Preoperative 18F-FDG PET/CT predicts disease-free survival in patients with primary invasive ductal breast cancer

Abstract

Background

Previous studies have correlated the maximum standardized uptake value (SUV_max) of breast cancer lesions with histological and biological characteristics such as tumor size, histologic grade, or hormonal receptor expression status. However, controversy remains concerning the prognostic value of SUV_max in breast cancer.

Purpose

To determine if the SUV_max of a tumor on 18F-fluorodeoxyglucose positron emission tomography (18F-FDG PET) is associated with disease-free survival in patients with primary invasive ductal breast cancer.

Material and Methods

The institutional review board of our hospital approved this retrospective study. From 2009 to 2011, 508 women (mean age, 53.6 years; age range, 26–85 years) with newly diagnosed invasive ductal breast cancer who had undergone preoperative 18F-FDG PET/CT followed by surgery were identified. Clinicopathological variables and FDG uptake quantified by SUV_max were analyzed. The Cox proportional hazards model was used to evaluate the association between SUV_max and disease-free survival after controlling for clinicopathological parameters.

Results

There were 21 recurrences at a median follow-up of 46 months. The mean SUV_max of the primary tumor was significantly higher in patients with a recurrence than those who remained disease-free (9.5 ± 3.5 vs. 6.6 ± 4.2, P < 0.001). A receiver operating characteristic curve indicated that a SUV_max of 5.95 was the optimal cut-off value to predict disease-free survival. Multivariate analysis identified a high SUV_max (≥5.95) and high T stage as independent significant variables associated with worse disease-free survival.

Conclusion

A high primary tumor SUV_max on 18F-FDG PET/CT was an independent factor associated with worse disease-free survival in patients with primary invasive ductal breast cancer.

Keywords

Breast neoplasms disease-free survival positron emission tomography fluorodeoxyglucose F18 multivariate analysis

Introduction

The advent of mammographic screening in conjunction with therapeutic improvements has led to a significant increase in breast cancer survival over the last few decades (1). Nevertheless, breast cancer is the leading cause of cancer death among women worldwide (2). In addition, locoregional recurrences are unfortunately common in breast cancer and about 20% of patients diagnosed with local disease will develop metastases (3 –5). A better understanding of tumor biology is needed to develop rational, individualized treatments.

18F-fluorodeoxyglucose positron emission tomography (18F-FDG PET) has the ability to identify the increased tissue glucose metabolism of malignant tumors. 18F-FDG PET is a unique, non-invasive method of tissue characterization that is important for the diagnosis and follow-up of cancer (6). In breast cancer, tumor metabolism measured by maximum standardized uptake value (SUV_max) varies considerably among primary sites (7). Thus, the relationship between tumor biology and glucose metabolism has been investigated for several years. Some studies demonstrate that the FDG uptake in breast cancers is significantly correlated with traditional prognostic factors including tumor size, histological grade, and proliferation index (8 –10). Furthermore, FDG uptake is correlated with tumor subtypes, which are known to be associated with tumor prognosis (11 –14). In more recent studies, the association between tumor blood flow and glucose metabolism has been evaluated in patients with breast cancer. These studies have shown that FDG uptake is associated with tumor angiogenesis, as measured by immunohistochemistry bioassay, and suggested that 18F-FDG PET may aid in the development of an optimal biological and functional profile of a tumor as well as individualized breast cancer therapies (15,16).

However, the prognostic value of 18F-FDG PET in predicting clinical outcomes has not been assessed fully (17 –20). Therefore, the purpose of this study was to evaluate whether tumor SUV_max on 18F-FDG PET/CT was associated with disease-free survival in patients with primary invasive ductal breast cancer.

