Comparative study of the value of dual tracer PET/CT in evaluating breast cancer

Abstract

The present study was conducted to assess the relationship between tumor uptake and pathologic findings using dual‐tracer PET/computed tomography (CT) in patients with breast cancer. Seventy‐four patients with breast cancer (mean age 54 years) who underwent ¹¹ C‐choline and 2‐[¹⁸ F]fluoro‐2‐deoxy‐d‐glucose (¹⁸ F‐FDG) PET/CT prior to surgery on the same day were enrolled in the present study. Images were reviewed by a board‐certified radiologist and two nuclear medicine specialists who were unaware of any clinical information and a consensus was reached. Uptake patterns and measurements of dual tracers were compared with the pathologic findings of resected specimens as the reference standard. Mean (±SD) tumor size was 5.9 ± 3.2 cm. All primary tumors were identified on ¹⁸ F‐FDG PET/CT and ¹¹ C‐choline PET/CT. However, ¹⁸ F‐FDG PET/CT demonstrated focal uptake of the primary tumor with (n = 38; 51%) or without (n = 36; 49%) diffuse background breast uptake. Of the pathologic findings, multiple logistic regression analysis revealed an independent association between fibrocystic change and diffuse background breast uptake (odds ratio [OR] 8.57; 95% confidence interval [CI] 2.86–25.66; P < 0.0001). Tumors with higher histologic grade, nuclear grade, structural grade, nuclear atypia, and mitosis had significantly higher maximum standardized uptake values (SUV _max) and tumor‐to‐background ratios (TBR) for both tracers. Multiple logistic regression analysis revealed that only the degree of mitosis was independently associated with a high SUV _max (OR 7.45; 95%CI 2.21–25.11; P = 0.001) and a high TBR (OR 5.41; 95%CI 1.13–25.96; P = 0.035) of ¹¹ C‐choline PET/CT. In conclusion, ¹¹ C‐choline may improve tumor delineation and reflect tumor aggressiveness on PET/CT in patients with breast cancer.

Positron emission tomography/computed tomography (PET/CT) with the glucose analog 2‐[¹⁸F]fluoro‐2‐deoxy‐d‐glucose (¹⁸F‐FDG) is recognized as an important tool in initial tumor evaluation, including staging, in the evaluation of treatment response, and in the assessment of recurrent disease for breast cancer.1, 2 It has been reported that PET/CT adds incremental diagnostic confidence to PET in 60% of patients and in >50% of regions with increased ¹⁸F‐FDG uptake.3 Tatsumi et al.4 concluded that PET/CT was preferable in evaluating breast cancer lesions in view of the level of diagnostic confidence that it allows. Regardless of the exact type of PET/CT fusion technique, ¹⁸F‐FDG uptake in non‐malignant conditions often leads to high background uptake on breast imaging.5

Histological changes are the cause of considerable variations and false‐positive findings on breast imaging. Fibrocystic changes (FCC) are the most common of these conditions that can affect the assessment of imaging features on mammography6, 7 and MRI.8, 9 Similarly, there is evidence in the literature that ¹⁸F‐FDG PET and accelerated glucose metabolism as a result of FCC lead to false‐positive findings and difficulty in determining the boundary of specificity.10

Choline is an essential component of the cell membrane and choline uptake is upregulated by choline kinase‐α, which catalyses the phosphorylation of choline.11, 12 In mammary epithelial cells, levels of phosphocholine metabolites increase due to overexpression of choline kinase‐α, which is regulated by the mitogen‐activated protein kinase (MAPK) pathway.11, 12, 13 Recent clinical studies in patients with breast carcinoma undergoing molecular‐targeted therapy suggest that ¹¹C‐choline uptake is 10‐fold higher in aggressive breast carcinoma phenotypes and that the uptake of ¹¹C‐choline on PET is correlated with tumor grade.13 Thus, ¹¹C‐choline is considered a promising radiotracer for the evaluation of breast cancer in the clinical setting prior to treatment.

