Intraoperative radiocolloid injection for sentinel node biopsy postneoadjuvant chemotherapy

j o u r n a l o f s u r g i c a l r e s e a r c h 1 9 8 ( 2 0 1 5 ) 1 4 9 e1 5 4

Available online at www.sciencedirect.com

ScienceDirect journal homepage: www.JournalofSurgicalResearch.com

Intraoperative radiocolloid injection for sentinel node biopsy postneoadjuvant chemotherapy Huan N. Vu, MD,a,* Rebecca R. Shoemaker, MD,b Patricia F. O’Connor, MD,c Wen Wan, PhD,d and Melvin J. Fratkin, MDe a

Department of Surgery and Massey Cancer Center, Virginia Commonwealth University, Richmond, Virginia Department of Physical Medicine and Rehabilitation, Wayne State University Detroit Medical Center, Detroit, Michigan c Department of Dermatology, Virginia Commonwealth University, Richmond, Virginia d Department of Biostatistics and Massey Cancer Center, Virginia Commonwealth University, Richmond, Virginia e Department of Radiology, Virginia Commonwealth University, Richmond, Virginia b

article info

abstract

Article history:

Background: This study sought to determine significance of radiocolloid injection timing for

Received 13 January 2015

sentinel node biopsy (SNB) after neoadjuvant chemotherapy (NAC).

Received in revised form

Material and methods: A retrospective comparison of intraoperative (IRCI) and preoperative

3 May 2015

(PRCI) radiocolloid injection for SNB was performed in breast cancer patients who had

Accepted 13 May 2015

completed NAC. The sentinel node identification rate (SNIDR) was tested for noninferiority

Available online 21 May 2015

by a two-proportion z-test. The differences between clinical demographics, pathologic demographics, and SNIDR were evaluated by Fisher exact test. The difference in the

Keywords:

number of sentinel nodes removed was analyzed by two-sample t-test.

Breast cancer

Results: In the 6-y study period, 120 SNB were performed after NAC: 84 received PRCI and 36

Sentinel node biopsy

received IRCI. The two groups were similar except there were fewer clinical T2 and more

Radiocolloid injection

clinical T3 and T4 with IRCI (P ¼ 0.0008). The SNIDR was 92.9% with PRCI and 80.6% with

Neoadjuvant chemotherapy

IRCI. By two-proportion z-test, IRCI was not “noninferior” (P ¼ 0.5179). By Fisher exact test, the SNIDR of the two groups did not differ. The SNIDR differs only in patients who experience T downstaging (100% versus 80%, P ¼ 0.0173). The mean number of lymph nodes removed was higher with IRCI: 3.38 versus 2.49 nodes (P ¼ 0.0068). There were more positive SNB with IRCI: 32.1% versus 55.2%, (P ¼ 0.0432). The incidence of nontherapeutic axillary dissection was similar between the two groups (3.6% for PRCI versus 5.6% for IRCI). Conclusions: IRCI for SNB after NAC may be inferior to PRCI. Published by Elsevier Inc.

1.

Background

For breast cancer, the axillary nodal status has been crucial in determining the prognosis and need for adjuvant systemic therapy. Routine axillary nodal staging with axillary dissection, however, has not been shown to affect survival but has been associated with increased risk for arm lymphedema. To

avoid nontherapeutic axillary dissection and reduce the incidence of lymphedema, axillary sentinel node biopsy (SNB) has become the standard of care for axillary nodal staging of breast cancer. With SNB, only a few axillary nodes are selected and excised for pathologic evaluation rather than an anatomically directed axillary dissection. The selection of the sentinel node frequently requires a preoperative injection of

* Corresponding author. Department of Surgery and Massey Cancer Center, Virginia Commonwealth University, PO Box 980011, Richmond, VA 23298. Tel.: þ1 804 828 9322; fax: þ1 804 828 4808. E-mail address: [email protected] (H.N. Vu). 0022-4804/$ e see front matter Published by Elsevier Inc. http://dx.doi.org/10.1016/j.jss.2015.05.020