Material and Methods

Study population

The Institutional Review Board of our hospital approved this retrospective study, and the requirement for informed consent was waived. A review of medical records from January 2009 to December 2011 at our institution identified 618 consecutive women with newly diagnosed invasive ductal breast cancer who had undergone whole-body 18F-FDG PET/CT for staging and subsequently underwent curative surgery with a clear resection margin. Among these 618 patients, 110 were excluded: 37 women underwent surgical excision or ultrasound-guided vacuum-assisted excision before 18F-FDG PET/CT, 52 had no focal uptake on 18F-FDG PET/CT, 14 had distant metastasis at the time of initial diagnosis, and seven did not have available immunohistochemical data. The remaining 508 breast cancer patients comprised our study population. Of these 508 patients, 351 women had presented with breast symptoms and 157 women had screening-detected cancer. The mean age of the patients was 53.6 years (age range, 26–85 years). The mean interval between 18F-FDG PET/CT and surgery was 14.5 days (range, 2–35 days).

18F-FDG PET/CT

All patients fasted for at least 8 h and had serum glucose levels <120 mg/dL at the time of 18F-FDG injection (5.2 MBq/kg of body weight) and were examined in the supine position with arms elevated. 18F-FDG PET/CT scans were obtained from the skull base to the upper thigh 1 h after an intravenous administration of 18F-FDG using a Biograph instrument (Siemens Medical Solution, Hoffman Estates, IL, USA), which is based on a full-ring PET scanner and a dual-slice helical CT scanner. A CT scan was obtained for attenuation correction (tube voltage 120 kVp, current intensity 80–170 mAs, and slice thickness of 3 mm) prior to the PET scan without the use of contrast agent. A PET scan was then acquired in three-dimensional mode for 3 min per bed position. PET data were obtained using a high-resolution whole body scanner with an axial field of view of 21.6 cm. Attenuation correction was performed for all patients with iterative reconstruction.

The PET/CT images were interpreted by two nuclear medicine physicians (both with more than 10 years of experience in PET/CT imaging) in consensus. Regions of interest were manually placed over the area of maximal activity on slices of the primary breast tumors in attenuation-corrected images, and the SUV_max within the region of interest was obtained. The SUV_max was calculated as follows: SUV_max = maximum activity concentration in region of interest (MBq/g)/[injected dose (MBq)/body weight (g)].

Clinicopathological evaluation

Clinicopathological information, including age, tumor size, axillary nodal status, histologic grade, nuclear grade (assessed using the Elston and Ellis method (21)), and type of surgery, was obtained from electronic medical records or surgical pathology reports. The expression status of the estrogen receptor (ER), progesterone receptor (PR), and human epidermal growth factor receptor 2 (HER2) were determined based on the surgical specimen. ER and PR positivity were defined by an Allred score ≥3 (22), which were expressed as the sum of the proportion score and intensity score of positively stained tumor cells, and is in the range of 0–8. The intensity of HER2 staining was scored as 0, 1+, 2+, or 3+ (23). Tumors with a 3+ score were deemed HER2 positive, and tumors with a 0 or 1+ score were deemed HER2 negative. Tumors with a 2+ score were submitted for fluorescence in situ hybridization (FISH) testing to determine HER2 status. If the ratio of the HER2 score to the chromosome 17 probe signals was greater than 2.2, the tumor was considered HER2 positive.

Breast cancers were classified into four subtypes based on immunohistochemistry or FISH findings of ER, PR, HER2, and Ki-67 as follows: luminal A (ER positive and/or PR positive, HER2 negative, Ki-67 <14%), luminal B (ER positive and/or PR positive, HER2 positive, Ki-67 ≥14%), triple-negative (ER negative, PR negative, and HER2 negative), and HER2 positive (ER negative, PR negative, and HER2 positive) (24).

Statistical analysis

The relationships between 18F-FDG uptake of the primary tumor and the clinicopathological parameters were calculated using the Kruskal-Wallis and Mann-Whitney U test with Bonferroni correction.

The primary outcome was disease-free survival, defined as the time interval from the date of surgery to the date of first disease recurrence. Breast cancer recurrence was defined as the first reappearance, including either local (limited to ipsilateral breast or chest wall and/or axillary, infraclavicular, or supraclavicular lymph nodes) or distant (metastasis to other parts of the body). The last date of data collection was 31 January 2014, and women for whom no event had occurred or who were lost to follow-up were censored accordingly.

