PATONE’S RAD-ONC REVIEW NOTES 2008

Breast Cancer
General
Anatomy
• Mammary gland: Lies over pec major and extends from 2nd to 6th rib vertically and from the sternum
to ant or mid-axillary line laterally.
• Cooper’s Ligament: Helps suspend the breast. Can be seen on mammogram.
• LN: Level I lateral to pectoralis, II deep to pectoralis, III medial to pectoralis

Epidemiology
• In 2004, 217,000 new cases, 41,000 deaths in U.S.
o 1,500 cases in males with 470 deaths
• Lifetime risk of developing breast ca (SEER data): 13% White female, ∼ 10% AA female.
• 1-2% rate of bilat ca at presentation and 5-8% metachronous bilat breast ca rate
• Risk factors: Increasing age (>50) and female sex are most important
o Personal/family history of BrCa, nulliparous, FCB after 30 yo, benign breast ds
o Exposure to ionizing radiation (e.g., HD pts getting mantle field have increased risk)
ƒ 56% for women ≤ 19 yrs of age at RT
ƒ 7% for women b/ 20-29 yrs
ƒ 1% for women ≥ 30 yrs
o EtOH, possibly obesity, HRT (Lancet 362, 2003)
o Protective: Less lifetime estrogen exposure and breast feeding (Lancet 360, 2002)
• Determining risk: Gail Model (Age at menarche, first live birth, # of prior bxs, h/o atypical
hyperplasia, and # of 1st degree relatives w/ breast ca) and Claus Model (Incorporates 1st and 2nd
degree relatives).

Genetics
• 5-10% of breast ca may have a major inherited component and 2/3 (66%) BRCA related. 50%
sporadic
• Familial syndromes: Li Fraumeni (50% p53 mutation, 25% b/l breast cancer, others ca like sarcoma,
colon, brain, leukemia), Cowden ds (Ch 10q, rare, other things like high arched palate, brain ca,
and thyroid ca) and Ataxia Telangiectasia (Ch 11, 4% of breast ca <40, other things like
oculocutaneous telangectasias, cerebellar ataxia, immune deficiency, leukemia and lymphoma).
• Her2/Neu (Human Epidermal growth factor Receptor 2, aka Neu or c-erbB-2 – localized to Chr 17,
TK receptor of EGFR or HER family) amplified in 20-40% of all breast cancers. Amplification
conveys worse prognosis (resistance to systemic chemo?). Herceptin (trastuzumab) being used.
• Bcl-2 (also a proto-oncogene) prevents apoptosis; p53 (tumor suppressor) mutation most frequent
in breast cancer
• Genetic testing for familial breast ca (≥ 2 breast ca < 50 yo or ovarian ca at any age), jewish/polish
ancestry w/ breast ca, relatives of known mutation carriers, male breast ca, multiple primaries
(breast and ovarian), and fallopian tube ca.

BRCA1/2
• BRCA1/2 – both tumor suppressor genes. 1 in 800 in general population carry BRCA1
mutation, but is 1 in 50 in Ashkenazi Jews (higher incidence of bilateral cancer)
• After initial dx of breast ca in BRCA carrier:
o Risk of contralat breast ca = ∼ 3%/yr (Reduced to 1.5%/yr w/ Tam or ovary rsx)
• BRCA1 (Chr 17q21): Familial breast (∼85% lifetime risk) & ovarian ca (>50% risk), Ashkenazi
Jews, BrCa in ~½ of BRCA1+ women before the age of 50

DR’S BREAST CANCER REVIEW NOTES  PAGE 1 
 PATONE’S RAD-ONC REVIEW NOTES 2008

• BRCA2 (Chr 13q12-13): Increased risk female and male breast ca (not ovarian,<25% risk)

Probability of a BRCA mutation in a pt dx w/ breast cancer

Pts age of breast Family History
cancer dx No breast ca < 50 1 relative w/ breast ca > 1 relative w/ Breast ca < 50 and
or any ovarian ca < 50 but no ovarian breast ca < 50 but no ovarian ca at any
ovarian age
Age < 50 yo 8% 18.6% 33.7% 46.4%
Age > 50 yo 2.5% 9.1% 12.2% 21.8%

• BCT—normal response of normal tissues to RT, similar rate of local recurrence. Higher rate of 2nd
breast cancers and ovarian cancer. 10-yr CBC rate was 40%.
• Prevention options: bilateral mastectomy/oophorectomy/Tamoxifen

Screening
• Screening is a form of 2º prevention (early detection). 1º prevention (prevents the ds from occurring)
includes prophylactic mast, oophorectomy, Tam and Selenium (? ↓ ch breakage)
• Screening mammography in all females 50-70 has been shown (Nystrom Lancet 2002) to ↓ breast ca
mortality. MRI probably better in younger pts (< 50 bec their dense breast tissue is not an issue
for MRI) and is therefore often used in high-risk (i.e. BRCA mutation) pts.
• MRI screening for high-risk pts (Kriege NEJM 2004) is more sensitive but less specific than mammos for
detecting breast cancer but less so for DCIS. Sensitivity of PE vs. mammo vs. MRI was 17.9, 33.3
and 79.5% respectively. Specificity was 98.1, 95 and 89.8% respectively. Also compared to mammo
MRI found more tumors < 1 cm (43% vs. 14%) and before they resulted in + nodes/micromet (21%
vs. 52%). MRI though, led to 2x more unneeded tests and 3x more unneeded biopsies.
• Screening Guidelines (MRI added to mammo in women w/ dense breast, usually young)
o Average risk
ƒ Baseline mammo at age 35 yrs
ƒ 40-50 yrs: mammo q 2 yrs
ƒ > 50 yrs: mammo q 1 yr
o High risk
ƒ Baseline mammo at age 30 yrs
ƒ >40: mammo q 1 yr
ƒ If inherited susceptibility gene: monthly self exam, clinical exam 1-2/yr and q 1 yr
mammo beginning b/ the age of 25-30 yrs.

Surgical technique
• Breast Conserving Therapy (BCT)

Absolute Contraindication Relative Contraindication Non-Factors

Persistent + margins after reasonable # of attempts Large tumor in small breast (poor cosmesis) + axillary nodes
≥ 2 gross tumors in separate quadrants (multicentric) Multiple gross tumors in same quadrant Location of tumor
(multifocal)
Diffuse, malignant appearing micro-calcs
Pregnancy (except 3rd trimester → RT post delivery Concern over 2nd malignancies (not ↑
[much] by RT)
Prior RT → unacceptable dose to breast Large breast, making set-up, reproducibility, Concern over evaluating for recurrence
and dose homogeneity problematic
Active SLE or scleroderma (relative per NCCN) Collagen Vascular disease

• Modified radical mastectomy (entire breast, underlying pectoralis major fascia, and level I/II ALND has
replaced the Halstead Radical mastectomy (entire breast, skin, pectoralis major & minor, and level

DR’S BREAST CANCER REVIEW NOTES  PAGE 2 
 PATONE’S RAD-ONC REVIEW NOTES 2008

I-III ALND). Skin Sparing mastectomy is generally done for immediately reconstructed pts because it
spares a majority of the pts natural skin.

• Total Mastectomy w/ Level I, II ALND +/- reconstruction

o Essentially, a modified radical mastectomy
ƒ Total (or Simple) M: Removal of entire breast, but not muscles or Ax LN
ƒ Modified Radical M: Removal of entire breast + Ax LND

Modified radical mastectomy

A. Shaded area = tissue removed

B. Level I Axillary LN
C. Level II Axillary LN
- Beneath pectoralis minor m.
D. Level III LN (not excised)

• LNs at risk include the axillary LNs (interpectoral nodes (“Rotter’s nodes”, Level II, posterior to pec
minor) and nodes along the axillary vein & its tributaries; Level I lateral and Level III medial to pec
minor) → SC fossa nodes (Level IV; lat/sup border defined by omohyoid muscle, medial border by IJ
vein, inf border by the clavicle/SC vein) → IMNs (Level V)

Staging
• Updated in 2002
• IHC or molecular staining only positive LN’s (cN0 and pN0 (i+, mol+)) do not currently change tx

T Stage T Stage N Stage (Clinical)

Tis: CIS or Paget’ds of nipple w/o tumor cN1 Movable ipsilateral axillary LN
cN2a Matted ipsilateral axillary LN or
T1: Tumor < 2 cm in greatest dimension b Clinically apparent ipsilateral IMN without axillary LAN
cN3a Mets in ipsilateral infraclavicular LN
Mic microinvasion ≤ 1mm
b Mets in ipsilateral IMN and axillary LN
a > 0.1 but < 0.5 cm c Mets in ipsilateral SC LN
b > 0.5 but < 1.0 cm
c >1.0 but < 2.0 cm N Stage (Pathologic)
pN0 Not seen on H&E (These don’t change clinical management)
T2: Tumor > 2 cm but < 5 cm in greatest dimension (i-) Neg IHC stain
(i+) Pos IHC stain only, no cell cluster > 0.2 mm diameter
T3: Tumor > 5 cm in greatest dimension (mol-) Neg molecular stain (RT-PCR)
(mol+)Pos molecular stain, no cell cluster > 0.2 mm diameter
T4: Invasive and/or involves skin
a Invades chest wall (Not pect muscle) pN1
b Edema (peau d’orange) or ulceration mi Micromets (> 0.2 mm but ≤ 2.0 mm)
a 1-3 + axillary LN (at least one > 2.0 mm)
of skin or satellite nodules in breast b Micromets in IMN on SLND – not clinically apparent
c Both a and b c 1-3 + axillary LN and micromets in IMN on SLND
d Inflammatory breast cancer
pN2a 4-9 + axillary LN (at least one > 2.0 mm)
b Mets in clinically apparent IMN

