Background
Malignant phyllodes tumors (MPTs) are a rare type of breast tumor, accounting for only 0.5% of all breast malignancies.1 MPTs typically exhibit an insidious onset and subsequent rapid growth pattern. Most patients experience swift and temporary mass enlargement involving a solitary, unilateral, painless breast lump. In some cases, patients may present with sizable breast masses accompanied by signs such as skin thinning, superficial varicose veins, and even skin erosions and ulcers. Surgical intervention remains the primary therapeutic treatment for MPTs according to the NCCN Clinical Practice Guidelines in Oncology (NCCN Guidelines) for Breast Cancer.2 However, the efficacy of radiotherapy and chemotherapy remains inconclusive. Despite the lack of prospective clinical studies,3,4 the absence of randomized controls hinders the establishment of robust adjuvant treatment guidelines for MPTs. The most critical prognostic aspect of MPTs is their pronounced susceptibility to both local recurrence and distant metastasis. According to WHO statistics, the local recurrence rate for MPTs can reach 23% to 30%, and there are even reports in the literature that the local recurrence rate of MPTs can reach 40%.5,6 The predominant sites of metastasis are the lungs and bones, with a distant metastasis rate reaching 22%.7,8 Patients facing local recurrence or distant metastasis have an exceedingly poor prognosis, with reported average survival times ranging from 4 to 17 months.3,9–12
Due to the rarity and aggressive clinical nature of MPTs, effective treatment options for local recurrence and distant metastases following surgery remain elusive. This underscores the importance of assessing predictive risk factors in patients with MPTs and gauging the potential efficacy of adjuvant therapy. This multicenter study was designed to investigate prognostic risk factors and evaluate the feasibility of adjuvant chemotherapy as a treatment strategy for patients with MPTs without distant metastasis.
Patients and Methods
Inclusion Criteria
From 1999 to 2023, we performed a comprehensive retrospective analysis of medical records pertaining to patients diagnosed with MPTs who underwent surgical intervention at 4 medical institutions in China: Sun Yat-Sen Memorial Hospital of Sun Yat-sen University, Peking University Shenzhen Hospital, the Second Affiliated Hospital of Xi’an Jiaotong University, and the Affiliated Hospital of Guizhou Medical University. This retrospective study included patients who were first diagnosed with MPTs without distant metastasis and received R0 resection surgery. Patients with metastatic disease at the initial visit, those without complete clinicopathologic feature data, and those lost to follow-up were excluded.
Materials and Statistical Analysis
Comprehensive clinical and pathologic information of patients diagnosed with MPTs was extracted from electronic medical records. The diagnoses for all patients were made through pathologic examination. The pertinent clinicopathologic attributes of patients with MPTs at the time of diagnosis were compiled, encompassing factors such as age, tumor laterality, and tumor size. Additionally, details regarding surgical procedures, axillary lymph node involvement, and the administration of radiotherapy or chemotherapy were also collected.
The median and range were calculated for continuous variables. The study’s primary endpoint was local recurrence–free survival (LRFS), which represents the time between tumor removal surgery and the occurrence of local recurrence, death, or the last follow-up. The secondary endpoint was overall survival (OS), reflecting the duration from tumor removal to death due to any cause or the final follow-up. The final follow-up time was determined by the date of the last telephone follow-up conducted by the respective center’s follow-up office or the date of the last hospital visit for examination. To construct survival curves and calculate 5-year LRFS and 5-year OS, the Kaplan-Meier method was used. Prognostic risk factors associated with LRFS and OS were determined through both univariate and multivariate Cox proportional hazard regression models, accompanied by the reporting of hazard ratio (HR) and 95% confidence interval. Propensity score matching (PSM) was used to mitigate selection bias and create cohorts with balanced baseline characteristics, as detailed in the Supplementary Materials, available online with this article.
Propensity Score Matching
In our investigation, patients were split into 2 groups—chemotherapy (CTx) and nonchemotherapy (non-CTx)—based on whether adjuvant chemotherapy was administered following the initial surgery. To mitigate the impact of confounding factors and systematic bias on both chemotherapeutic efficacy and tumor prognostic outcomes within the CTx and non-CTx groups, we used PSM, thereby establishing groups with well-balanced baseline characteristics. The propensity scores were calculated by logistic regression analysis using SPSS Statistics, version 26.0 (SPSS Corp). The propensity score covariates in this study included age at diagnosis, maximum tumor diameter, tumor laterality, surgery type, and lymph node status. A caliper of 0.02 was used, and subsequently, a 1:1 match was performed between the CTx and non-CTx groups using the nearest propensity score without replacement. The most common diagnostic for propensity scores is the hypothesis test,13 which is also used in this study.