The optimal cut-off value was determined using a receiver operating characteristic (ROC) curve analysis to dichotomize SUV_max values for survival analysis. Survival curves were estimated using Kaplan-Meier analysis, and the survival differences were analyzed by the log-rank test. The clinicopathological variables (age, tumor stage, axillary lymph node, histologic grade, nuclear grade, ER status, PR status, HER2 status, and type of surgery) and the SUV_max were analyzed in a univariate analysis of disease-free survival with a Cox proportional hazards model. Variables with P values less than 0.1 in the univariate analysis were entered into the multivariate model for the prediction of outcome.

The SPSS software (version 18.0, SPSS, Chicago, IL, USA) was used for all the data analyses except the ROC curve analysis, which was performed using the MedCalc software (version 10.3.0.0, MedCalc Software, Mariakierke, Belgium). A P value of less than 0.05 indicated statistical significance.

Results

Patients’ characteristics and recurrence outcome

The majority of patients (66.7%, 339 of 508) underwent breast-conserving surgery; the remainder (33.3%, 169 of 508) underwent total mastectomy. None of the patients received primary systemic therapy before surgery. All patients who had breast-conserving surgery received adjuvant radiation therapy. The mean tumor size at surgical histology was 2.5 ± 2.0 cm (range, 0.2–12.5 cm). The mean SUV_max of the 508 tumors was 5.7 ± 3.2 (range, 1.0–24.5). Tumor SUV_max values according to clinicopathological parameter are shown in Table 1. The mean SUV_max values were 5.3 ± 3.0 in luminal A, 7.1 ± 4.3 in luminal B, 9.5 ± 5.2 in triple-negative, 8.1 ± 4.0 in HER2 positive, with significant differences among the groups (P < 0.001) (Table 1) (Fig. 1a).

Fig. 1.

SUV_max according to tumor subtypes (a) and disease recurrence status (b) using box and whisker plots. Whiskers denote the maximum and minimum data points or 1.5× the intequartile range. Open circles represent outlier and asterisks represent extreme values.

Table 1.

Relationship between SUV_max and clinicopathological parameters.

Variables	n (%)	SUV_max	Range	P value
Age (years)				0.036
<45	108 (21.3)	7.8 ± 5.1	1.1, 24.5
≥45	400 (78.7)	6.4 ± 3.8	1.2, 23.5
T stage				<0.001
T1	256 (50.4)	5.1 ± 3.3	1.1, 22.1
T2	209 (41.1)	8.2 ± 4.3	1.2, 24.5
T3	43 (8.5)	9.1 ± 4.6	1.5, 18.6
Lymph node status				<0.001
Negative	290 (57.1)	6.1 ± 4.0	1.1, 24.5
Positive	218 (43.9)	7.6 ± 4.2	1.2, 22.1
Histologic grade				<0.001
Grade 1	123 (24.2)	4.6 ± 2.6	1.2, 16.2
Grade 2	241 (47.4)	6.2 ± 3.5	1.1, 17.0
Grade 3	144 (28.3)	9.5 ± 4.8	1.3, 24.5
Nuclear grade				<0.001
Grade 1	19 (3.7)	3.4 ± 1.8	1.2, 7.8
Grade 2	248 (48.8)	5.4 ± 3.0	1.1, 16.9
Grade 3	241 (47.4)	8.4 ± 4.7	1.5, 24.5
ER status				<0.001
Negative	173 (34.1)	8.9 ± 4.7	1.5, 24.5
Positive	335 (65.9)	5.6 ± 3.3	1.1, 17.9
PR status				<0.001
Negative	212 (41.7)	8.4 ± 4.7	1.2, 24.5
Positive	296 (58.3)	5.6 ± 3.2	1.1, 17.1
HER2 status				0.001
Negative	371 (73.0)	6.4 ± 4.1	1.1, 24.5
Positive	137 (27.0)	7.6 ± 4.1	1.2, 22.1
Tumor subtype				<0.001
Luminal A	271 (53.3)	5.3 ± 3.0	1.1, 16.9
Luminal B	78 (15.4)	7.1 ± 4.3	1.2, 17.9
Triple-negative	100 (19.7)	9.5 ± 5.2	1.5, 24.5
HER2 positive	59 (11.6)	8.1 ± 4.0	2.2, 22.1
Type of surgery				<0.001
BCS	339 (66.7)	6.3 ± 3.9	1.2, 24.3
Mastectomy	169 (33.3)	7.7 ± 4.4	1.1, 24.5
Disease group				<0.001
Recurrence	21 (4.1)	9.5 ± 3.5	4.1, 16.9
Disease-free	487 (95.9)	6.6 ± 4.2	1.1, 24.5