N3a > 10 ipsilateral axillary LN or + infraclavicular LN

b + Ax LN and clinically apparent IMN or >3 + ax LN and IMN
DR’S BREAST CANCER REVIEW NOTES  micromet PAGE 3 
c Positive supraclavicular LN
 PATONE’S RAD-ONC REVIEW NOTES 2008

Stage (AJCC 2002)

I T1N0 Stage I: ≤ 2cm

IIa T0-1N1, T2N0 Stage IIa: ≤ 5cm or ≤ 2cm w/ 1-3+ ALN and/or +IMN on SLN
IIb T2N1, T3N0 IIb: ≤ 5cm w/ 1-3+ ALN and/or +IMN on SLN or > 5 cm

IIIa T0-3N2, T3N1 Stage IIIa: Any size mass, w/ 4-9+ ALN or 1-3+ ALN and/or +IMN on SLN or c+IMN
IIIb T4N0-2 IIIb: CW, edema, ulceration, satellite nodules and nodes as for IIIa
IIIc T0-4N3 IIIc: Any size mass or T4 findings and/or ≥ 10+ ALN or + infraclavicular LN or
+ Ax LN and clinically + IMN or >3 + ax LN and IMN micromet or + S’Clav LN

IV M1 Stage IV: Distant mets, cervical LN, contra IMN or SC

DOUG’S BREAST CANCER REVIEW NOTES  PAGE 4 
 PATONE’S RAD-ONC REVIEW NOTES 2008

Ductal Carcinoma In Situ (Intraductal carcinoma)

• Incidence increased > 500% b/ 1973-1992 (mammograms), now 50,000/yr in US
• DCIS = 20% of breast cancers, detected by screening mammo and 90% nonpalpable
• Mammo: Shape, size and shape of microcalcifications. Cluster microcalcs = ≥ 5 calcs each under
1mm in dia isolated in a small volume of the breast and projected w/in a 1cm2 area on mammo
• Histology: Comedo and Non-comedo (Cribriform, micropapillary, papillary, solid)
o Low-grade (No atypia or necrosis, cribriform or micropapillary)
o Intermediate-grade (Either atypia or necrosis, solid or cribriform or micropapillary)
o High-grade (atypia, necrosis, comedo)
• Mastectomy: curative in 95-100%
• Excision + RT: LR rate is 7-20%, ~1/2 of recurrences invasive, but no obvious OS diff (Ernster 2000)
compared to mastectomy
o Most at or near the original site, w/ path features similar to index lesion
• Recurrence predictors: Margin status (< 1mm) highly predictive and comedonecrosis
o USC/VNPI = Path data (grade, +/- necrosis) + Margin + Tumor size + Age (Updated paper)
o 1-3 points/group: Low risk 4-6, intermediate 7-9, high 10-12.
o Original Silverstein data (Cancer 1996) His recs based on total score of VNPI:

-3, 4: Lumpectomy alone

1 2 3
Tumor size ≤ 1.5 cm 1.6 – 4 cm > 4 cm -5,6,7: Lump + RT (RT reduced LR by 17%)
Histology Low grade LG w/ necrosis High grade w/ necrosis
Margin width ≥ 1cm 1-9 mm < 1mm -8,9: Mast (High [20%] LR despite RT)

• Contralateral recurrence (invasive or DCIS), similar to pts with primary invasive ds – 0.5-1%/yr
• ALND/SLNDt: Risk for all comers < 1%, therefore routine LN sampling not recommended.
However, many high-risk DCIS pts (i.e. large >4cm lesion, palpable, diffuse high-grade) have a
component of microinvasive ds and thus may consider SLND. One study (Klauber-DeMore, Ann Surg Oncol
200) looking at pts w/ high-risk DCIS found a 12% + sentinel LN rate
• No RCT of mastectomy vs. BCT, but several RCT for invasive ds
• M alone (no BCT) in pts w/ diffuse microcalcs, multicentic ds, poor comesis, and collagen vascular ds
• 3 RCT trials comparing excision +/- RT (see Table and fig from Burstein NEJM Review article 2004)
o RT reduces IBTR about 50% (i.e. 5yr IBTR 16% (excision alone) vs. 8% (BCT))
o Absolute risk continues to increase over time however (see fig)
o Tamoxifen reduces IBTR and Contralateral tumor about another 50%
o All went to 50 Gy (Boost above 50 Gy not standard, almost all did not get it)

Excision +/- RT
NSABP-B17 (Fisher NEJM 1993 → Cancer 1995 → JCO 1998 → Cancer 1999 (8-yr f/u) → Semin Oncol
2001 (12-yr f/u))
• Excision with negative margin (“no ca at ink”) required (80% mammo detected)Æ +/- 50 Gy RT
• Every pt subset benefited from RT. No OS benefit.
8-yr f/u (∼ 800 pts) L alone L + RT Relative reduction Absolute benefit
Noninvasive IBRT 13.4 % 8.2 % 40 % 5%
Invasive IBRT 13.4 % 3.9 % 70 % 10%
LR 27 % 12 % 55 % 15%
• LF factors: Uncertain or involved margins. Only independent predictor for IBTR in nonirradiated
pts was mod-marked comedonecrosis (40 vs. 14% w/ RT)
• The longest f/u of all the trials (at least 12-yrs) thus far. Lessons learned: The 10-yr risk of IBTR
w/o adjuvant therapy after lumpectomy is about 30% (about 50% DCIS and 50% Invasive

DOUG’S BREAST CANCER REVIEW NOTES  PAGE 5 
 PATONE’S RAD-ONC REVIEW NOTES 2008

cancer). RT decreases the risk of both by 50%, but no OS benefit shown in any study. Tam
decreases that risk by another 50%.

EORTC 10853 (Julien Lancet 2000 → Bijker JCO 2001)

• Excision (< 5cm DCIS, 70% mammo detected) Æ +/- 50 Gy
• LF (4-yr): 16% surgery alone vs. 9% with RT (p=S)
• LF factors: Age < 40, solid, close or involved or unspecified margins

UKCCCR (Houghton (UK, Australia, NZ) Lancet 2003)

• Negative margin excision Æ +/- 50 Gy RT +/- Tam
• 2x2 factorial design (Excision +/- 50 Gy RT or Excision +/- Tam)
• LF: 14% surgery alone vs. 6% with RT
• RT + Tam had lowest rate of recurrence, but additional benefit small (p=ns, but trend)

RTOG 98-04 – ongoing trial in low-risk pts ( L → Observation vs. RT → +/- Tam)
• ≤ 2.5 cm, low or int. grade, noncomedo, ≥ 3 mm from inked margin, LN negative, ∼ 50 Gy (no boost)

BCT +/- Tamoxifen

NSABP-B24 (Fisher Lancet 1999 → Semin Oncol 2001, median f/u 83 mos)
• 1804 pts with excision (-margin or microscopic + only) Æ 50 Gy RT +/- tamoxifen
• Overall decrease in breast ca events from 13.4 Æ 8.2% with tamoxifen (p=S) found in all groups
• Significant ↓ of invasive recurrences 7.2% vs. 4.1% and non-significant ↓ in recurrent DCIS (6.2%
w/o, 4.2% with Tam↓). No OS benefit.
• About 40-50% relative reduction of IBTR, but only about a 3% absolute reduction
• Tam ↓ IBTR in pts w/ tumor + microscopic margins to a level approaching tumor – margin pts
receiving placebo
• Allred (Breast ca res treat abstract 2002) subgroup analysis of 628 pts in B-24 w/ known ER status. At
median f/u of 104 mos, ER + pts received a 59% reduction in all breast cancer events (p=s) w/
Tam, but only a 20% reduction in ER – pts (p=0.51, ns)

UKCCCR (Houghton (UK, Australia, NZ) Lancet 2003)

