Abstract
Background/Aim: Intraoperative blood loss (IBL) during the surgical treatment of various cancers affects complication rates and prognosis. However, few studies have examined the importance of minimal IBL in breast cancer surgery. We used factor analysis to examine the prognostic importance of IBL in breast cancer. Patients and Methods: One hundred ninety-seven patients who underwent mastectomy plus axillary lymph node dissection (level II) after preoperative chemotherapy between June 2007 and June 2021 were included. Pearson’s Chi-square test was used to confirm the relationships between different factors. Kaplan–Meier survival curves and the log-rank test were used to examine prognosis. Logistic regression was performed using a Cox proportional hazards model. Results: The median IBL was 55.0 g (range=5.0-420.0 g). IBL was <100 g in 143 patients (72.5%), 100-200 g in 39 patients (19.8%), and >200 g in 15 patients (7.6%). Logistic regression analysis showed that patients with IBL ≥200 g had a significantly worse prognosis (disease-free survival: p=0.003, log-rank test; overall survival: p<0.001, log-rank test). Factor analysis revealed that HER2-negative status (p=0.015), non-pathological complete response (p=0.031), obesity (p=0.001), heavy smoking (p=0.047), and diabetes mellitus (p=0.004) were significantly associated with increased IBL. Conclusion: IBL in breast cancer was correlated with various clinicopathological factors associated with a poor prognosis, suggesting that increased IBL may be associated with poor prognosis in breast cancer as well.
Surgery is a treatment strategy aimed at radically curing most carcinomas. Surgical intervention is also essential for resectable breast cancer. The amount of intraoperative blood loss (IBL) is an indicator of invasion to the patient, and surgeons are making efforts to reduce the amount of IBL.
Some studies have reported that the amount of IBL affects the prognosis of gastric cancer, colorectal cancer, pancreatic cancer, hepatocellular carcinoma, and prostate cancer (1-13). However, to the best of our knowledge, no studies have investigated IBL and prognosis in breast cancer. Previous studies have reported that differences in the surgical method cause a difference in the amount of IBL (2-5, 7, 14). Therefore, we examined the correlation between IBL and clinicopathological factors and the effect of IBL on the prognosis of patients with breast cancer who underwent mastectomy and axillary lymph node dissection (ALND) (level II) after preoperative chemotherapy (POC) of the same regimen. We hypothesized that IBL would be associated with prognosis, similar to other forms of cancer (1-13).
Patients and Methods
Patient background. From June 2007 to June 2021, 214 patients with breast cancer underwent mastectomy and ALND (level II) after POC at Osaka City University Hospital. However, since ten patients with large primary tumors underwent additional flap plasty or skin grafting, and seven patients whose amount of IBL was unknown were excluded, 197 patients were included in this study. In these eligible cases, none of the patients required blood transfusions during or after surgery. All breast cancers were pathologically diagnosed by core needle biopsy or vacuum-assisted biopsy, and the expression of estrogen receptor (ER), progesterone receptor (PgR), human epidermal growth factor receptor 2 (HER2), and Ki67 was also evaluated by immunostaining. Based on the results, we classified them into three subtypes according to previous reports (15). Breast cancer with negative ER, PgR, and HER2 expression was defined as triple-negative breast cancer (TNBC), and breast cancer with negative ER and PgR expression negative but HER2 expression positive was defined as HER2-enriched breast cancer (HER2BC). Other breast cancers, that is, breast cancers positive for ER and/or PgR, were defined as hormone receptor-positive breast cancer (HRBC). Breast cancer progression was assessed prior to POC using computed tomography (CT), ultrasonography (US), and bone scintigraphy. Before POC, the smoking status of all patients was confirmed through interviews and each pack-year was calculated. In addition, all patients were confirmed to have hypertension or diabetes as a comorbidity. Height and weight were measured before POC, and the body mass index (BMI) was calculated as weight (kg) divided by the squared height (m) (kg/m2). The BMI cut-off value for this study was 25, which was divided into normal weight and overweight according to the World Health Organization (WHO) category.
Regarding POC, FEC100 (500 mg/m2 fluorouracil, 100 mg/m2 epirubicin, and 500 mg/m2 cyclophosphamide) was administered in four courses every 3 weeks in the first half, and 80 mg/m2 paclitaxel was administered in 12 courses weekly in the second half (16-18). For patients with HER2-positive breast cancer, additional weekly (2 mg/kg) or tri-weekly (6 mg/kg) trastuzumab was administered during the paclitaxel treatment period. After POC, imaging tests were performed again to evaluate the therapeutic effect according to the Response Evaluation Criteria in Solid Tumors (19). The therapeutic effects evaluated as clinical partial response (cPR) and complete response (cCR) were defined as “Responder” in objective response rate (ORR), and those evaluated as clinical stable disease (cSD) and clinical progressive disease (cPD) were defined as “non-responders”. Subsequently, a mastectomy and ALND (level II) were performed. Patients diagnosed with axillary lymph node metastases before POC underwent ALND, even if the lymph node metastases disappeared with POC. ALND was not performed in patients without axillary lymph node metastasis before and/or after POC. Patients who underwent breast-conserving surgery or non-Level II ALND were excluded from the study. The definition of a pathological complete response (pCR) follows the National Surgical Adjuvant Breast and Bowel Project B-18 protocol, that is, “the complete disappearance of the invasive components of the lesion with or without intraductal components, including those in the lymph nodes” (20). After surgery, standard adjuvant therapy was administered according to each subtype and the surgical procedure. The median follow-up time was 1,620 days (range=62-4,817 days) after surgery.