Results
Clinicopathologic Characteristics of Patients With MPTs (Before Matching)
The study involved 4 separate institutions where a total of 145 patients diagnosed with MPT were treated between 1999 and 2023. The flowchart of this study is illustrated in Supplementary Figure S1. All patients underwent surgical therapy and were categorized into 2 groups based on their receipt of adjuvant chemotherapy after their initial surgery: CTx (n=31; 21.4%) and non-CTx (n=114; 78.6%).
Table 1 summarizes the clinicopathologic characteristics of the individuals from both groups. Among the participants, most (61.4%) were aged ≤45 years, with a median age of 43 years (range, 16–77 years). The distribution of tumor incidence was nearly equal between the left (48.3%) and right (51.7%) sides, showing no significant difference (P=.675). The median tumor size was 6.5 cm (range, 1.5–25.0 cm), and no substantial difference in tumor size was observed between the 2 groups (P=.219). Mastectomy was the preferred surgical choice for most patients (69.7%), whereas a smaller proportion of patients underwent breast-conserving surgery (BCS) (30.3%). There was no significant difference in the rate of breast conservation between the CTx and non-CTx groups (22.6% vs 32.5%; P=.289). Axillary lymph node surgery was performed on 70 (48.3%) patients, of whom 52 (74.3%) had sentinel lymph node biopsy (SLNB) and 18 (25.7%) underwent axillary lymph node dissection (ALND). Interestingly, the CTx group exhibited a greater frequency of ALND than the non-CTx group (7 [22.5%] vs 11 [9.7%], respectively; P=.028). The percentage of positive axillary lymph nodes was 3.4%. Notably, the CTx group displayed a greater percentage of positive axillary lymph nodes than the non-CTx group (3 [9.7%] vs 2 [1.8%], respectively; P=.032).
Clinicopathologic Characteristics of Patients With Malignant Phyllodes Tumors
The individual chemotherapy regimens of 31 chemotherapy patients, the majority of whom received 4 to 8 cycles of systemic therapy, are shown in Supplementary Table S1. Among the patients, 35.4% (n=11) opted for anthracycline + ifosfamide (AI) as the initial postoperative adjuvant chemotherapy regimen, a recommended approach for advanced soft tissue sarcoma. The first chemotherapy regimens consisting of anthracycline + cyclophosphamide (AC) and anthracycline + cyclophosphamide/docetaxel (AC-T) were chosen by 25.8% (n=8) of the patients. Other initial chemotherapy regimens that were applied sparingly included docetaxel/cyclophosphamide (TC), docetaxel/carboplatin (TCb), and cyclophosphamide/methotrexate/fluorouracil (CMF). Unfortunately, the table does not include information about the dose or duration of chemotherapy due to incomplete data.
Effect of Adjuvant Chemotherapy on Local Recurrence or Distant Metastasis in Postoperative Patients
The median follow-up period was 54.5 months (range, 3.4–294.2 months). The incidence of local recurrence and distant metastasis is presented in Supplementary Table S2. Among all patients, 37 experienced local recurrence, resulting in an overall local recurrence rate of 25.5%. The median time to local recurrence was 11.7 months (range, 0.6–57.6 months), and up to 73.0% of patients experienced local recurrence within 2 years after surgery. There was no statistically significant difference in local recurrence rates between the CTx and non-CTx groups (8% [25.8%] vs 29% [25.4%], respectively; P=.967). Similarly, no statistically significant difference in the rate of distant metastasis was found between the CTx and non-CTx groups (6% [19.4%] vs 18% [15.8%], respectively; P=.636). Of the patients with distant metastasis, 18 experienced lung metastases, accounting for 75% of all metastatic cases. Bone metastases were found in 5 patients (20.8%), whereas a few individuals displayed metastases to the adrenal glands, liver, and heart.