• As above

DOUG’S BREAST CANCER REVIEW NOTES  PAGE 6 
 PATONE’S RAD-ONC REVIEW NOTES 2008

DOUG’S BREAST CANCER REVIEW NOTES  PAGE 7 
 PATONE’S RAD-ONC REVIEW NOTES 2008

LCIS
• Always an incidental finding (0.5-3.5% incidence), nonpalpable, no specific mammo findings
• Difficult to distinguish b/ atypical lobular hyperplasia and LCIS. Generally ER+, Her2/Neu -, and
shows loss of E-Cadherin
• Multicentric (60-80%), frequently involves contralateral breast (35-60%)
• Really more of a marker for breast cancer (RR 3-11!) occurring elsewhere. Not felt to worsen risk
of LR after BCT; one conflicting report from Fox Chase (29 vs. 6% LR if LCIS at margin)
• At 20-25 years f/u, 10-35% risk of invasive cancer (Most are invasive ductal carcinoma not lobular)
o Risk equivalent for contralateral breast
• Treatment options
o Prophylactic bilateral simple mastectomy – particularly in pts w/ genetic/family predispostion
o Bx and life long close observation +/- Tam (annual mammos)
ƒ Tamoxifen: Decreases risk of malignancy by ∼50% at 5 years
ƒ WLE and negative margins are not needed
ƒ Cancer development risk is 1%/yr
o No role for RT

Paget’s Disease
• Described by Sir James Paget as chronic inflammatory changes of the nipple areolar complex
• 1-4% of all breast cancers, unilateral, and present as chronic eczematous eruptions of the nipple-
areolar complex and/or nipple discharge/erosion/pain/bleeding.
• Mammogram: Retroareolar spread or branching calcifications
• Treatment options
o Mastectomy is most widely accept treatment
o BCT emerging as acceptable tx in properly selected pts (unifocal w/o diffuse microcalcs)
ƒ Central segmentectomy +/- nipple-areolar reconstruction + RT
ƒ EORTC 10873 (Bijker Cancer 2001) had 61 pts undergo cone excision and/or nipple
areolar excision w/ negative margins → Whole breast RT (50Gy in 25)
• Median f/u 6.4 yrs, w/ only 6.5% recurrence rate

DOUG’S BREAST CANCER REVIEW NOTES  PAGE 8 
 PATONE’S RAD-ONC REVIEW NOTES 2008

Invasive Breast Cancer

• Primarily in females (200:1 females:males) and is the MC female cancer in the US (> 200,000/yr, 41,
000 deaths)
• Lifetime risk is 1 in 9. Age is the single most important risk factor.
• Pts w/ inherited mutations (BRCA-1 or BRCA-2) have highest risk. Cumulative risk in women w/
BRCA-1 is 3% by 30, 50% by 50, 54% by 60, and 85% by 70
• Though they generally present w/ more advanced ds, male breast cancer pts have similar survival stage for
stage (Scott-Conner Surgery 1999) and treatment approach (most choose mastectomy though)
• Staging: T1a-c (≤ 2cm), T2 (2.1-5cm), T3 (≥ 5.1cm), T4a-d (a=Invades CW (pectoralis not sufficient),
b=edema or skin ulceration or satellite nodules, c=a+b, d=inflammatory); N1 (c:movable ipsilat; p:1-3
nodes), N2a-b (c:fixed (2a), IMN w/o axillary (2b); p:4-9), N3a-c (c:infra (3a), IMN and axillary (3b),
or supraclav (3c); p:>10). Mutimultifocal (w/in the same ductal system and/or quadrant or < 5 cm of
the primary lesion) or multicentric (in different quadrants or > 5 cm from the primary lesion) tumor.
• Axillary nodal status is the most important predictor of survival. Nemoto Cancer 1980 has good
figure showing survival decreasing w/ increasing nodes + (5-yr OS w/ 0 nodes (70%), 1-3 nodes
(60%), 4 nodes (50%), 6-10 nodes (40%), 11-20 nodes + (30%)
• If LN negative, next most important prognostic is tumor size

Natural History
• Most common sites of origin: UOQ (39%), Central (29%), UIQ (14%), LOQ (9%), LIQ (5%)
• Incidence of + Axillary LN increases with size of primary tumor > grade > LVI > histology

Tumor size < 0.5 cm 0.6-1 cm 1.2-2 cm 2.1-3 cm 3.1-5 cm

+ Ax LN <10% 15% 30% ~40% ~50%

• SC LN + in ~ 4%
o Higher for high histologic grade, >4 Ax LN, + level II/III Ax LN
• IMN involvement
o Handley (?Ann R coll Surg Engl 1975): Surgical study of % IMN involvement as f{tumor
location, ax LN status}

Axillary Location of Primary Tumor

Involvement Upper Inner Lower Inner Central Upper Outer Lower Outer
Axilla not involved 14% 6% 7% 4% 5%
Axilla Involved 45% 72% 46% 22% 19%
Total 27% 31% 32% 14% 13%

o Veronesi: % IMN involvement as a f{pt age, ax LN status}, tumors > 2cm

Age < 40 and > 2cm 41-50 and > 2cm ≤ 2 cm

+ Ax LN 41% 33% 27%
- Ax LN 16% 11% 8%

Pathology
• Tumor markers: CA 15-3: Elevated w/ bony mets. CA 27-29: Can predict post-Tx recurrence.
• Unfavorable histologies: Ductal, lobular, mucoepidermoid, adenocystic, invasive micropapillary,
centrally necrotizing. Favorable histologies: Tubular, medullary, mucinous, and papillary.
• Ductal ca (infiltrating): Most common (>>50%)
• Lobular ca (infiltrating)

DOUG’S BREAST CANCER REVIEW NOTES  PAGE 9 
 PATONE’S RAD-ONC REVIEW NOTES 2008

o Singly or small clusters in a targetoid or single-file pattern

o Frequently ER+ (> 80% of time)
o Prognosis depends strongly on subtype (5)
ƒ Best for tubulopapular, worst for solid variant
• Mucoepidermoid ca: Very rare, high-grade – poor prognosis
• Adenocystic ca: Rare, same behavior as salivary gland
• Invasive micropapillary ca: Involve skin and LN, high-grade, ER +
• Centrally necrotizing ca:
o Well circumscribed, nodules w/ extensive central necrosis and narrow rim of viable cells
o ER/PR negative, high incidence of LN mets
• Tubular ca
o Nonaggressive: Low incidence of LN mets, 14% N+ overall (7% in pure tubular ca, 25%
mixed tubular). PPRO: In low-risk tubular ca (< 1 cm) ALND may be omitted
o Most (84%) are ER+, 5-yr OS >90%
o RFS and OS much higher in N+ tubular ca pts than other histologies
o In small tumors, chemotherapy can probably be omitted, just use adjuvant Tam/AI
• Medullary ca
o Microscopically, large cell w/ pleomorphic nuclei w/ lymphoid infiltrates. Grossly, well
circumscribed. Paradoxically almost always ER/PR negative.
o LN mets infrequent, generally good prognosis
• Mucinous (aka mucoid or colloid) ca
o Older women, slow growing w/ long duration of symptoms
o Low frequency of axillary LN mets (4%), thus much better prognosis than IDCa
o In small tumors, chemotherapy can probably be omitted, just use adjuvant tam/AI
• Papillary ca
o Rare, frondlike projections w/ fibrous stalks
o Nonexistent fibrovascular cord, nuclear hyperchromatism
• Angiosarcoma (see below, nonepithelial malignancies of breast)
o Grossly: large, soft, ill-defined
o Microscopically: uneven, dilated vascular channels
o Rare, but much higher rates in:
ƒ Irradiated breasts: Cumulative incidence of angiosarcoma 15 yrs post diagnosis 0.9 per
1,000 for cases receiving radiation (SE = 0.2) vs 0.1 per 1,000 for cases not
receiving radiation (Yap IJROBP 2002)
ƒ Setting of lymphedema (Stewart-Treves)
o Generally poor prognosis, “unpredictable” behavior
ƒ LR high s/p simple excision alone
ƒ MRM vs BCT + XRT
ƒ Consider chemotherapy (?)

Early Stage Disease/Breast-conserving therapy

In General
• Roughly, T1-2 N0-1, T3 N0
• 6 or 7 RCT’s of mastectomy vs. BCT show no OS difference despite slightly higher rate of IBTR
Studies include: NSABP B-06 (Fisher NEJM 2002, 20-yr f/u data), Veronesi/Milan I (NEJM 2002, 20-yr f/u data),
EORTC 10801 (Van Dongen JNCI 2000), WHO (Arriagada JCO 1996), US NCI (Jacobson NEJM 1995), Denmark 82TM
(Blichert-Toft JNCI Monogr 1992)
• Lumpectomy +/- RT: six additional studies—show ~2/3 reduction in local recurrence w/ RT. NSABP-
06: even in smallest tumors there was a local recurrence benefit with RT
• Joint center tried no RT with best subgroup: <2 cm, neg. nodes, no LVI, 1 cm margin, no EIC and
recurrence rate was still 23% at 66 mos. (5% with DM’s).

DOUG’S BREAST CANCER REVIEW NOTES  PAGE 10 
 PATONE’S RAD-ONC REVIEW NOTES 2008

• RT is still SOC following lumpectomy, no subset that has not benefited.

