Benign Phyllodes Tumors: Comparison of Prognosis among Three Different Surgical Approaches

Purpose To evaluate the prognosis of patients with benign phyllodes tumors (PTs) treated by different surgical methods and to explore the influencing factors of local recurrence. Methods We retrospectively analyzed 215 benign PTs from 193 patients who underwent surgery at Chinese PLA General Hospital between October 2008 and December 2020. We stratified our analysis according to surgical factors and explored the clinicopathological factors to influence local recurrence. Results Among 193 patients, a total of 17 (8.8%, 17/193) recurred during follow-up. There were 89 patients in the US-VAE group, of whom 6 (6.7%) recurred; 8 of 57 patients (14%) in the local lumpectomy group recurred, while 3 of 47 patients (6.4%) in the extended lumpectomy group recurred (P=0.252). Multivariate logistic regression analysis showed that tumor diameter, mitosis, and history of breast myoma were independent risk factors for tumor recurrence (P=0.005, P=0.006, and P=0.004, respectively). The intraoperative blood loss, operation time, and scar length of the US-VAE group were shorter than those of the other two groups (P < 0.05). Conclusion Negative surgical margins of benign PTs can obtain similar prognosis as negative surgical margins >10 mm. Therefore, we recommend that a follow-up observation policy be adopted for patients with unexpected benign PTs, rather than unnecessary open surgical resection. Patients' maximum tumor diameter, mitosis, and fibroadenoma history were independent predictors for recurrence of benign PTs.


Introduction
Benign phyllodes tumors (PTs) of the breast are rare fbroepithelial tumors, accounting for less than 1% of all primary breast tumors [1].Te World Health Organization (WHO) classifes PTs into benign, borderline, and malignant according to the pathological characteristics of tumors.Te proportions of the three subtypes are 60%-75%, 13%-26%, and 10%-20%, respectively [2].For women, it mainly occurs between the ages of 35 and 55 years [3].Surgery is the main treatment for this disease.Te surgical methods include ultrasound-guided vacuum-assisted resection (US-VAE), local mass resection, and extended mass resection.Both previous fndings and the national comprehensive cancer network (NCCN) recommend that the negative surgical margin of PTs must be no less than 10 mm regardless of the surgical approach [4].
Te low incidence of PTs and the lack of large-scale clinical data lead to our current limited understanding of it.Te recurrence rate of benign PTs is 0-20.8%,which is lower than that of borderline or malignant PTs [5,6].Since benign PTs have a better prognosis, can patients be spared negative margin surgery larger than 10 mm?In view of this, this study provides the frst systematic comparison of the three surgical modalities for benign phyllodes tumors.Can emerging minimally invasive US-VAE techniques be comparable to traditional open surgery?What are the risk factors that afect the local recurrence of benign tumors?Tis research will answer the abovementioned questions separately.

Patient Enrollment. Tis retrospective study was approved by the Medical Ethics Committee of PLA General
Hospital and waived the requirement for informed consent.We evaluated the pathological data and clinical outcomes of patients who underwent breast mass surgery in our hospital from October 2008 to December 2020, including age, menopause, history of fbroadenoma, tumor size, ultrasound fndings, histological characteristics, tumor recurrence, and operation type.Te inclusion criteria for the enrolled patients were as follows: (a) patients underwent one of the following three surgeries: US-VAE, local mass resection, or extended mass resection.(b) Te pathology of breast mass resection was benign PTs.Exclusion criteria were as follows: (a) including carcinoma in situ and invasive breast cancer and (b) previous breast cancer history and unavailable follow-up data.Tis study stratifed patients according to surgical factors and analyzed the clinicopathological risk factors of recurrence in patients with benign PTs.

