Application of intraoperative frozen section examination in the management of female breast cancer in China: a nationwide, multicenter 10-year epidemiological study

Background Intraoperative frozen section examination (IFSE) during breast cancer surgery can partly reflect the status of surgical treatment since the surgical method used directly determines the purpose of IFSE use in disease management. This study aims to investigate the application of, changing trends in, and factors influencing IFSE in the management of female breast cancer in China. Methods We collected the sociodemographic and clinical data of 4,211 breast cancer patients between 1999 and 2008 and statistically analyzed these data using χ2 or Fisher’s exact tests. Results A total of 2,283 (54.22%) patients with breast cancer underwent IFSE. During the 10-year study period, IFSE use was associated with an increase in the number of sentinel lymph node biopsies (SLNB) and breast-conserving surgeries (BS) performed, with significant regional differences noted in this trend (P <0.05). Patients’ education, occupation, age, tumor size estimated by preoperative palpation, and the use of imaging examinations affected the purpose of IFSE use (P <0.05). Conclusions Our results show that the purpose of IFSE in the surgical treatment of breast cancer in China is gradually approaching that in developed countries. We believe that policymakers must address the differences in breast cancer treatment based on the socioeconomic status of patients. Lastly, the use of IFSE for determining tumor characteristics should be avoided as far as possible, and patient education and breast cancer screening programs tailored to the Chinese population should be established. Our findings may guide the formulation of breast cancer control strategies in China and other low-income countries.

method used directly determines the purpose for which IFSE is used [6].
Although breast cancer is the most common cancer among women in China [7], few studies have analyzed the total population of breast cancer patients in the country. Our previous studies demonstrated, for the first time, the pathological characteristics of breast cancer in Chinese women [8,9]. The current study is part of a series of studies concerning the application and temporal tendencies of and regional differences in IFSE use in the diagnosis and treatment of female breast cancer in China during the period 1999-2008. We also examined factors affecting the purpose for which IFSE is used in breast cancer management.

Subjects
This series of studies was approved by the Institutional Review Board of the Cancer Foundation of China. The approval covered data collection from all the participating centers. Female patients with breast cancer who had received treatment (surgery, medical oncology, radiotherapy) were enrolled from seven first-class, grade III hospitals or referral centers in north, northeast, central, south, east, northwest, and southwest China [8]. Since there was no risk associated with participating in the study, patients' informed consent was not obtained. The identity of all patients was kept anonymous.

Data collection
According to the pre-designed recruitment protocol, at least 50 female patients with breast cancer were sampled from each hospital or referral center per year. The sampling method used could largely avoid selection bias. Two data input clerks from each site independently double-entered data from the patients' record into a computer database. Quality control was performed during each step of data collection. Finally, all completed double-entry databases were sent to the Cancer Hospital/Institute, Chinese Academy of Medical Sciences, for validation using the statistical software EpiData (http://www.epidata.dk/). Details of the methods have been described elsewhere [8].
Patient characteristics collected were general information, risk factors, imaging examination results, treatment mode, and clinicopathological characteristics of the tumors. Histological subtyping of tumors was based on the 1981 and 2003 WHO histological classification criteria [10,11]. Breast cancer staging was performed according to the American Joint Committee on Cancer tumor-node-metastasis staging system of 1997 and after [12,13].

Grouping criteria
Subjects were assigned to two groups depending on the reason for using IFSE. In Group A, IFSE had been used for diagnosing the type of breast tumor, while in Group B, it had been used to assess metastasis to a sentinel lymph node (SLN) and/or to identify the resection margin during breast-conserving surgery (BS) but not to identify the type of primary tumor.
The seven geographic regions mentioned were classified as high-or low-socioeconomic status (SES) areas, according to previously described criteria [14]. Four socioeconomic indicators were recorded: gross domestic product per capita, percentage of health-service expenditure in the regional/provincial public affairs general budget, ratio of urban to rural population, and percentage of illiteracy among females aged 15 and over. They were used to establish the SES of each area with the selected hospital or referral center.
According to the commonly used classification in China, the ages of the patients at diagnosis of breast cancer were divided into four groups: <35 years, 35-49 years, 50-64 years, and >64 years.

Statistical analysis
All statistical analyses were performed using the statistical software SPSS version 16.0 (SPSS Inc., Chicago, IL, USA). The trend χ 2 test was employed to examine the temporal changes in IFSE application and its frequency during the period 1999 to 2008. Factors affecting the use of IFSE were analyzed using the χ 2 test. The effect of preoperative examination methods on the frequency of IFSE use was evaluated using the χ 2 test and Fisher's exact test. A P value <0.05 was considered statistically significant.