150

j o u r n a l o f s u r g i c a l r e s e a r c h 1 9 8 ( 2 0 1 5 ) 1 4 9 e1 5 4

radiocolloid (PRCI). A hand-held radiation probe is then used intraoperatively to identify the sentinel nodes for excisional biopsy. To permit sufficient time for the radiocolloid to travel from the injection site to the axillary sentinel node, the injection is usually performed before the patient is anesthetized for the operation. The injection is directed into the ipsilateral breast, frequently into the subareolar space and can be uncomfortable for patients. Many have reported intraoperative injection of radiocolloid (IRCI) is equivalent to PRCI in patients with breast cancer [1e8]. Although SNB has become the standard of care for nodal staging of breast cancer, there remains debate over the role and timing of SNB for patients who receive neoadjuvant chemotherapy (NAC). At many institutions, SNB is performed before NAC, while at others SNB is delayed until the definitive operative management of the breast. The NSABP B-27 trial has supported SNB for breast cancer patients post-NAC [9]. But when patients presented with positive axillary node, the ACoSOG Z1071 trial suggested SNB as an alternative to routine axillary dissection post-NAC may yield an unacceptably high rate of false negative SNB at 12.6% [10]. All these studies have used PRCI. None of the IRCI studies have included patients after NAC [1e8]. This study sought to determine whether IRCI would be comparable with PRCI for sentinel node identification rate (SNIDR) in patients who have received NAC.

surgeons postanesthesia and subareolar before surgical incision for visual localization of sentinel node. Intraoperative audio localization of the sentinel node uses the C-Trak Automatic Analyzer system and the OmniProbe with collimator (Care Wise Medical Products Corp, Morgan Hill, CA).

2.3.

SNIDR is defined as the identification of at least one sentinel node among the patients who underwent SNB. Sentinel nodes are defined as hot and blue, hot but not blue, not hot but blue, and any abnormal node palpable on SNB. “Hot” is defined as nodal tissue with greater than 10% of the ex vivo count of the “hottest” sentinel node. All sentinel nodes are removed as part of the procedure. The clinical T stage (cT) and clinical N stage (cN) were obtained from the preoperative clinical records. The pathologic T stage (pT) and pathologic N stage (pN) were obtained from the pathology record of the surgical specimen. Tumor (T) downstaging is defined as a reduction of one or more T stage values from clinical T stage to the pathologic T stage. Clinically positive axillary node (cNþ) was defined as palpable and suspicious per the surgeon. Fine needle aspiration of clinically or radiographically abnormal nodes was frequently used for confirmation but was not required. Node (N) downstaging is defined as conversion from cN þ disease before NAC to pN0 stage.

2.4.

2.

Material and methods

2.1.

Study design

This is a noninferiority retrospective study. The primary objective was to compare the SNIDR between those patients with IRCI versus PRCI. The hypothesis is that the identification rate (p1) of sentinel node in patients with IRCI is not lower than the rate (p0) in those with PRCI. After institutional review board approval was obtained to collect data, eligible patients were identified through the Massey Cancer Center patient registry. Inclusion criteria included all patients undergoing SNB for breast cancer at Virginia Commonwealth University Health System (VCUHS) from January 1, 2005eDecember 31, 2010. Only patients who had received NAC before their SNB were included. Patients who were treated for breast cancer without SNB were excluded.

2.2.