Although both data from ¹⁸F‐FDG and ¹¹C‐choline PET/CT allow more precise evaluation of the primary breast cancer, direct comparisons of these two tracers in breast cancer have not been made. In the present study, we sought to confirm and extend previous findings of ¹¹C‐choline PET/CT studies by investigating the association between histological findings and the results of ¹⁸F‐FDG PET/CT investigations in patients with breast cancer.

Materials and Methods

Patients

Seventy‐four patients (mean age 54 years; range 25–89 years) with breast carcinoma were enrolled in the present retrospective dual PET/CT study between March 2008 and March 2010. Patients were eligible for inclusion in the study if they met the following criteria: (i) performance status 0 or 1; (ii) no concomitant malignancy; (iii) histologically proven breast carcinoma diagnosed by biopsy at least 1 month before; and (iv) no history of hormone therapy. All patients were required to provide written informed consent. A regimen of 5‐fluorouracil, epirubicin, and cyclophosphamide (FEC) plus paclitaxel was used as neoadjuvant chemotherapy in 32 patients (43%). As a rule, hormone therapy was introduced after completion of imaging studies if needed. Our institutional review board (National Cancer Center Hospital, Tokyo, Japan) approved the present study, which complied with the Health Insurance Portability and Accountability Act. The clinical records of all patients were available for review. All patients received surgery after imaging studies.

Phantom study

A phantom study of PET/CT was performed prior to the clinical study at two institutions to clarify the optimum conditions for data acquisition and to ensure quality control.14 Studies were performed with a whole‐body PET/CT scanner (Aquiduo PCA‐7000B; Toshiba Medical Systems, Tochigi, Japan). The CT component of the scanner has a 16‐row detector. We used an NEMA image quality (IQ) phantom (NU 2‐2001) for cross calibration, because this type of phantom is used in many institutions and data regarding the estimation of the optimum time are available. The radioactivity concentration of the background was set at 2.6 ± 0.2 kBq/mL ¹⁸F‐FDG, similar to that in clinical settings. The radioactivity concentration of the hot portion was fourfold greater than that of the background. Data were collected over a period of 2–5 min in the dynamic acquisition mode and for 30 min in the static acquisition mode. The data acquired, including normalization data, cross‐calibration data, blank scan data, and transmission data, were assessed for visual inspection, phantom noise equivalent count (NEC_phantom), percentage contrast (Q_H,10 mm) and percentage background variability (N_10 mm). The preferred parameters pertinent to the clinical condition were NEC_phantom > 10.4 (counts), N_10 mm < 6.2%, and Q_H10 mm/N_10 mm > 1.9%. After a review of the data analyses, the optimum conditions for the PET/CT were determined as follows: data acquisition, 180 s for one bed; field‐of‐view, 500 mm; iteration, 4; subset, 14; matrix size,128 × 128; filter, Gaussian 8 mm in full width at half maximum; reconstruction, ordered‐subsets expectation maximization (OSEM).

Data acquisition

¹¹C‐Choline was synthesized using a commercially available module, as described by Hara et al.15 Prior to the ¹¹C‐choline PET/CT study, patients fasted for at least 6 h. Immediately after they had evacuated their bladder, patients were placed in a supine, arm‐up position. For the PET/CT, low‐dose CT data were first acquired at 120 kVp using an autoexposure control system (beam pitch 0.875 or 1 and 1.5 or 2 mm × 16‐row mode). Data acquisition was performed for each patient from the top of the skull to the mid‐thigh. Patients maintained normal shallow respiration during the three‐dimensional acquisition of CT scans. No iodinated contrast material was administered. Acquisition of emission scans from the head to the mid‐thigh was started 5 min after intravenous administration of a mean ¹¹C‐choline dose of 475 MBq (range 469–491 MBq). The ¹⁸F‐FDG PET/CT study was performed 1 h after the ¹¹C‐choline PET/CT study in all patients. Patients received an intravenous injection of 311 MBq (range 197–397 MBq) ¹⁸F‐FDG with an uptake phase at 64 ± 5 min.

