Available online at www.sciencedirect.com

ScienceDirect

EJSO xx (2013) 1e6 www.ejso.com

Can FDG-PET/CT predict early response to neoadjuvant chemotherapy in breastcancer?

W.P. Andrade a,*, E.N.P. Lima b, C.A.B.T. Os�orio c, M. do Socorro Maciel a, G. Baiocchi d,A.G.V. Bitencourt b, M.F. Fanelli e, A.S. Damascena f, F.A. Soares c

aDepartment of Breast Surgery, A.C. Camargo Cancer Hospital, Sao Paulo, BrazilbDepartment of Radiology and Nuclear Medicine, A.C. Camargo Cancer Hospital, Sao Paulo, Brazil

cDepartment of Pathology, A.C. Camargo Cancer Hospital, Sao Paulo, BrazildDepartment of Gynecologic Oncology, A.C. Camargo Cancer Hospital, Sao Paulo, Brazil

eDepartment of Medical Oncology, A.C. Camargo Cancer Hospital, Sao Paulo, BrazilfA.C. Camargo Cancer Hospital, Sao Paulo, Brazil

Accepted 23 August 2013

Available online - - -

Abstract

Purpose: Neoadjuvant chemotherapy (NAC) in breast cancer is currently used not only for locally advanced tumors, but also for large oper-able tumors when breast preservation is considered. It also provides the opportunity to evaluate chemotherapy tumor response. Our aim wasto correlate the relative change in the standardized uptake value (SUV) of 18F-2-fluoro-2-deoxy-D-glucose positron emission tomography(FDG-PET/CT) with pathologic response after NAC.Methods: We prospectively evaluated 40 patients with invasive ductal breast carcinomas from February 2010 to December 2011. FDG-PET/CTwas performed at baseline and after the second cycle of NAC. All patients underwent surgery after NAC. Pathologic response was eval-uated according to Residual Cancer Burden (RCB) index.Results: The mean age was 41.9 years. Median primary tumor size was 6 cm. Pathologic complete response (pCR) was obtained in 12(30%) patients. The tumor baseline mean maximum SUV (SUVmax), and after second cycle were: 8.97 (sd.4.3) and 4.07 (sd.3.2), respec-tively. The relative change (DSUV) after the second course of NAC was significantly higher for patients with pCR (�81.58%) whencompared to the non-pCR patients (�40.18%) ( p ¼ 0.001). The optimal DSUV threshold that discriminates between pCR and non-pCR was �71.8% (83.3% sensitivity; 78.5% specificity). Moreover, the optimal DSUV threshold to discriminate between NAC respondersand non-responders was �59.1% (68% sensitivity; 75.0% specificity).Conclusions: Our data suggest that the FDG-PET/CT DSUV after the second course of NAC can predict pathological response in ductalbreast carcinomas, and potentially identify a subgroup of non-responding patients for whom ineffective chemotherapy should be avoided.Synopsis: Breast cancer is the most frequently diagnosed cancer in women. The indications for neoadjuvant chemotherapy are increasing.Early information on chemotherapy response is crucial and methods that predict the therapeutic effectiveness might avoid potentially inef-fective chemotherapies in non-responding patients.� 2013 Elsevier Ltd. All rights reserved.

Keywords: Breast cancer; Neoadjuvant chemotherapy; Treatment monitoring; 18F-FDG; PET/CT

* Corresponding author. Departamento de Mastologia, Hospital AC Ca-

margo, Rua Antonio Prudente, 211, 01509-010 S~ao Paulo, Brazil. Tel.:

þ55 11 2189 5110; fax: þ55 11 2114 6072.

E-mail addresses: wesley.andrade@hotmail.com,

wesleypereiraandrade@gmail.com (W.P. Andrade).

