Radiation-Induced Sarcoma of the Breast: A Systematic Review

A systematic review of the published English-language literature was performed to evaluate evidenced-based practices for radiation-induced sarcoma treatment. Surgery with wide negative margins remains the primary treatment; the role of adjuvant and neoadjuvant chemotherapy is uncertain.

Learning Objectives

After completing this course, the reader will be able to:

1. Evaluate the level of evidence to inform decision making for the treatment radiation induced sarcoma of the breast.

2. Explain diagnostic criteria for radiation-induced sarcoma.

3. Describe the effectiveness of surgery, chemotherapy and radiation therapy for radiation induced sarcoma of the breast.

This article is available for continuing medical education credit at CME.TheOncologist.com

Abstract

Introduction.

Radiation-induced sarcoma (RIS) is a rare, aggressive malignancy. Breast cancer survivors treated with radiotherapy constitute a large fraction of RIS patients. To evaluate evidenced-based practices for RIS treatment, we performed a systematic review of the published English-language literature.

Methods.

We performed a systematic keyword search of PubMed for original research articles pertaining to RIS of the breast. We classified and evaluated the articles based on hierarchal levels of scientific evidence.

Results.

We identified 124 original articles available for analysis, which included 1,831 patients. No randomized controlled trials involving RIS patients were found. We present the best available evidence for the etiology, comparative biology to primary sarcoma, prognostic factors, and treatment options for RIS of the breast.

Conclusion.

Although the evidence to guide clinical practice is limited to single institutional cohort studies, registry studies, case–control studies, and case reports, we applied the available evidence to address clinically relevant questions related to best practice in patient management. Surgery with widely negative margins remains the primary treatment of RIS. Unfortunately, the role of adjuvant and neoadjuvant chemotherapy remains uncertain. This systematic review highlights the need for additional well-designed studies to inform the management of RIS.

Introduction

Radiotherapy is an integral component of the treatment of patients with primary breast cancer; in particular, prospective, randomized trials have established the safety of breast-conserving surgery (BCS) [1]. Radiation-induced sarcoma (RIS), a rare iatrogenic malignancy, can occur after radiotherapy and is associated with poor outcomes. After radiotherapy, the cumulative RIS incidence is 3.2 per 1,000 at 15 years (versus 2.3 per 1,000 for primary sarcoma in a population without radiotherapy) [2]. The occurrence rate of RIS is low: over a 10-year period, 0.03%–0.2% [3–8]. RIS comprises about 3% of all soft-tissue sarcomas [9].

The first reported case occurred in the early 1920s [10], yet little is known about the molecular biology of this disease; consequently, no targeted therapy is available. RIS typically occurs 10 years after the index breast cancer, but the latency period can be as long as 20 years [11–13] or even as short as 6 months [14]. Radiation-induced cutaneous angiosarcoma (RIA) typically has a shorter average latency period (4 years) [15, 16]. Establishing an accurate diagnosis of RIS is critical before embarking on treatment. To be classified as RIS, Cahan and colleagues proposed that: (a) there must be evidence of an initial malignant tumor of a different histology than the putative RIS, (b) development of the sarcoma must occur in an irradiated field, (3) there must be a prolonged latency period (typically >4 years) between the two malignancies, and (d) the second malignancy must histopathologically be a sarcoma [17–19].

Methods

To evaluate evidenced-based practices for RIS treatment, we performed a systematic review of the published English-language literature. Our PubMed search used the keywords “radiation-induced sarcoma of the breast,” “radiation-induced angiosarcoma of the breast,” “second primary malignancy of the breast,” “angiosarcoma and breast,” “radiotherapy and angiosarcoma of the breast,” “radiotherapy and sarcoma,” “radiation therapy and angiosarcoma,” and “malignant fibrous histiocytoma and radiation.” By searching references within those articles, we obtained further articles.

We identified a total of 124 original articles, which included 1,831 patients reported in 1970–2010. We excluded review articles that did not include original research, studies exclusively of sarcoma of the bone, and articles not specifically related to breast cancer.