Ultrasound-Guided Vacuum-Assisted Excision Procedure.
Te US-VAE procedures were mainly performed by a wellorganized team headed by Zhi Li Wang, using the iU22 ultrasound system (Philips Medical Systems, Andover, Massachusetts), with an L12-5 linear array probe, to determine the position and boundary of the lesions.Te optimal insertion site and the needle track are designed based on ultrasound examination.Local infltration anesthesia around the tumor was performed.Ten, we used the No. 7 EnCor Probe ® system (EnCor ® inserts MR, SenoRx, Allso Viejo, CA) directly into the tumor or behind the tumor through a 2-3 mm incision, cut, and automatically transported the tissue sample to the collection basket.When the ultrasound examination confrms that there is no obvious lesion image, we stop cutting.Ten, we take out the rotating cutting needle and apply partial pressure to prevent bleeding.We then immediately performed an ultrasound scan after surgery to ensure there are no residual lesions or hematomas.An adhesive elastic band was used for 24-48 h.Te treatment time depends on the size of the lesion and is defned as the period from the frst incision to the end of the last incision.

Local Lumpectomy
Procedure.We determine the location and size of the tumor by palpation.After the local anesthesia is completed, the skin and subcutaneous tissue are incised.Te tumor is completely removed by carefully separating the surrounding area of the tumor along the breast tissue.Before suturing the skin, we thoroughly rinse the surgical area and stop bleeding and store instruments and dressings.Finally, an elastic bandage was used to compress the postoperative incision.

Extended Lumpectomy
Procedure.After palpating to determine the location and size of the tumor, local anesthesia (or general anesthesia) is performed.Te skin and subcutaneous tissue were incised on the surface of the breast tumor, carefully separated from the pectoralis major fascia along 10 mm around the tumor, and the tumor was completely removed.Before suturing the skin, we fushed the surgical feld, stopped bleeding thoroughly, and counted the instruments and dressings without error.Ten, we used an elastic bandage to compress the postoperative incision.
2.5.Follow-up.Te follow-up was censored in December 2020, and the follow-up of all patients was performed by reviewing the electronic system, text messages, or telephone calls (one of the methods or more) of the outpatient clinic of our hospital.Patients' follow-up time and follow-up results (physical examination + ultrasound or MRI examination) were recorded.

Data and Statistical Analysis.
Te age used for each patient was the age at the time of diagnosis.Lesion size on US was defned by the maximum diameter.Te follow-up period was defned as the time from the confrmation of benign PTs by surgery to the last clinical follow-up.Recurrence was defned as clinical and sonographic tumor recurrence within the previously resected area.
All statistical analyses were performed using SPSS 23.0, standard version (SPSS Inc., Chicago, IL).Statistical comparisons were performed using the χ 2 -test or Fisher's exact test for categorical variables and the Kruskal-Wallis test for continuous variables.Te relationship between each clinical factor and local recurrence was screened by univariate analysis.Te logistic multiple regression analysis was subsequently used to calculate and analyze the independent infuencing factors of local recurrence.P values <0.05 were considered statistically signifcant.

Clinicopathologic Features.
A total of 243 benign PTs in 221 female patients were screened out.Among these cases, 28 without follow-up were excluded.Terefore, a total of 215 benign PTs (Figure 1) from 193 patients were fnally included, 89 of which underwent US-VAE (Figure 2), 57 underwent local lumpectomy, and 47 underwent extended lumpectomy.Te demographic characteristics, including age, history of breast fbroids, image examination (including sonographic fndings), and surgical records, were extracted from the original resumes.Te clinicopathological characteristics are listed in Table 1.

Recurrence
Rate.Among the 193 patients, 17 cases (8.8%, 17/193) recurred during the follow-up period; in the US-VAE group, 6 out of 89 patients (6.7%) relapsed; in the local lumpectomy group, 8 out of 57 patients (14%) recurred, while 3 out of 47 tumors in the extended lumpectomy group (6.4%) recurred (P � 0.252).Recurrent cases are summarized in Table 2.All recurrent lesions were treated surgically and pathologically diagnosed.One of the recurrences was a malignant phyllodes tumor, and 16 cases were benign PTs.

2
Te Breast Journal Among these 17 relapsed patients, no distant metastasis occurred during the subsequent follow-up period.Univariate analysis showed that tumor diameter, mitosis, and history of breast fbroma may be potential risk factors for recurrence (Table 3; P < 0.05).Age, menopause, lump location, echo pattern, lesion edge, blood fow, BI-RADS categories, and surgical methods were not signifcantly diferent between the groups with and without recurrence (P � 0.350, 0.218, 0.163, 0.177, 0.409, 0.365, 0.861, and 0.252, respectively).Multivariate logistic regression analysis revealed that the diameter of the lump, mitoses, and history of breast fbroids are independent risk factors for tumor recurrence.Te results are presented in Table 4.