0748-7983/$ - see front matter � 2013 Elsevier Ltd. All rights reserved.

http://dx.doi.org/10.1016/j.ejso.2013.08.025

Please cite this article in press as: Andrade WP, et al., Can FDG-PET/CT predict

Oncol (2013), http://dx.doi.org/10.1016/j.ejso.2013.08.025

Introduction

Breast cancer is the most frequently diagnosed cancer inwomen, and it remains the second most important cause ofcancer death in women worldwide.1

Although initially performed only for locally advancedbreast cancer, the indications for neoadjuvant chemo-therapy (NAC) increased and are also currently used for

early response to neoadjuvant chemotherapy in breast cancer?, Eur J Surg

2 W.P. Andrade et al. / EJSO xx (2013) 1e6

patients with operable but large tumors and unfavorable tu-mor/breast size index. NAC not only increases the likeli-hood of breast conservative surgery, but also provides theopportunity to evaluate the chemotherapy responsein vivo.2e4

Approximately 70% of patients demonstrate clinicalresponse after NAC, but only 20e30% achieve pathologiccomplete response (pCR). pCR after NAC is an importantprognostic factor for better disease-free and overall survivalwhen compared to patients with non-pathologic completeresponse (non-pCR).5

Nevertheless, the effectiveness of NAC can only bedetermined by surgery after the completion of chemo-therapy, as only the pathological exam of the resected spec-imen can differentiate pathological responders from non-responders.5,6

Early information on chemotherapy response is crucialand may guide to better therapeutic strategies. Methodsthat predict the therapeutic effectiveness might drive tomore individualized treatments and avoid potentially inef-fective chemotherapies in non-responding patients (non-pR).6e8

Clinical examination and conventional imaging modal-ities, such as mammography, echography, and magneticresonance are methods currently used to evaluate tumorresponse after NAC. However, they are not able to distin-guish between viable tumor and fibrotic scar, as they pro-vide only anatomical and morphological data.9,10

Previous studies have shown that the tumor metabolicchanges starts just after the first cycle of NAC and thesemodifications precede the tumor size reduction. Glucosemetabolism is increased in most malignant tumors, andcan be identified by FDG-PET/CT. FDG-PET/CT noninva-sively evaluates the tumor metabolism and measures the tis-sue proliferation by the standard uptake value (SUV).Therefore, the sequential tumor analyses with FDG-PET/CT may provide a sensitive method to evaluate NACresponse in breast cancer.10e12

Our aim was to correlate the relative change in the SUVof 18F-FDG-PET/CT with the pathologic response afterNAC in patients with breast cancer.

Patients and methods

Study design

This prospective analysis included 40 individuals withductal invasive breast cancer who underwent NAC followedby surgery, at the Department of Breast Surgery, AC Ca-margo Cancer Hospital, from February 2010 to December2011. The Institutional Review Board approved the studyand a written informed consent was obtained from allpatients.

The inclusion criteria were: 1) cT2 with unfavorable tu-mor/breast size index; 2) cT3 and cT4 tumors; 3) cN2/N3disease.13 The exclusion criteria were: pregnancy, breast-

Please cite this article in press as: Andrade WP, et al., Can FDG-PET/CT predict

Oncol (2013), http://dx.doi.org/10.1016/j.ejso.2013.08.025

feeding, age under 18 years, inability to undergo serialFDG-PET/CT and ineligibility for surgery.

The histopathologic diagnosis of breast cancer wasconfirmed by core-needle biopsy. Physical examination,mammography, breast ultrasound, breast magnetic reso-nance imaging, chest and abdomen images, bone scansand FDG-PET/CT studies were obtained for all patientsat baseline.

Three FDG-PET/CT were performed for each patient:the first at baseline before chemotherapy, the second afterthe second chemotherapy cycle (immediately before thethird cycle), and the third before surgical treatment. AfterNAC completion, all patients underwent appropriate onco-logic surgery that included primary breast tumor removaland axillary lymph nodes dissection.