Image interpretation

Dedicated software (Vox‐base SP1000 workstation; J‐MAC Systems, Sapporo, Japan) was used to review all PET, CT, and coregistered PET/CT images in all standard planes. Images were analyzed visually and quantitatively by two independent reviewers, who recorded their findings after reaching a consensus. A region of interest (ROI) was outlined within areas of increased uptake and measured on each slice. When the lesion was extensively heterogeneous, the ROI was set so as to cover all the components of the lesion. The diffuse pattern of breast was assigned to the breast that shows homogeneous accumulation greater than aortic blood except for the primary lesion. For quantitative interpretations, the standardized uptake value (SUV) was determined according to the standard formula, with activity in the ROI recorded as Bq/mL per injected dose (Bq) per weight (kg), but time decay correction for whole‐body image acquisition was not performed. The maximum SUV (SUV_max) was recorded using the maximum pixel activity within the ROI. The tumor‐to‐background ratio (TBR) was calculated with reference to uptake in the contralateral breast.

Pathologic analysis

All patients underwent surgery. Each tumor was staged according to the TNM classification of the International Union against Cancer.16 Resected specimens were fixed in 10% buffered formalin and embedded in paraffin wax. Then, 4‐μm sections were obtained in a plane perpendicular to the long axis of the breast. Paraffin‐embedded microslices were stained with H&E. Tissue grading, nuclear grading, and structural grading were done using the grading system of Elston and Ellis.17 Estrogen receptor (ER) and progesterone receptor status was evaluated using the H‐scoring system of McCarty et al.18 Human epidermal growth factor‐2 (HER‐2/neu) was evaluated by immunostaining with 4B5 primary antibody. Evaluation of the primary lesion was based on the following pathologic findings: FCC, differentiation, subtype, location, diameter of the invasive component, diameter of the non‐invasive component, ratio of the invasive component in the tumor (%), tissue grading, nuclear grading, structural grading, nuclear atypia, mitosis, necrosis, fat invasion, cutaneous invasion, muscular invasion, ER status, progesterone receptor status, and HER‐2/neu status. In the present study, “non‐invasive component” referred to ductal carcinoma in situ (DCIS).

Statistical analysis

The Chi‐squared test or Fisher's exact probability test were used to compare pathologic findings associated with PET/CT findings. In addition, the Wald test and 95% confidence intervals (CI) were used to evaluate the statistical significance of individual variables. To determine relationships of SUV and TBR between the two tracers, we used Spearman rank correlation. Comparisons of mean values between groups were made using Student's t‐test or analysis of variance (anova) with Bonferroni's adjustment for multiple comparisons. Parsimonious univariate and multivariate logistic regression models were used to measure independent associations with PET/CT findings. Statistical tests used a two‐sided significance level of 0.05. Statistical analyses were performed using PASW Statistics 19 (IBM, Tokyo, Japan).

Results

In all, 74 patients completed the study procedures. The demographic data for all patients are given in Table 1. There were 66 patients (89%) with invasive tumors, 60 of which were ductal carcinoma and six lobar carcinoma. Eight patients (11%) had non‐invasive ductal carcinoma.

Table 1.

Patient demographics

Age (years)	54 ± 13 (24–78)
Tumor side
Right	44 (59)
Left	30 (41)
Tumor size (cm)	5.9 ± 3.2 (1.8–12.0)
Main location
Medial upper quadrant	8 (11)
Medial lower quadrant	8 (11)
Lateral upper quadrant	46 (62)
Lateral lower quadrant	6 (8)
Central	2 (3)
Invasive tumor	66 (89)
Non‐invasive tumor	8 (11)

Data are given as the mean ± SD, with the range in parentheses, or as the number of patients in each group with percentages in parentheses.

All primary tumors were identified on ¹⁸F‐FDG PET/CT and ¹¹C‐choline PET/CT (Fig. 1). The SUV_max of ¹¹C‐choline PET/CT was significantly lower than that of ¹⁸F‐FDG PET/CT (P = 0.002; Table 2). Conversely, the TBR of ¹¹C‐choline PET/CT was significantly higher than that of ¹⁸F‐FDG PET/CT (P < 0.0001; Table 2). Using ¹⁸F‐FDG PET/CT, focal uptake of the primary tumor with (n = 38 [51%]; Fig. 2) or without (n = 36 [49%]) diffuse background breast uptake was demonstrated. Conversely, ¹¹C‐choline PET/CT showed only focal uptake of the primary tumor in all patients. There were significant differences between patients with or without diffuse background breast uptake on ¹⁸F‐FDG PET/CT for TBR of ¹⁸F‐FDG (Table 2). There was no interaction between ¹¹C‐choline uptake and background breast uptake patterns on ¹⁸F‐FDG PET/CT. There were significant differences for the diameter of the non‐invasive component and the percentage invasive component between patients with and without diffuse background breast uptake on ¹⁸F‐FDG PET/CT (Table 2).