Prognostic Factors Impacting LRFS and OS
Factors affecting local recurrence and OS in patients with MPTs were evaluated in this study. The results of univariate and multivariate Cox analyses for LRFS and OS are presented in Tables 2 and 3, respectively. Univariate analysis revealed that LRFS was associated with tumor size (HR, 1.076; 95% CI, 1.025–1.129; P=.003), surgery type (HR, 2.694; 95% CI, 1.122–6.467; P=.027), and axillary lymph node status (HR, 8.606; 95% CI, 3.001–24.675; P<.001) (Supplementary Figure S2). OS was associated with age (HR, 4.678; 95% CI, 2.188–10.000; P<.001), tumor size (HR, 3.148; 95% CI, 1.464–6.766; P=.003), surgery type (HR, 3.192; 95% CI, 1.208–8.434; P=.019), axillary lymph node status (HR, 10.789; 95% CI, 3.685–31.584; P<.001), local recurrence (HR, 8.545; 95% CI, 3.932–18.568; P<.001), and distant metastasis (HR, 33.897; 95% CI, 13.639–84.245; P<.001) (Supplementary Figure S3). Therefore, older age, larger tumors, mastectomy, positive axillary lymph nodes, local recurrence and distant metastasis were significantly associated with worse OS. However, in the multivariate analysis, LRFS was solely associated with axillary lymph node status (HR, 7.763; 95% CI, 2.334–25.824; P=.001), whereas OS was correlated with age (HR, 5.308; 95% CI, 1.868–15.085; P=.002), tumor size (HR, 5.073; 95% CI, 2.044–12.593; P<.001), axillary lymph node status (HR, 7.447; 95% CI, 1.941–28.574; P=.003), local recurrence (HR, 2.477; 95% CI, 1.980–6.264; P=.045), and distant metastasis (HR, 17.671; 95% CI, 5.981–52.204; P<.001). No significant differences in LRFS or OS were found between the 2 groups regarding chemotherapy according to univariate or multivariate analysis.
Univariate and Multivariate Cox Analyses of Prognostic Factors for LRFS
Univariate and Multivariate Cox Analyses of Prognostic Factors for OS
Effect of Postoperative Chemotherapy on LRFS and OS
Prior to PSM, the Kaplan-Meier curves showed that patients who received postoperative adjuvant chemotherapy (CTx group) did not have improved LRFS (HR, 1.09; 95% CI, 0.50–2.40; P=.821) or OS (HR, 1.17; 95% CI, 0.50–2.73; P=.711) compared with those who did not receive chemotherapy (non-CTx group) (Figure 1). The 5-year LRFS and 5-year OS rates were 71.56% and 79.04%, respectively, in the non-CTx group, whereas in the CTx group, the rates were 70.61% and 81.61%, respectively.
Before PSM, Kaplan-Meier curves of (A) LRFS and (B) OS for patients with MPT who did and did not receive postoperative adjuvant chemotherapy.
Abbreviations: CTx, chemotherapy; HR, hazard ratio; LRFS, local recurrence–free survival; MPT, malignant phyllodes tumor; OS, overall survival; PSM, propensity score matching.
Citation: Journal of the National Comprehensive Cancer Network 22, 7; 10.6004/jnccn.2024.7023
After PSM at a 1:1 ratio, the baseline characteristics of the CTx and non-CTx groups were balanced, mitigating the potential influence of variations in clinicopathologic features on the analytical outcomes (Table 1). After PSM, 28 patients were assigned to both the CTx and non-CTx groups. The median follow-up duration was 91.6 months (range, 4.3–294.2 months). The median age of the patients was 37 years (range, 19–58 years), and the median tumor size was 6.5 cm (range, 2.0–25.0 cm). Similar to the findings before PSM, the Kaplan-Meier curves following PSM also indicated that patients in the CTx group did not exhibit better 5-year LRFS (HR, 1.05; 95% CI, 0.35–3.14; P=.934) or 5-year OS (HR, 1.95; 95% CI, 0.51–7.44; P=.328) than patients in the non-CTx group (Figure 2). In the CTx group, the 5-year LRFS and 5-year OS rates were 74.31% and 86.64%, respectively, whereas in the non-CTx group, the rates were 73.89% and 88.86%, respectively.
After PSM, Kaplan-Meier curves of (A) LRFS and (B) OS for patients with MPT who did and did not receive postoperative adjuvant chemotherapy.
