Relationship between 18-FDG-PET / CT and Clinicopathological Features and Pathological Responses in Patients with Locally Advanced Breast Cancers

Objective: We investigated the relationship between the maximum standardized uptake value (SUVmax) of whole-body positron emission tomography/computed tomography (PET/CT) performed before treatment and the demographical and histopathological features in locally advanced breast cancer (LABC), as well as the role of PET/CT in the evaluation of pathological complete response (pCR) after neoadjuvant chemotherapy (NAC).


Introduction
Locally advanced breast cancer (LABC) comprises a heterogeneous group of patients with slowgrowing tumors in addition to those with a biologically aggressive disease.Because of high locoregional recurrence and metastasis risks in such patients, intensive treatment becomes imperative [1].Neoadjuvant chemotherapy (NAC) is accepted as the standard treatment for LABC.NAC increases the effective resection rates and breast-preserving surgery rates [2,3].Furthermore, it aims toward achieving the in vivo evaluation of chemotherapy (CT) response [4,5].
It has been shown that the pathological complete response (pCR) after NAC maintains the survival advantage in certain subtypes of breast cancers.Therefore, pCR may be used as a marker to evaluate treatment results [6,7].Nonetheless, breast cancers have different biological features.While recurrence occurs in some patients with pCR, certain patients without pCR have a good prognosis [8].Therefore, new predictive and prognostic factors in addition to pCR after NAC need to be determined.
Fluorodeoxyglucose positron emission tomography/computed tomography (FDG PET/CT) is a valuable instrument in staging, restaging after recurrence, and evaluation of CT response for breast cancer patients [9].The maximum standardized uptake value (SUV max ), expressed as the level of F18-FDG uptake, has been reported to be related with aggressive tumor biology in many studies.There are studies showing that histolopathological parameters such as tumor size, histo-Relationship between 18-FDG-PET/CT and Clinicopathological Features and Pathological Responses in Patients with Locally Advanced Breast Cancers logical grade, or hormone receptor expression status are correlated with the FDG uptake level in breast cancers [10][11][12].Further, studies have suggested that F18-FDG PET/CT is beneficial in the preoperative evaluation and prognosis of breast cancers and may be used in the decision toward CT [13,14].
FDG PET/CT has advantages as compared to conventional imaging studies.There are studies showing that it is a sensitive method in the evaluation of an early response to NAC [15].However, studies have also shown that its sensitivity is inadequate for the evaluation of pCR [16].Therefore, the use of FDG PET/CT in NAC is controversial.In this study, we investigated the relationship between the demographical and histopathological features of LABC and SUV max values of PET/CT before treatment, as well as the role of PET/CT in the evaluation of pCR after NAC.

Patients
For the period between 2011 and 2015, a file search was performed involving 453 patients treated for breast cancer in our unit.Here 51 breast cancer patients with stage 2 to stage 3 diseases that received NAC were retrospectively evaluated.Basal PET/CT was applied to all the patients before NAC.Patients with distant metastasis and bilateral breast cancer were excluded.All the patients received 4-8 cycles of CT.All the patients underwent surgery after undergoing controlled PET/CT following NAC.Bone metastasis occurred in one patient, and this patient underwent a mastectomy operation.The study was approved by the local ethics committee.Written informed consent was obtained from patients who participated in this study.

Clinicopathological Evaluations
Clinical data and pathological findings for all the patients were recorded by the investigation of polyclinic follow-up files and hospital archives.Age, menopausal status, clinical stage, and tumor localization of the patients were recorded.All the patients were diagnosed by a Tru-cut biopsy.Tumors were graded according to the modified Scarff-Bloom-Richardson grading system.Estrogen (ER), progesterone (PR), HER2, and Ki67 expression statuses were analyzed from the routine pathological evaluation results conducted in our center.Here, 1% or more positively stained nuclei in 10 high-power fields was considered to be positive for ER and PR [17].HER2 was analyzed by immunohistochemistry, and the staining intensity was reported as 0, 1+, 2+, or 3+.A score of 3+ was denoted as HER2 positive, and a score of 0 or 1+ was denoted as HER2 negative.If the score was 2+, then gene amplification using fluorescence in situ hybridization was used.When the ratio of HER2 gene copies to chromosome 17 signals was found to be greater than 2.2, HER2 was considered to be positive [18].After NAC, modified radical mastectomy (MRM) along with axillary dissection (AD) or breast-conserving surgery (BCS) was performed in all the patients.Tumor size; histological grade; presence of vascular, lymphatic, or perineural invasion; and lymph nodal involvement were assessed by means of surgical materials.The precise definition of pCR in breast carcinoma is uncertain.We adopted a pCR of ypT0/is ypN0, which means there is no residual invasive cancer in the breast and axillary lymph nodes, but noninvasive breast residuals are permitted.

