a paucity
of data on the appropriate radiation regimen in an IBC pop-
ulation who largely receive taxane-containing regimens.
At our institution, over the past decade, IBC patients have
been treated in a consistent fashion with anthracycline and
taxane-containing chemotherapy, modified radical mastec-
tomy (MRM), and locoregional radiation. This purpose of
this study was to examine locoregional outcomes resulting
from our technique using standard fractionation radiation to
50 Gy with daily skin bolus, a considerably lower radiation
dose compared with other regimens recently reported. In
addition, the use of daily skin bolus with no modification
by skin reaction at our institution is considered aggressive.
We therefore also aimed to assess toxicities and tolerability
of this regimen among this specific population.
METHODS AND MATERIALS
The charts of patients diagnosed with IBC between January 1995
and March 2006 were retrospectively evaluated. Only patients irra-
diated in our department with curative intent were included.Patients
with metastatic disease or disease recurrence before RT were
excluded. Patients with a pathologic finding of tumor emboli invad-
ing the dermal lymphatics in their mastectomy specimen but who
lacked clinical findings of IBC were excluded as well. Of 359 charts
reviewed, 107 patients were identified who met these criteria,
including 1 patient with synchronous bilateral IBC.
The majority of patients received their mastectomy and chemo-
therapy treatments at Memorial Sloan-Kettering Cancer Center
sultation, generally by an MSKCC surgeon or medical oncologist.
Several patients were diagnosed and initially treated with chemo-
therapy and/or mastectomy at an outside institution before receiving
radiationinour department. ThediagnosisofIBC inthesecases was
made based on review of the patient’s history, pathology findings,
and outside physician documentation. All pathologic specimens
were reviewed by the MSKCC pathology department confirming
the diagnosis of invasive carcinoma. A skin biopsy (to assess the
presence of tumor emboli in the dermal lymphatics) was performed
in 65% of the study population.
Patient and tumor characteristics are listed in Table 1.
The sequencing of chemotherapy, MRM, and RT is outlined in
The majority of patients (95%) received anthracycline and tax-
ane-containing chemotherapy. The various regimens administered
are shown in Table 2. The AC-T regimen (doxorubicin (A)
Before 2004, T was usually administered in the adjuvant setting
after MRM, and after 2004, all of the chemotherapy, including T,
was usually administered neoadjuvantly. Furthermore, after 2004,
the majority of patients received neoadjuvant dose-dense (dd) che-
motherapy, that is, AC-T chemotherapy delivered every 2 weeks
rather than every 3 weeks. For the purpose of analysis, only patients
who received all planned cycles of dd AC-T in the absence of other
Eleven patients received less common chemotherapy regimens
categorized as ‘‘other’’ in Table 2. The regimens listed as ‘‘other’’
included: dd AC-T with gemcitabine; dd AC-T with gemcitabine
and capecitabine; dd AC followed by weekly paclitaxel; dd AC
followed by docetaxel; paclitaxel alone; docetaxel and cyclo-
phosphamide; docetaxel followed by capecitabine; paclitaxel
followed by cyclophosphamide, epirubicin, and 5-fluorouracil;
Table 1. Patient characteristics
Skin punch biopsy
Did not receive
Hormone receptor status
Her 2 status unknown
Nodes positive (pathologic)
Margin status unknown
At least partial response to chemotherapy
Less than partial response
pCR in the axillary lymph nodes
pCR in the breast
2002 AJCC Stage (clinical)
Abbreviation: pCR = pathologic complete response.
Characteristics are given as median (range) for continuous vari-
ables and n (%) for categorical variables.
* Includes only those patients who received NA chemotherapy.
1106I. J. Radiation Oncology d Biology d PhysicsVolume 77, Number 4, 2010
cyclophosphamide, methotrexate, and 5-fluorouracil; carboplatin
with paclitaxel. One patient with initially chemo-refractory disease
received doxorubicin and vinorelbine, then vinblastine, cyclo-
phosphamide, and doxorubicin, then gemcitabine, epirubicin, and
Clinical response to chemotherapy was documented by the med-
ical oncologist and surgeon before mastectomy. The majority of
patients achieved at least a partial response determined by clinical
examination. Only 5 patients had less than a partial response (2
patients, no clear clinical evidence of response; 3 patients, local
progression of clinical signs and symptoms during chemotherapy).