Figure 1.

Results for a 51‐year‐old woman with invasive ductal carcinoma of the left breast. (a,c) ¹⁸ F‐fludeoxyglucose (FDG) PET/computed tomography (CT) images (fusion image: a; PET alone: (c) reveal a focal hypermetabolic focus in the primary tumor (arrows). The maximum standardized uptake value (SUV _max) was 5.5 and the tumor‐to‐background ratio (TBR) was 47.0. (b,d) Transverse ¹¹ C‐choline PET/CT images (fusion image: b; PET alone: d) also reveal a focal hypermetabolic focus in the primary tumor (arrows). The SUV _max was 5.0 and the TBR was 137.5. On microscopy, the tumor contained 100% invasive component.

Table 2.

Computed tomography (CT)/PET measurements and pathologic components with or without diffuse background breast uptake on ¹⁸ F‐fludeoxyglucose PET/CT

	Total	With diffuse uptake	Without diffuse uptake	P‐value
11C‐Choline uptake of tumor
SUVmax (g/mL)	3.7 ± 2.9	3.6 ± 3.5	3.8 ± 2.0	0.789
TBR	8.0 ± 6.0	7.7 ± 9.9	8.3 ± 9.8	0.709
18F‐FDG uptake of tumor
SUVmax (g/mL)	4.4 ± 3.1	4.6 ± 3.6	4.2 ± 2.5	0.571
TBR	3.7 ± 2.7	3.2 ± 2.4	4.5 ± 2.9	0.016
Diameter of invasive tumor (cm)	4.1 ± 3.5	4.2 ± 4.0	4.0 ± 3.0	0.800
Diameter of non‐invasive tumor (cm)	1.8 ± 2.3	2.6 ± 2.9	0.9 ± 1.0	0.002
% Invasive component	66.0 ± 36.5	54.0 ± 41.4	78.6 ± 25.4	0.003

FDG, fludeoxyglucose; SUV_max, maximum standardized uptake value; TBR, tumor‐to‐background ratio.

Figure 2.

Results for a 47‐year‐old woman with invasive scirrhous carcinoma of the left breast. (a,c) ¹⁸ F‐fludeoxyglucose (FDG) PET/computed tomography (CT) images (fusion image: a; PET alone: c) reveal a focal hypermetabolic focus (arrows) of the primary tumor with diffuse background breast uptake. The maximum standardized uptake value (SUV _max) was 5.4 and the tumor‐to‐background ratio (TBR) was 35.7. (b,d) Transverse ¹¹C‐choline PET/CT images (fusion image: b; PET alone: (d) reveal only a focal hypermetabolic focus in the primary tumor (arrows). The SUV _max was 5.0 and the TBR was 125.0. On microscopy, the tumor contained 15% invasive component. Diffuse fibrocystic changes were found in the background breast.

The pathologic findings and background breast uptake patterns on ¹⁸F‐FDG PET/CT are listed in Table 3. Patients with diffuse background breast uptake had significantly different values for percentage invasive component, FCC, necrosis, and triple negative tumor compared with patients without diffuse background breast uptake. There were no significant differences between the two groups in histologic grade, nuclear grade, structural grade, nuclear atypia, mitosis, fat invasion, or cutaneous invasion. Nor were there any significant differences in hormone receptor status between the two groups, specifically HER‐2/neu, ER, and progesterone receptors. Only FCC showed an independent association with diffuse background breast uptake on multiple logistic regression analysis (OR 8.57; 95% CI 2.86–25.66; P < 0.0001).

Table 3.