ER, estrogen receptor; HER2, human epidermal growth factor receptor 2; PR, progesterone receptor.

During a median follow-up period of 46 months (range, 16–60 months), 21 patients (4.1%, 21 of 508) experienced a recurrence. There were 10 local, seven distant, and four combined local and distant recurrences. The median disease-free survival was 33 months (range, 10–59 months). Patients with a cancer recurrence had a significantly higher primary tumor SUV_max than those without recurrence (9.5 ± 3.5 vs. 6.6 ± 4.2; P < 0.001) (Table 1) (Fig.1b).

Disease-free survival analysis

The ROC curve demonstrated that a SUV_max of 5.95 was the optimal cut-off for predicting disease-free survival (area under the curve, 0.734; standard error, 0.042) (Fig. 2). Based on SUV_max, patients were divided into two groups, and the disease-free survival was compared between patients with high (≥5.95) and low (<5.95) SUV_max. Kaplan-Meier survival analysis indicated that patients with a high SUV_max (≥5.95) had a significantly lower disease-free survival rate than those with a low SUV_max (<5.95) (P < 0.001) (Fig. 3).

Fig. 2.

The ROC curve for predicting disease-free survival.

Fig. 3.

Disease-free survival curves of patients according to SUV_max for the primary tumor.

Univariate and multivariate analyses

Univariate analysis demonstrated that young age (<45) (hazard ratio (HR) = 2.415, 95% confidence interval (CI) = 1.017–5.735, P = 0.046), high tumor stage (T3) (HR = 4.103, 95% CI = 1.585–10.623, P = 0.004), and high SUV_max (≥5.95) (HR = 7.340, 95% CI = 2.159–24.958, P = 0.001) were significant predictors of cancer recurrence (Table 2). Variables with a P value less than 0.1 in the univariate analysis were entered into the multivariate model. Because ER and PR status were highly correlated, only ER status was included in the multivariate analysis. Multivariate analysis using the Cox proportional hazard model revealed that a high primary tumor SUV_max (≥5.95) (adjusted HR = 5.541, 95% CI = 1.524–20.141, P = 0.009) and high tumor stage (T3) (adjusted HR = 3.504, 95% CI = 1.283–9.564, P = 0.014) remained independent variables associated with worse disease-free survival (Table 3).

Table 2.

Univariate Cox proportional hazards analysis of variables with disease-free survival outcomes.

Variables	Patients (n = 508)	Events (n = 21)	Hazard ratio	95% CI	P value
Age (years)					0.046
<45	108	6	2.415	1.017, 5.735
≥45	400	15	1
Tumor stage					0.004
T3	43	6	4.103	1.585, 10.623
T1 or T2	465	15	1
Axillary lymph node					0.475
Positive	218	8	0.724	0.299, 1.756
Negative	290	13	1
Histologic grade					0.082
Grade 3	144	8	2.194	0.905, 5.317
Grades 1 or 2	364	13	1
Nuclear grade					0.805
Grade 3	241	9	0.897	0.377, 2.131
Grades 1 or 2	267	12	1
ER status					0.061
Negative	173	9	2.298	0.963, 5.483
Positive	335	12	1
PR status					0.051
Negative	212	8	2.401	0.995. 5.796
Positive	296	13	1
HER2 status					0.58
Negative	371	17	1.361	0.457, 4.054
Positive	137	4	1
SUV_max					0.001
≥5.95	235	18	7.34	2.159, 24.958
<5.95	273	3	1
Type of surgery					0.409
BCS	339	7	1.471	0.588, 3.680
Mastectomy	169	14	1

BCS, breast-conserving surgery; ER, estrogen receptor; HER2, human epidermal growth factor receptor 2; PR, progesterone receptor.