The NAC regimens were divided in: 1) AC e Adria-mycin 60 mg/m2 plus cyclophosphamide 600 mg/m2 eevery 21 days for 4 cycles; 2) AC-T e Adriamycin60 mg/m2 plus cyclophosphamide 600 mg/m2 followedby docetaxel 75 mg/m2 or paclitaxel 80 mg/m2 weekly(for 12 weeks); 3) AC-TH e Adriamycin 60 mg/m2 pluscyclophosphamide 600 mg/m2 followed by docetaxel75 mg/m2 or paclitaxel 80 mg/m2 weekly (for 12 weeks)plus Trastuzumab (first dose 8 mg/kg followed by 6 mg/kg); 4) FEC-TH e Fluorouracil 500 mg/m2, epirubicin100 mg/m2 plus cyclophosphamide 600 mg/m2 followedby docetaxel 75 mg/m2 plus Trastuzumab (first dose8 mg/kg followed by 6 mg/kg); and 5) TCH e Docetaxel75 mg/m2 plus carboplatin AUC 6 plus Trastuzumab (firstdose 8 mg/kg followed by 6 mg/kg).

18F-FDG PET imaging and analysis protocol

The images were obtained on a Gemini PET/CT (PhilipsMedical Systems) with whole-body PET scanner. The pa-tients fasted for at least 6 h prior PET imaging. Serumglucose level was measured and it had to be lower than200 mg/dL for all patients. After that, the patients receivedan intravenous injection of 5.0 megabecquerels per kilo-gram (MBq/Kg) of 18F-FDG, and the first images were ac-quired approximately 90 min after the radiotracer injection.The patients were laid in supine position during the study,and were comfortably positioned on the scanner table withboth arms at their side.

The regions of interest (ROIs) were manually drawn onthe slice with the highest radioactivity concentration. Thelesions were analyzed using the maximum standardized up-take value (SUVmax) method, which was defined as themaximum tissue concentration of FDG in the ROI. TheSUVmax was calculated by the formula: tissue concentra-tion (MBq/g)/injected dose (MBq)/body weight (g). TheSUVmax was measured at the breast tumor and at the ipsi-lateral axillar sites.

An experienced nuclear medicine physician interpretedthe whole-body PET images, and was blinded for patient’shistory, clinical findings, and conventional imaging.

early response to neoadjuvant chemotherapy in breast cancer?, Eur J Surg

3W.P. Andrade et al. / EJSO xx (2013) 1e6

Pathological assessment

Samples from core needle biopsy were studied at diag-nosis for histological and biological characteristics thatincluded histological type, tumor grading, nuclear grading,number of mitoses, hormonal receptor status (estrogen orprogesterone receptor), Ki-67 index, c-erb B2 status and tu-mor intrinsic subtype (luminal-A, luminal-B, HER2 and tri-ple negative). HER-2 overexpression was considered afterimmunohistochemistry if 3þ, or after amplification by fluo-rescence in situ hybridization (FISH).

Tumor intrinsic subtype was classified as: luminal-A(hormonal receptor positive, no HER2 over- expressionand Ki-67 <14%); luminal-B (hormonal receptor positivewith HER2 over expression or Ki-67 >14%); HER 2 (hor-monal receptor negative with HER2 over expression); andtriple negative (when hormonal receptor and HER2 areall negative).14,15

The surgical specimens (breast and axillary lymph no-des) were evaluated by the same pathologist. Pathologicresponse was classified according to residual cancer burden(RCB) protocol: RCB-0 (pCR), RCB-I (minimal residualdisease), RCB-II (moderate residual disease) and RCB-III(extensive residual disease).16 This protocol combines his-topathological components of residual disease (cellularity,overall diameter, number and extent of nodal involvement)and suggests a numerical index for the residual cancerburden.

Pathologic complete response (pCR) was defined as theabsence of malignant cells (no invasive ductal carcinomaand no ductal carcinoma in situ) in neither the primary tu-mor site or in the axillary lymph nodes. It corresponds toRCB-0. Pathologic non-responders (RCB-III) were classi-fied as non-pR.