Pathologic characteristics and background breast uptake on ¹⁸ F‐fludeoxyglucose PET/computed tomography

	No. patients		P‐value
	With diffuse uptake	Without diffuse uptake
Invasive component
>30%	20	32	0.001
<30%	18	4
Fibrocystic change
Present	24	6	<0.0001
Absent	14	30
Histologic grade
1 or 2	16	22	0.102
3	22	14
Nuclear grade
1 or 2	16	22	0.102
3	22	14
Structural grade
1 or 2	14	18	0.253
3	24	18
Nuclear atypia
1 or 2	16	20	0.247
3	22	16
Mitosis
1 or 2	26	20	0.254
3	12	16
Necrosis
Present	22	8	0.002
Absent	16	28
Fat invasion
Present	22	24	0.437
Absent	16	12
Cutaneous invasion
Present	4	4	0.163
Absent	32	32
HER‐2/neu receptor
Positive	22	14	0.102
Negative	16	22
Estrogen receptor
Positive	28	21	0.163
Negative	10	15
Progesterone receptor
Positive	28	21	0.163
Negative	10	15
Triple negative
Yes	4	11	0.032
No	34	25

There was a modest correlation between the diameter of the invasive tumor and SUV_max (P < 0.0001) or TBR (P = 0.006) on ¹⁸F‐FDG PET/CT (Table 4). Similar trends were found between the diameter of the invasive tumor and SUV_max (P < 0.0001) and TBR (P < 0.0001) on ¹¹C‐choline PET/CT (Table 5). The TBR on ¹¹C‐choline PET/CT also showed a modest correlation with the percentage invasive component (P = 0.047). The diameter of the non‐invasive tumor was not correlated with SUV_max or TBR on either ¹⁸F‐FDG or ¹¹C‐choline PET/CT.

Table 4.

Relationship between ¹⁸ F‐fludeoxyglucose uptake and invasive or non‐invasive tumor components

	18F‐FDG
	SUVmax	P‐value	TBR	P‐value
Diameter of invasive tumor	0.381	<0.0001	0.318	0.006
Diameter of non‐invasive tumor	−0.058	0.625	−0.14	0.234
% Invasive component	0.126	0.286	0.189	0.089

FDG, fludeoxyglucose; SUV_max, maximum standardized uptake value; TBR, tumor‐to‐background ratio.

Table 5.

Relationship between ¹¹C‐choline uptake and invasive or non‐invasive tumor components

	11C‐Choline
	SUVmax	P‐value	TBR	P‐value
Diameter of invasive tumor	0.425	<0.0001	0.537	<0.0001
Diameter of non‐invasive tumor	0.038	0.745	−0.066	0.575
% Invasive component	0.125	0.29	0.232	0.047

SUV_max, maximum standardized uptake value; TBR, tumor‐to‐background ratio.

Pathologic characteristics and tracer uptake are summarized in Table 6. Tumors with a higher histologic grade, nuclear grade, structural grade, nuclear atypia, and mitosis showed significantly higher SUV_max and TBR for both¹⁸F‐FDG and ¹¹C‐choline PET/CT. Tumors without expression of hormone receptors, including ER and progesterone receptors, and triple negative tumors showed significantly higher SUV_max and TBR for both¹⁸F‐FDG and ¹¹C‐choline PET/CT. Tumors expressing FCC and fat invasion were more likely to have high SUV_max and TBR on ¹¹C‐choline PET/CT, but these differences were not identified in the TBR of ¹⁸F‐FDG PET/CT. In addition, tumors with necrosis and cutaneous invasion were found to have greater SUV_max and TBR only on ¹¹C‐choline PET/CT. There was no significant association between the SUV_max or TBR and the percentage of invasive component or the HER‐2/neu status for both tracers. After adjusting for age and tumor size, multiple logistic regression analysis revealed that the degree of mitosis was independently associated with high SUV_max (OR 7.45; 95% CI 2.21–25.11; P = 0.001) and high TBR (OR 5.41; 95% CI; 1.13–25.96; P = 0.035) of ¹¹C‐choline PET/CT.

Table 6.