• The risk of developing a contralateral breast cancer is 0.5-1% per yr. Tam decreases the risk of an
IBTR and a contralateral second primary breast cancer by a relative rate of 50%.
• Unlike IBTR (in or near tumor bed), which generally happen within 2-5 yrs, 50% of contralateral second
primary tumors occur p 5 yrs as well as “distant” or elsewhere IBTR (see B-04).
• Absolute contraindications to BCT: 1st or 2nd trimester pregnancy (if 3rd, do lumpectomy then RT after
delivery), >2 tumors in separate quadrants, diffuse microcalcifications (mammogram evaluation, thus,
very important), EIC with + margins, locally advanced tumors, prior breast RT that would lead to
excessively high total dose, systemic collagen vascular ds (active SLE or scleroderma)
• Relative contraindications: large tumor to breast ratio, morbid obesity, psychological attitude of patient
• No good evidence for an increased risk of contralateral breast cancers or 2nd malignancies with BCT

Radical Mastectomy (Halstead) vs. Total (“simple”) Mastectomy

• NSABP B04 set precedent for lesser surgery. No systemic chemo in trial. Also important because
established that axillary/SC nodal RT is as effective as LND in cN+ pts
• Radical Mastectomy = entire breast removal, pectoralis muscles, and axillary contents.
• Total (“simple”) Mastectomy = entire breast removal, but not muscles or axillary contents
• Modified Radical mastectomy = entire breast removal + axillary contents, but not muscles
• NSABP-B04 (Fisher Cancer 1977 → NEJM 1985 → NEJM 2002): RM vs. TM + RT vs. TM
o T1-3 cN0 pts (Over 1000 pts): RM vs. TM w/o ALND but with regional RT vs. TM w/o
regional RT but w/ ALND for recurrence (RM vs. TM +RT vs. TM)
o T1-3 cN+ pts (Almost 600 pts): RM vs. TM w/o ALND but w/ regional RT (RM vs. TM+RT)
o RT: 50 Gy CW tangents (N0) + 10-20 Gy boost (N+). All pts got 45 Gy to SC and IMNs
o At 25-yr f/u, no differences in DFS, RFS, Distant DFS, or OS.
o Node negative pts: LR ↑ in TM w/o RT vs. TM+RT but did not effect regional or distant
recurrences or survival. N+ pts: No diff. in any outcome measures.
o The only survival difference was b/ N0 or N+ pts.
o Regardless of nodal status and tx arm, most recurrences were DM
o See my Journal club handout from 9/02

Mastectomy vs. BCT (L + RT)

• 6 or 7 RCT’s of mastectomy vs. BCT show no OS difference despite slightly higher rate of IBTR
• 2 of 3 meta-analysis (Morris et al, Winchester et al) actually suggest a small OS benefit of BCT over M
• Note: B-06 looks at M vs. L + RT and L alone vs. L + RT

Time Systemic
Trial # Pts Stage Surgery
Period Therapy
Melphalan,
NSABP 06 1219 76-84 I-II WE, MRM
5FU
Milan I 701 73-80 I Q, RM CMF
NCI 237 79-87 I-II LE, MRM AC
EORTC 10801
868 80-86 I-II LE, MRM CMF
IGR 179 72-80 I WE, MRM None
DBCG 904 83-89 I-III WE, Q, MRM CMF, T

Local Recurrence, % Survival, %

Trial F/U, yrs
BCT M BCT M
NSABP 06 12 8 63 59 15
Milan I 7 4 65 65 18
NCI 16 6 77 75 10
DOUG’S BREAST CANCER REVIEW NOTES  PAGE 11 
 PATONE’S RAD-ONC REVIEW NOTES 2008

EORTC 10801 20 12 65 66 10
IGR 9 14 73 65 15
DBCG 5 6 79 82 6

• In Veronesi study, 20-yr IBTR 8% vs. 2%.

Lumpectomy alone vs. L + RT

• At least six studies now done – show ~2/3 reduction in local recurrence w/ RT.
• Meta-analysis (Vinh-Hung J Natl Cancer Inst 2004) suggest OS benefit to adding RT
• NSABP B06 (Fisher NEJM 1985 → World J Surg 1994 → NEJM 2002): Also M vs. L+/-RT (see above.
Even in smallest tumors there was a LR benefit with RT
o 2000 pts w/ Stage I and II (< 4 cm tumors) IBC
o Mastectomy (M) vs. L + RT vs. Lumpectomy (L) alone
o RT associated with a marginally significant decrease in breast ca deaths, offset by increase in
deaths from other causes (pre3D CRT)
o 20-yr IBTR rate 39% L alone vs. 14% L + RT, but most recurrences in 1st 5 yrs
o ∼70% of recurrences (distant or local) in the first 10-yrs and only ∼10% after 10-yrs
o After 20 yrs of f/u, DFS (∼ 35%) slightly improved with RT (p=.07) compared to L alone,
but OS identical (∼ 45%) in all groups (argument against LC resulting in decreased DM
and thus decreased OS, but again done in pre3D era)

Tumor Size Total

Study / Nodal Randomized Surgery XRT
Status (- / + XRT)
NSABP-06 T < 4 cm / 1450 Lumpectomy 25 x 2 Gy tangents;
(5) N0 or N1 (719/731) No boost
T < 2.5 cm / 579 Quandrantectomy + Axillary 25 x 2 Gy + 5 x 2 Gy
Milan III (6)
N0 or N1 (280/299) LND boost
T < 2 cm / 381 Lumpectomy +Axillary LND 27 x 2 Gy tangents;
Uppsala (7)
N0 (197/184) or sampling No boost
T < 4 cm / 837 Lumpectomy + Axillary LND 16 x 2.5 Gy tangents
Ontario (8)
N0 (421/416) + 5 x 2.5 Gy boost
T < 4 cm / 585 Lumpectomy + Axillary LND 20-25 x 2.5-2 Gy tang
Scotland (9)
N0 or N1 ( 294/291 ) or sampling + 10-15 Gy boost

DOUG’S BREAST CANCER REVIEW NOTES  PAGE 12 
 PATONE’S RAD-ONC REVIEW NOTES 2008

8 + RT, Annual Incidence No RT, Annual Incidence 40

+ RT, Cum ulative Incidence No RT, Cum ulative Incidence

Cumulative Incidence IBTR (%)

Annual Incidence IBTR (%)

6 30

4 20

2 10

0 0
0 1 2 3 4 5 6 7 8 9
Time (years)
• At 8 years f/u, the above 5 RCT demonstrate a four-fold reduction in the rate of IBTR
o ~ 4%/yr without XRT vs. 1%/yr with XRT
o In NSABP-06 relative reduction less at 20 yrs: 39% vs. 14% (2% vs. 0.7%)
• None show a significant improvement in OS

L +/- Tam and/or RT

• NSABP B21 (Fisher JCO 2002)
o L + Tam only vs. L + RT only vs. L + Tam + RT combined
o Invasive tumors < 1 cm randomized to RT alone, tamoxifen alone or both
o Cumulative 8-yr IBTR was 16.5% with Tam, 9.3% with RT and 2.8% with RT/Tam
o Tamoxifen treated patients had significant reduction in contralateral breast cancers
o No OS differences in any of the 3 groups (~93%)

• CALGB 9343 (Hughes NEJM 2004)

o L + Tam +/- RT in Stage I ER + pts > 70 yrs of age (median age 75)
o LRR (5) improved w/ RT (4% vs. 1%) but no OS

• Canadian Trial (Fyles NEJM 2004)

o L + Tam +/- RT in T1-2 N0 (I-IIA) pts > 50 yrs of age (median age 68)
o IBTR (8) improved w/ RT (17.6 vs. 3.5%) and axillary recurrence (2.5 vs. 0.5%) as well
o DFS (5) improved w/ RT (91 vs. 84%) but no OS (93.2 vs. 92.8%)
o Pts aged 50-59 had a greater risk of relapse than pts > 60 as well as pt w/ T2 tumors

L and RT +/- tumor bed boost

EORTC (Bartelink H, NEJM 2001;345:1378)
• 2,657 pts w/ Stage I/II dz s/p L +ALND + 25x2 Gy +/- 8x2 Gy boost (50 vs. 66 Gy)
• 5-yr LR 4.3 vs 7.3% +/- boost (p = 0.001)
o Pts < 40 yo benefited most: 10 vs 20% LR
o LF increased risk of mets (by 8.3%?)
Lyon Breast Cancer Trial (Romestaing P, JCO 1997;15:963)
• 1,024 pts w/ Stage I/II dz s/p L +ALND + 25x2 Gy +/- 5x2 Gy boost (50 vs. 60 Gy)
• 5-yr LR 3.6 vs 4.5% +/- boost (p=0.044)

DOUG’S BREAST CANCER REVIEW NOTES  PAGE 13 
 PATONE’S RAD-ONC REVIEW NOTES 2008

o Difference more pronounced after 7 years

Sentinel lymph node dissection

• In experienced surgeons, false negative rate is < 5%. RCTs testing SLND almost complete.
• If SLN +, all pts probably need completion LND. Retrospective studies (Viale Annal of Surg 2005 and
Menes J Am Coll Surg 2005) looking at pts even w/ the smallest SN mets (< 2mm), revealed that there
was still a 15-20% chance of finding additional + LNs on completion ALND.