Statistical analysis

The database was generated in SPSS, version 16.0(SPSS, Inc., Chicago, IL) for Windows. The ManneWhit-ney test was used to compare SUVs between pCR and non-pCR group and between pR and non-pR. For all tests, analpha error up to 5% ( p < 0.05) was considered significant.

Receiver operating curve (ROC) analyses determined theoptimal cut-off values of DSUV that better discriminate pCRfrom non-pCR, and pR from non-pR patients. Sensibility,specificity, positive predictive value (PPV), negative predic-tive value (NPV) and accuracy were calculated.

Results

The patients’ clinical, pathological data and pCR rateare summarized in Table 1.

The mean age was 41.9 years (range: 27e64). Mediantumor size was 6 cm (range: 2.5e17). The initial NACwas uniformly administered, as nearly all patients (95%)received adriamycin plus cyclophosphamide regimen

Please cite this article in press as: Andrade WP, et al., Can FDG-PET/CT predict

Oncol (2013), http://dx.doi.org/10.1016/j.ejso.2013.08.025

between baseline and second FDG-PET/CT (first and sec-ond course of NAC). Regarding HER-2 positive patients,only one did not receive trastuzumab during NAC, but itwas administered after surgery for 12 months.

Pathological response after NAC

After completion of NAC, all patients underwent breastsurgery (mastectomy or breast-conserving surgery) andaxillary lymphadenectomy. We observed pCR (RCB-0) in12 (30%) patients. Furthermore, RCB-I, RCB-II, andRCB-III were found in 5 (12.5%), 11 (27.5%), and 12(30%) patients, respectively.

FDG-PET/CT and DSUV threshold value

For all patients, the mean baseline FDG-PET/CT of theprimary tumor SUVmax was 8.97 (standarddeviation � sd.4.3). Moreover, the mean SUVmax of thepatients with pCR (11.26, sd. 4.18) was significantly higherthan with non-pCR (7.98, sd. 4.04) ( p ¼ 0.040) (Table 2).

In the second FDG-PET/CT, the mean SUVmax of theprimary tumor was 4.07 (sd. 3.2). At this time, the meanSUVmax of the patients with pCR (2.73, sd. 1.13) wassignificantly lower than the non-pCR (4.64, sd. 3.63)( p ¼ 0.048).

Furthermore, the relative change (DSUV) after the sec-ond course of NAC (Fig. 1) was significantly higher forthe pCR group (�81.58%) when compared to the non-pCR group (�40.18%) ( p ¼ 0.001).

The optimal DSUV threshold that discriminate betweenpCR and non-pCR was: �71.8% (83.3% sensitivity; 78.5%specificity; 91.7% positive predictive value; 80% negativepredictive value; and 62.5% accuracy) (Fig. 2). Moreover,the optimal DSUV threshold that discriminate betweenNAC responders (pR) and non-responders (non-pR/RCB-III) was �59.1% (68% sensitivity; 75.0% specificity; 50%positive predictive value; and 70% negative predictivevalue; and 86.3% accuracy) (Table 3).

Discussion

In breast cancer, the major indication of NAC is thetreatment of locally advanced tumors. Although onlyapproximately 30% of patients achieve pCR, it is an impor-tant prognostic factor for better disease-free and overallsurvival.17e19 Yet, pathologic analysis of the breast andaxillary lymph node remains as the standard method fortherapy response evaluation.6,10,16

Thus, it is important to look not only for more effectivetherapies, but also for new methods that can better monitorthe therapy efficacy. Clinical decisions such as continua-tion, modification, or cessation of chemotherapy may beinfluenced by early identification of drug response. Further-more, the treatment may be more individualized and

early response to neoadjuvant chemotherapy in breast cancer?, Eur J Surg

Table 1 (continued )

Variables N� of patients (%)

NAC regimens in the 1st and 2nd courses

AC 38 95.0

4 W.P. Andrade et al. / EJSO xx (2013) 1e6

ineffective chemotherapies may be avoided, as well as theirtoxicities and worsened quality of life.6e8,10

FDG-PET/CT performed at baseline and after 1 or 2 cy-cles of NAC could help to identify patients that response to

Table 1

Clinical and pathological characteristics of the 40 patients with breast can-

cer submitted to neoadjuvant chemotherapy.