Abbreviations: CTx, chemotherapy; HR, hazard ratio; LRFS, local recurrence–free survival; MPT, malignant phyllodes tumor; OS, overall survival; PSM, propensity score matching.
Citation: Journal of the National Comprehensive Cancer Network 22, 7; 10.6004/jnccn.2024.7023
Discussion
Breast phyllodes tumors are uncommon fibroepithelial neoplasms of the breast that arise from fibrous and connective tissue with dual differentiation. Among them, MPTs are even more uncommon, accounting for only 0.5% of all breast malignancies.1 The NCCN Guidelines recommend extended local resection, with a negative margin of at least 1 cm, for MPTs.2 However, there is a lack of definitive recommendations for postoperative adjuvant therapy. Despite the successful attainment of wide negative margins, the local recurrence rate following surgery can reach 40%,5,6 underscoring the urgency for breast specialists to explore effective postoperative adjuvant therapy strategies for MPTs.
Barth et al3 demonstrated that postoperative adjuvant radiotherapy significantly reduced the local recurrence rate in patients with negative surgical margins. Similarly, data from another extensive retrospective study14 indicated that adjuvant radiotherapy lowered local recurrence rates but did not exert a significant impact on disease-specific survival or overall survival. A meta-analysis conducted by our team revealed that postoperative radiotherapy effectively controlled disease recurrence and prevented metastasis.15 However, the role of adjuvant radiation therapy remains uncertain due to inconsistent findings, primarily attributed to the lack of prospective randomized studies and the limited inclusion of a small number of patients receiving adjuvant radiation in most published series.16 Few studies exist regarding the effects of adjuvant chemotherapy for phyllodes tumors. A prospective study4 conducted in Mexico included 28 patients with MPTs, 17 of whom received adjuvant chemotherapy using the doxorubicin and dacarbazine regimen. At a median follow-up of 15 months, no significant disparity in the 5-year recurrence-free survival rate was observed between the chemotherapy and observation groups (58% vs 86%; P=.17). In addition, most available data derive from case reports or studies involving chemotherapy for metastatic PTs.17–21 Two retrospective studies19,20 specifically addressing systemic treatment of metastatic phyllodes tumors revealed that patients with metastatic disease who underwent anthracycline-based (AI/anthracycline alone/other anthracycline regimens) chemotherapy regimens, particularly the AI regimen, experienced enhanced survival benefits. According to Palassini et al,19 66.1% (37/56) of patients with metastatic MPTs opted for anthracycline-based chemotherapy regimens, comparable to the 64% (32/50) according to Parkes et al.20 Similarly, most (74.2%) patients in our study received anthracycline-based chemotherapy.
The purpose of our research was to assess the effectiveness of adjuvant chemotherapy in patients with MPTs without distant metastasis and treated with R0 resection surgery. Our investigation revealed an overall local recurrence rate of 25.5% and an overall distant metastasis rate of 16.6% for MPTs. The lung emerged as the most prevalent site for distant metastases, consistent with prior research findings. Notably, no significant difference was observed in terms of local recurrence or distant metastasis rates between the CTx and non-CTx groups (Supplementary Table S2). This finding prompted the conclusion that postoperative adjuvant chemotherapy does not appear to confer benefits in terms of controlling either local recurrence or distant metastasis.
Because phyllodes tumors are more likely to recur than other tumors, numerous studies have investigated prognostic risk factors for breast phyllodes tumors. Factors such as age, tumor size, tumor grade, histologic features, surgery type, and surgical margin have been shown to be associated with local recurrence.22–26 Our univariate analysis revealed that tumor size, surgery type, and axillary lymph node status were associated with local recurrence. Interestingly, mastectomy is a risk factor for local recurrence, possibly reflecting a relationship to the large size of the tumor itself and the need for mastectomy; large tumors are also risk factors for local recurrence. However, in our multivariate analysis, we found that LRFS in MPTs was only significantly associated with positive axillary lymph nodes. Additionally, older age, larger tumors, positive axillary lymph nodes, local recurrence, and distant metastasis were risk factors for OS in both univariate and multivariate analyses, whereas surgery type was correlated with OS in univariate analysis but not in multivariate analysis. According to the NCCN Guidelines, axillary lymph node surgery is not routinely recommended for patients with MPTs, and research has indicated that <5% of patients with phyllodes tumors exhibit positive axillary lymph nodes.27 Our study is consistent with these findings, demonstrating that positive axillary lymph nodes were present in only 3.4% of patients with MPTs. In patients with positive lymph nodes, there is a significant association between positive axillary lymph nodes and local recurrence, with 80% of such patients experiencing local recurrence. Therefore, we propose that routine axillary staging is unnecessary for patients with MPTs unless preoperative assessments indicate axillary lymphadenopathy. For patients with positive axillary lymph nodes, we recommend intensified postoperative follow-up. Notably, our findings indicated that surgical type had no impact on LRFS or OS. Therefore, considering adequate surgical margins, breast-conserving surgery could be preferred to ensure postoperative cosmetic breast outcomes and alleviate the psychosocial impact on patients. Overall, our study revealed that postoperative adjuvant chemotherapy did not offer protective benefits in terms of improving LRFS or OS.
Due to the low incidence of MPTs, most studies exploring survival have included a small number of patients, sometimes <10. There are even fewer studies on the effectiveness of adjuvant chemotherapy, almost none; therefore, a meta-analysis could not be performed. This has led experts to question the generalizability of their conclusions. Our study revealed a 5-year overall LRFS of 71.13% and a 5-year overall OS of 79.43%. This outcome aligns with findings from our prior meta-analysis.15 Importantly, the rates of LRFS and OS demonstrated no significant differences between the CTx and non-CTx groups, both before and after PSM. These results collectively indicate that postoperative adjuvant chemotherapy has limited value in improving the prognosis and survival of patients with MPTs.
All of the aforementioned data demonstrated that postoperative adjuvant chemotherapy for MPTs without distant metastasis did not improve survival or lower the rate of local recurrence or distant metastasis. However, it is crucial to note that these findings rely solely on our retrospective data. Our study is the first multicenter clinical study to evaluate the effectiveness of postoperative adjuvant chemotherapy for MPTs without distant metastasis with R0 resection utilizing PSM, but a key limitation of our study is that the analysis of chemotherapy efficacy did not include consideration of the chemotherapy regimen. Additionally, the sample size of patients who received chemotherapy was relatively small, and not all patients were uniformly subjected to the AI regimen, which is comparatively impactful for metastatic disease. Therefore, large-scale prospective randomized controlled clinical trials should be performed to obtain more impartial and objective findings on the effectiveness of postoperative adjuvant chemotherapy in MPTs. However, given the low incidence of MPTs, conducting a randomized controlled trial with a large sample size is challenging.
Conclusions
In this retrospective series of MPTs without distant metastasis treated with R0 resection surgery, we observed that patients who received postoperative adjuvant chemotherapy with various regimens, most of whom received 4 to 8 cycles of systemic therapy, experienced no significant improvement in survival.
References
- 1.↑
Zhang H, Tang S, Biskup E, et al. Long-term survival after diverse therapeutic modalities in malignant phyllodes tumors of the breast. Technol Cancer Res Treat 2022;21:15330338221121086.
- 2.↑
Gradishar WJ, Moran MS, Abraham J et al. NCCN Clinical Practice Guidelines in Oncology: Breast Cancer. Version 4.2023. Accessed March 23, 2023. To view the most recent version, visit https://www.nccn.org
- 3.↑
Barth RJ Jr, Wells WA, Mitchell SE, et al. A prospective, multi-institutional study of adjuvant radiotherapy after resection of malignant phyllodes tumors. Ann Surg Oncol 2009;16:2288–2294.
- 4.↑
Morales-Vásquez F, Gonzalez-Angulo AM, Broglio K, et al. Adjuvant chemotherapy with doxorubicin and dacarbazine has no effect in recurrence-free survival of malignant phyllodes tumors of the breast. Breast J 2007;13:551–556.
- 5.↑
Belkacémi Y, Bousquet G, Marsiglia H, et al. Phyllodes tumor of the breast. Int J Radiat Oncol Biol Phys 2008;70:492–500.
- 6.↑
Asoglu O, Ugurlu MM, Blanchard K, et al. Risk factors for recurrence and death after primary surgical treatment of malignant phyllodes tumors. Ann Surg Oncol 2004;11:1011–1017.
- 7.↑
Tan BY, Acs G, Apple SK, et al. Phyllodes tumours of the breast: a consensus review. Histopathology 2016;68:5–21.
- 8.↑
Onkendi EO, Jimenez RE, Spears GM, et al. Surgical treatment of borderline and malignant phyllodes tumors: the effect of the extent of resection and tumor characteristics on patient outcome. Ann Surg Oncol 2014;21:3304–3309.
- 9.↑
de Roos WK, Kaye P, Dent DM. Factors leading to local recurrence or death after surgical resection of phyllodes tumours of the breast. Br J Surg 1999;86:396–399.
- 10.↑
Kapiris I, Nasiri N, A’Hern R, et al. Outcome and predictive factors of local recurrence and distant metastases following primary surgical treatment of high-grade malignant phyllodes tumours of the breast. Eur J Surg Oncol 2001;27:723–730.
- 11.↑
Abdalla HM, Sakr MA. Predictive factors of local recurrence and survival following primary surgical treatment of phyllodes tumors of the breast. J Egypt Natl Canc Inst 2006;18:125–133.
- 12.↑
Reinfuss M, Mituś J, Duda K, et al. The treatment and prognosis of patients with phyllodes tumor of the breast: an analysis of 170 cases. Cancer 1996;77:910–916.
- 13.↑
Granger E, Watkins T, Sergeant JC, et al. A review of the use of propensity score diagnostics in papers published in high-ranking medical journals. BMC Med Res Methodol 2020;20:132.
- 14.↑
Gnerlich JL, Williams RT, Yao K, et al. Utilization of radiotherapy for malignant phyllodes tumors: analysis of the National Cancer Data Base, 1998–2009. Ann Surg Oncol 2014;21:1222–1230.
- 15.↑
Chao X, Chen K, Zeng J, et al. Adjuvant radiotherapy and chemotherapy for patients with breast phyllodes tumors: a systematic review and meta-analysis. BMC Cancer 2019;19:372.
- 16.↑
Boutrus RR, Khair S, Abdelazim Y, et al. Phyllodes tumors of the breast: adjuvant radiation therapy revisited. Breast 2021;58:1–5.
- 17.↑
Yonemori K, Shimizu C, Hasegawa T, et al. Effectiveness of MAID therapy against metastatic malignant phyllodes tumors and stromal sarcoma of the breast. Breast Care (Basel) 2006;1:194–197.
- 18.↑
Mituś J, Reinfuss M, Jakubowicz J, et al. Malignant phyllodes tumor of the breast: treatment and prognosis. Breast J 2014;20:639–644.
- 19.↑
Palassini E, Mir O, Grignani G, et al. Systemic treatment in advanced phyllodes tumor of the breast: a multi-institutional European retrospective case-series analyses. Breast Cancer Res Treat 2022;192:603–610.
- 20.↑
Parkes A, Wang WL, Patel S, et al. Outcomes of systemic therapy in metastatic phyllodes tumor of the breast. Breast Cancer Res Treat 2021;186:871–882.
- 21.↑
Neron M, Sajous C, Thezenas S, et al. Impact of metastasis surgery and alkylating-agent-based chemotherapy on outcomes of metastatic malignant phyllodes tumors: a multicenter retrospective study. Ann Surg Oncol 2020;27:1693–1699.
- 22.↑
Kim S, Kim JY, Kim DH, et al. Analysis of phyllodes tumor recurrence according to the histologic grade. Breast Cancer Res Treat 2013;141:353–363.
- 23.↑
Ditsatham C, Chongruksut W. Phyllodes tumor of the breast: diagnosis, management and outcome during a 10-year experience. Cancer Manag Res 2019;11:7805–7811.
- 24.↑
Ben Hassouna J, Damak T, Gamoudi A, et al. Phyllodes tumors of the breast: a case series of 106 patients. Am J Surg 2006;192:141–147.
- 25.↑
Wei J, Tan YT, Cai YC, et al. Predictive factors for the local recurrence and distant metastasis of phyllodes tumors of the breast: a retrospective analysis of 192 cases at a single center. Chin J Cancer 2014;33:492–500.
- 26.↑
Chaney AW, Pollack A, McNeese MD, et al. Primary treatment of cystosarcoma phyllodes of the breast. Cancer 2000;89:1502–1511.
- 27.↑
Chen WH, Cheng SP, Tzen CY, et al. Surgical treatment of phyllodes tumors of the breast: retrospective review of 172 cases. J Surg Oncol 2005;91:185–194.