Margins
• No standardized definition of adequate margin; direct correlation between margin width and
extent of tumor cells near the margin and the ultimate local control rate
• Joint center study: < 1mm did the same as negative (7% LR)
• Margins extensively + (involvement in >3 LPF) and focally positive (1-3 LPF) had recurrence
rates of 27% and 14% respectively
• Amount of tumor close to the margin is important (Beaumont)
• Would like to see increased margin with younger patients
• ~50% rate of residual tumor in patients with positive margin that have re-excision; ~6-20% in
patients with 1 mm margin who have re-excision.
• Some have shown increased rate of DM’s with positive margins (LR also probably impacts on
survival)
• Importance of post-op mammogram to ensure that all calcification have been removed

EIC
• JCRT retrospective study showed 21% LR with EIC, 6% without. Defined EIC as 25% or more
of the primary tumor is IDC, and IDC is seen outside (adjacent to) the infiltrating tumor
border
• More often have + margins and more locally extensive. 88% + reexcision vs. 48% for EIC
negative tumors at JCRT.
• Holland mastectomy data showed 59% of EIC + tumors had residual ds > 2 cm from primary
tumor (vs. 29% of EIC negative). 21% @ 6 cm.
• EIC w/ + margin – higher relapse. EIC w/ - margin – does not effect LR, so not a contraindication
to BCT in this case (may need larger volume of resection)

Sequence of chemo and RT

• Recht NEJM 1996: Included N0 and N1 pts. Improved rate of DM’s with chemo first; 10-year report:
no different
o More LR with CT first, more DM’s with RT first (p=NS)
o Close margins: CT-first did worse; + margins: both groups did poorly
o Conclusions: with prompt RT, close and negative margins have similarly low LR; with
delayed RT, close margins appear to have a high risk of LR. Stated though, that for pts at
high risk of DM, chemo before RT best.
• Retrospective studies suggest that a delay of RT > 4 months increases local failure risk

Nodal irradiation
• Negative sentinel node treated like negative axillary dissection
• If + sentinel node, standard of care is full dissection (generally level I/II)
• ~ 60% of patients with 1-3 positive sentinel nodes have no further mets at completion dissection
(study of need for further surgery underway)
• Axillary radiation: may be considered instead of sampling or instead of dissection if positive
sentinel node (investigational)

DOUG’S BREAST CANCER REVIEW NOTES  PAGE 14 
 PATONE’S RAD-ONC REVIEW NOTES 2008

• > 4 nodes (or incompletely dissected axilla with positive nodes) definitely an indication for RT to
supraclav/?IMN’s.
• Nodal recurrence patterns (Recht) give little justification for regional nodal irradiation in pts with
1-3 involved nodes if nodal failure is utilized as the end point
• Comprehensive nodal irradiation may have a survival benefit (based on post-mastectomy data).
Sterilize occult sites of ds that lead to distant mets—is the justification that is given.
• cN0: risk of axillary recurrence is ~1-5% with RT alone (retrospective data and B-04)
• No value of nodal RT for pNO—only increases lymphedema risk
• No LR benefit shown for axillary RT with nodes (? Without) ECE

Toxicity
Lymphedema risk
• 37% with full axillary dissection plus RT; 6% with level I/II dissection
• JCRT 4% actuarial risk at 6 years with axillary RT only
Brachial plexopathy
• TD 5/5 classically 60 Gy at 2 Gy/fx, but emerging data suggest lower doses safer (e.g. < 54 Gy)
• Risks increased w/ concurrent chemo, younger pts (have longer time to manifest sxs)
• Sxs include paresthesias, hypesthesias, weakness, and impaired reflexes. Pain assoc. w/ recurrent
tumor, but is uncommon with radiation plexopathy (<20% have pain)

Reduced duration of treatment

Accelerated fractionation
• Canadian studies using 2.5-2.75 Gy/d to 40-44 Gy with ~5% 5-yr local failure (no difference
in LR and cosmesis vs. 2 Gy to 50).
• ? late toxicity

Partial breast irradiation

• Historically, done as boost (e.g. LDR catheters)
• In modern area, generally synonymous w/ accelerated partial breast RT

Accelerated partial breast RT

• Improve utilization of BCT, decrease time and cost
• May be able to give before systemic chemotherapy (? Better LC)
• 80-90% of LR after BCT occur in tumor bed region
• NSABP/RTOG 03-14 (Vicini RCT study) – just opened
• Whole breast vs. partial breast (mammosite HDR or interstitial cath HDR or EBRT), 3.4
Gy BID to 34 Gy total. Will accrue 3000 pts.
• LM method: 4-field (Sup-medial oblique, Inf-medial oblique, lat tangent, and en-face electron
field). 15x/6x for photon fields. 2.85 Gy (photons) + 1 Gy (electrons) = 3.85 Gy/fx, BID,
therefore 38.5 Gy in 5 days.

Brachytherapy:
• One small RCT from Hungary now w/ 7-yr f/u shows no diff from WBRT.
• LDR: Vicini 50 Gy, Krishman 20-25 Gy, Lavenda 50-60 Gy, all w/ LC > 90%
• British study of LDR 55Gy over 5.5 days had a LR rate of 15% and 96% excellent cosmesis.
• Beaumont criteria: no multicentric ds (optimal mammo evaluation), only N0, < 3 cm, margins
> 2 mm, no residual calcifications, RT within 8 weeks of last breast surgery.
o Cover cavity plus 1-2 cm margin. LDR: 50 Gy in 96 hours; HDR: 32 Gy (4 Gy BID x
4 days)

DOUG’S BREAST CANCER REVIEW NOTES  PAGE 15 
 PATONE’S RAD-ONC REVIEW NOTES 2008

o 5-yr actuarial results currently 1.2%; no difference in LC, DM’s or OS c/w matched
EBRT patients
• HDR: Kuske, RTOG 95-17; 3.4 x 10—5 days. 99 patients enrolled.
• MammoSite—2-yr results with 88% excellent/good cosmesis and no recurrences. Cosmesis
worse with < 7 mm spacing from skin. 34 Gy (3.4 Gy BID x 5 days). Rx to 1 cm from
balloon surface (IJROBP 2003)

EBRT (3D conformal, electrons, intraoperative RT)

• Beaumont: tumor bed + 1.5 cm (treats about ¼ of breast to full dose); typically 4 fields
• Veronesi: electron IORT 21 Gy single fraction after quadrentectomy. Current phase III trial
vs 60 Gy EBRT
• London trial with intraop 5 Gy x 1 with kV device KM saw at SABCS; RCT vs. standard
EBRT

IMRT
• Skin toxicity associated with V105% or greater
• Able to generate more consistent/uniform treatment plans
• Can improve tumor bed dose, better toxicity, decrease acute toxicity and lung/heart dose

DOUG’S BREAST CANCER REVIEW NOTES  PAGE 16 
 PATONE’S RAD-ONC REVIEW NOTES 2008

Post-mastectomy RT
General
• In the 1960s and 70s five RCTs were published showing that PMRT reduces LRR but does not affect OS
(e.g. Rutqvist (Stockholm trial) IJROBP 1992). These however were in the era of no systemic therapy
and using what is regarded today as cardiotoxic RT (i.e. “hockey-stick” field)
• In the 1980s and 90s 15 RCTs of PMRT were conducted in pts receiving adjuvant systemic therapy, but
these were all underpowered (50-300 pts). Not until 1997(see below) was an OS benefit shown.
• With chemo, frequency of LRR as a first site of relapse increased. LRR may decrease OS by acting as a
source for “re-seeding” of DM, thus PMRT improves OS by improving LC.
• Recht (ECOG data) JCO 1999: Assessed patterns of failure in node + pts treated w/ mastectomy and
chemo, but w/o RT. 10-yr rate of LRF w/ or w/o DF was 28.7% if ≥ 4 + nodes vs. 12.9% if 1-3.
• MDACC; tumors > 5 cm with negative nodes had LR rate of 29%; most common sites of recurrence were
chest wall (68%) and supraclavicular fossa (41%)
• LR rate higher with < 10 nodes axillary sampled vs. > 10
• Absolute OS difference w/ PMRT is about 10%

Overgaard (Danish BCG 82b) NEJM 1997

• 1708 pre-menopausal node + or T3/4 N0 pts Æ CMF +/- RT (50 Gy)
o Total mastectomy + level I/II ALND (median 7 nodes)
o Randomized to 8 cycles CMF + RT vs. 9 cycles CMF
o Comprehensive nodal tx (Ax/SC/IMNs). IMN tx w/ electrons.
• LR and OS improved in all groups (node negative, 1-3 and 4 or more groups)
• 10-yr LRR (32 vs. 9%), DFS (34 vs. 48%), and OS (45 vs. 54%) benefit with RT (all p <.001).
• All nodal groups had an OS benefit from RT: N0 (70 to 82% w/ RT), N1-3 (54 to 62%), and N4+
(20 to 32%). OS benefit ∼ 10% w/ RT!
• Criticized for low numbers of axillary nodes removed and no Adria. Also LR rate in 1-3 group was
30% in this study vs. only 12.9% in Recht’s study
• Separate study w/ Tamoxifen +/- RT (Overgaard, Danish BCG 82c Lancet 1999): 10-yr OS benefit
in post-menopausal pts (45 vs. 36%) given RT (no OS benefit in T3N0 in this study)
• No difference in cardiac-related mortality or morbidity at 10-yrs

Ragaz (British Columbia Trial) NEJM 1997

• 318 pre-menopausal node positive pts Æ CMF +/- RT (37.5 Gy, 2.34/fx)
o All received CMF and level I/II ALND (median 11 nodes)
o Co-60 RT w/ 35 Gy to Ax/SC/IMN via photons
• 15-yr LRR (33% vs. 13%, p=.003), DFS (33% vs. 50%, p=.007) and OS (46% vs 54%, p=.07) all
improved with RT
• Same advantage 1-3, 4 or more nodes
• In 1-3 with ECE the OS was 39% without RT vs. 65% with RT without a difference in local
failure (suggests reduction in metastatic spread).