Table 3.

Mutivariate Cox proportional hazards analysis of disease-free survival outcomes.

Variables	Adjusted hazard ratio	95% CI	P value
Age (years)		0.892, 5.293	0.088
<45	2.172
≥45	1
Tumor stage		1.283, 9.564	0.014
T3	3.504
T1, T2	1
Histologic grade		0.465, 3.406	0.651
Grade 3	1.258
Grade 1, 2	1
ER status		0.635, 4.318	0.302
Negative	1.656
Positive	1
SUV_max		1.524, 20.141	0.009
≥5.95	5.541
<5.95	1

ER, estrogen receptor.

Discussion

In this study, patients whose tumors had high FDG uptake on pretreatment 18F-FDG PET/CT, as indicated by a high SUV_max, were at increased risk of cancer recurrence compared to those with low FDG uptake tumors. On multivariate Cox analysis, a high SUV_max (adjusted HR = 5.541, 95% CI = 1.524–20.141, P = 0.009) of the primary tumor was one of the most important independent factors influencing disease-free survival of breast cancer patients, along with high tumor stage (adjusted HR = 3.504, 95% CI = 1.283–9.564, P = 0.014).

Several authors have investigated the usefulness of 18F-FDG PET/CT in predicting the clinical outcomes of breast cancer and proposed several SUV_max cut-off values (17 –20). In particular, Ueda et al. separated low SUV_max (≤4.0) and high SUV_max (>4.0) groups in a series of 152 patients (3 ductal carcinoma in situ, 136 invasive ductal carcinoma, 9 invasive lobular carcinoma, and 4 other special subtype), and found that tumors with a high SUV_max had increased 10-year relapse and 10-year mortality rates (17). Song et al. used a SUV_max cut-off of 6.6 in 55 patients with invasive ductal breast cancer (20). However, these previous studies were based on small numbers of cancer cases or tumors of varied histological subtypes. The mean SUV_max of breast cancer is known to differ among histologic subtype, and the histologic subtype of breast cancer is also a prognostic factor (7,10). Thus, we only analyzed tumors from 508 primary breast cancer patients with an invasive ductal subtype and concluded that a SUV_max cut-off of 5.95 was optimal for predicting disease-free survival from a primary tumor. Furthermore, while 18F-FDG PET is not currently recommended for routine use in primary breast cancer (25), at our institution, 18F-FDG PET/CT was routinely performed preoperatively for staging in patients with invasive breast cancer during our study period. Therefore, the strengths of our study were both the large sample size and enrollment of patients with a specific pathology. We believe our study will help to clarify the prognostic utility of FDG uptake for patients with primary invasive ductal carcinoma.

Breast cancer is a heterogeneous disease. Several biological features – such as ER, PR, and HER2 status – are routinely used to identify patients at increased risk of recurrence or those likely to benefit from systemic therapies. Triple-negative and HER2 positive breast cancers have more aggressive biological features and are associated with poor clinical outcomes when compared to luminal subtypes (26 –29). In our study, the mean SUV_max values of the triple-negative and HER2 positive breast cancers were 9.5 ± 5.2 and 8.1 ± 4.0, respectively, which was higher than luminal A and luminal B subtypes (5.3 ± 3.0 and 7.1 ± 4.3, respectively). This finding is compatible with a previous report (14) of FDG uptake according to the immunohistochemically defined breast cancer subtypes recommended by the St. Gallen International Expert Consensus (24). In this previous study, triple-negative (9.8 ± 6.0) and HER2 positive (7.4 ± 4.7) tumors had 1.67- and 1.27-fold higher SUV_max values than luminal A (4.7 ± 3.5) tumor after adjustment for confounding factors (14). We believe FDG uptake could be useful in predicting tumor aggressiveness and therefore, tumor prognosis and treatment response may be predicted preoperatively using 18F-FDG PET/CT.