Statistical analysis. All statistical analyses were performed using SPSS version 28.0 (IBM, Armonk, NY, USA). The relationship between each clinicopathological feature was evaluated using Pearson’s chi-square test. Prognostic analyses, such as disease-free survival (DFS) and overall survival (OS), were performed using the Kaplan–Meier method and log-rank test. The hazard ratios (HRs) and 95% confidence intervals (CIs) were calculated using the Cox proportional hazards model, and significant factors were entered into a Cox regression model. The odds ratios (OR) and 95% CIs) were calculated using the logistic analysis. Multivariate analysis was performed using a multivariable logistic regression model. A p-value <0.05 was considered statistically significant.
Results
Clinicopathological features. Table I shows the clinicopathological factors of the 197 patients who underwent mastectomy and ALND (level II) after POC. The median age was 59 years (range=30-75 years), and the median tumor diameter was 37.3 mm (range=9.2-119.8 mm). Breast cancer invaded the skin in 75 patients (38.1%). All patients were diagnosed with axillary lymph node metastasis prior to POC. The results of immunohistochemical staining were ER positive in 92 patients (46.7%), PgR positive in 38 patients (19.3%), HER2 positive in 70 patients (35.5%), and high Ki67 in 130 patients (66.0%). As a result, 95 patients (48.2%) were classified as HRBC, 47 patients (23.9%) were classified as HER2BC, and 55 patients (27.9%) were classified as TNBC. After POC, 161 patients (81.7%) were evaluated as responders in ORR. The median operative time was September 2014. Pathological examination of resected specimens revealed pCR in 44 patients (22.3%). The median BMI before POC was 21.8 kg/m2 (14.3-37.6 kg/m2). There were 141 never-smokers (71.6%), while 22 smokers (11.2%) had more than 20 pack years. Before POC, 15 patients (7.6%) were being treated for diabetes mellitus, and 53 patients (26.9%) were being treated for hypertension.
Clinicopathological features of 197 patients who underwent a mastectomy and axillary lymph node dissection after preoperative chemotherapy.
The median IBL amount was 55.0 g (range=5.0-420.0 g). In 143 patients (72.5%), the IBL amount was <100 g, named the low-IBL group, whereas in 15 patients (7.6%), the IBL amount was >200 g, named the high-IBL group. Thirty-nine patients (19.8%) that did not belong to either category were named the medium-IBL group.
Impact of intraoperative blood loss on prognosis. The log-rank test was used to compare prognosis between participants grouped according to IBL. The high-IBL group had significantly worse DFS (p=0.003), and OS (p<0.001) compared to the other groups (Figure 1). In a univariate analysis, HER2 (p=0.021, HR=0.511), HER2BC (p=0.010, HR=0.380), TNBC (p=0.011, HR=1.842), ORR (p=0.010, HR=0.499), pCR (p=0.005, HR=0.348), and IBL (p=0.001, HR=2.776) were significantly associated with IBL (Table II and Figure 2). Multivariate analysis revealed that TNBC (p=0.008, HR=2.100), ORR (p=0.049, HR=0.575), and pCR (p=0.022, HR=0.396) were independent predictors of prognosis. However, IBL was not an independent risk factor (p=0.062, HR=1.851). Univariate analysis for OS showed significant differences in ORR (p=0.011, HR=0.390), diabetes mellitus (p=0.003, HR=3.810), and IBL (p<0.001, HR=3.851); however, multivariate analysis revealed no independent risk factors for prognosis, including IBL (p=0.150, HR=1.922) (Table III) (Figure 3).
Log-rank tests were performed for prognosis after dividing into three groups by intraoperative blood loss (IBL). The High-IBL group had a significantly worse prognosis for both disease-free survival (DFS) (A) and overall survival (OS) (B) (DFS: p=0.003, log-rank; OS: p<0.001, log-rank).
Univariate and multivariate analyses concerning disease-free survival.
Forest plot. Univariate analysis for disease-free survival (DFS), HER2 (p=0.021, HR=0.511), HER2-enriched breast cancer (HER2BC) (p=0.010, HR=0.380), triple negative breast cancer (TNBC) (p=0.011, HR=1.842), objective response rate (ORR) (p=0.010, HR=0.499), pathological complete response (pCR) (p=0.005, HR=0.348), and intraoperative blood loss (IBL) (p=0.001, HR=2.776) showed significant differences.
Univariate and multivariate analyses concerning overall survival.
Forest plot. Univariate analysis for overall survival (OS) showed significant differences in objective response rate (ORR) (p=0.011, HR=0.390), diabetes (p=0.003, HR=3.810), and intraoperative blood loss (IBL) (p<0.001, HR=3.851), but multivariate analysis revealed no independent factors, including IBL (p=0.150, HR=1.922).