Internal Mammary Nodes

• Nodes in first 3 interspaces most likely to be involved
• Veronesi: 29% rate of IM involvement in 556 patients with axillary involvement
• Randomized trials of surgical treatment of IMN have shown no benefit, except in Lacour trial of
T1/2 medial tumors with positive axillary nodes. However, most were underpowered to show
small OS benefit. Remember Ragaz and Overgaard studies did include the IMNs
• Arriagada 1988 multi-variate analysis of 1195 patients noted an OS advantage for treatment of
IMN’s with medial tumors and positive axillary nodes.

DOUG’S BREAST CANCER REVIEW NOTES  PAGE 17 
 PATONE’S RAD-ONC REVIEW NOTES 2008

• Incidence of clinical failures is low (1-2%), but benefit may be in survival with decreased mets as
mentioned above.
• Pierce IJROBP 2002 paper looks at the dosimetric effect of various techniques used to treat the
IMNs
• Pending phase III trial of CW + SC +/- IMN RT after mastectomy: Romestaing (French SFRO)
Radiother Oncol Abstract 1998. Also EORTC (included BCT pts as well and Canadian study
(BCT pts only)

PMRT technique:
• Volume: CW, IMNs, S’clav nodes. Axilla generally not treated if level I/II dissection done
• Dose: CW: 50 Gy/25fx. IMN: 50 Gy/25 fx. SC: 46 Gy/23 fx. Scar: 10 Gy/5 fx boost.
• Techiques & Borders: CW: tangents (6x). SC/Axilla: AP (d=3-5 cm). IMN: PWTFs or separate
IMN (2:1 or 3:1 E:P, e.g. 9mEv-12 80% 32 Gy/16 fx: 6x 3-5 cm 16 Gy/8 fx). Scar boost:
electrons.
• Thoms IJROBP 1989: Importance of erythema on CW. Noted 30% CW failure in pts w/ minimal
erythema vs. only 13% in pts w/ brisk erythema or desquamation.
• Special considerations: Bolus 1 cm to CW q.o.d. Heart block: Consider electron patch if used
• 3D with PWTF appears to be best dosimetric approach. ? Role of breath-hold techniques to
decrease heart in field (treat at inspiration).

DOUG’S BREAST CANCER REVIEW NOTES  PAGE 18 
 PATONE’S RAD-ONC REVIEW NOTES 2008

Systemic therapy

Adjuvant chemotherapy

Early Breast Cancer

Early Breast Ca Trialists’ collaborative group (Lancet 1998, Oxford overview)
• Chemo improved RFS for < 50 and 50-69; few > 70 studied
• Significant improved OS for < 50 and 50-69; recurrence reductions during first 5 years and OS
throughout 10 years
• Proportional reductions in risk were similar for node-negative and node-positive
• < 50 node negative, 10-yr OS increased from 71Æ78% (7% absolute survival diff)
• < 50 node positive, 10-yr OS increased from 42Æ53% (11%)
• 50-69 node negative, 10-yr OS increased from 67Æ69% (2%)
• 50-69 node positive, 10-yr OS increased from 46Æ49% (3%)
• 20% reduction in contralateral breast tumors
• No significant increase in non-cancer deaths
• No benefit to more than 3-6 months of chemo
• Anthracycline-containing regimens slightly better than Cyclophos/Mtx/5FU (CMF) (72 vs. 69% 5-
yr OS, p=S). 2-3% absolute survival benefit!
• Small improvement in LC beyond that from RT as well (other studies)
• Older women with ER poor tumors who do not receive Tam get same relative benefit of chemo as
in younger women.

NSABP B-20 (Fisher JNCI 1997):

• 2300 pts w/ LN negative, ER positive breast cancers. Randomized to adjuvant Tam alone vs.
Tam + Chemo (CMF or MF).
• LC, DM and OS improved w/ Chemo in all subgroups, though the reduction was greatest in pts
< 49 yo.

Toxicity:
• addition of chemo to RT may increase rate of pneumonitis, brachial plexopathy, edema and rib
fractures (variable effect in different studies).
• May slightly reduce cosmesis vs. RT alone.
• Concurrent chemo-RT definitely more toxic—esp. MTX and Adria.
• Cardiotoxicity-4% CHR/MI in pts with adriamycin and Lt sided tangents (Harrington-
Hardenbergh). Taxol and Herceptin may also be a problem. Minimize dose to the heart.

Locally advanced breast cancer:

• There are at least 6 RCTs looking at adjuvant C. Most did not show OS benefit but are small
trials. Larger studies (Derman IJROBP 1999, Rubens Eur J Can Clin (EORTC 10792) 1989, and Klefstron
Cancer (Helsinki III) 1987) however do suggest an OS benefit.
• Preop vs. Postop C trials are at least equivalent w/ some showing a DFS and OS benefit w/ preop
C. Preop C allows BCT and tests tumor chemoresponsiveness, therefore C for LABC is
generally given preop.

Neoadjuvant chemotherapy
• 60-80% tumor response to neoadj chemo (10-20% achieve a CR and 50-60% a PR). In general tumor
size decreases by 50% and a 10% pCR is obtained.
• Have clip placed prior to therapy to aid in resection later if pCR

DOUG’S BREAST CANCER REVIEW NOTES  PAGE 19 
 PATONE’S RAD-ONC REVIEW NOTES 2008

• Comparisons against adjuvant C do not suggest a clear survival benefit, but NAC does increase the
BCT rate and allows evaluation of ca response to the chemo
• Some however are advocates for mastectomy in LABC pts because the risk of recurrence might be
higher in larger tumors that required regression prior to receiving BCT (e.g. downstaging prior to
BCT)—need larger margins than usual to feel comfortable (? 2 mm).
o In NSABP B-18 (see below): IBTR rate was twice as high in pts who only became
BCT candidates after responding to NA chemo compared to BCT candidates at dx
(14.5% vs. 6.9% , p=.04)
• Conflicting data about accuracy of SLN mapping after chemo, therefore most feel it is prudent to do
sentinel nodes before neoadjuvant therapy or axillary dissection if neodjuvant therapy given before
the nodes are sampled
• Even with cCR after neoadj chemo in early/small tumors, probably can’t avoid surgery. Paper from
Royal Marsden (Ring JCO 2003) looked at treating 136 early staged pts after cCR to chemo. Half
had XRT only (no surgery) and half S+RT. Though no OS or DFS difference, there was a trend
toward worse 5-yr LRC with RT only (21% vs. 10%, p=.09)

NSABP B-18 (Fisher JCO 1998 → Cancer 2002)

• 1,532 pts w/ T1-3N0-1 BrCa randomized to 4 cycles pre vs. post-op AC
• Included MRM and BCT pts
• 1° endpoints: OS and DFS. 2° endpoints: Preop chemo ability to downstage the primary tumor and
involved axillary LNs, comparison of lumpectomy rates and rates of ipsilateral breast tumor
recurrence (IBTR), and correlation b/ primary tumor response and outcome.
• Through 9 years of f/u, no differences in OS (70% postop vs. 69% preop, p=.8) or DFS (53% vs.
55%, p =.50).
• Strong correlation b/ primary tumor response and outcome (In preop pts, those with pCR had best
outcome). This correlation has only become statistically stronger with longer follow-up.
• Preop pts received notably more lumpectomies (68% BCT with preop vs. 60% post-op, p=S) than
postop patients, especially patients w/ tumors > 5 cm at study entry.
o Rate of IBTR slightly higher in preop gp (10.7% vs 7.6%), but Δ not significant?
• Marginally statistically significant treatment-by-age interactions emerging for OS and DFS suggest
that younger pts benefit from preop tx, while older pts benefit from postop tx.