NAC
A large majority of patients received four cycles of anthracycline-based regimens and four cycles of taxane regimens.Only one patient was evaluated after receiving three cycles of dosedense anthracycline-based treatment.There was CR in PET/CT imaging; after surgery, three more cycles were administered.The operation was performed in one patient after six cycles of docetaxel/doxorubicin/cyclophosphamide (TAC) regimen.BCS or MRM with AD was performed in all the patients after the completion of NAC.All the patients received anti-HER2 treatment, radiotherapy, and endocrine therapy, if necessary, according to standard guidelines.
18F-FDG PET/CT Imaging Patients were fasted overnight at least 12 h before whole-body PET/CT imaging was performed.Patients with a fasting blood sugar level above 120 mg/dl were excluded.After an intravenous injection of approximately 12 mCi (444 MBq) of 18F-FDG, the patient waited in a silent room for about 1 h; thereafter, the patient was imaged using an integrated PET/CT camera, which comprised a 16-slice CT gantry, integrated with an LSO-based full-ring PET scanner (Siemens Biograph 16; Siemens, Knoxville, TN, USA).CT was applied with 120-200 mAs at 120 kV adjusted to the body weight of the patient.Scanning was performed proximal to the thigh from the base of the skull.The PET images were reconstructed by a repetitive method (ordered-subset expectation maximization: two iterations, eight subsets) to achieve attenuation correction and image fusion.After reconstruction, the PET, CT, and fused images were visualized by using commercially available software (e-soft/VSIM, Siemens Medical Solutions) along the axial, coronal, and sagittal planes.The maxi-mum intensity projection for the PET images and co-recorded PET/CT images were obtained and analyzed by a nuclear medicine practitioner.This was calculated by using the following formula: measured activity concentration [Bq/mL] × body weight [kg]/injected activity [Bq].The SUV max value was determined by drawing the region of interest around the primary tumor on the transaxial slices.

Statistical Analysis
The data were presented as mean, standard deviation, median, minimum, maximum, percentage, and number.The normal distribution of the constant variables was analyzed with the Shapiro-Wilk test.While comparing two groups, independent samples t test was used if the normal distribution condition was achieved; otherwise, the Mann-Whitney U test was used.While comparing more than two groups, the ANOVA test was used if the normal distribution condition was achieved; otherwise, the Kruskal-Wallis test was used.Pearson correlation was used for numerical data.Spearman' s correlation was used for nominal data.To assess diagnostic accuracy, we performed the receiver operating characteristic (ROC) curve analysis.Subsequently, the area under the ROC curve (AUC) was estimated.Here, p<0.05 was considered to be statistically significant.The Statistical Package for the Social Sciences (SPSS) software version 20 (IBM Corp.; Armonk, NY, USA) was used for all the performed statistical analyses.

Patient Characteristics
The clinicopathological features of the patients are summarized in Table 1.A total of 51 patients were analyzed.Only one patient was male.The mean age of the patients was 49 (32-69) years.Out of these patients, 52.9% (n=27) were premenopausal.Further, 92.2% (n=47) had stage 3 disease and the tumor was localized in the right breast in 54.9% (n=28).Out of these patients, 90.2% (n=46) had IDC, 5.9% had ILC, 2% (n=1) had mucinous carcinoma, and 2% (n=1) had phyllodes tumor as the histological subtype.The following histologic grades were used: grade 1 for 1 patient (2%), grade 2 for 22 patients (43%), and grade 3 for 8 patients (15.6%).ER was positive in 43 patients (84.3%), and PR was positive in 39 patients (76.4%).When the ER and/ or PR receptors were simultaneously analyzed, the hormone receptors in 45 patients (88.2%) were found to be positive.HER2 positivity was observed in 60.8%, and 9.8% had the triple-negative molecular subtype.Out of these patients, 39.2% (n=20) achieved a complete response, 54.9% (n=28) achieved a partial response, 2.9% (n=2) exhibited a stable disease, and progressive disease was observed in 2% (n=1) in PET/ CT that was performed after NAC.
All the patients underwent surgery.One patient had BCS, and one patient who underwent negative sentinel lymph node sampling underwent a simple mastectomy operation.All the other patients (96%) had MRM with AD.The patient exhibiting bone metastasis was defined and this patient also underwent surgery after NAC, but did not receive radiotherapy to the breast: this patient died during the second year of treatment.In the follow-up period, recurrence was determined in 15.7% (n=8) patients with bone metastasis as the most common site (50%, n=4).The second-most-common site of recurrence was the brain.Out of the three patients with brain metastasis, two died related to the disease.Acute leucosis occurred in one patient who was receiving adjuvant endocrine therapy, because of which the patient died.In the followup period, 7.8% (n=4) patients died.The treatments of other patients with recurrence are continuing.