Our study had several limitations. First, this was a retrospective study with a relatively short-term follow-up period (median duration, 46 months; range, 16–60 months). We are preparing a prospective study to confirm the currently reported results, and to determine if preoperative PET/CT has predictive prognostic value. Second, all invasive ductal breast cancers with positive FDG uptake on PET/CT were included in this study, regardless of size. The SUV_max of small tumors may be underestimated due to the partial volume effect (30,31). Finally, we did not evaluate the status of adjuvant chemotherapy or adjuvant hormonal therapy, which could have affected the survival outcomes.

In conclusion, a high primary tumor SUV_max on preoperative 18F-FDG PET/CT and high tumor stage were independent significant factors associated with worse disease-free survival in patients with primary invasive ductal breast cancer. Although a SUV_max cut-off of 5.95 was identified as the most valuable predictive factor in the present study, large, prospective studies with longer follow-up are warranted for confirmation. The results of our study suggest that the SUV_max of a primary tumor could be used as a biomarker for risk stratification of patients with primary invasive ductal breast cancer.

Footnotes

Conflict of interest

None declared.

Funding

This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors.

References

Berry

Cronin

Plevritis

. Effect of screening and adjuvant therapy on mortality from breast cancer. N Engl J Med 2005; 353: 1784–1792.

Jemal

Bray

Center

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

Van Dongen

Voogd

Fentiman

. Long-term results of a randomized trial comparing breast-conserving therapy with mastectomy: European Organization for research and treatment of cancer 10801 trial. J Natl Cancer Inst 2000; 92: 1143–1150.

Clarke

Collins

Darby

. Effects of radiotherapy and of differences in the extent of surgery for early breast cancer on local recurrence and 15-year survival: an overview of the randomized trials. Lancet 2005; 366: 2087–2106.

Brewster

Hortobagyi

Broglio

. Residual risk of breast cancer recurrence 5 years after adjuvant therapy. J Natl Caner Inst 2008; 100: 1179–1183.

Jadvar

Alavi

Gambhir

. 18F-FDG uptake in lung, breast, and colon cancers: molecular biology correlates and disease characterization. J Nucl Med 2009; 50: 1820–1827.

Avril

Menzel

Dose

. Glucose metabolism of breast cancer assessed by 18F-FDG PET: histologic and immunohistochemical tissue analysis. J Nucl Med 2001; 42: 9–16.

Bos

van Der Hoeven

van Der Wall

. Biologic correlates of (18) fluorodeoxyglucose uptake in human breast cancer measured by positron emission tomography. J Clin Oncol 2002; 20: 379–387.

Groheux

Giacchetti

Moretti

. Correlation of high 18F-FDG uptake to clinical, pathological and biological prognostic factors in breast cancer. Eur J Nncl Med Mol Imaging 2011; 38: 426–435.

10.

Buck

Schirrmeister

Kühn

. FDG uptake in breast cancer: correlation with biological and clinical prognostic parameters. Eur J Nucl Med Mol Imaging 2002; 29: 1317–1323.

11.

Basu

Chen

Tchou

et al . Comparison of triple-negative and estrogen receptor-positive/progesterone receptor-positive/HER2-negative breast carcinoma using quantitative fluorine-18 fluorodeoxyglucose/positron emission tomography imaging parameters: a potentially useful method for disease characterization. Cancer 2008; 112: 995–1000.

12.

Osborne

Port

Gonen

. 18F-FDG PET of locally invasive breast cancer and association of estrogen receptor status with standardized uptake value: microarray and immunohistochemical analysis. J Nucl Med 2010; 51: 543–550.

13.

Specht

Kurland

Montgomery

. Tumor metabolism and blood flow as assessed by positron emission tomography varies by tumor subtype in locally advanced breast cancer. Clin Cancer Res 2010; 16: 2803–2810.