EORTC 10902
• 698 pre/postMP pts w/ T1-T4b N0-1 BrCa randomized to 4 cycles pre vs. post-op FEC
o 4 cycles FEC→ surgery + XRT → Tam (if over 50)
o Surgery → 4 cycles FEC + XRT → Tam (if over 50)
• At 56 months f/u, improved BCT rate but no difference in OS
o BCT rate 38% in pre-op gp vs. 21% in post-op gp (p=s)
o 4-yr DFS 65% in pre-op vs 70% in post-op gp (p= NS)
o 4-yr OS 82% in pre-op vs 84% in post-op gp (p= NS)

NSABP B-27
• 2,189 pts w/ T1-3 N0-1 BrCa randomized to preop AC vs. preop AC → Taxotere
o 4 cycles AC → surgery → Tam vs.
o 4 cycles AC → 4 cycle docetaxel (taxotere) → surgery → Tam
• Neoadjuvant A/C followed by docetaxel significantly increased the primary tumor cCR (65%
vs 40%, p < .001) and pCR rates (25.6% vs 13.7%, p < .001) and decreased the positive
axillary nodes rate (40.5% vs 48.5%, p = .01).

DOUG’S BREAST CANCER REVIEW NOTES  PAGE 20 
 PATONE’S RAD-ONC REVIEW NOTES 2008

Adjuvant hormonal therapy

Tamoxifen (from Lancet 1998 overview)

• Given for 5 years after surgery for early, ER + breast ca reduces the risk of death by ~25% in all
patients
• The absolute improvement in 10-yr survival was 10% for node positive (50.5Æ61.4%) and 5% for
node-negative (73.3Æ78.9%). P=S.
• Benefits irrespective of age, menopausal status, use of chemo.
• ∼50% contralateral breast cancer reduction.
• Decrease local failure with RT in BCT; does not affect the cosmetic result.

Aromatase inhibitors
Background
• Suppress plasma estrogen levels in post-menopausal women by inhibiting or inactivating aromatase
(enzyme responsible for synthesis of estrogens from androgenic substrates like estrone from
androstenedione and estradiol from testosterone—peripheral conversion)
• Exemestane (Aromasin), Anastrozole (Arimidex), Letrozole (Femara)
• Aromatase is present in SQ, liver, fat and breast-cancer tissue (10x plasma)
• Not useful in premenopausal pts (causes upregulation of ovaries)

Compounds
• 1st generation—aminoglutethemide, nonsteroidal. “Medical adrenalectomy”—limited by rash and
drowsinessÆ led to development of new agents
• Type 1 inhibitors are steroidal analogues of androstenedione and bind irreversibly (2nd generation
is Formestane and 3rd is Exemestane (Aromasin)).
• Type 2 inhibitors are nonsteroidal and bind reversibly (2nd generation is Fadrozole and
Rogletimide, 3rd are Anastrozole (Arimidex), Letrozole (Femara) and Vorozole).
• 3 generation are very specific for aromatase, no effect on cortisol or aldosterone levels
rd

o Anastrozole and letrozole can be given orally, once-daily

Indications
• Adjuvant therapy: Small benefit of anastrozole over tamoxifen in ATAC trial. Anastrozole better
if h/o DVT and contraindications to tam.
o 2005 ASCO recs: They now recommend it as first line therapy
• Neoadjuvant therapy: to downstage or try for BCT instead of mastectomy. RCT showed letrozole
better than tam in large ER/PR + tumors.
• 1 line therapy in advanced (e.g. metastatic) disease: Letrozole clearly superior to tamoxifen in
st

advanced disease, anastrozole is at least as good as tamoxifen

Adverse reactions
• Generally tolerated well; slightly less hot flashes and vaginal dryness/bleeding than tamoxifen
• More osteoporosis/fx (need pretx bone density) and musculoskeletal symptoms than tam

Data
• All RCT data in postmenopausal pts, also no OS benefit in any trial
• Baum (ATAC trial) Lancet 2002: (Arimidex (anatrazole), Tamoxifen Alone or in Combination;
5-yrs adjuvant tx)
o Almost 9400 postmenopausal receptor + pts, endpoint

DOUG’S BREAST CANCER REVIEW NOTES  PAGE 21 
 PATONE’S RAD-ONC REVIEW NOTES 2008

o89% 3-yr RFS for anastrozole vs. 87% for tamoxifen (p=.013). No OS difference to
date. Combination of tam and arimidex not better than tam alone.
o Arimidex had lower contralateral breast ca rate and was better tolerated in terms of
endometrial ca, CVA, DVT, vaginal bleeding/discharge and hot flashes.
o Arimidex had higher rate of fractures and musculoskeletal disorders (joints, etc)
• Goss NEJM 2003: (Tam 5 yrs → +/- Femara (letrozole) 5 yrs; 10-yrs of adjuvant tx)
o 5200 receptor + pts, endpoint DFS. Adding letrozole (Femara) improves 4-yr DFS
from 87 to 93%
o Hot flashes, vaginal bleeding, arthritis, myalgia, and arthralgia were higher in the
Letrozole group compared to placebo after Tam
• Coombes (Intergroup Exemestane Study) NEJM 2004: (Tam 2-3yrs → Tam or Aromasin
(Exemestane) 2-3 yrs; 5-yrs adjuvant tx)
o Double blind RCT with 4742 ER +/unknown postmenopausal breast ca pts that
received Tam (20 mg qd) x 2-3 yrs → randomization to 2-3 more yrs Tam vs. 2-3
yrs Exemestane (25 mg qd) for a total of 5 yrs.
o 32% risk reduction or 4.7% absolute DFS benefit switching to Aromasin
o Only a OS trend, but contralateral breast ca benefit (20 tam vs. 9 AI, p=.04)
Chemo Prevention

NSABP-P1 (Fisher J NCI 1998): Placebo vs. Tam (20mg/d x 5 yrs)

• Since tamoxifen reduced contralateral tumors in adjuvant setting (B-21), looked at for prevention
• 13,388 felt to be at increased risk: 60 or older, 35-59 with 5-yr predicted risk of at least 1.66%
(Gail model), or h/o LCIS
• Result: Tam ↓ risk of invasive breast ca by ~50% in all groups, DCIS also reduced by 50%
• ER + tumors reduced by 69%, no difference in ER – tumors
• No difference in heart ds, but Tam ↑ CVA, PE, DVT in > 50 y.o., and bone density (fxs ↓)
• Endometrial Ca (all early stage) increased with Tam > 50 y.o. (RR 2.5)

NSABP-P2 (STAR)
• Study of Tam And Raloxifene (Evista = a selective estrogen receptor modulator (SERM)
• Study underway - hope to decrease endometrial ca risk

DOUG’S BREAST CANCER REVIEW NOTES  PAGE 22 
 PATONE’S RAD-ONC REVIEW NOTES 2008

Locally advanced and Inflammatory Breast Ca (T3 or T4)

Background
• Locally Advanced Non-Inflammatory BrCa (Non-IBC) = “locally advanced”
o Resectable: IIIA (T3 N1), also IIB (T3 N0)
o Unresectable: IIIB (T0-3 N2 or T4a-c N0), IIIC (any T N3)
• Locally Advanced Inflammatory BrCa (IBC) = “inflammatory”
o Unresectable: IIIB (T4d N0-3)
• AJCC 1997: Stage III A (T3 N+), B (any T N2-3), C (T4)
• 15% of BrCa pts are locally advanced or inflammatory (1-4%)
• Definitions ambiguous—skin involvement from prolonged growth/neglect vs. rapidly fulminant
neoplasm. Prognosis is pretty much the same.
• Inflammatory may be clinical finding (warmth, red, edema) or pathologic (dermal lymph tumor
emboli) or both
• Cancers of ductal or lobular origin that invade superficially into the dermal lymphatics and travel
the plane just beneath the skin surface which is “path of least resistance” and therefore quickly
spread.
• Classically (in order of worsening ds): extensive skin erythema, peau d’orange (edema from
intramammary lymphatic involvement), skin fixation, satellite nodules and en cuirasse (plaquelike
skin that is yellowish/red/gray).