No patient who developed distant metastases while undergoing che-
motherapy was included in this study.
All patients were treated with MRM. Of the patients, 94 (88%)
had their surgeries performed at MSKCC. Six patients had immedi-
ate breast reconstruction, 3 of these at an outside institution. The
3 patients who received breast reconstruction at MSKCC had com-
plete clinical response to chemotherapy documented by the surgeon
at the time of the operation.
Pathologic complete response (pCR) in the breast indicated the
absence of any tumor cells in the removed breast. We defined
pCR in the axillary lymph nodes (pCR in the nodes) as no residual
microscopic or macroscopic disease in the lymph nodes, indepen-
dent of disease status of the breast. The percentages of patients
achieving pCR are listed in Table 1.
Any patient who developed distant metastases or local progres-
sion after surgery and before RT was excluded.
Radiation technique for chest wall and regional lymphatics
The median time from surgery to RT was 137 days (range, 110–
192 days) for the 57 patients who received adjuvant chemotherapy
(Fig. 1, Groups B and D), and 50 days (range, 39–66) for the 50
patients who did not receive adjuvant chemotherapy (Fig. 1, Groups
A and C).
Radiation therapy was delivered to an unreconstructed chest wall
(CW) in 100 patients, to an intact (premastectomy) breast in 1 pa-
tient, and to a reconstructed breast in 6 patients. The target volume
included the entire CW and the draining regional lymphatics. The
median dose to the CW was 5040 cGy (range, 1080–5040 cGy).
adequate coverageoftheskinbytheprescription isodoseline.Bolus
remained on the skin for the entirety of RT. Boosting the
Fig. 1. Treatment groups.
Table 2. Systemic treatment characteristics
Sequential tamoxifen + AI
Abbreviations: AC-T = doxorubicin/cyclophosphamide followed
by paclitaxel; A-CMF = doxorubicin followed by cyclophos-
phamide/methotrexate/5-fluorouracil; AI = aromatase inhibitor;
phamide; CAF-T = cyclophosphamide/doxorubicin/5-fluorouracil
followed by paclitaxel; EC-T = epirubicin/cyclophosphamide fol-
lowed by paclitaxel; TAC = docetaxel/doxorubicin/cyclophospha-
Locoregional outcomes of inflammatory breast cancer d S. DAMAST et al. 1107
mastectomy scar was not routinely performed. Radiation treatment
characteristics are further detailed in Table 3.
on patient anatomy, physician discretion, and available technology
of the time, including the availability of CT-simulation, which
became routine at MSKCC in April 1998. Our technique for post-
mastectomy CT-based electron beam radiotherapy has been
described in detail previously (23).
The electron-based CW technique always included the internal
mammary nodes (IMNs) in the target volume. The IMNs were not
always targeted in the photon-based CW technique, and their inclu-
sion depended on the clinical judgment of the treating physician.
When the IMNs were included in a photon CW plan, one of two
techniques was used. Either a separate en face electron beam was
added or, alternatively, the photon tangential fields were widened
to incorporate the IMNs in the first four intercostal spaces.
Time to locoregional recurrence (LRR), distant failure, first fail-
ure, and overall survival (OS) were estimated using Kaplan-Meier
methods. All times were estimated from the date of diagnosis,
except for the analysis of response to neoadjuvant chemotherapy,
which was estimated at the date of surgery, when pathologic
response was assessed. For locoregional and distant failure, patients
who died without a recurrence were censored at the time of death,
and patients who were alive and free of each specific recurrence
were censored at last follow-up. Cox proportional hazards regres-
sion was used to assess associations between time to LRR and var-
ious clinical and treatment characteristics. Fisher’s exact test was
used to assess relationships among these characteristics.
The median follow-up for survivors was 3.8 years (range,
1.2–11.0 years). Results are shown in Figs. 2 to 5. Four
patients had disease progression while undergoing radiation
(2 local, 2 distant). Patients who failed during the course of
RT were included in the survival analysis and were censored
at the time of first failure or death.