SNB techniques

The Division of Surgical Oncology manages all surgical care for breast cancer at VCUHS. The injection sites of radiocolloid for sentinel node localization vary among the five active surgical oncologists of the division. Sentinel node mapping at VCUHS consists of PRCI with 1 mCi in 0.5 mL of filtered (0.22 m) Tc-99m sulfur colloid. All surgeons inject in the subareolar location; some surgeons also include peritumoral or intradermal injection. All but one surgeon usually inject the sulfur colloid agent preanesthesia in the preoperative holding area. One surgeon routinely injects after induction of anesthesia in the operating room and subareolar. Five milliliters of Lymphazurin (Covidien, Mansfield, MA) is also routinely injected by all

Definitions

Statistical analysis

The difference of the identification rate between the two groups (p1  p0) was estimated for noninferiority with a two-sided 95% confidence interval and a prespecified noninferiority margin of 10% was applied. If the lower bound of the 95% confidence interval for the estimated difference was determined to be above 10%, the IRCI would be considered noninferior to the PRCI. The noninferiority test through a two-proportion z-test was also used. If the P value was <0.05, then IRCI would be considered not worse than the PRCI group in terms of the identification rate. The Fisher exact test was used to compare the two groups in terms of demographic characters, cT and cN stages, pT and pN stages, SNIDR, identification rate by pT and pN stages, the overall rate of positive sentinel nodes, the rate of positive sentinel nodes by pT stages, and axillary dissection findings. The two-sample t-test was used to compare the number of sentinel nodes by pT stage and numerical demographic characters. A two-sided type 1 error of 0.05 was used for each test to define statistical significance. Statistical analysis was performed using SAS 9.3 (SAS Institute Inc, Rockville, MD) and GraphPad Prism 5.0 (GraphPad Software Inc, La Jolla, CA). Statistical analysis was reviewed and confirmed by a biostatistician.

3.

Results

3.1.

Study population

A total of 904 SNBs were performed at our institution from January 1, 2005eDecember 31, 2010. Of these, eleven SNBs had incomplete records and were excluded from analysis. Of the

151

j o u r n a l o f s u r g i c a l r e s e a r c h 1 9 8 ( 2 0 1 5 ) 1 4 9 e1 5 4

Figure e Sentinel node identification rate by pT stages.

remaining, 120 SNBs were in women who had undergone NAC before the SNB. Of these, 84 underwent PRCI for localization and 36 underwent IRCI. The average age of the study population was 50.8 y; 51.6 y for PRCI and 48.7 for IRCI. All patients were women. There was equivalent distribution of ethnicities between the two groups. The incidence of estrogen receptor, progesterone receptor, and Her2/Neu positivity was also equivalent between the two groups. There were more cT2 (61.9% versus 38.9%) and fewer cT3 (33.3% versus 50.0%) and cT4 (4.8% versus 11.1%) in the PRCI group (P ¼ 0.0008). There was similar incidence of clinically positive nodal disease in both groups statistically. The distribution of pT stages, T downstaging, and N downstaging was similar between the two study groups. In both groups, there were similar distributions of all pathologic T stages. The rate of tumor downstaging was the same for both groups (52.4% for PRCI versus 58.3% for IRCI; P ¼ nonsignificant [NS]). Similarly, the rate of nodal downstaging was statistically the same for both groups (53.1% for PRCI versus 35.0% for IRCI). There was a higher incidence of pathologic positive axilla after NAC in the IRCI group (32.1% for PRCI versus 58.3% for IRCI; P ¼ 0.0088; Table 1).

3.2.

Sentinel node identification

The overall SNIDR was 92.9% for PRCI and 80.6% for IRCI. The difference of identification rate between the two groups was 12.3% and is obviously larger than 10%. The noninferiority test through a proportion z-test was NS (P ¼ 0.5179). These values indicate IRCI is inferior to PRCI with the noninferiority margin of 10%. However, using the Fisher exact test (a superiority test) to compare the identification rate between the two groups, P value was 0.0589. The SNIDR were similar for patients with tumor downstaging, 84.6% for PRCI versus 81.8% for IRCI. For patients without tumor downstaging, the identification rate was 100% for PRCI versus only 80.0% for IRCI; P ¼ 0.0173. The identification rates remain equivalent for both groups when analyzed by pT stage of the tumor. When evaluated by cN stage (before NAC), there was also no difference between the two groups. There was also no difference in the identification rate for patients who had nodal downstaging (88.2% versus 81.3%) or who did not have nodal downstaging (86.7% versus 76.9%; Table 2 and Figure).