Results

We found no prospective, randomized trials on RIS of the breast; therefore, no level 1 or level 2 evidence is available to guide treatment of RIS. Using the hierarchal levels of scientific evidence proposed by Hadorn et al. [20] (Table 1), we classified and evaluated the 124 original articles. We also subdivided the number of RIS patients (n = 1,827) and key clinical data by level 3 evidence (Table 2), level 4 evidence (Table 3), and level 5 evidence (Table 4).

Table 1.

Hierarchal levels of evidence

Table 2.

Level 3 (a and b) evidence

Table 2a.

(Continued)

Abbreviation: RIS, radiation-induced sarcoma.

Table 3.

Level 4 evidence

Abbreviation: RIS, radiation-induced sarcoma.

Table 4.

Level 5 evidence (a, b, and c)

Table 4a.

(Continued)

Abbreviation: RIS, radiation-induced sarcoma.

Etiology of RIS

Given the rarity of RIS of the breast, risk factors are difficult to elucidate. With the improved survival rates of patients with primary breast cancer, the incidence of RIS may be increasing. Breast cancer and non-Hodgkin's lymphoma were the most common antecedent malignancies [9, 21]. Cozen et al. [13] found a significantly elevated risk for RIA in breast cancer patients. Previous radiotherapy has always been considered the principle risk factor for RIS, although the incidence rate is only about 0.2% in breast cancer patients treated with radiation [6, 22–25]. Breast cancer patients treated with radiotherapy had a fivefold higher risk for RIA than controls who received no radiation [26]. They were also found to have a 33% higher risk for any second malignancy [27]. Both radiotherapy and chemotherapy for breast cancer were believed to increase the risk for RIS [28]. Only eight studies examined the risk for RIS after BCS versus mastectomy, but no apparent differences were found in the rates of RIS [18, 28–34].

In breast cancer patients, several studies found a dose–response effect between the amount of radiation and the risk for RIS [35, 36]. Karlsson et al. [37] reported a direct dose–response effect, with the risk for sarcoma increasing as the integral dose increased up to 150–200 joules; after that point, the relationship was unclear. A dose of 10 Gy was reported as the minimum total required to induce RIS [38]. Clarke et al. [32] noted a 20% higher risk for a second primary malignancy after radiotherapy. However, the overall risk for RIS was small compared with the potential benefit of radiotherapy [2, 39].

It is not known how new radiotherapy techniques, such as the delivery of intensity-modulated radiation therapy (IMRT), may influence the risk for RIS [40]. Whereas IMRT can increase the volume of tissue exposed to low doses of radiation [41], and thus increase the risk for radiation-induced carcinogenesis, it has also been shown to decrease the volume of tissue exposed to high doses of radiation [42], which could exert a beneficial impact on the risk for RIS. Further, it is not known if hypofractionated regimens, which administer a higher dose per fraction but a lower total dose, will modify the risk for RIS [43]. It is plausible that hypofractionation could impact the risk for RIS either favorably or unfavorably in comparison with conventional fractionation.

Other factors have been suggested in the pathogenesis of RIS, including having a deleterious BRCA-1 mutation [44]. This defect in the DNA repair mechanism may theoretically increase radiosensitivity, increasing susceptibility to carcinogenic effects in surviving cells. Other factors possibly associated with RIS include hereditary diseases (such as Li-Fraumeni syndrome), the radiation site for the initial tumor, and the use of alkylator-based chemotherapy [10].

Our systematic review found no evidence that chemotherapy is a contributing risk factor for RIS. A randomized study by Valagussa et al. [45] involving 845 women with resectable breast cancer found that chemotherapy (cyclophosphamide, methotrexate, and fluorouracil) alone—with no postoperative radiotherapy—was not associated with a higher risk for a second malignancy.

Comparative Biology of RIS Versus Primary Sarcoma of the Breast

Both RIS of the breast and primary sarcoma of the breast had a low incidence rate of <1% of the general population [35]. A primary sarcoma presents as a firm mass within the breast without a prior history of breast cancer; RIS is slightly more difficult to identify by physical examination because of changes after radiotherapy [7, 46]. Both varieties have a wide range of histopathologic subtypes, including leiomyosarcoma, malignant fibrous histiocytoma (the most common histopathologic subtype of RIS), liposarcoma, fibrosarcoma, and angiosarcoma [37, 47, 48]. Chondrosarcoma, osteosarcoma, and other less common mesenchymal neoplasms have also been described. Ancillary studies, including immunohistochemistry, are routinely performed, especially in patients with high-grade lesions whose tumor lineage is not histologically apparent. No studies directly compared the relative incidence of different histopathologic subtypes of RIS with those of primary sarcoma. Primary sarcoma and secondary sarcoma of the breast appear morphologically identical by histopathologic analysis [4, 22]. RIS is usually a high-grade tumor; its size at excision can vary widely [49–51]. The histopathologic features of RIS may consist of spindle-shaped tumor cells, hemorrhagic tumor nodules, conspicuous mitotic figures, and necrosis [49–52].

RIS frequently presents at an advanced clinical stage; the most common sites of metastasis include the lungs and the lymph nodes [53]. On imaging, the appearances of RIS and the primary sarcoma are similar, making diagnosis difficult [54]. The most common RIS location is the chest wall [9]. Sporadic sarcomas occur most commonly in the parenchyma [55].

Gladdy et al. [56] found that the disease-free survival rate was much lower in RIS patients than in sporadic soft-tissue sarcoma patients, independent of sarcoma histologic type. In addition, toxicity is much more of a concern in RIS patients because of their previous adjuvant therapy [56]. Because RIS patients may have already received the maximum safe dose of radiation (and may have received cytotoxic chemotherapy), fewer treatment options are typically available.

Typically, RIA of the breast appears in cutaneous areas [55, 57]. Secondary angiosarcoma is more difficult to identify clinically because it presents as skin thickening and discoloration, often causing a delay in diagnosis [58, 59]. The differential diagnosis for RIA also includes inflammatory breast carcinoma, edema of the breast, fibrous histiocytoma, and cellulitis [55, 60–62]. The most common RIA is hemangiosarcoma [13].

Diagnosis

The levels of evidence for RIS ranged from level 3a to level 5c (Tables 2–4). Like other breast neoplasms, RIS is diagnosed by physical examination of the patient and by appropriate imaging modalities, which guide the histologic confirmation of RIS by tissue diagnosis. For RIA, skin changes such as skin discoloration (ranging from red to purple), elevated skin, and skin thickening are common and indicative of disease [7, 16, 38, 55, 63–69]. During a skin biopsy, the specimen should be taken from the darkest and most infiltrated area [70]. Bruising of the skin may also be evident [71]. However, skin alterations and fibrosis are typical in radiated areas. During examination of the skin, the clinician must be alert to interval changes, which, unfortunately, are not typically documented systematically (as by serial photography). A high index of suspicion for the types of changes indicative of RIS is warranted, with a low threshold for biopsy.

In addition to physical examination, radiologic imaging is the backbone of longitudinal follow-up. Moore et al. [64] suggested comparing all prior mammograms and emphasized close attention to images of the quadrant of the breast where the patient has noticed skin changes. Of note, false-negative results of mammography make magnetic resonance imaging (MRI) and incisional biopsies valuable when evaluating RIS [31, 72]. Mammography results may be negative even after skin changes have been noticed [62, 67, 72, 73]. RIS can be distinguished from Stewart-Treves syndrome, which is the occurrence of angiosarcoma in patients with lymphedematous extremities after breast cancer surgery [74].

Ultimately, histologic evaluation confirms the diagnosis of RIS (Fig. 1). As noted above, skin biopsy may be sufficient for tumors with a predilection for cutaneous involvement, such as angiosarcoma. Deeper lesions require other techniques. RIS can be identified using fine-needle aspiration (FNA); however, RIS is difficult to distinguish from recurrent carcinoma given the typically small sample volume, its morphologic similarity to carcinoma cells, and the paucity of vasoformative elements in RIS [75]. FNA cannot be expected to provide definitive information about the histologic architecture of RIS nor can FNA provide adequate material for an immunohistochemical evaluation. Therefore, core needle or incisional biopsies are preferred: they provide accurate and satisfactory sampling [51, 58, 60, 65, 76, 77]. If FNA results are negative but clinical suspicion of RIS is strong, a core needle biopsy or even an incisional biopsy that includes skin and s.c. tissue may be required to achieve a tissue diagnosis [7, 16, 68].

Figure 1.

Histologic micrograph of a radiation-induced sarcoma (RIS) of the breast. This image shows a radiation-induced malignant fibrous histiocytoma displaying a storiform growth pattern with highly atypical (often bizarre) neoplastic cell nuclei. This histologic image of RIS shows a striking low-power appearance as a result of cellularity, a vague whorled pattern, and marked nuclear atypia. The patient, a 52-year-old woman with ductal carcinoma in situ, was treated with breast-conserving surgery and adjuvant radiotherapy of 5,000 Gy. Six years later, she returned with a lump at her lumpectomy site; an incisional biopsy revealed atypical spindle-cell fibroblast proliferation consistent with a radiation-induced secondary malignancy.

MRI as well as three-dimensional reconstructions can help facilitate excellent preoperative planning of the extent of resection [30, 78]. MRI scans can reveal the spread of a tumor and predict chest wall involvement, facilitating operative intervention (Fig. 2) [78, 79].

Figure 2.

Magnetic resonance imaging (MRI) scan of a radiation-induced sarcoma (RIS) of the breast. A breast MRI scan shows gadolinium in a patient with RIS. Tumor involvement with the chest wall is present. Bulky residual disease is demonstrated between 9 o'clock and 12 o'clock along the posterior margin of the seroma cavity and at 6 o'clock on the inferior margin of the seroma cavity 18 mm from the chest wall. The more posterior residual disease directly infiltrates the pectoralis major muscle.

Prognosis

The 5-year survival rate of RIS patients was in the range of 27%–48% [5, 80]. The 5-year disease-free survival rate was 35% [7]. The number of skin lesions was an important prognostic factor. For RIA patients with multiple skin lesions, the 2-year survival rate was 0% [81]; with a single lesion, it was 50% [3]. For RIS patients overall, the median survival time was 23 months [82].

Tumor size appeared to be an important prognostic factor. For RIS patients with tumors <2 cm, the median survival time was 80 months; with tumors >5 cm, the survival time was only 20 months [30]. Blanchard et al. [18] reported that patients with a smaller mean tumor size had fewer local recurrences than patients with larger tumors. Local recurrence rates after surgical therapy alone were high, in the range of 50%–68% [5].

Tumor grade is an important prognostic factor for RIS, as reflected by its inclusion in the seventh edition of the American Joint Committee on Cancer (AJCC) Staging Manual [81]. Most RIS tumors were high-grade, with some having heterogeneity in their grade [71]. De Smet et al. [10] reported that 80% of RIS tumors, but only 40% of primary sarcoma tumors, were high grade (grade 3 per the AJCC staging system). Low-grade tumors conferred better outcomes in terms of the disease-free survival interval [82, 83]. Patients with high-grade tumors had a higher stage (per the AJCC staging system) at presentation and a 5-year survival rate of only 18%; patients with lower stage tumors had a 5-year survival rate of 47%.

Many, but not all, studies noted no lymph node involvement in RIS patients [49, 68, 69, 84], underscoring that sarcoma does not tend to metastasize via a lymphatic route. Therefore, a surgical nodal evaluation procedure (such as sentinel lymph node biopsy) is not recommended for RIS patients. Brady et al. [85] noted three important factors that have an unfavorable impact on patient mortality rate: presentation with metastatic disease, incomplete or no surgical resection, and tumor size >5 cm.

Treatment

Although surgery remains the most effective treatment for RIS patients, surgery alone often was lacking in terms of overall survival and local recurrence rates, emphasizing the need for better multidisciplinary therapy [30, 86]. Many studies demonstrated that neither chemotherapy alone nor radiotherapy alone was sufficient therapy for RIS [18, 30, 87, 88]. Complete surgical excision was critical [89], specifically, wide local excision with negative surgical margins, either by segmental or total mastectomy. Positive margins had a significantly higher risk for local recurrence [28, 30, 90]. Thijssens et al. [28] found that R0 resections (microscopically free of tumor) conferred significantly higher survival rates than either R1 (microscopically positive for tumor) or R2 (macroscopically positive for tumor) resections. No significant differences in survival rates were found between patients with R1 and R2 resections. Retrospective studies noted that surgical margins of 2–4 cm were necessary for proper disease clearance in RIS patients [66, 91, 92], although another study reported that margins of 1 cm provided successful local control [73].

A proper surgical approach is crucial to decreasing the risk for recurrence; one study suggested that, for optimal margin control, the surgeon should extirpate one plane beyond the anatomic plane of known disease [5]. En bloc chest wall excision was necessary in some patients, such as those with tumors in previously radiated fields, with extensive disease infiltration [30]. Souba et al. [93] revealed good outcomes with complete excision and chest wall resection in RIS patients. Despite poor overall survival outcomes for patients undergoing chest wall resection, this may be justified given the low postoperative morbidity and the lack of other effective treatment options [94]. No reports were available regarding the appropriateness of BCS in RIS patients.

Physical exam findings for RIA are shown in Figure 3. In all RIA patients, a total mastectomy should be performed expediently after the diagnosis [63, 84, 89, 95, 96]. Colville et al. [97] recommended a margin of 5 cm. Turner and Greenall performed a radical mastectomy with axillary clearance [98]. Axillary dissection should be avoided to prevent lymphedema [30]. Sarcoma is not expected to metastasize to regional lymph nodes; therefore, axillary lymph node dissection has no role in RIS patients [30].

Figure 3.

Radiation-induced angiosarcoma. (A): Radiation-induced angiosarcoma occurring in a patient previously radiated after mastectomy. (B): Radiation-induced angiosarcoma occurring in a patient previously radiated after lumpectomy. Photos courtesy of Dr. Chan Raut, M.D., Department of Surgery, Brigham and Women's Hospital, Boston, Massachusetts.

Role of Chemotherapy

The role of chemotherapy for RIS remains ambiguous. No level 1 or 2 studies are available to address this question for RIS because of the rarity of this disease. Most retrospective studies addressing systemic therapy in RIS patients were thus subject to significant selection bias in their comparison and treatment groups. The literature on soft-tissue sarcoma treatment can be used for general guidance, although the assumption that RIS behaves the same as primary soft-tissue sarcoma might not be entirely accurate because RIS displays more aggressive biology.

A number of studies have reported poor survival and local recurrence rates even after combining surgery with standard adjuvant chemotherapy [28, 85, 87, 99]. Lagrange et al. [82] reported no difference in the survival rate between patients treated with surgery and those treated with surgery plus chemotherapy. However, other studies have reported that, following resection with widely negative margins, adjuvant therapy may be beneficial [29, 30, 56, 84, 86, 94, 100]. Barrow et al. [30] reported favorable results after adjuvant chemoradiotherapy. Angiosarcomas have also been shown to respond to the combination of radiation therapy and surgery [77, 100]. Rosen et al. [29] observed that adjuvant chemotherapy had a beneficial effect on the disease-free survival rate for high-grade angiosarcoma patients who received adjuvant therapy versus those who did not.

Kuten et al. [87] described treatment with cyclophosphamide, vincristine, doxorubicin, and DTIC (dacarbazine) therapy. Patients in that case series were treated with surgery with this combination chemotherapy or chemotherapy alone. Unfortunately, every patient died within 6–36 months from the time of their initial diagnosis, leading them to conclude that this regimen was ineffective [87]. Brady et al. [85] identified no difference in the 5-year survival rate between patients receiving chemotherapy as part of their treatment and patients not receiving chemotherapy. Yap et al. [101] revealed that agents such as methotrexate, 5-fluorouracil, and doxorubicin combined with DTIC may give a favorable response for lymphangiosarcoma. Another study reported the sensitivity of RIA to docetaxel and paclitaxel [102].

The use of adjuvant chemotherapy in the treatment of soft-tissue sarcoma is controversial. For certain sarcoma subtypes, such as Ewing's sarcoma, rhabdomyosarcoma, and osteosarcoma, chemotherapy plays a key role in disease management. For other subtypes, adjuvant therapy has not been proven to be reliably beneficial in randomized trials. Several meta-analyses have been conducted studying randomized trials of adjuvant chemotherapy for soft-tissue sarcomas. These suggest benefit for the addition of adjuvant chemotherapy, particularly for larger, high-grade soft-tissue sarcomas of the extremities [103]. The combination of doxorubicin and ifosfamide appears to be associated with the greatest benefit [104]. The overall efficacy of this approach is modest and it is associated with significant toxicity. Whether or not RIS patients would gain similar benefits from adjuvant chemotherapy is unknown because of the rarity of this disease, limiting the ability to perform a focused trial in RIS patients.

Neoadjuvant therapy in this setting may have several advantages over adjuvant administration. Primary tumors may regress, allowing a higher probability of R0 resection, less radical resection, or even resection of previously unresectable lesions. Chemosensitivity can be assessed, with early discontinuation of inactive therapy, sparing patients from potential treatment toxicity and allowing assessment of response to alternate agents. Micrometastatic disease may also be potentially treated using this approach, although no published reports are available regarding survival outcomes with this approach for RIS. The only randomized trial of neoadjuvant therapy for soft-tissue sarcoma was halted early as a result of slow accrual [105]. Among the 134 evaluable patients enrolled, there was no difference observed in the relapse-free or overall survival outcomes. At present, neoadjuvant therapy finds it primary use in patients with unresectable or marginally resectable disease. Quadros et al. [80] reported favorable outcomes for neoadjuvant chemotherapy followed by surgery to treat RIS. However, this was the only published article identified in our systematic literature search discussing outcomes of neoadjuvant chemotherapy in RIS patients. With neoadjuvant therapy, the tumor burden may be reduced such that surgical excision is facilitated, resulting in potentially less morbidity. However, complications with tissue quality may result because of radiation exposure. In this regard, such therapy may be relevant in the treatment of breast RIS, but decisions regarding the appropriateness of neoadjuvant chemotherapy should be made on an individual basis by the multidisciplinary team.

With adjuvant radiation therapy, histologic examination of surgical pathology specimens and the design of postoperative radiation therapy are more straightforward than after neoadjuvant radiation therapy. However, postoperative radiation therapy usually results in greater toxicities [106]. Additionally, radiation is believed to be less effective for RIS because of fibrotic changes resulting in an inadequate blood supply [87].

Treatment of metastatic or unresectable disease is, at present, inadequate. Such patients are routinely enrolled in trials of therapy for primary soft-tissue sarcomas. At present, single-agent anthracycline therapy is generally viewed as standard in patients with advanced soft-tissue sarcoma. A large meta-analysis of anthracycline-based therapy indicated a response rate of 26% and median overall survival time of 51 weeks [107].

Ifosfamide, an alkylating agent, may be used as salvage therapy or may be combined with doxorubicin. Ifosfamide-containing regimens yield results similar to those of single-agent doxorubicin, with a response rate of 25%, median progression-free survival time of 19 weeks, and median overall survival time of 54 weeks [108].

Considerable controversy exists regarding the benefit of combination therapy for advanced soft-tissue sarcomas with anthracyclines (primarily doxorubicin) and ifosfamide, but such regimens also have greater toxicity [108–110]. The studies published to date have been criticized for the relatively low doses of ifosfamide used.

Single-agent doxorubicin remains the standard first-line therapy for patients with advanced soft-tissue sarcomas. Ifosfamide may be added if a possible improvement in an individual patient's response justifies the greater toxicity and logistical complexity. Both drugs are associated with significant toxicity: dose-dependent cardiotoxicity, in the case of anthracyclines, and hemorrhagic cystitis and neurotoxicity, in the case of ifosfamide [109].

Several smaller, retrospective analyses indicated that breast RIS displays sensitivity to those agents [111–115]. RIA of the breast may be susceptible to taxane-based regimens, thus justifying their use in patients with metastatic disease [116]. Until more detailed molecular analysis allows personalized selection of therapy for breast RIS patients, the relative rarity of breast RIS suggests that treatment of advanced disease should be guided by general results obtained with soft-tissue sarcoma treatment. However, the overall role of chemotherapy in RIS treatment remains to be determined by additional clinical trials. With the development of novel chemotherapeutic agents, a combined approach may ultimately increase the probability of survival for RIS patients [5, 113, 117].

Role of Radiotherapy

Few studies evaluated radiotherapy and hyperthermia in RIS patients [118, 119]. Feigenberg et al. [100] reported that neoadjuvant hyperfractionated radiotherapy was successful in RIA patients, enhancing local control and allowing removal of as much twice-radiated tissue as possible. A recent study by Palta et al. [120] found similar results with hyperfractionated and accelerated radiation doses in RIA patients; the overall 5-year survival rate in 14 patients was 86%. Notably, radiotherapy in RIS patients had less impact than chemotherapy because of the patients' tolerance of the radiation dosage from their previous radiotherapy for their breast cancer [121, 122]. A second course of radiotherapy raises concerns about toxicities such as rib fracture, pneumonitis, and soft-tissue necrosis. We classified the evidence in support of radiotherapy in RIS patients as level 3a–5c.

Discussion

The optimal management of RIS—a rare and aggressive disease—is controversial [30, 88]. As with primary sarcoma patients, surgery remains crucial as the primary therapy for RIS [28]. For patients with nonmetastatic RIS, aggressive resection with widely negative margins is necessary [5, 30]. If the tumor infiltrates beyond the breast into the chest wall, then chest wall excision may also be required to achieve an R0 resection [30].

Low survival rates underscore the need for better detection and treatment. With regard to detection, the relatively long latency period for many RIS patients underscores the importance of long-term oncologic follow-up, exceeding 5 years after radiation, performed by skilled oncologists well versed in the evaluation of the radiated breast and chest wall. With regard to treatment, adverse prognostic variables suggesting the need for neoadjuvant or adjuvant therapy include tumor size, grade, and histologic type. Complete resection may not be possible because of extensive infiltration [28]. However, the effectiveness of chemotherapy in RIS patients is uncertain. We found no prospective randomized controlled trials of systemic therapy (radiotherapy or chemotherapy) in RIS patients. Long-term surveillance of skin changes (often sites of RIS) in the radiated field is important [123]. Much of the information on RIS that we found came from case reports and retrospective, single-institution cohort studies. Such evidence is low on the hierarchal levels of evidence, limiting our ability to reach a definitive conclusion on the best evidence-based practices in the treatment of RIS. Prospective, randomized trials, registry studies, and well-designed retrospective cohort studies are needed to better inform the treatment of RIS after breast cancer.

Conclusion

This systematic review of RIS of the breast evaluated evidenced-based treatment options for this rare disease. No published articles had level 1 or 2 evidence. Therefore, only level 3–5c evidence is available to guide clinicians. Patients with RIS have a worse prognosis than primary soft-tissue sarcoma patients. Surgery with widely negative margins remains the primary treatment of RIS. Unfortunately, the role of adjuvant and neoadjuvant chemotherapy remains uncertain; systemic therapy decision making for RIS is typically based on studies in primary sarcoma because of the rarity of this disease. Additionally, the role of reirradiation for RIS remains unclear because of the paucity of studies addressing this issue. This systematic review highlights the need for additional well-designed studies to inform the management of RIS.

Author Contributions

Conception/Design: Julie E. Lang

Provision of study material or patients: Julie E. Lang, Lee D. Cranmer

Collection and/or assembly of data: Julie E. Lang, Grishma R. Sheth, Lauren Grasso-LeBeau

Data analysis and interpretation: Julie E. Lang, Grishma R. Sheth, Lee D. Cranmer, Lauren Grasso-LeBeau

Manuscript writing: Julie E. Lang, Grishma R. Sheth, Lee D. Cranmer, Benjamin D. Smith, Lauren Grasso-LeBeau

Final approval of manuscript: Julie E. Lang, Grishma R. Sheth, Lee D. Cranmer, Benjamin D. Smith, Lauren Grasso-LeBeau