In total, 12 patients developed LRR, defined as patholog-
ically confirmed recurrences in the ipsilateral CW and/or
regional lymph nodes either during, or after the completion
of radiation. Seven patients had failures isolated to the skin
or CW, and 4 patients had regional nodal failures (1 supracla-
vicular, 1 axilla, 2 IMNs). One patient had experienced fail-
ure in both the CW and ipsilateral supraclavicular nodal
Unadjusted hazard ratios for the impact of clinical and
treatment characteristics on LRR are shown in Table 4. As
there were only 12 locoregionalfailuresrecorded, a multivar-
iate analysis was not performed. Number of positive axillary
were significant variables (p < 0.05). Kaplan-Meier analysis
of LRR-free survival according to NR is shown in Fig. 6. A
value of NR > 0.5 was also predictive of distant metastases–
Preoperative taxanes use was associated with a lower like-
lihood of NR > 0.5. Of the 61 patients who received taxane
Table 3. Radiation therapy (RT) treatment characteristics of
RT duration (days)
Regional lymphatic groups targeted
Internal mammary nodes
Axilla (posterior axillary boost)
Scar boost performed
Fraction size (cGy)
Electrons + photons
Characteristics are given as median (range) for continuous vari-
ables and n (%) for categorical variables.
Fig. 2. Overall survival: at 3 years, 82%; at 5 years, 61%.
Fig. 3. Local recurrence–free survival: at 3 years, 90%; at 5 years,
1108 I. J. Radiation Oncology d Biology d Physics Volume 77, Number 4, 2010
before mastectomy, 16 (26%) had an NR > 0.5 vs. 21 (46%)
of the 46 patients who did not receive preoperative taxanes (p
< 0.05). Despite this finding, preoperative taxane use was not
a significant predictor for LRR, as shown in Table 4.
No patient who received dose-dense chemotherapy or
a scar boost experienced LRR. Therefore, hazard ratios
were not calculated for these factors. The use of dose-dense
chemotherapy did not significantly affect DMFS or OS.
All patients developed acute dermatitis (88% experienced
a Grade 2 or higher skin reaction). Of the patients, 68%
developed moist desquamation (Grade 2–3) during RT.
Bolus on the skin was maintained throughout treatment as
tolerable. Treatment breaks were discouraged but necessary
in some circumstances because of skin toxicity.
Eleven patients received less than prescription dose to the
CW secondary to intolerable moist desquamation. In this
group, the median dose to the CW was 4,800 cGy (range,
3,800–4,980 cGy), with only 1 patient receiving less than
4,600 cGy. Thirteen patients received less than the prescrip-
tion dose to the regional lymph nodes because of acute toxic-
ity with a median dose of 4,680 cGy to the supraclavicular
region (range, 4,500–4,980 cGy). Altogether, 90 patients
(84%) completed treatment to full dose as prescribed. Of
note, no patient who stopped treatment early because of
skin toxicity developed LRR. In all cases, the skin was noted
to have healed in subsequent follow-up visits.
Late toxicities were documented at least 90 days after radi-
ation using the Common Terminology Criteria for Adverse
Fig. 4. Distant metastases–free survival: at 3 years, 61%; at 5 years,
Fig. 5. Disease-free survival: At 3 years, 54%; at 5 years, 40%.
Table 4. Unadjusted analysis of time to local failure among
CharacteristicHazard ratio (95% CI) p Value
>50% Nodes positive
Nodes positive (continuous)
Duration of RT
Time from surgery to RT
Response to neoadjuvant chemoy
pCR = pathologic complete response; NS = not significant; RT =
* Excludes the 6 patients who were Stage NX.
yTaken fromdate of surgery,after response to chemotherapy was
assessed, includes only those patients who received neoadjuvant
chemotherapy (n = 101).
zExcludes 5 patients who received surgery first and 5 patients
who did not receive taxane.
Locoregional outcomes of inflammatory breast cancer d S. DAMAST et al. 1109