Pathologic characteristics and tracer uptake

	11C‐Choline				18F‐FDG
	SUVmax	P‐value	TBR	P‐value	SUVmax	P‐value	TBR	P‐value
% Invasive component		0.979		0.432		0.934		0.79
>30%	3.7 ± 2.3		8.5 ± 5.6		4.4 ± 3.1		3.9 ± 2.4
<30%	3.7 ± 3.5		7.4 ± 6.4		4.4 ± 3.1		4.1 ± 3.2
Fibrocystic change		<0.0001		<0.0001		0.002		0.429
Present	5.4 ± 3.5		11.4 ± 6.7		5.7 ± 3.5		4.3 ± 2.2
Absent	2.5 ± 1.6		5.7 ± 4.0		3.5 ± 2.5		3.8 ± 3.1
Histologic grade		<0.0001		<0.0001		<0.0001		<0.0001
1 or 2	2.2 ± 1.1		4.4 ± 2.5		2.9 ± 2.0		2.9 ± 2.3
3	5.3 ± 3.3		11.8 ± 6.2		6.1 ± 3.3		5.2 ± 2.8
Nuclear grade		<0.0001		<0.0001		<0.0001		<0.0001
1 or 2	2.2 ± 1.1		4.4 ± 2.5		2.9 ± 2.0		2.9 ± 2.3
3	5.3 ± 3.3		11.8 ± 6.2		6.1 ± 3.3		5.2 ± 2.8
Structural grade		<0.0001		<0.0001		<0.0001		0.011
1 or 2	2.3 ± 1.5		4.7 ± 3.5		3.0 ± 2.4		3.1 ± 2.7
3	4.8 ± 3.2		10.5 ± 6.3		5.5 ± 3.2		4.7 ± 2.7
Nuclear atypia		<0.0001		<0.0001		<0.0001		0.012
1 or 2	2.1 ± 1.1		4.2 ± 2.4		2.8 ± 2.0		2.8 ± 2.3
3	5.2 ± 3.3		11.6 ± 6.1		5.9 ± 3.2		5.1 ± 2.7
Mitosis		<0.0001		<0.0001		<0.0001		<0.0001
1 or 2	2.1 ± 1.1		4.5 ± 2.5		2.8 ± 1.8		2.6 ± 2.1
3	6.4 ± 3.0		13.8 ± 5.5		7.1 ± 3.0		6.3 ± 2.1
Necrosis		0.046		0.001		0.051		0.529
Present	4.5 ± 3.2		10.6 ± 7.1		5.3 ± 3.3		4.2 ± 2.6
Absent	3.1 ± 2.6		6.2 ± 4.3		3.8 ± 2.9		3.8 ± 2.9
Fat invasion		0.003		0.002		0.024		0.061
Present	4.5 ± 3.3		9.6 ± 6.5		5.0 ± 3.4		4.4 ± 2.8
Absent	2.4 ± 1.5		5.3 ± 3.6		3.4 ± 2.2		3.2 ± 2.6
Cutaneous invasion		0.004		<0.0001		0.133		0.706
Present	6.4 ± 3.1		14.8 ± 7.3		6.0 ± 4.7		4.3 ± 2.2
Absent	3.4 ± 2.7		7.2 ± 5.3		4.2 ± 2.8		3.9 ± 2.8
HER‐2/neu receptor		0.53		0.772		0.518		0.766
Positive	3.5 ± 2.6		8.2 ± 6.0		4.2 ± 2.7		4.1 ± 3.1
Negative	3.9 ± 3.2		7.8 ± 6.0		4.7 ± 3.5		3.9 ± 2.5
Estrogen receptor		<0.0001		<0.0001		<0.0001		<0.0001
Positive	2.3 ± 1.4		5.4 ± 3.4		3.1 ± 1.8		2.9 ± 2.9
Negative	6.3 ± 3.3		13.1 ± 6.6		7.0 ± 3.5		6.1 ± 2.5
Progesterone receptor		<0.0001		<0.0001		<0.0001		<0.0001
Positive	2.4 ± 1.5		5.6 ± 3.7		3.1 ± 1.9		2.9 ± 2.9
Negative	6.2 ± 3.4		12.7 ± 6.8		7.0 ± 3.5		6.1 ± 2.5
Triple negative		<0.001		<0.0001		<0.0001		0.002
Yes	6.7 ± 3.3		12.9 ± 6.5		7.4 ± 3.8		6.2 ± 2.1
No	2.9 ± 2.2		6.7 ± 5.2		3.7 ± 2.4		3.4 ± 2.7

FDG, fludeoxyglucose; SUV_max, maximum standardized uptake value; TBR, tumor‐to‐background ratio.