Variables N� of patients (%)

N 40 100.0

Age (years)

<50 26 65.0

�50 14 35.0

Tumor size (T)

�5 cm 13 32.5

>5 cm 27 67.5

Lymph node involvement (N)

Negative 6 15.0

Positive 34 85.0

Histological type

Ductal 40 100.0

Lobular 0 0

Tumor grading

Grade I 2 5.0

Grade II 14 35.0

Grade III 23 57.5

Missing 1 2.5

Nuclear grading

Score I 0 0

Score II 5 12.5

Score III 35 87.5

Number of mitoses

Score I 18 45.0

Score II 10 25.0

Score III 12 30.0

Estrogen receptor status

Negative 16 40.0

Positive 24 60.0

Progesterone receptor status

Negative 21 52.5

Positive 19 47.5

HER-2/neu-overexpressing

Negative 28 70.0

Positive 12 30.0

Tumor intrinsic subtype

Luminal A 5 12.5

Luminal B 20 50.0

HRþ HER 2þ 6 15.0

HRþ KI 67 > 14% 14 35.0

HER 2 6 15.0

Triple negative 9 22.5

pCR rate 12 30.0

pCR by tumor intrinsic subtype

Luminal A 1/5 20.0

Luminal B 4/20 20.0

HRþ HER 2þ 3/6 50.0

HRþ KI 67 > 14% 1/14 7.0

HER 2 3/6 50.0

Triple negative 4/9 44.4

NAC regimens

AC 2 5.0

AC-T 27 67.0

AC-TH 9 22.5

FAC-TH 1 2.5

TCH 1 2.5

FAC 1 2.5

TCH 1 2.5

NAC e neoadjuvant chemotherapy; 1st e first; 2nd e second; A e adria-

mycin;Ce cyclophosphamide;Te taxane (docetaxel or paclitaxel);He tras-

tuzumab; F e fluorouracil; E � epirubicin.

Please cite this article in press as: Andrade WP, et al., Can FDG-PET/CT predict

Oncol (2013), http://dx.doi.org/10.1016/j.ejso.2013.08.025

treatment. Some authors suggested that the best time toperform the early evaluation was after the second cycleand immediately before the third cycle of NAC.6,20e22

Berriolo-Riedinger et al.5 analyzed 47 patients submit-ted to NAC and found that relative decrease in FDG uptake(DSUV) of �60% after the first cycle predicted the pCRwith a sensitivity of 91%, specificity of 86% and accuracyof 87%. The DSUV after the first course of NAC was alsosignificantly higher in the pCR group than in the non-pCRgroup ( p < 0.0001). Buchbender et al.23 analyzed 26 pa-tients that had a baseline FDG-PET/CT scan, and in thisstudy the subsequent FDG-PET/CT were performed afterthe second cycle of NAC. A higher optimal cut-offDSUV of �88% was suggested to discriminate betweenpCR and non-pCR.

Table 2

FDG PET/CT results at baseline and after 2 cycles of NAC, along with

change in SUV (in percent).

Group Mean Standard

deviation

p

Baseline

FDG PET/CT (SUVmax)

Non-pCR 7.98 4.04 0.040

pCR 11.26 4.18

After Course 2 NAC

FDG PET/CT (SUVmax)

Non-pCR 4.64 3.63 0.048

pCR 2.73 1.13

Decrease in SUV

DSUV (%)

Non-pCR �40.18 39.51 0.001

pCR �81.58 13.92

SUVmax e maximum standardized uptake value; NAC e neoadjuvant

chemotherapy; non-pCR e non-pathologic complete response; pCR epathologic complete response.