Patient characteristics
A total of 4,211 eligible breast cancer patients were enrolled in this series of studies, which comprised 9.3% of the total number of breast cancer cases encountered at the hospitals or referral centers [8]. Included in the 4,211 patients were 2,283 patients (54.22%) who had undergone IFSE. In terms of age, 139 patients (6.09%) were aged under 35 years at diagnosis, 1,145 (50.15%) were aged between 35 and 49 years, 793 (34.73%) between 50 and 64 years, and 206 (9.02%) above 64 years. As mentioned, the seven geographic regions were grouped into low-and high-SES areas according to the area-based SES. High-SES areas were north, northeast, east, and south China, with 1,227 subjects (53.75%); low-SES areas were northwest, central, and southwest China, with 1,056 subjects (46.25%).

Temporal changes in the use of IFSE
The total number and percentage of the patients who underwent IFSE during the 10-year study period did not change significantly year-on-year, as shown in Table 1. There was, however, a change in the reason for using IFSE (Table 2 and Figure 1). The number and percentages of IFSEs performed for identifying the type of primary tumor declined gradually, while the corresponding figures for assessment of metastasis to SLNs or identification of the resection margin increased gradually (P <0.05). Overall, during the study period, 94.61% (2,160/ 2,283) of IFSEs were used to identify the type of primary tumor, 2.98% (68/2,283) for sentinel lymph node biopsy (SLNB), and 9.46% (216/2,283) for BS. In most cases, IFSE was used exclusively for identifying the type of primary tumor (2,014/2,283; 88.22%), whereas in 4.91% of cases (112/2,283), it was performed to identify both the type of primary tumor and the resection margin for BS; fewer cases involved IFSE use exclusively for SLNB or identification of the resection margin for BS. Thus, IFSE for identifying the type of primary tumor (Group A) accounted for 94.61% of cases (2160/2,283), while that for other reasons accounted for 5.39% of cases (123/ 2,283) (Group B; Table 2).

Comparison of regional differences in the frequency of IFSE
The frequency of IFSE use for various reasons differed among the different SES regions as well. In low-SES areas, the percentage of IFSE used for assessment of metastasis to SLNs and/or identification of the resection margin (Group B) was very low (1.52%; 16/1,056). In contrast, in high-SES areas it was significantly higher (8.72%; 107/1,227; P <0.05) ( Figure 2). Table 3 shows the effects of patients' sociodemographic characteristics on indications for IFSE. Most patients who underwent IFSE were educated up to junior and senior high school (48.96%; 638/1,303), and the main occupation was manual work, accounting for 51.07% (1,024/2,005) of the total patient population. A little more than half the study population was in the 35 to 49 years age range (50.15%; 1,145/2,283). Statistical analyses showed that the level of education, occupation, and age all affected the indications for IFSE, as follows: i) Group A gradually reduced with advanced education level, while Group B increased significantly (P = 0.001).

Effect of sociodemographic characteristics on the frequency of IFSE use
ii) The purpose of IFSE use also varied according to patients' occupations. The highest percentage of Group A   (96.09%) was observed among manual workers, while the highest percentage of Group B (8.37%) was found among intellectual workers, and the between-group difference was significant (P <0.001). iii) Lastly, Group A showed a gradually increasing tendency with age, while Group B exhibited a gradually declining tendency with age (P <0.001).

Effects of preoperative examinations on purpose of IFSE use
The effects of preoperative examinations on the indications for IFSE are shown in Table 4. Both the tumor size, determined by preoperative palpation, and the performance of molybdenum-target X-ray or ultrasonography examinations prior to surgery affected the purpose of IFSE use (P <0.05). IFSE was performed in a higher percentage for Group B purpose in patients who had tumor sizes ≤20 mm or who had undergone molybdenumtarget X-ray or ultrasonography examinations (P <0.05); on the contrary, the percentage of Group A was significantly lower in these patients (P <0.05). The lymph node status, determined by preoperative palpation, was not significantly correlated with the purpose of IFSE use (P >0.05).