Discussion

The present study examined the association between dual‐tracer uptake and histological background in breast cancer. Despite positive correlations for SUV_max or TBR with ¹⁸F‐FDG and ¹¹C‐choline, mitosis was found to be correlated with ¹¹C‐choline uptake only, which reflects tumor aggressiveness reported in the previous study of patients with breast cancer.19 The results also reveal that diffuse background breast uptake on ¹⁸F‐FDG PET/CT depends on FCC and this pattern of uptake was not identified in any patients on ¹¹C‐choline PET/CT. Our findings suggest that ¹¹C‐choline may be feasible for the imaging of breast cancer particularly for patients with underlying FCC in whom mammography and ¹⁸F‐FDG PET/CT are limited.

Our observation of a positive correlation between mitosis and ¹¹C‐choline uptake supports results reported in previous studies.13, 19 This phenomenon was not affected by the underlying histological background because comparative correlation coefficients of SUV_max and TBR were similar on ¹¹C‐choline PET/CT. Furthermore, the association between mitosis and ¹⁸F‐FDG uptake was not observed, regardless of positive correlation between ¹⁸F‐FDG and ¹¹C‐choline uptake. This discrepancy in terms of mitosis and tracer uptake in our patients is presumably caused by differences in the degree of tracer uptake.

The present study demonstrated that there were significant differences in the diameter of the non‐invasive component and the percentage invasive component between patients with and without diffuse background breast uptake on ¹⁸F‐FDG PET/CT. However, the SUV_max and TBR of both tracers were similar between patients with or without diffuse background breast uptake on ¹⁸F‐FDG PET/CT. These results suggest that the non‐invasive component of breast cancer, which refers to the DCIS component in the present study, cannot be depicted by both tracers. These findings are consistent with that of another study that suggested DCIS could not be precisely visualized by PET.1 Neubauer et al.20 suggested that the DCIS component could be detected by dynamic contrast‐enhanced MRI, but the specificity was unfavorable because of an overlap in kinetic curve appearance. A major limitation of previous studies, as well as the present study, is that whole‐body PET/CT scanners were used to evaluate primary lesion of the breast.

Fibrocystic changes are the most common diffuse benign condition of the breast related to changes in responses to estrogen and progesterone. The histology of FCC varies considerably and includes cysts, apocrine metaplasia, fibrosis, calcification, ductal hyperplasia, adenosis, and fibroadenomatous changes.21, 22 Because of its diverse appearances and kinetic features, FCC is major cause of false‐positive findings on MRI.23, 24, 25 As for PET studies, Yutani et al.26 have previously explored the ¹⁸F‐FDG uptake of FCC in 38 patients with breast cancer, providing evidence that diffuse ¹⁸F‐FDG uptake caused by accompanying FCC obscures uptake by the primary tumor. Palmedo et al.27 have confirmed that FCC is a major cause of reduced specificity in the detection of primary breast cancers on ¹⁸F‐FDG PET. Furthermore, Kole et al.28 compared the detectability of primary lesions between ¹⁸F‐FDG PET and ¹¹C‐tyrosine PET in patients with breast cancer and concluded that the visual assessment and delineation of the primary tumor were complicated only on ¹⁸F‐FDG PET when the contralateral breast tissue served as the control because FCC is a bilateral disease. As far as we were aware, the present study is the first that has been designed to evaluate the primary lesion of breast cancer using the dual tracers of ¹⁸FDG and ¹¹C‐choline. However, considering the high incidence of FCC, PET tracers including ¹¹C‐tyrosine and ¹¹C‐choline in addition to ¹⁸F‐FDG are more likely to fulfill specificity expectations.