Figure 1. Change in SUV following chemotherapy.

early response to neoadjuvant chemotherapy in breast cancer?, Eur J Surg

Figure 2. Receiver operating curve (ROC) analyses for the prediction of pathological response of breast cancer using FDG PET. (A) Differentiation between

complete pathological response pCR and partial pathological response (non-pCR). (B) Differentiation between pathological responders (pR) and pathological

non-responders (non-pR).

5W.P. Andrade et al. / EJSO xx (2013) 1e6

In our series, we performed a baseline FDG-PET/CT andthe second FDG-PET/CT after 2 cycles of NAC. We foundthat DSUV after the second cycle of NAC was significantlyhigher for the pCR group when compared to the non-pCRgroup ( p ¼ 0.001). We additionally demonstrated that theoptimal threshold of DSUV that discriminate betweenpCR and non-pCR after the second cycle of NAC was�71.8%, with a positive predictive value of 91.7%, andnegative predictive value of 80.0% (Fig. 2).

However, the optimal DSUV to identify the non-responder group is still unknown. Only few studies ad-dressed this issue and nearly all evaluated the DSUV after2 cycles of NAC and found DSUV that ranges from�40% to �55% (Table 3). Schelling et al.20 found in 22 pa-tients a DSUV cutoff of �55%, with an accuracy of 91%.Rousseau et al.21 study included 64 patients and suggestedthat DSUV of �40% was the optimal threshold to separateresponders from non-responders, with a positive predictivevalue, negative predictive value and accuracy of 89%,

Table 3

Studies evaluating different thresholds for optimal differentiation between respon

after 2 cycles.

Study n Tumor size (cm) Histological

scale

Martoni et al. (2010)22 34 NAa Miller-Payne

Kumar et al. (2009)7 23 Mean 7.10 cm

(range: 4.10e12.0)

Sataloff

Rousseau et al. (2006)21 64 Median 4.0 cm

(range: 1.0e10.0)

Sataloff

Schwarz-Dose et al. (2006)12 104 Mean 4.9 cm

(range: 3.0e12.0)

Honkoop

Schelling et al. (2000)20 22 Median 5.5 cm

(range: 3.5e12.0)

Honkoop

Andrade et al. (2012) 40 Mean 6.7 cm

(range: 2.5e17.0)

Symmans

NAC e neoadjuvant chemotherapy; PPV e positive predictive value; NPV e nea Tumor size related by TNM classification.

Please cite this article in press as: Andrade WP, et al., Can FDG-PET/CT predict

Oncol (2013), http://dx.doi.org/10.1016/j.ejso.2013.08.025

85.5% and 87%, respectively. Martoni et al.22 studied 34patients and the DSUV threshold with optimal negative pre-dictive value to predict pathologic response was �50%. Ku-mar et al.7 also found a DSUV cutoff of �50% withaccuracy of 87%. Schwarz-Dose et al.12 published a multi-center trial with 104 patients where 81 were evaluated afterthe second cycle, and DSUVof �55% predicted pathologicresponse with a sensitivity of 69%, specificity of 63%, andnegative predictive value of 89%.

Our study suggests the optimal threshold of DSUVs of�59.1% to discriminate between pR and non-pR (orRCB-III) after the second cycle of chemotherapy, with pos-itive predictive value of 50.0%, negative predictive value of70.0% and accuracy of 86.3% (Fig. 2). The patients withnon-pathologic response constitute the group that shouldavoid ineffective chemotherapy regimes and further beingconsidered for treatment strategy change, which mayinclude other chemotherapy regimens or earlier oncologicsurgery.

ders and nonresponders during NAC for breast cancer with FDG PET/CT

Optimal SUV

threshold

Sensitivity Specificity PPV NPV Accuracy

�50% 100% 30% 27% 100% 44%

�50% 93% 75% 88% 86% 87%

�40% 89% 95% 89% 85% 87%

�55% 69% 63% 32% 89% 64%

�55% 83% 94% NA NA 91%

�59.1% 68% 75% 50% 70% 86.3%

gative predictive value; NA e not available.

early response to neoadjuvant chemotherapy in breast cancer?, Eur J Surg

6 W.P. Andrade et al. / EJSO xx (2013) 1e6

The differences between the SUV thresholds in thestudies may be explained by the lack of consensus on path-ological response definition. Although the pathologic anal-ysis after NAC completion is the standard method forassessing the treatment response, several different criteriawere described.4,12,20,24e26 We used the criteria describedby Symmans et al.,16 where pCR was defined as no malig-nant cells (no invasive ductal carcinoma and no ductal car-cinoma in situ) present in sections from both primary tumorand axillary lymph node.

Overall, our series is comparable in size to themost signif-icant studies on this topic and contributes valuable data. Inour trial we have a homogeneous sample with all patientswith ductal invasive breast cancer, NAC regimens were uni-form, and we used an objective criteria for tumor response.16

In conclusion, our data suggest that the FDG-PET/CTDSUV after the second course of NAC can predict patho-logical response in breast cancer, and potentially identifya subgroup of non-responding patients for whom ineffec-tive chemotherapy should be avoided.

Conflict of interest statement

To the best of our knowledge, none of authors have norelevant financial relationships or any other conflict ofinterest.

References

1. Jemal A, Bray F, Center MM, Ferlay J, Ward E, Forman D. Global

cancer statistics. CA Cancer J Clin 2011;61:69–90.

2. Lee MC, Newman LA. Management of patients with locally advanced

breast cancer. Surg Clin North Am 2007 Apr;87(2):379–98. ix. Review.

3. Mathew J, Asgeirsson KS, Cheung KL, Chan S, Dahda A,

Robertson JF. Neoadjuvant chemotherapy for locally advanced breast

cancer: a review of the literature and future directions. Eur J Surg On-

col 2009 Feb;35(2):113–22. Review.

4. Groheux D, Giacchetti S, Espi�e M, Rubello D, Moretti JL, Hindi�e E.

Early monitoring of response to neoadjuvant chemotherapy in breast

cancer with 18F-FDG PET/CT: defining a clinical aim. Eur J Nucl

Med Mol Imaging 2011 Mar;38(3):419–25.

5. Berriolo-Riedinger A, Touzery C, Riedinger JM, et al. [18F]FDG-PET

predicts complete pathological response of breast cancer to neoadju-

vant chemotherapy. Eur J Nucl Med Mol Imaging 2007 Dec;34(12):

1915–24.

6. Avril N, Sassen S, Roylance R. Response to therapy in breast cancer.

J Nucl Med 2009 May;50(Suppl. 1):55S–63S.

7. Kumar A, Kumar R, Seenu V, et al. The role of 18F-FDG PET/CT in

evaluation of early response to neoadjuvant chemotherapy in patients

with locally advanced breast cancer. Eur Radiol 2009 Jun;19(6):

1347–57.

8. Jacobs MA, Ouwerkerk R, Wolff AC, et al. Monitoring of neoadjuvant

chemotherapy using multiparametric, 23Na sodium MR, and multimo-

dality (PET/CT/MRI) imaging in locally advanced breast cancer.

Breast Cancer Res Treat 2011 Jul;128(1):119–26.

Please cite this article in press as: Andrade WP, et al., Can FDG-PET/CT predict

Oncol (2013), http://dx.doi.org/10.1016/j.ejso.2013.08.025

9. Kim SJ, Kim SK, Lee ES, Ro J, Kang S. Predictive value of [18F]FDG

PET for pathological response of breast cancer to neo-adjuvant

chemotherapy. Ann Oncol 2004 Sep;15(9):1352–7.

10. Wang Y, Zhang C, Liu J, Huang G. Is 18F-FDG PET accurate to pre-

dict neoadjuvant therapy response in breast cancer? A meta-analysis.

Breast Cancer Res Treat 2012 Jan;131(2):357–69.