Relationship between Demographical and Histopathological features and PET/CT Pretreatment
The mean SUV max value of 51 patients was 8.53.There was no statistically significant difference between the mean SUV max value and age (> 50 and ≤ 50 years), menopausal status (premenopausal or postmenopausal), tumor localization, clinical stage, tumor Ki67 index, and tumor grade of the patients.The mean SUV max value is higher in the triple-negative group than those in the HER2 positive and luminal groups.The lowest SUV max value was observed in the luminal group.Although there was a relationship between the mean SUV max value and the molecular subtypes, the axillary SUV max values were statistically significant (Table 2).
A correlation analysis was formulated between the age, stage, menopausal status, localization, grade, Ki67, subtypes breast/axillary with after/ before CT primary tumor, and axillary SUV max values among all the patients, but no statistically significant relationship was revealed.

Role of PET/CT in pCR Evaluation
PET/CT FDG uptake after NAC was observed in 3 out of the 15 patients in which pCR was achieved.Furthermore, pCR could not be achieved in 8 out of 20 patients that did not show any abnormal FDG uptake in PET/CT.There was no significant difference between the pretreatment mean SUV max value of the patients with or without pCR.However, there was a significant difference between the SUV max values of these groups after NAC (SUV max breast 2.92 vs. 0.30, p=0.01;SUV max axillary 1.5 vs. 0.0, p=0.02) (Table 3).
ROC analysis was performed between the post-CT primary tumor and axillary SUV max with/without pCR.The sensitivity was 75% and specificity was 75% for the primary tumor, and the sensitivity was 55% and specificity was 100% for the axillary lymph nodes (Table 4) (Figure 1).
While none of the patients with a complete response in PET/CT exhibited recurrence, an abnormal FDG uptake in control PET/CT after NAC continued in all the patients in which metas-  tasis occurred during the follow-up period.DFS and OS analyses were not performed because of the short follow-up period of the patients.