Clinical
• T3 N1 – resectable, N2 – borderline, T4 or N3 (IMN) - unresectable
• Staging: T4a-chest wall, T4b-edema, ulceration of skin or satellite skin nodules confined to the
same breast, T4c-both chest wall involvement and skin involvement or edema, T4d-inflammatory
carcinoma (tumor embolization of dermal lymphatics)
• Columbia Clinical Classification (Haagensen clinical criteria of operability, Ann Surg 1943)
A No edema, ulceration or skin fixation. No nodes.
B No edema, ulceration or skin fixation. Nodes, but < 2.5cm and without fixation.
C Any one of the 5 Grave Signs (operable but 5-yr OS 23%, 5-yr LF 27%)
1. edema <1/3 of breast
2. skin ulceration
3. fixation to chest wall (not pectoralis)
4. nodes > 2.5cm
5. fixation of nodes (skin or deep)
D All more advanced (categorically inoperable, 5-yr OS 0%, 5-yr LF 53%) – Neoadj tx first
1. 2 or more grave signs 2. edema >1/3 of breast
3. satellite skin nodules 4. Inflammatory carcinoma
5. SC nodes 6. IM nodes
7. arm edema 8. distant mets

Treatment
• Historically (Bloom, BMJ, 1962) – no treatment gave median survival of 2.7 yrs and 18% (5-yr) for
locally advanced breast ca (T3/4). Poor QOL. 73% w/ marked breast ulceration & 21% w/
pleural surface exposure.
• Locally advanced (operable): 20-30% 5-yr OS with RT alone and ~ 40-50% with mastectomy +
RT. For inflammatory cancer: RT only (70 Gy) < 3 % and mastectomy < 10% 5-yr. Median
survival < 2 years.
• Multimodality tx now SOC: chemotherapy, hormonal, RT and surgery. Generally start chemo 1st
bec. of rapid spread of inflammatory cancer.
• Algorithm: Adriamycin-based chemo (typically > 60-80% CR/PR rate) Æ assess for resectability,
mastectomy if yes/RT if no then reevaluate for mastectomy Æ chemo/hormones

DOUG’S BREAST CANCER REVIEW NOTES  PAGE 23 
 PATONE’S RAD-ONC REVIEW NOTES 2008

• ∼ 80% of borderline/inoperable tumors tx w/ chemo, regress enough to become operable

• Chemo: pCR typically 10%, cCR 15-20%
• Mastectomy only for chemo responders. If no chemo response, all are dead by 2 years.
• 5-yr OS with trimodality treatment ~30-50% (median survival ~4 years).
• Can consider BCT for T3 ds with good response (~55% of these will avoid mastectomy after
neoadjuvant chemo)
• MDACC treats BID—similar LC, but less toxicity giving higher doses. (Ueno Cancer Chemo
Pharm 40:321-329, 1997).

RT after Neoadjuvant systemic therapy

• Optimal role unclear after down-staging. We know that LR related to number of nodes after
MRM, margin in BCT.
• MDACC data of presurgical chemo + MRM: No positive nodes-- LRR 10% (vs. 4% with
surgery as initial treatment), 1-3—17% (vs.10%), 4-9--47% (vs. 21%).
• Rates of LRR after neoadjuvant chemo are higher stage for stage (pathologic, when clinical
stage used not much difference).
• Probably best to do sentinel node bx prior to neoadjuvant chemo in cN0 to get accurate
staging prior to treatment
• BCT after NAC for LABC probably ok but would not do for IBC, + margins, tumors which
responded poorly to NAC, or multicentric ds.

CALGB (Perloff JCO 1998): Preop chemo → RT vs. S

• 113 pts w/ Stage III BrCa tx w/ neoadjuvant FAC and Vincr/Prednis x 3, followed by MRM or
definitive RT
• No diff in LF, DFS, or OS.
• This is an argument for RT alone after a NA chemo but for most pts SOC is still likely
proceeding w/ S followed by adjuvant RT.

DOUG’S BREAST CANCER REVIEW NOTES  PAGE 24 
 PATONE’S RAD-ONC REVIEW NOTES 2008

Recurrent Breast Ca
Regional Nodal Recurrence after BCT
• In early-stage ds, the risk is only about 1-5% at 8-yrs (lowest in pts w/ pN0 ds and as the # of
nodes removed increases), but most (2/3 of these pts) develop metastatic ds.
• Most common site of recurrence is the axilla, then SC, them IMN, then multiple sites. Pts w/
isolated axillary recurrences have best prognosis of those with RNR.
• Risk of RNR is lowest in pts w/ < 4 nodes + and a complete ALND (or – SLND), therefore RN
irradiation remains controversial (at Duke w/ irradiate all N+ pts). NCI of C (MA-20 trial) is
addressing this issue.

In Breast Tumor Recurrence (IBTR) after BCT

• Pts c an IBTR after BCT have only a slightly worse prognosis than a pt c de novo ds. They are
usually treated with mastectomy (ie increased morbidity with reirradiation and surgery) c 85-95%
LC, 47-68% 10-yr DFS and 60-70% 10-yr OS
• NSABP B-06: At 20 yrs, there was a marginally significant decrease in deaths due to breast cancer
(p=.04) with L + RT compared to L alone, but OS the same (? LC → DM)
• Risk factors for IBTR: Young age, + micro margins, gross multifocality, EIC, LVI
• Differentiating true recurrence vs. new primary difficult, but recurrences tend to happen sooner (2-
5 yrs p RT) and have worse prognosis
• Many advocate systemic adjuvant therapy p IBTR. This is being addressed in a RCT by the
International Breast Cancer Study group and NSABP, but at present, seems warranted at least for
pts with an invasive tumor > 1-2 cm that occurs within 3-4 years of the primary
Chest Wall recurrence
• By contrast, pts c a CW recurrence after mastectomy, have a much poorer prognosis (LP states it
is a marker for metastatic ds), but warrant RT
o Halverson IJROBP 1990 – Mallinckrodt study looking at 244 pts c isolated local-
regional recurrence after mastectomy
o Pts were irradiated c curative intent
o 5-yr and 10-yr OS was 43% and 25%
o The 5-year local-regional tumor control was best for patients with isolated chest wall
recurrences (63%), intermediate for nodal recurrences (45%), and poor for
concomitant chest wall and nodal recurrences (27%)
• Treatment: Resection, CW RT and adjuvant systemic chemotherapy

DOUG’S BREAST CANCER REVIEW NOTES  PAGE 25 
 PATONE’S RAD-ONC REVIEW NOTES 2008

Nonepithelial Malignancies of the Breast

Background
• Make up < 5% of all breast neoplasms, therefore most reported series are small and diff. centers use
disparate classification schemes obscuring true prevalence and clinical course.
• Epithelial tumor general categories: Ductal (i.e. invasive, tubular, etc.), Lobular, and Carcinoma NOS.
• Nonepithelial tumors can be divided in 4 main categories:
o Primary breast sarcomas
o Secondary or tx related breast sarcomas
o Phyllodes tumors
o Primary breast lymphomas
• Primary sarcomas: Malignant tumor arising from the connective tissue of breast. Often exhibit more
rapid growth than epithelial tumors. Usu. Lack calcification and spiculation, thus can be mistaken for
benign lesions like fribroadenoma. Types vary but in one large series (90 pts, excluded phyllodes) MFH
(70%) and Angiosarcoma (10%) are the MC. Prognosis similar to sarcomas elsewhere, namely size
and grade are most important. Like other sarcomas, tend to spread locally or hematogenously and not
typically to regional LNs.
• Angiosarcomas (aka lymphangiosarcomas or hemangiosarcomas) are a unique subtype because they
tend to be more aggressive. Although in general rare sarcoma, they tend to occur mostly in the breast.
Most important PI is grade. High grade go anywhere, even LNs
• Therapy related sarcomas: Well documented but rare. Types similar to above, with higher proportion
being angiosarcomas and associated with RT and chronic lymphedema. Overall prognosis is generally
poor c 5-yr OS 27-35%.
• Phyllodes = “leaflike”(aka cystosarcoma phyllodes): Kind of misnomer bec. rarely have cystic
components and are not true sarcomas by origin or behavior. Grossly, described as a head of
cauliflower, large encapsulated tumor. Prognosis depends on whether it is benign (majority),
malignant or borderline. Even in malignant tumors metastasis is rare (< 5%). Excision +/- RT used.
• Breast Lymphomas: Arise from intramammary LNs and involve axillary LNs at presentation about 30-
40% of time. 10% have bilateral breast involvement. Histology similar to lymphomas elsewhere, thus
½ of pts have DLBCL, with prognosis and treatment similar to lymphomas of the same stage and
histology in other locations. (One variant occurring in the breast is notable. It occurs in young women
who present with an aggressive, high-grade, burkitt-type lymphoma, often assoc. with
pregnancy/lactation, is usu. fatal despite chemo). Unlike the above tumors, surgery is not the mainstay
of therapy. Tx should be based on cell type and ds extent, but at present IFRT alone used for Stage I
indolent ds and chemo/rt for aggressive or widespread ds.

Treatment
• Similar for most nonepithelial tumors except lymphomas.
• Breast sarcomas are treated similarly to soft tissue sarcomas that occur elsewhere in histology,
prognosis, and therapy.
• Surgery: Mainstay of treatment. Goal is wide excision c negative margins. Retrospective studies
have not shown a diff. b/ BCT and mastectomy. Axillary dissection is rarely necessary, but some
degree of sampling may be considered (i.e. angiosarcoma, palpable nodes)
• RT: Rationale for use extrapolated from sarcomas at other sites. Namely, that adjuvant RT
improves LC for both low and high grade tumors but without OS benefit.
• Chemotherapy: Controversial as it is at other sites, but should probably be considered in pts with
high-grade large tumors.

DOUG’S BREAST CANCER REVIEW NOTES  PAGE 26