Complications.
Complications include hematoma, wound oozing, infection, and ecchymosis.A total of 8 patients in the US-VAE group developed complications, of which 6 developed hematomas and 2 developed skin ecchymosis.A total of 6 patients in the local lumpectomy group had complications, of which 1 had hematoma, 2 had wound exudative infection, and 3 had skin ecchymosis.Seven patients in the extended lumpectomy group had complications, 2 patients had ecchymosis and hematoma, and 3 patients had wound exudative infection.(P � 0.572, Table 1) Te intraoperative blood loss, operation time, and wound length in the group US-VAE were shorter than those in the other two groups (P < 0.05), as detailed in Table 5.

Discussion
Te study aimed to systematically compare the efects of three surgical methods on the recurrence of benign PTs.NCCN guidelines recommend a negative surgical margin distance of no less than 10 mm for all pathologic types of PTs.However, few datasets support this practice.In our study, after a median follow-up of 39 months, 17 patients had local recurrence.Among them, there were 16 benign PTs and 1 malignant PT.Te diference in recurrence of benign PTs between diferent surgical approaches was not statistically signifcant, which illustrated that a negative resection margin greater than 10 mm could not reduce the recurrence rate of benign PTs.An analysis [7] of 1,702 patients enrolled in 12 studies also showed that the recurrence rate of benign PTs after surgery was very low and that 112 cases of 1,052 benign phyllodes tumors recurred.When the margin is positive, the recurrence rate increases.Graña López et al. [8], by analyzing 160 benign breast nodules (11 lesions corresponded to benign PTs) resected from 2007 to 2016, concluded that resection of benign PTs by vacuum-assisted percutaneous excision did not increase the risk of recurrence compared with surgery.It should be noted that the 89 patients who underwent US-VAE included in this study Te Breast Journal were not diagnosed with phyllodes tumors preoperatively, so these lesions were benign PTs accidentally diagnosed by US-VAE.Te second International Consensus Conference on lesions of uncertain malignant potential in the breast (B3 lesions) also recommended vacuum-assisted biopsy system treatment for the lesion which lacks high-risk cytological features, although more prolonged observation is necessary [9,10].Terefore, although our sample size was small, these results suggest that surgical margin status does not appear to be predictive of local recurrence and that increasing the negative distance to the surgical margin does not confer greater beneft to patients with breast PTs.At the same time, we recommend a follow-up observation policy for patients  PTs of the breast may lead to the degradation or upgradation of pathological types in the process of recurrence and metastasis.In our study, it was observed that after local recurrence, the pathological type of one patient was upgraded from benign to malignant.Tis is similar to Tan's previous report.In his study of 353 cases of breast phyllodes, he found that the pathology of 14 cases of breast phyllodes upgraded from benign to marginal after recurrence [11].Terefore, in clinical practice, we should try to avoid the recurrence of patients to avoid pathological upgrading.At present, the causes of pathological changes after recurrence or metastasis of breast PTs are not clear, and its mechanism and pathways need to be further studied.Tere may be new genetic changes in recurrent phyllodes.In US-VAE: ultrasound-guided vacuum-assisted excision.Te Breast Journal a previous study, only 3 of the 9 benign phyllodes showed histological upgrading at the time of recurrence, but 6 (67%) of the 9 histological benign tumors obtained new gene changes related to borderline/malignant phenotypes (+1q, +7p, −9p, and −13) [12].Diferent authors found a broad variety of predictors, although with varying statistical power.In the present study, multivariate logistic regression analysis revealed that the diameter of the lump, mitoses, and history of breast fbroids are independent risk factors for tumor recurrence.First, although the prediction of local recurrence rates of PTs based on tumor size remains controversial, in recent studies, smaller PTs still show lower recurrence rates than larger PTs [13,14].Te results of our study suggest that the median tumor size of the US-VAE group was 2.27 cm, which is signifcantly smaller than that of the other two groups.Te reason for the diference in sizes is when the benign breast tumor is small, patients