The exact mechanism of ¹¹C‐choline uptake by tumor cells is largely unknown; however, ¹¹C‐choline has been proposed as a marker of the extracellular receptor kinase/MAPK pathway, exhibits significant uptake in tumor tissues, and is regarded as a favorable tracer for breast cancer.13 ¹¹C‐Choline uptake may occur via a choline‐specific transporter protein that is overexpressed in the cell membranes of breast cancer. ¹¹C‐Choline is phosphorylated by choline kinase, which is upregulated in tumor cells for the synthesis of phosphatidylcholine, and is retained within tumor cells.11, 12 Phosphatidylcholine is an essential component of cell membranes and is involved in the modulation of transmembrane signaling by carcinogenesis. Therefore, ¹¹C‐choline metabolism is accelerated in cell proliferation and is enhanced with increasing tumor grade of breast cancer. In the present study, tumors with higher histologic grade, nuclear grade, structural grade, nuclear atypia, and mitosis showed significantly higher SUV_max and TBR for ¹¹C‐choline PET/CT. These results are in accord with those of previous in vivo and in vitro studies.19, 29

¹¹C‐Choline PET/CT has been introduced as feasible method for the evaluation of breast cancer. In the present study, tumors without ER or progesterone receptors and triple negative tumors showed greater uptake of ¹¹C‐choline compared with control groups. This suggests that ¹¹C‐choline uptake reflects tumor aggressiveness. In a study of 32 patients with pathologically proven breast cancer expressing ER, no association was found between ¹¹C‐choline uptake and hormone receptor status.19 The apparent discrepancy between the present study and those of the previous study19 may be due, in large part, to differences in the patient populations studied.

In the present study, tumors exhibiting fat invasion were more likely to have a high SUV_max and TBR on ¹¹C‐choline PET/CT, but these differences were not identified in the TBR of ¹⁸F‐FDG PET/CT. This appeared to be associated with diffuse ¹⁸FDG uptake of breast caused by accompanying FCC, which may obscure tumor delineation. The presence of necrosis or cutaneous invasion was also found to have an association with SUV_max and TBR on ¹¹C‐choline PET/CT. Overall, our results are consistent with those reported in in vivo and in vitro studies, in which ¹¹C‐choline uptake was found to reflect tumor aggressiveness of breast cancer.13, 19

The present study design had limitations. First, the present study was designed to assess tumor uptake of dual tracers prior to surgery. The results from a breast cancer patient population of will not fully explain the detectability of advanced or recurrent disease. Second, the present study was an observational study and not a clinical trial, which raises the possibility of confounding factors affecting the results. Third, although ¹¹C‐choline is clearly a possible PET tracer for tumor localization in patients with breast cancer, its short half‐life restricts its practical application. However, ¹⁸F‐choline is a tracer with a longer half‐life than that of ¹¹C‐choline, and so ¹⁸F‐choline may improve the accuracy of tumor localization. Additional comparative studies regarding detectability and pathologic correlation are needed to validate the findings of the present study. Although we found that ¹¹C‐choline uptake reflected tumor aggressiveness in patients with breast cancer, we did not have any data regarding nodal status and follow‐up management of the patients. Further studies are needed to clarify the relationship between ¹¹C‐choline uptake and patient outcome with a long follow‐up period.

In conclusion, the results of the present study suggest that ¹¹C‐choline PET/CT allows for the evaluation of tumor aggressiveness and improves delineation of primary tumors compared with ¹⁸F‐FDG PET/CT in patients with breast cancer. The results demonstrate the advantages and potential of ¹¹C‐choline, but clinical evaluation with a long follow‐up period is warranted to clarify the exact role of this technique and how it affects patient outcome.

Abbreviations

CT

computed tomography

DCIS

ductal carcinoma in situ

ER

estrogen receptor

FCC

fibrocystic change

¹⁸F‐FDG2‐[¹⁸F]

fluoro‐2‐deoxy‐d‐glucose

HER‐2/neu

human epidermal growth factor‐2

MAPK

mitogen‐activated protein kinase

SUV_max

maximum standardized uptake value

TBR

tumor‐to‐background ratio

Disclosure Statement

The authors declare that they have no conflicts of interest.

(Cancer Sci, 2012; 103: 1701–1707)