Discussion
18F-FDG PET/CT-a noninvasive imaging modality-reveals both tumor metabolism and anatomical details.Higher 18F-FDG uptake in malignant cells makes the tumor visible.SUV max is a semiquantitative value, and it indicates the 18F-FDG uptake.The increased proliferation activity in tumor cells is related to the density of 18F-FDG uptake of the tumor [19].A large space exists between the 18F-FDG uptake values in our study (SUV max : 1.13-27.07),and this situation reflects the heterogeneity of the glucose metabolism of breast carcinomas.
The 18F-FDG uptake value has been found to be related to tumor biology in various maligni-ties in several studies [20].There are studies comparing the histopathological parameters in breast cancer with 18F-FDG uptake.Several studies that have supported this relationship have shown that the 18F-FDG uptake values are lower in invasive lobular carcinomas than that in invasive ductal carcinomas [11].The authors defined that the lower intensity of tumor cells in lobular carcinomas, lower expression of GLUT1, lower proliferation rates, and diffuse infiltrative tumor growth patterns of the surrounding tissue could be defined by this relationship [21].
A comparison could not be made in this study owing to the lower number of patients with invasive lobular carcinoma (three patients).
No relationship was determined between the demographical features of the patient and SU-V max .No relationship was observed between patient age and SUV max in several studies in the literature [8,12], similar to our study.Further, no significant relationship was observed between the menopausal status of the patient and SUV max .Although the study revealed that the 18F-FDG uptake value was 1.3 times higher in premenopausal patients [11], another study revealed that menopausal status and tumor SU-V max were independent [22].
The tumor grade is a significant predictive factor in breast carcinomas.Pathologists classify breast carcinomas according to the modified Bloom-Richardson grading system.In this system, parameters such as nuclear grade, tubule formation, and mitotic rate are scored from one to three.The calculation of the final score yields the histological grade.A strong positive correlation has been revealed between the histological grade and 18F-FDG uptake levels in certain studies [10,14].Further, a relationship between 18F-FDG uptake and nuclear pleomorphisms and mitotic activity has been reported by Berriolo-Riedinger et al. [23], but they found no relationship with tubular formation.This may be because of the fact that the role of mitotic count and nuclear pleomorphism is much more in the glycolytic pathway and glucose consumption.Although the mean SUV max values of patients with a grade 3 disease was higher in our study, no relationship between the grade and SUV max value was observed, possibly because of the limited number of patients.
Ki67 is a marker that indicates proliferation activity and is used in various cancers.High Ki67 value was found to be related with poor prognosis in patients [24].The major disadvantage of Ki67 value is the high variability between the observers evaluating this value [25].There are articles that have revealed a positive correlation   between the SUV max values and Ki67 proliferation index and lymphovascular invasion [10,12].However, no correlation has been shown between Ki67 and SUV max in our study.Because lymphovascular invasion was not defined in the pathological reports of certain patients, it was not evaluated in this study.
Prognosis is better in ER (+) patients than ER (-) ones [26].Data regarding the relationship between 18F-FDG uptake and hormone receptor status is unclear.While no relationship was reported between the receptor positivity and FDG uptake in certain studies [21], higher SUVmax values were determined in ER (-) patients in many studies [11].Further, higher SUV max values were established in receptor-negative patients than receptor-positive patients in our study.
Furthermore, cerbB2 overexpression is associated with the aggressiveness of tumor and poor prognosis.Although Ueda et al. [14] reported a correlation between cerbB2 oncogene overexpression and SUV max , this was not shown in many other studies [23].Further, a significant relationship with the primary tumor could not be shown in our study.Groheux et al. [11] suggested that there may be a minor role of the HER2 gene in the glycolytic pathway.
Triple-negative breast cancer (TNBC) comprises approximately 15% of all the invasive breast cancers.It has a high recurrence rate and poor OS time [27].This poor prognosis is a result of the aggressive characteristic of the disease and lack of effective targeted therapy options [28].
Because of the high SUV max uptake, PET-CT is found to be more useful in TNBC [29].The 18F-FDG uptake of TNBC patients was found to be high in our study, supporting many other studies [10][11][12].
PET-CT FDG uptake can facilitate the identification of metabolic changes earlier than conventional imaging methods during or at the end of treatment.It is found to be more sensitive (87%) in the evaluation of a treatment response than clinical examination (39%) and CT imaging (56%) [1].Therefore, it becomes a guideline for the effectiveness of CT and treatment processes.Wahl et al. [30] showed that PET-CT facilitated the determination of a decrease in glucose metabolism in the tumor and pathological tumor response after the effective treatment of primary breast cancers.Tateishi et al. [16] showed that the sensitivity of PET-CT in the evaluation of pCR was not acceptable, but its specificity was high.There is a wide range of variation in the sensitivity (39%-100%) and specificity (74%-100%) of PET-CT in the literature.In this study, the sensitivity was 75% and specificity was 75% for the primary tumor, and the sensitivity was 55% and specificity was 100% for the axillary lymph nodes.In our study, false positivity with PET-CT was also established in three out of 15 patients with pCR.Furthermore, out of the 20 patients with a complete response in PET-CT, 12 exhibited true positivity and 8 exhibited false positivity.There is no significant difference between the mean pretreatment SUV max values of the patients with or without pCR.However, there is a significant difference between the SUV max values between these two groups after NAC (SUV max breast 2.92 vs. 0.30, p=0.01;SUVmax axilla 1.5 vs. 0.0, p=0.02).
Our study has several limitations.Firstly, since this was a retrospective study and the number of patients was low, its results should be interpreted with caution.In some studies, it was revealed that the ∆SUV max measurements used to evaluate the metabolic response after NAC correlated with pCR.In our study, ∆SUV max measurements were not performed.The association between the histological subtypes and pCR was not assessed because of the low number of patients in the subgroups.Finally, we did not attempt to analyze the relationship between SUV max and survival as the follow-up period of the study population was relatively short.
Consequently, no relationship between PET-CT SUV max values and demographical features of the patients could be established in LABC.However, it seems that there may be a relationship with certain histopathological prognostic factors.Moreover, there is a significant relationship between the post-treatment SUV max value and pCR.PET/CT may be useful in predicting the prognosis in NAC-subjected patients with LABC because none of the patients with a complete response in PET/CT exhibited recurrence.In order to support these results, there is a need for randomized prospective studies that involve a large number of patients and a longer follow-up period.

Figure 1 .
Figure 1.ROC curve analysis of PET2 SUV max for prediction of pCR.SUV2T: SUV max in the primary tumor after chemotherapy; SUV2N: SUV max in the lymph nodes after chemotherapy.

Table 1 .
Clinical and pathologic characteristics of the patients

Table 2 .
Metabolic characteristics of the primary tumor according to the clinical and histopathological status

Table 3 .
Tumor metabolic characteristics according to achievement of pCR

Table 4 .
Sensitivity, specificity, AUC, cut-off and asymptotic significance of ROC analysis of parametersTo assess the diagnostic accuracy, we performed receiver operating characteristic (ROC) curve analysis.The area under the ROC curve (AUC) was then estimated.AUC: area under curve; SUV2T: SUV max in the primary tumor after chemotherapy; SUV2N: SUV max in the lymph nodes after chemotherapy.