11. Smith IC, Welch AE, Hutcheon AW, et al. Positron emission tomog-

raphy using [(18)F]-fluorodeoxy-D-glucose to predict the pathologic

response of breast cancer to primary chemotherapy. J Clin Oncol

2000 Apr;18(8):1676–88.

12. Schwarz-Dose J, Untch M, Tiling R, et al. Monitoring primary sys-

temic therapy of large and locally advanced breast cancer by using

sequential positron emission tomography imaging with [18F]fluoro-

deoxyglucose. J Clin Oncol 2009 Feb 1;27(4):535–41.

13. Edge SB, Byrd DR, Carducci MA, et al. AJCC cancer staging manual.

7th ed. New York: Springer; 2009.

14. Sorlie T, Perou CM, Tibshirani R, et al. Gene expression patterns of

breast carcinomas distinguish tumor subclasses with clinical implica-

tions. Proc Natl Acad Sci U S A 2001;98:10869–74.

15. Cheang MC, Chia SK, Voduc D, et al. Ki67 index, HER2 status, and

prognosis of patients with luminal B breast cancer. J Natl Cancer Inst

2009;101:736–50.

16. Symmans WF, Peintinger F, Hatzis C, et al. Measurement of residual

breast cancer burden to predict survival after neoadjuvant chemo-

therapy. J Clin Oncol 2007;25:4414–22.

17. Fisher B, Bryant J, Wolmark N, et al. Effect of preoperative hemother-

apy on the outcome of women with operable breast cancer. J Clin On-

col 1998;16:2672–85.

18. Bear HD, Anderson S, Smith RE, et al. Sequential preoperative or

postoperative docetaxel added to preoperative doxorubicin plus cyclo-

phosphamide for operable breast cancer: National Surgical Adjuvant

Breast and Bowel Project Protocol B-27. J Clin Oncol 2006;24:

2019–27.

19. Von Minckwitz G, K€ummel S, Vogel P, et al. Intensified neoadjuvant

chemotherapy in early-responding breast cancer: phase III randomized

GeparTrio study. J Natl Cancer Inst 2008;100:552–62.

20. Schelling M, Avril N, N€ahrig J, et al. Positron emission tomography

using [(18)F]fluorodeoxyglucose for monitoring primary chemo-

therapy in breast cancer. J Clin Oncol 2000;18:1689–95.

21. Rousseau C, Devillers A, Sagan C, et al. Monitoring of early response

to neoadjuvant chemotherapy in stage II and III breast cancer by [18F]

fluorodeoxyglucose positron emission tomography. J Clin Oncol 2006

Dec 1;24(34):5366–72.

22. Martoni AA, Zamagni C, Quercia S, et al. Early (18)F-2-fluoro-2-

deoxy-d-glucose p�ositron emission tomography may identify a subset

of patients with estrogenreceptor-positive breast cancer who will not

respond optimally to preoperative chemotherapy. Cancer 2010 Feb;

116(4):805–13.

23. Buchbender C, Kuemmel S, Hoffmann O, et al. FDG-PET/CT for the

early prediction of histopathological complete response to neoadjuvant

chemotherapy in breast cancer patients: initial results. Acta Radiol

2012 Jul;53(6):628–36.

24. Sahoo S, Lester SC. Pathology of breast carcinomas after neoadjuvant

chemotherapy: an overview with recommendations on specimen pro-

cessing and reporting. Arch Pathol Lab Med 2009;133:633–42.

25. Sataloff DM, Mason BA, Prestipino AJ, Seinige UL, Lieber CP,

Baloch Z. Pathologic response to induction chemotherapy in locally

advanced carcinoma of the breast: a determinant of outcome. J Am

Coll Surg 1995 Mar;180(3):297–306.

26. Honkoop AH, van Diest PJ, de Jong JS, et al. Prognostic role of clin-

ical, pathological and biological characteristics in patients with locally

advanced breast cancer. Br J Cancer 1998 Feb;77(4):621–6.

early response to neoadjuvant chemotherapy in breast cancer?, Eur J Surg