Discussion
With continual advances in science and technology and evolution of breast cancer treatment, IFSE plays an increasingly important role in the surgical treatment of this disease. Simultaneously, the principal purpose of IFSE has shifted from identification of tumor type to identification of SLN metastasis and assessment of the resection margin during BS. In our previous nationwide multicenter retrospective clinical epidemiological studies of female breast cancer in China, we described the pathological characteristics of this condition [8,9]. The present study, which is part of a series of studies, investigated the application of IFSE and factors that influence its use in the management of female breast cancer. An objective of this study was to guide the formulation of breast cancer control strategies in China and other lowincome countries. First, we analyzed the indications for IFSE and changing trends in its use. The results showed that IFSE was still mainly used to identify the type of breast tumor. However, its overall use for this purpose has declined over the 10-year study period, while its use for assessment of metastasis to SLNs and identification of the  resection margin for BS has gradually increased. The indications for IFSE in China differ from those in developed countries, where SLNB and BS are the major indications for surgical treatment of breast cancer [15][16][17][18][19][20][21][22][23][24][25][26][27]. Correspondingly, the purpose of IFSE use has changed. IFSE is increasingly being applied in SLNB and BS and less so for identifying the type of primary tumor [15]. Both the Edge report from the 28 th Annual San Antonio Breast Cancer Symposium [28] and the results of the NSABP B-32 clinical trial [29] indicated that open surgical resection combined with IFSE is no longer an acceptable treatment for breast cancer in developed countries. However, differences in economic levels may necessitate differences in treatment modes for breast cancer across countries [30][31][32]. The overall level of economic development is significantly lower in China than in developed countries, and this could affect diseasecontrol strategies. In the present area-stratified analysis, we found that regional differences in economic development affected the use of IFSE: regions with high-SES tended to use IFSE for assessment of metastasis to SLNs and identifying the resection margin during BS more frequently than those with low-SES. We believe that in developing countries or regions, the use of preoperative biopsy techniques should be increased in order to decrease the use of IFSE for intraoperative diagnosis of breast tumor so that we can choose the more appropriate treatment pattern before operation. Our results are encouraging in that we found that the use of IFSE for SLNB and BS has significantly increased over the 10year study period, while its use for identifying the type of primary tumor has decreased. This trend suggests that surgical treatment of breast cancer in China is gradually approaching that of developed countries. We believe that the influence of regional differences in economic levels on the use of IFSE and breast cancer control strategies deserves more attention as a means of benefiting more patients. We also evaluated the effects of patients' sociodemographic characteristics on the use of IFSE for surgical treatment of breast cancer. We found that the purpose of IFSE use was closely associated with patients' education, occupation, and age. This association may be attributed to the following factors: i) The time to diagnosis is longer in the case of patients of lower educational level, lower occupational income, and older age [33][34][35], because of which they have advanced disease and the opportunity to perform SLNB and BS is poor. ii) Subjects with higher educational levels and income are more willing to accept new concepts and technologies, such as SLNB and BS [36,37]. iii) Subjects with lower incomes cannot afford post-surgical radiotherapy, and most live in poor areas where radiotherapy is seldom performed [37]. iv) Younger patients are more likely to care about the esthetics of surgery and hence may be more willing to accept BS and SLNB [38,39], while older patients are more prone to reject these two types of surgery [40]. On the basis of these findings, we believe that better education among individuals of low-SES should be emphasized in order to help them accept SLNB and BS. Finally, we assessed the effects of preoperative examinations on the purpose of IFSE use. Our previous study showed that imaging examinations affected the clinicopathological characterization of breast cancer at diagnosis [9]. The current study found that the frequency of imaging examinations also affected the rate of IFSE use. Greater implementation of imaging examinations led to greater use of IFSE for SLNB and BS. The following reasons may explain this phenomenon: i) Breast cancer can be detected at earlier stages in patients who have undergone molybdenum-target X-ray and ultrasonographic examinations; therefore, their tumors may be smaller and the possibility of SLN metastasis is lower [41,42], and, consequently, these patients are more likely to receive SLNB and BS. ii) Second, preoperative molybdenum-target X-ray and ultrasonographic examinations can better predict the safety of SLNB and BS [43][44][45]. In the present study, we also found that tumor size, as estimated by preoperative palpation, affected the indications for IFSE, in that IFSE was used more frequently for SLNB and BS in patients with tumors ≤20 mm. This finding may demonstrate that preoperative estimation of tumor size is a key factor in determining whether BS will be performed [46,47], as BS is safest when tumors are small [48]. Conversely, patients with large tumors have a higher probability of having axillary lymph node metastasis [49,50] and a greater risk of complications from a SLNB [51,52].
Our study has a limitation, which must be acknowledged: since subjects were enrolled from hospitals or referral centers in seven traditional geographic regions, a selection bias may have been present. In China, most patients with breast cancer visit high-grade hospitals. All hospitals or referral centers involved in the present study are grade III and have the requisite physician resources and equipment for performing IFSE. However, grade I and II hospitals lack the resources to perform comprehensive breast cancer therapy, including IFSE. Nonetheless, since our study had a broad base and included diverse populations, the selection method used is probably the most suitable to reflect the diverse nature of present-day China.

Conclusions
In summary, in the present study, we analyzed the use of IFSE in the diagnosis and treatment of breast cancer in China during the 10-year period from 1999 to 2008. From our observations, we concluded that IFSE is still used mainly to identify the type of breast tumor. Ideally, the use of preoperative biopsy techniques should be improved so that IFSE use for tumor diagnosis during breast cancer operations can be decreased. In line with the introduction of newer concepts in breast cancer treatment and treatment modes, the purpose of IFSE in surgical treatment of breast cancer in China is gradually approaching that of developed countries. Nevertheless, regional differences in IFSE use require the attention of policymakers. Finally, patients' sociodemographic characteristics and the availability and use of preoperative imaging examinations significantly affect the use of IFSE in breast cancer treatment. Therefore, better education and breast cancer screening suitable for China's needs may help prevent breast cancer and aid the system in serving those with this disease.