14.

Koo

Park

Kang

. 18F-FDG uptake in breast cancer correlates with immunohistochemically defined subtypes. Eur Radiol 2014; 24: 610–618.

15.

Groves

Shastry

Rodriquez-Justo

. ¹⁸F-FDG PET and biomarkers for tumour angiogenesis in early breast cancer. Eur J Nucl Med Mol Imaging 2011; 38: 46–52.

16.

Vinh-Hung

Everaert

Lamote

. Diagnostic and prognostic correlates of preoperative FDG PET for breast cancer. Eur J Nucl Med Mol Imaging 2012; 39: 1618–1627.

17.

Ueda

Tsuda

Asakawa

. Clinicopathological and prognostic relevance of uptake level using 18F-fluorodeoxyglucose positron emission tomography/computed tomography fusion imaging (18F-FDG PET/CT) in primary breast cancer. Jpn J Clin Oncol 2008; 38: 250–258.

18.

Song

Lee

Jeong

. 18F-FDG uptake by metastatic axillary lymph nodes on pretreatment PET/CT as a prognostic factor for recurrence in patients with invasive ductal breast cancer. J Nucl Med 2012; 53: 1337–1344.

19.

Inoue

Yutani

Taguchi

. Preoperative evaluation of prognosis in breast cancer patients by [(18)F]2-Deoxy-2-fluoro-D-glucose-positron emission tomography. J Cancer Res Clin Oncol 2004; 130: 273–278.

20.

Song

Hong

Lee

. Prognostic value of primary tumor uptake on F-18 FDG PET/CT in patients with invasive ductal breast cancer. Nucl Med Mol Imaging 2011; 45: 117–124.

21.

Elston

Ellis

. Pathological prognostic factors in breast cancer. I. The value of histological grade in breast cancer: experience from a large study with long-term follow-up. Histopathology 1991; 19: 403–410.

22.

Allred

Harvey

Berardo

. Prognostic and predictive factors in breast cancer by immunohistochemical analysis. Mod Pathol 1998; 11: 155–168.

23.

Moeder

Giltnane

Harigopal

. Quantitative justification of the change from 10% to 30% for human epidermal growth factor receptor 2 scoring in the American Society of Clinical Oncology/College of American Pathologists guidelines: tumor heterogeneity in breast cancer and its implications for tissue microarray based assessment of outcome. J Clin Oncol 2007; 25: 5418–5425.

24.

Goldhirsch

Wood

Coates

. Strategies for subtypes-dealing with the diversity of breast cancer: highlights of the St. Gallen international expert consensus on the primary therapy of early breast cancer 2011. Ann Oncol 2011; 22: 1736–1747.

25.

Fletcher

Djulbegovic

Soares

. Recommendations on the use of 18F-FDG PET in oncology. J Nucl Med 2008; 49: 480–508.

26.

Dent

Trudeau

Pritchard

. Triple-negative breast cancer: clinical features and patterns of recurrence. Clin Cancer Res 2007; 13: 4429–4434.

27.

Gianni

Dafni

Gelber

. Treatment with trastuzumab for 1 year after adjuvant chemotherapy in patients with HER2-positive early breast cancer: a 4-year follow-up of randomized controlled trial. Lancet Oncol 2011; 12: 236–244.

28.

Malorni

Shetty

De Angelis

. Clinical and biological features of triple-negative breast cancer in a large cohort of patients with long-term follow-up. Breast Cancer Res Treat 2012; 136: 795–804.

29.

Voduc

Cheang

Tyldesley

. Breast cancer subtype and the risk of local and regional relapse. J Clin Oncol 2010; 28: 1684–1691.

30.

Kumar

Chauhan

Zhuang

. Clinicopathologic factors associated with false negative FDG-PET in primary breast cancer. Breast Cancer Res Treat 2006; 98: 267–274.

31.

Soret

Bacharch

Buvat

. Partial-volume effect in PET tumor imaging. J Nucl Med 2007; 48: 932–945.