are more inclined to choose minimally invasive surgery.US-VAE has a high complete resection rate and can leave smaller scars [15].When the tumor diameter is relatively large, US-VAE is less suitable as a treatment modality.It may be because those with relatively large tumor volume require multiple excisions, which may easily cause gas accumulation in the tumor cavity, afect Bultrasound observation, and increase the risk of tumor recurrence [16].Terefore, US-VAE is more suitable for benign PTs with a maximum diameter of <3 cm.Others have a similar opinion about the tumor size cutof when treating benign PTs with US-VAE, while some scholars also believe that the cutof value of tumor size using the US-VAE method should be 3.3 cm [17,18].Patients with tumor size ≥3 cm and <3 cm were evaluated separately, and there was a statistically signifcant diference in the local recurrence rate, indicating that patients with breast tumors ≥3 cm are prone to recurrence.Terefore, for benign PTs with larger diameters, extended lumpectomy seems to be a more efective method for reducing the local recurrence rate than US-VAE and local lumpectomy.
To date, the number of mitotic fgures also determines the prognosis of PTs to some extent.In this study, patients with high mitosis fgures had an OR value of 10.881 compared with those with low mitoses' fgures.Tis shows that for benign PTs, as the number of mitotic images increases, the possibility of recurrence is greater.At the same time, we found that benign PTs with a history of fbroadenoma had a higher recurrence rate than those without a history of fbroadenoma.Terefore, when patients have a history of fbroadenoma, a regular review should be adhered to guard against secondary or even multiple recurrences.Given that fbroadenoma and PTs belong to the same category as fbroepithelial breast tumors and have similar clinical symptoms, imaging manifestations, and histomorphology.Relevant studies have suggested that they may be partially correlated [19].In a recent study, exome sequencing of 22 PTs and targeted sequencing of 100 fbroepithelial neoplasms revealed the genetic profle of fbroepithelial neoplasms, with MED12 (73%) and RARA (32%) mutations common in fbroadenoma patients and in all pathologically classifed PTs' patients [20].Notably, mutations in FLNA (28%), SETD2 (21%), and KMT2 (9%) were observed only in PTs, suggesting a role in driving PTs' development.Terefore, we speculated that some patients with a history of fbroadenoma had stromal cell mutation during recurrence, and fbroadenoma was upgraded to PTs due to tumor disease.Tis study also shows that US-VAE treatment of PTs is superior to the conventional open surgery in terms of wound length, blood loss, operation time, postoperative scar size, and breast beauty.Tere was no statistically signifcant diference in the incidence of postoperative complications among the three surgical methods.Tese conclusions have also been confrmed in previous studies [21,22].
Tis study has some limitations.First, this is a singlecenter retrospective study, and the relatively small number of patients recruited will lead to inevitable inherent bias.For example, patients treated with US-VAE are more likely to have smaller tumors.Although the data were included continuously, we also considered these factors in the multivariate analysis, and the study still included other factors that could not be adjusted.Second, in the present study, this exclusion may have led to biased recurrence rate estimates when loss to follow-up did not occur at random or was associated with a hazard of death.Finally, a further study with a larger sample size and longer follow-up is necessary.

Conclusions
Negative surgical margins of benign PTs can obtain similar prognosis as negative surgical margins >10 mm.Terefore, we recommend that a follow-up observation policy be adopted for patients with unexpected benign PTs, rather than unnecessary open surgical resection.Patients' maximum tumor diameter, mitosis, and fbroadenoma history were independent predictors for recurrence of benign PTs.

Figure 1 :Figure 2 :
Figure 1: Flowchart illustrating the selection process of the study population.

Table 2 :
Patient and tumor characteristics of the 17 recurred cases.

Table 1 :
Clinicopathological characteristics of patients.

Table 3 :
Univariate analysis of infuencing factors of postoperative local recurrence.

Table 4 :
Multivariate logistic regression analysis of infuencing factors of postoperative local recurrence.

Table 5 :
Comparison of operation-related indexes between the three groups.