Correlation between intraoperative blood loss and clinicopathological factors. The correlations between IBL and clinicopathological factors are summarized in Table IV. Increased IBL was significantly associated with HER2 negativity (p=0.015), a non-PCR (p=0.031), higher BMI (p=0.001), heavy smoking (p=0.047), and comorbid diabetes mellitus (p=0.004). Trend-level associations between increased IBL and higher Ki67 (p=0.079), TNBC status (p=0.092) and a longer surgical history (p=0.063) were found, but these were not statistically significant (p<0.05). When entered in a multivariable logistic regression model, the OR could not be calculated, because none of the patients with HER2BC or those who achieved pCR had excessive IBL (Table V). However, obesity was a significant risk factor for IBL (p=0.010, OR=5.756).
Comparison of clinicopathological features by intraoperative blood loss.
Univariate and multivariate analyses concerning higher intraoperative blood loss.
Discussion
To date, it has been reported that the amount of IBL adversely affects the prognosis of gastric cancer, colorectal cancer, pancreatic cancer, hepatocellular carcinoma, and prostate cancer; however, this study is the first of its kind to examine the relationship between IBL and prognosis in breast cancer (1-13). Previous reports have investigated the factors that affect IBL and the causes of increased IBL that adversely affect prognosis. The amount of IBL differs depending on the year of surgery (9), and this tendency was also observed in our study. The reasons for this are that the devices used in surgery have changed, and the surgeon’s technical skills have improved, and we believe that this is the same in this study. The reason for the higher IBL in non-PCR cases may be that the residual tumor made it easier to bleed during resection, but this is just our hypothesis. Previous studies have reported that IBL is significantly correlated with cancer progression (1, 4, 14). One study concluded that this is the main reason why IBL affects prognosis (14). However, these reports did not unify the surgical procedures. In this study, there was no significant difference in the degree of cancer progression by unifying the surgical methods. Furthermore, it has been reported that the amount of IBL affects prognosis even after adjusting for the degree of progression in patients with gastric cancer (5). A previous study reported that IBL was correlated with BMI and surgical procedures (5). There is also a report on colorectal cancer in which IBL correlated with BMI and surgical procedures (2). In this study, a strong correlation was found between IBL and BMI. Obesity has been reported to be a poor prognosis factor for breast cancer in many reports (21-24). Smoking and diabetes mellitus, which are both correlated with IBL, have also been identified as poor prognostic factors in patients with breast cancer (25-30). It is speculated that IBL is correlated with many of these poor prognostic factors described for breast cancer, which is one of the causes of an unfavorable prognosis.
A limitation of this study is the lack of consideration of factors that may affect the amount of IBL, such as the surgeon’s skill or patient’s ability to coagulate. However, we found that IBL was correlated with various poor prognostic factors described for breast cancer, indicating that an increase in IBL is associated with a poor prognosis. Furthermore, in previous studies which examined other carcinomas, there are various reasons why an increase in IBL may have contributed towards a poor prognosis. Some studies have suggested that increased IBL may promote tumor spillage and hematogenous spread during surgery (31-34). It is also possible that increased IBL causes a decrease in host immunity (11). Animal experiments have shown that, as the amount of IBL increases, the activity and cytotoxicity of natural killer cells decrease, and the growth rate of tumors increases (35, 36). In addition, a previous study reported that IBL in upper gastrointestinal tract cancer reduces postoperative natural killer cell activity and adversely affects patient survival (37). A study of IBL in colorectal cancer also suggested that excessive bleeding, rather than transfusion, may lead to immunosuppression, and thus adversely affect long-term survival (1). In breast cancer surgery, immunosuppression due to IBL may lead to an increased risk of postoperative recurrence, and it is necessary to make efforts to reduce the amount of IBL as much as possible.
In conclusion, IBL in breast cancer was correlated with various clinicopathological factors associated with poor prognosis, suggesting that increased IBL is associated with a poor prognosis in breast cancer as well.
Acknowledgements
The Authors thank Tomomi Okawa (Department of Breast Surgical Oncology, Osaka Metropolitan University Graduate School of Medicine) for the helpful advice regarding data management. This study was funded by grants from the Japan Society for the Promotion of Science (KAKENHI, Nos. 20K08938, 26461957, and 17K10559) to Shinichiro Kashiwagi.
Footnotes
Authors’ Contributions
All Authors were involved in the preparation of this manuscript. KT collected the data and wrote the manuscript. SK, NI, RK, AY, WG, YA, YT, KO, and TM performed the operation and designed the study. SK, MS, and HT summarized the data and revised the manuscript. KM provided a substantial contribution to the study design, performed the operation, and revised the manuscript. All Authors read and approved the final manuscript.
Conflicts of Interest
All of the Authors have no conflicts of interest to disclose regarding this study.
- Received November 3, 2022.
- Revision received November 14, 2022.
- Accepted November 15, 2022.
- Copyright © 2023 International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved.