Novel Form of Breast Intraoperative Radiation Therapy with CT-Guided High-Dose-Rate Brachytherapy: Interim Results of a Prospective Phase-II Clinical Trial

Abstract

BACKGROUND:

Precision breast intraoperative radiation therapy (PB-IORT) is a novel method of IORT that uses customized CT-based treatment plans and high-dose-rate (HDR) brachytherapy. We conducted a phase-II multi-institution trial to evaluate the efficacy of PB-IORT.

STUDY DESIGN:

Between 2015 and 2022, 3 centers enrolled women aged 45 years and older with invasive or in situ carcinoma measuring 3 cm or smaller and N0 status (n = 358). Breast-conserving surgery was performed, and a multilumen balloon catheter was placed in the lumpectomy bed. CT images were used to create customized HDR brachytherapy plans that delivered 12.5 Gy to the tumor bed. The primary outcome assessed was the 5-year rate of index quadrant tumor recurrence. An interim analysis was conducted after one-third of eligible participants completed 5 years of follow-up. This trial is registered with clinicaltrials.gov (NCT02400658).

RESULTS:

The cohort comprised 153 participants with a median age of 64 years and median follow-up time of 5.9 years. The estimated 5-year index quadrant tumor recurrence rate and overall survival were 5.08% (95% CI 2.23 to 9.68) and 95.1%, respectively. Locoregional (ipsilateral breast and axilla) and distant recurrence rates were each 1.96%. Seven deaths occurred during the first 5 years of follow-up, with only 1 attributable to breast cancer. Overall, 68.6% of patients experienced any adverse effects, and 4 cases of breast-related severe toxicities were observed.

CONCLUSIONS:

This study presents the results of a planned interim analysis of a phase-II trial investigating PB-IORT and demonstrates the efficacy and safety of single-fraction, CT-based, HDR brachytherapy after breast-conserving surgery. These findings provide valuable insights into the use of PB-IORT as a treatment modality.

The current standard of care for early-stage breast cancer consists of breast-conserving surgery (BCS) and adjuvant radiation therapy, typically in the form of whole breast irradiation (WBI). However, as many as 25% of women who undergo BCS do not receive the full course of adjuvant radiation due to the time commitment required for daily treatments with standard WBI.^1,2 Additionally, more than 90% of ipsilateral breast tumor recurrences occur in the same breast quadrant as initial cancer,³ which questions the necessity of WBI. As a result, clinical trials have explored approaches to minimize radiation volume (partial breast irradiation), shorten treatment duration (hypofractionated irradiation), or employ a combination of both (accelerated partial breast irradiation [APBI]).^4,5 The increased use of these alternative methods has already been shown to increase radiation treatment compliance.^6,7 Additionally, APBI is now incorporated into the National Comprehensive Cancer Network guidelines for suitable patient populations.⁸

Intraoperative radiation therapy (IORT) is a form of APBI that delivers a single fraction of radiation during BCS, maximizing patient convenience and minimizing the exposure of healthy tissue to radiation. Two randomized clinical trials have evaluated outcomes after IORT. The Targeted Intraoperative Radiation Therapy Trial (TARGIT) compared conventional breast IORT (CB-IORT) to WBI.^9–12 In this trial, CB-IORT was performed in a standard operating room, delivering 20 Gy to the tumor bed and 5 to 7 Gy at a depth of 1 cm.⁹ After a median follow-up of 2.4 years, the 5-year local recurrence rate was higher in the IORT group but found to be noninferior to WBI (3.3% vs 1.3%) based on prespecified definitions.¹⁰ Another study, the Intraoperative Irradiation Therapy for Early Breast Cancer with Electrons Trial (ELIOT), used electrons to deliver 21 Gy to the lumpectomy cavity and reported a 5-year ipsilateral breast recurrence rate of 4.4% for ELIOT compared with 0.4% for WBI (p < 0.0001).¹³ With a median follow-up time of 12.4 years, long-term results for the ELIOT trial showed a more drastic difference between 15-year recurrence rates for the 2 groups (ELIOT, 12.6% vs WBI, 2.4%) but no difference in overall survival between the groups.¹⁴

These trials showed that IORT had acceptable but higher 5-year local tumor recurrence rates than WBI and was a viable patient-centered treatment option for a select population of women with early-stage breast cancer. However, these studies were not without technical limitations. The low dose delivered (5 to 7 Gy) at depths of 1 cm or more from the applicator surface in TARGIT has been criticized for resulting in suboptimal dosimetry that is likely insufficient for sterilizing residual cancer cells. Additionally, 15.21% of the participants randomized in TARGIT received WBI in addition to CB-IORT, possibly affecting the rate of recurrence in that group.¹⁰ ELIOT’s generalizability is limited by the use of a more extensive surgical procedure than typical BCS.¹⁵ Additionally, the high long-term recurrence rates seen in ELIOT call into question its broad inclusion criteria and even prompted the identification of a subgroup of participants who may be more suitable for IORT.¹⁴ Between these studies, TARGIT’s methods are more readily applicable to most centers: it used a standard operating room and methods more widely used in the US.

In general, CB-IORT is limited by the absence of image-guided treatment planning. To overcome these limitations, precision breast IORT (PB-IORT) was developed to create a maximally effective and efficient radiation treatment option. To determine the optimal dose and depth of radiation for PB-IORT, a dosimetry study was conducted using CT scans from patients previously treated with multilumen balloon–catheter, high-dose-rate (HDR) brachytherapy APBI at the University of Virginia (UVA).¹⁶ Comparison treatment plans were formulated to simulate the specifications of a 50-kV x-ray system for CB-IORT and to develop a volume-optimized HDR brachytherapy plan. The HDR brachytherapy plans exhibited superior or similar dosimetry parameters compared with CB-IORT, with the exception of a slightly higher (yet acceptable) cardiac dose. PB-IORT uses a multilumen balloon applicator, providing improved target volume coverage compared to CB-IORT while maintaining doses within normal tissue tolerance limits.¹⁶ The dosimetry advantages of PB-IORT compared with CB-IORT were attributed to the enhanced dose homogeneity index, which represents the smoothness of a radiation plan. It allows for the delivery of higher doses beyond the applicator surface in PB-IORT (12.5 Gy) compared to CB-IORT (5 to 7 Gy) while maintaining the same applicator surface dose (approximately 20 Gy). Seeking the highest dose possible with a 20-Gy surface dose to limit the risk of fat necrosis and match that of TARGIT,¹⁶ a prescription dose of 12.5 Gy at 1 cm from the applicator was recommended for PB-IORT.

In addition to using image-guided HDR brachytherapy for personalized higher-dose IORT, PB-IORT is the first breast IORT method to incorporate CT imaging for applicator position verification and computerized treatment planning.¹⁷ CT imaging enables breast surgeons to make individualized clinical decisions, including additional tumor resection or adjustment of applicator positioning within the lumpectomy cavity to enhance conformity with the applicator. A review of the first 103 PB-IORT treatment plans revealed that 78.6% of patients required dosimetry modifications based on CT findings to protect the skin, chest wall, or both.^18,19

A phase-I trial previously demonstrated the feasibility and safety of PB-IORT.¹⁷ The primary aim of this phase-II trial was to determine the efficacy of PB-IORT. The current study presents the results from a planned interim analysis of the phase-II trial.

METHODS

This prospective multicenter phase-II clinical trial enrolled 358 participants between 2015 and 2022 and is currently closed to accrual. The study received approval from the institutional review board and all trial participants provided informed consent. For this interim analysis, all enrolled patients were from the UVA or Thomas Jefferson University Hospital (TJUH). Enrollment at TJUH began in April 2016, explaining the difference in enrollment between the 2 centers.

Eligibility criteria included women aged 45 years or older with diagnosed invasive carcinoma or ductal carcinoma in situ measuring less than 3 cm and clinical N0 disease. Exclusion criteria included: pregnancy; history of ipsilateral breast cancer treated with radiation therapy; multicentric breast cancer; known pathogenic genetic mutations; breast implants; and cancer invading the skin or chest wall. The study protocol did not allow treatment with neoadjuvant hormone therapy or chemotherapy. Adjuvant systemic therapy recommendations were left to the discretion of the treating medical oncologist.

At UVA, BCS and PB-IORT were performed in a dedicated brachytherapy suite equipped with in-room CT imaging, full anesthesia, and HDR brachytherapy capabilities (Fig. 1A).¹⁷ After BCS, a multicatheter brachytherapy balloon (Contura multilumen; Hologic, Bedford, MA) was placed in the lumpectomy bed. CT images were obtained and reviewed by the surgeon and radiation oncology team. A customized HDR brachytherapy plan was created and a 12.5-Gy dose was delivered. Subsequently, the breast surgeon removed the catheter and closed the incision. At TJUH, participants initially underwent BCS at an outpatient surgery center. Catheter placement was done through a separate stab incision. Patients recovered before proceeding to the radiation suite for CT imaging, treatment planning, radiation delivery, and catheter removal (Fig. 1B). At UVA, participants with invasive cancer underwent sentinel lymph node biopsy before BCS and PB-IORT, and final trial eligibility depended on pathologically confirmed negative nodes. At TJUH, sentinel lymph node biopsy was performed concurrently with BCS and catheter placement.

Figure 1.

Precision breast intraoperative radiation therapy (PB-IORT) steps. (A) At the University of Virginia, PB-IORT takes place in 1 fully integrated brachytherapy suite from start to finish. (B) At other participating trial sites without a fully integrated suite, like Thomas Jefferson University Hospital, PB-IORT is split into 2 locations. BCS, breast-conserving surgery.

PB-IORT was administered either during BCS (ie the prepathology cohort) or as a separate procedure within 30 days of BCS (ie the postpathology cohort; see Fig. 2). We included a postpathology cohort, despite findings from TARGIT suggesting worse outcomes, because we hypothesized that the higher radiation dose and use of CT image guidance would mitigate differences in outcomes.¹⁰ If final histopathology revealed positive surgical margins, re-excision lumpectomy to achieve negative margins and WBI were recommended. TJUH patients with positive sentinel nodes were advised to undergo WBI.

Figure 2.

Trial schematic. Trial participants were either in the prepathology or postpathology group depending on the timing of their radiation therapy in relation to their breast-conserving surgery (BCS). The pathway that they followed, starting with BCS, is shown. IORT, intraoperative radiation therapy, SLNB, sentinel lymph node biopsy, WBI, whole breast irradiation.

The study’s primary end point is the rate of breast cancer recurrence in the same quadrant as the initial cancer, known as index quadrant tumor recurrence (IQTR). Secondary end points include the rate of locoregional recurrence (any biopsy-proven breast cancer in the axilla or ipsilateral breast outside the index quadrant) and the rate of distant recurrence (recurrent breast cancer outside the breast and axilla). Additional secondary outcomes include overall survival, incidence of general adverse events, incidence of major toxicities, and physician-and patient-assessed cosmetic outcomes. Toxicities were graded at each follow-up visit according to the National Cancer Institute Common Toxicity Criteria for Adverse Events.²⁰ The treated breast was assessed at each follow-up visit by the physician and patient, using the Harvard scale, and graded as identical, slightly different, clearly different, or seriously distorted compared to the untreated breast.²¹ Breast pigmentation, size, and shape were also evaluated. For the interim analysis, only physician scoring was analyzed.

Statistical considerations

The overall trial study was designed to include 1 interim analysis of the primary outcome (IQTR), which was to occur after one-third of eligible participants had been followed for 5 years. Adverse events were described using simple frequency counts, which indicated type, maximum observed grade, and attribution. Binary outcomes were presented as percentages with 95% CI for the rates. Considering death as a competing risk, cumulative incidence distributions were used to display and estimate the 5-year incidence of index quadrant recurrence. The product-limit method of Kaplan and Meier was employed to estimate overall survival and 5-year survival probabilities.

RESULTS

Cohort characteristics

A total of 358 participants were enrolled and treated with PB-IORT between 2015 and 2022 and an interim analysis was completed when at least one-third of the total enrolled patients had been followed for 5 years, as planned. There were 153 participants included in the analysis: 133 (86.9%) treated at UVA and 20 (13.1%) treated at TJUH. Median follow-up time for the cohort was 5.9 years (interquartile range 1.3 years). The mean age of the cohort was 64.1 ± 8.2 years, with participants ranging 45 to 83 years old. The majority of women were White and non-Hispanic. Most were treated in the prepathology cohort (n = 112 [73%]; see Table 1).

Table 1.

Patient and Tumor Characteristics

Characteristic	Data (n = 153)
Age, y, mean ± SD (range)	64.07 ± 8.19 (45–83)
BMI, kg/m2, mean ± SD (range)	29.98 ± 6.20 (18.58–56.92)
Race or ethnicity, n (%)
Black	17 (11.11)
White	132 (86.27)
Asian	1 (0.65)
Unknown	3 (1.96)
Hispanic or Latino, n (%)	4 (2.61)
Prepathology, n (%)	112 (73.20)
Tumor type, n (%)
Invasive ductal carcinoma	96 (62.75)
Invasive lobular carcinoma	13 (8.50)
Invasive mucinous carcinoma	2 (1.31)
Mixed	2 (1.31)
DCIS, n (%)
DCIS only	44 (28.76)
DCIS and invasive cancer	51 (33.33)
None	58 (37.91)
Pathologic tumor size, n (%)
<1 cm	92 (60.13)
1–2 cm	52 (33.99)
>2 cm	9 (5.88)
Grade, n (%)
Low	68 (44.44)
Intermediate	82 (53.6)
High	56 (36.60)
Receptor status, n (%)
ER+	144 (94.12)
PR+	101 (66.01)
HER2+	12 (7.84)
Lymphovascular invasion present, n (%)	4 (2.61)

DCIS, ductal carcinoma in situ, ER+, estrogen receptor–positive, HER2+, human epidermal growth factor receptor 2–positive, PR+, progesterone receptor–positive.

The average tumor size was 9.6 mm: 60.1% were smaller than 1 cm, 34.0% were between 1 and 2 cm, and 5.9% were larger than 2 cm. Most tumors were invasive ductal carcinoma (62.8%); 33% of tumors were a combination of invasive cancer and DCIS, 28.76% were DCIS alone. Most tumors were estrogen receptor (ER)– and progesterone receptor–positive and Her2/neu nonamplified (Table 1).

Adverse events

In this cohort, 7 participants had positive margins, and 5 of those patients had additional surgery to re-excise the margins. Two patients did not have additional surgery, although 1 did receive WBI. Overall, 8 participants (5.2%) received adjuvant WBI: 5 for positive margins and 3 for close margins (smaller than 2 mm for pure DCIS). One participant who was recommended for WBI and margin re-excision died from a pulmonary embolism before receiving additional treatment. On final pathology, 2 participants had tumors larger than 3 cm and no patients had positive lymph nodes (Table 2).

Table 2.

Adverse Events

Adverse event	Frequency, n (%)
Positive margins	7 (4.58)
Re-excision performed	4 (2.61)
Whole breast irradiation received	8 (5.23)
Tumor >3 cm	2 (1.31)
Positive nodes	0 (0.0)

Recurrence

The study’s primary end point is the breast cancer IQTR rate, of which there were 7 (4.6%). There were 3 locoregional recurrences (1.96%), defined as occurring in the ipsilateral breast or axilla (outside the index quadrant), and 3 distant recurrences (1.96%). However, 2 participants with locoregional recurrences developed distant metastases after their initial recurrence. Therefore, they were counted as both locoregional and distant recurrences, meaning that at 5 years, 11 (7.2%) participants experienced any recurrence. Two additional participants had recurrences in their sixth year after PB-IORT (Table 3). The prepathology cohort had 6 recurrences (5.4%): 3 IQTR, 2 locoregional, and 1 distant recurrence. The postpathology cohort had 5 recurrences (12.2%): 4 IQTR and 1 locoregional recurrence. There was no significant difference between the number of recurrences in pre- and postpathology cohorts (p = 0.147). Using a cumulative incidence curve and including death as a competing risk, we estimated the 5-year rate of IQTR to be 5.08% (95% CI 2.23 to 9.68) after PB-IORT (Fig. 3).

Table 3.

Recurrence Location and Initial Tumor Pathology

Recurrence location and initial tumor pathology	Participants with recurrence within 5 y of treatment	Participants with recurrence >5 y after treatment
Index quadrant
IDC	1	1
IDC and DCIS	3	—
DCIS	3	—
Locoregional
IDC and DCIS	3	1
Distant
IDC and DCIS	1	—
Total	11	2

DCIS, ductal carcinoma in situ; IDC, invasive ductal carcinoma.

Figure 3.

Cumulative incidence of index quadrant recurrence. The estimated 5-year rate of index quadrant recurrence for precision breast intraoperative radiation therapy is 5.08% (95% CI 2.23 to 9.68) when death is included as a competing risk. IORT, intraoperative radiation therapy.

Mortality

Seven deaths occurred within the first 5 years of follow-up. Only 1 of these deaths was due to metastatic breast cancer. Six participants died after their first 5 years of follow-up with only 1 being due to metastatic breast cancer. Other causes of death included COVID-19 pneumonia, a pulmonary embolism, and new nonbreast primary cancers. The Kaplan-Meier estimate of overall 5-year survival is 95.1% (Fig. 4).

Figure 4.

Overall survival. The Kaplan-Meier estimate of overall 5-year survival is 95.1%. IORT, intraoperative radiation therapy.

Endocrine adherence

In total, 144 participants were ER-positive and therefore recommended to take adjuvant endocrine therapy (AET). Adherence to AET was measured at each follow-up visit. There were 134 participants with ER-positive cancer who did not have a cancer recurrence. Of those, 42 (31.3%) never took AET, and 55 (41.0%) completed at least 5 years of therapy. Participants without recurrences, who took AET for at least 1 month, took AET for a mean of 54.9 ± 22.9 months.

Of the 11 participants with any recurrence, 10 had an initial cancer that was ER-positive. Of the 7 participants with IQTR, 2 (28.6%) declined AET, 2 (28.6%) stopped AET early due to side effects, and 3 (42.9%) were compliant up until their cancer recurred. The 2 participants with an IQTR who discontinued AET due to side effects stopped after 2 and 20 months, respectively. Of the 3 participants with locoregional recurrences, 1 participant’s initial tumor was ER-negative and the 2 other participants were compliant with AET. All 3 women with distant recurrences were compliant with AET at the time of their recurrence.

Toxicities

Overall, 105 of the 153 participants (68.6%) experienced at least 1 toxicity. Most participants (68.6%) experienced 2 or fewer toxicities; 31.4% experienced no toxicity (Table 4). The most common toxicities included pain, dermatitis, seromas, superficial fibrosis, and hyperpigmentation (Table 5).

Table 4.

Number of Toxicities Experienced Per Patient

Toxicity per patient	Frequency, n (%)
0	48 (31.37)
1	34 (22.22)
2	23 (15.03)
3	18 (11.76)
4	8 (5.23)
5	9 (5.88)
6	6 (3.92)
7	4 (2.61)
8	1 (.65)
9	2 (1.31)

Table 5.

Breast-Related Toxicity Grading

Characteristic	Grade 1	Grade 2	Grade 3	Grade 4
Breast pain	24 (15.69)	4 (2.61)	—	—
Dermatitis	26 (16.99)	5 (3.27)	—	—
Seroma	27 (17.65)	15 (9.80)	—	—
Deep fibrosis	2 (1.31)	—	—	—
Breast infection	1 (0.65)	12 (7.84)	2 (1.31)	—
Wound infection	—	2 (1.31)	1 (0.65)	—
Superficial fibrosis	27 (17.65)	5 (3.27)	—	—
Wound dehiscence	—	1 (0.65)	—	—
Pruritus	6 (3.92)	—	—	—
Hyperpigmentation	33 (21.57)	—	—	—
Hematoma	1 (0.65)	—	—	—
Lymphedema	2 (1.31)	—	—	—

Data presented as n (%).

Severe toxicity was defined as any toxicity graded 3 or greater with skin breakdown or delayed wound healing, hematoma requiring surgical evacuation, seroma requiring more than 3 aspirations, or wound infection requiring intravenous antibiotics or surgical intervention. There were 4 cases of severe breast toxicity, including 2 breast infections, 1 seroma, and 1 wound infection (Table 4). Non–breast-related grade-3 toxicities included lymphocytopenia, generalized muscle weakness, and thromboembolism. Of the patients receiving WBI after IORT, 1 had a seroma requiring multiple aspirations, therefore qualifying it as a severe adverse event. There were no grade-4 or −5 toxicities reported.

Cosmetic outcomes

Graded according to the Harvard Scale, 94.3% of participants with available data at 24 months had excellent (E) or good (G) cosmetic outcomes (95% CI 88.4 to 97.4). Harvard scale values were compared at 6 and 24 months. At these timepoints, 97.4% of patients had cosmetic results that were stable or improved, and only 3 patients experienced worse cosmetic outcomes at 24 months compared to 6 months after treatment. Cosmetic outcomes were largely stable between 6 and 24 months, as 90.4% of participants had no change in their Harvard scale score between these timepoints. Concerning breast pigmentation, 98.4% had little to no difference between breasts (95% CI 94.1 to 100.0) at 24 months. Most patients had little to no size (83.5%; 95% CI 76.0 to 89.0) or shape differences at 24 months (89.8%; 95% CI 83.2 to 94.0). Breast size varied the most between 6 and 24 months, as only 86.5% of patients had stable findings at these time points compared to 91.3% for shape and 92.9% for pigmentation.

DISCUSSION

This study presents results from a preplanned interim analysis of a phase-II trial designed to evaluate the efficacy of PB-IORT using HDR brachytherapy after BCS. These early results demonstrate promising outcomes after PB-IORT, with acceptable recurrence and toxicity rates and favorable cosmetic outcomes. With a median follow-up of 5.9 years, the absolute 5-year IQTR rate is 4.6%.

Because this study is a nonrandomized interim analysis (n = 153), comparisons to large randomized CB-IORT or APBI trials should be made with caution, particularly before our final analysis with the full clinical trial cohort (n = 358) is completed. However, our initial IQTR rate is similar to the 3.3% local recurrence rate found in the TARGIT analysis examining 5-year results.¹¹ This is particularly notable due to the higher rate of WBI after IORT in TARGIT (15.2%) compared with our PB-IORT trial (5.2%).¹⁰ Our rate of IQTR is currently greater than local recurrence rates found in non-IORT trials, such as the Florence trial (10-year cumulative incidence, 3.7%) and the Randomized Trial of Accelerated Partial Breast Irradiation (RAPID) trial (8-year cumulative incidence, 3.0%).^22,23 We hypothesize that this is partly due to the overall higher dose of radiation that patients in these APBI studies received. While the treatment regimens in these APBI studies are more convenient than conventional WBI, we believe that PB-IORT still represents a convenient and effective treatment for appropriate patients with early-stage breast cancer. Finally, the breast cancer mortality rate for the interim cohort was low (1.3%), similar to that found in TARGIT (2.6%).¹⁰ Our rate of deaths that were not breast cancer–related (7.8%) was higher than expected; however, 2 deaths were secondary to COVID-19-related pneumonia.

IORT is a radiation treatment method driven by patients’ desire for efficient, effective, and convenient care. When interpreting the results of all IORT studies, it is essential to remember that this is a patient-centered treatment option designed to meet patient preferences while facilitating acceptable oncologic outcomes. Studies have even shown that patients and healthcare professionals are willing to incur increased risk for simpler radiation delivery, supporting the non-inferiority design of randomized IORT trials.^24–27 Our trial accrued rapidly, further supporting the idea that IORT is often sought by patients looking for breast cancer treatment options that can be completed quickly with little hassle. We have previously shown that PB-IORT patients at UVA traveled further than those undergoing breast surgery without PB-IORT,²⁸ which underscores the strong patient preference for this treatment option.

A debate currently exists regarding which patients are most appropriate for APBI and CB-IORT. Inclusion criteria from previous CB-IORT trials are varied and there is no consensus on which patients are most suitable for CB-IORT. The inclusion criteria for this study were based on previous IORT trial data and the various published APBI guidelines available in 2015.^9,13,29–31 Patient selection and inclusion of women with greater baseline risks may have impacted the higher 5-year rate of IQTR in this study compared with other APBI studies.³² In this trial specifically, patients with DCIS only were included, which differs from the inclusion criteria used in both TARGIT and ELIOT.^9,13 Notably, 3 of 7 IQTRs (42.9%) were patients with DCIS only (Table 3). Further analysis at the study conclusion will be necessary to examine how the chosen inclusion criteria, notably including patients with DCIS only, contribute to the recurrence rate.

This trial included pre-and postpathology cohorts. Performing IORT as a stand-alone procedure after BCS allows for final margin and node status to be known before radiation treatment. TARGIT reported a significant difference in recurrence rates between the pre- and postpathology cohorts: the rate of local recurrence rate in the prepathology group was 2.1% compared with a 5.4% rate of recurrence in the postpathology group.¹⁰ Fortunately, this difference in recurrence rate was not found to affect long-term overall survival.¹² Previous studies have theorized that the wound microenvironment formed in response to surgery actually promotes tumor proliferation and that this process is disrupted by immediate IORT, which theoretically decreases the impact of postpathology IORT.^12,33,34 In addition, it is possible that difficulty discerning the lumpectomy cavity weeks after the initial surgery, particularly if an adjacent tissue transfer was performed at the index operation, may affect optimal catheter placement.¹² The authors of the TARGIT trial ultimately recommend that IORT be performed concurrently with BCS.¹² However, the decision was made to include both pre-and postpathology cohorts in the phase-II PB-IORT trial because the novel intraoperative assessment of catheter placement with CT imaging and higher dose at a 1-cm depth may negate the worst outcomes previously reported for postpathology cohorts. Additionally, a postpathology group allowed patients who specifically traveled to a tertiary care cancer center for radiation treatment after having BCS at an outside local hospital to participate in this trial. With our current small sample size, it is difficult to draw conclusions about the pre- and postpathology cohorts until final analysis; however, we have not found a statistically significant difference between recurrences for the 2 groups.

There were 4 (2.61%) breast-related severe toxicities after PB-IORT, all of which were infectious. There were no grade-4 or −5 toxicities. The NSABP-B39 trial observed higher rates of severe toxicity with APBI than with WBI (10% grade 3 for APBI vs 7% grade 3 for WBI)^.5 However, other trials have found less acute toxicity with APBI compared with WBI.^22,32 The overall radiation dose received with PB-IORT is lower than with APBI as delivered in NSABP-B39, likely accounting for our study’s lower rate of severe toxicities. In this PB-IORT trial, there was a 3.92% rate of grade-3 toxicities, similar to the rate of 3.3% reported in the TARGIT study after CB-IORT.⁹ This is particularly notable considering early concerns that the increased radiation dose delivered with PB-IORT (relative to CB-IORT) may cause increased local tissue damage and worse wound healing. The ability to tailor radiation to avoid normal tissue using a multilumen balloon catheter brachytherapy and image-guided treatment planning may mitigate the risks of higher radiation doses with PB-IORT.

PB-IORT was found to have little effect on long-term cosmetic outcomes when comparing the treated and non-treated breasts. Of note, cosmetic outcomes at 6 months were indicative of long-term cosmesis, as all studied cosmetic measures remained largely stable after 6 months. At 24 months, 94.3% of treated breasts were graded as nearly identical (excellent) or slightly (good) different from the untreated breast. Comparatively, Whelan and colleagues looked at hypofractionated and standard WBIs and found that about 70% of patients had excellent or good cosmetic outcomes at 10 years after treatment.⁴ Multiple studies have compared APBI with WBI and found that APBI had worse cosmetic outcomes 3 years posttreatment.^22,35 However, studies looking at cosmetic outcomes, specifically in IORT, have reported better cosmetic outcomes after IORT than WBI.³⁶ In this trial, cosmesis is assessed by the patient, physician, and an independent plastic surgeon. Future analyses are planned to evaluate differences in patient-reported and physician-assessed cosmetic outcomes.

There are several limitations of the current study. First, since this is a phase-II study, there is no comparison arm, making it difficult to draw conclusions about PB-IORT relative to other forms of CB-IORT. Second, as previously discussed, the lack of consensus and prospectively validated data regarding who is clinically appropriate for IORT makes it challenging to know who should be offered IORT with certainty. Third, the generalizability of PB-IORT performed at UVA is limited because few institutions have access to fully integrated brachytherapy suites that include full anesthesia capability, CT-on-rails, and HDR brachytherapy. As part of this trial, however, PB-IORT was safely performed at 2 additional institutions without these capabilities. Additionally, we previously performed a time-driven activity-based cost analysis and found that the cost of the brachytherapy applicator used for PB-IORT accounts for a significant portion of the procedure’s total cost.³⁷ This should be taken into consideration when identifying which applicator to use in future iterations of PB-IORT. A final limitation of IORT, in general, is the inability to know the final pathology at the time of radiation therapy. At UVA, this was mitigated by performing the sentinel lymph node biopsy at a separate surgery before BCS for most patients. Although this allows for more informed treatment decisions, undergoing 2 separate surgical procedures adds time and expense. Even so, the 2 procedures require significantly fewer visits than WBI and did not negatively impact accrual.

CONCLUSIONS

The results from our interim analysis show that PB-IORT is a promising radiation treatment option with a low breast cancer recurrence rate, minimal side effects, and excellent cosmetic outcomes. As participant accrual has recently been completed, we look forward to the final analysis with a longer follow-up period for the full cohort. Ultimately, we advocate for a patient-centered approach to breast cancer treatment regimens that respects patient preferences. PB-IORT is a novel, image-based method of breast IORT that allows for a single episode of high-dose conformal radiation therapy that may be appropriate for selected patients with early-stage breast cancer.

Abbreviations and Acronyms

AET

adjuvant endocrine therapy

APBI

accelerated partial breast irradiation

BCS

breast-conserving surgery

CB-IORT

conventional breast intraoperative radiation therapy

ELIOT

Intraoperative Irradiation Therapy for Early Breast Cancer with Electrons Trial

ER

estrogen receptor

HDR

high-dose-rate

IORT

intraoperative radiation therapy

IQTR

index quadrant tumor recurrence

PB-IORT

precision breast intraoperative radiation therapy

TARGIT

Targeted Intraoperative Radiation Therapy Trial

TJUH

Thomas Jefferson University Hospital

UVA

University of Virginia

WBI

whole-breast irradiation