3.3.

Sentinel nodes removed

The average number of sentinel lymph nodes removed was lower with PRCI than with IRCI: 2.44 nodes for PRCI versus 3.27 nodes for IRCI (P ¼ 0.0068). When analyzed by pT stages, the

Table 1 e Patient characteristics. Patient characteristics

PRCI

IRCI

P value

Number of patients Age Sex Female: male Ethnicity, n (%) Caucasian AfricaneAmerican Other Receptor, n (%) ER PR Her2/Neu Clinical T stage, n (%) cT2 cT3 cT4 Clinical N stage, n (%) cN0 cN1þ Post neoadjuvant Pathologic T stage, n (%) pT0 pTis pT1 pT2 pT3 pT4 T downstaging, n (%) Yes No Pathologic N stage, n (%) pN0 pN1þ N downstaging, n (%) Yes No

84 51.6

36 48.7

1.0000

84:0

36:0

52 (61.9) 28 (33.3) 4 (4.8)

14 (38.9) 18 (50.0) 4 (11.1)

45 (53.6) 32 (38.1) 21 (25.0)

15 (41.7) 13 (36.1) 10 (27.8)

63 (75.0) 17 (20.2) 4 (4.8)

15 (41.7) 14 (38.9) 7 (19.4)

52 (61.9) 32 (38.1)

16 (44.4) 20 (55.6)

8 5 32 27 9 3

3 2 13 7 8 3

1.0000 0.0557

0.3193 1.0000 0.8210 0.0008

0.1074

0.4153 (9.5) (6.0) (38.1) (32.1) (10.7) (3.6)

(8.3) (5.6) (36.1) (19.4) (22.2) (8.3) 0.6896

44 (52.4) 40 (47.6)

21 (58.3) 15 (41.2)

57 (67.9) 27 (32.1)

15 (41.7) 21 (58.3)

17 (53.1) 15 (46.9)

7 (35.0) 13 (65.0)

0.0088

0.2585

Statistically significant P values are set in bold.

152

j o u r n a l o f s u r g i c a l r e s e a r c h 1 9 8 ( 2 0 1 5 ) 1 4 9 e1 5 4

Table 2 e Sentinel node identification rate.

Table 4 e Sentinel node malignancy rate.

Sentinel node identification

PRCI, n (%)

IRCI, n (%)

P value

Overall identification rate T downstaging Yes No pT stage pT0 pTis pT1 pT2 pT3 pT4 cN stage cN0 or cNx cN1þ N downstaging Yes No

78/84 (92.9)

29/36 (80.6)

0.0589

38/44 (86.4) 40/40 (100)

17/21 (81.0) 12/15 (80.0)

0.7153 0.0173

6/8 5/5 29/32 26/27 9/9 3/3

3/3 2/2 11/13 5/7 5/8 3/3

(100) (100) (84.6) (71.4) (62.5) (100)

1.0000 1.0000 0.6174 0.1006 0.0824 1.0000

50/52 (96.2) 28/32 (87.5)

13/16 (81.3) 16/20 (80.0)

0.0807 0.6949

15/17 (88.2) 13/15 (86.7)

6/7 (85.7) 10/13 (76.9)

1.0000 0.6389

(75.0) (100) (90.6) (96.3) (100) (100)

number of sentinel nodes identified was similar except for T2 tumors (2.50 for PRCI versus 4.20 for IRCI; P ¼ 0.0175) and T4 tumors (1.33 for PRCI versus 5.67 for IRCI; P ¼ 0.0255). There was more sentinel nodes sampled in pathologically positive axilla with IRCI (2.59 for PRCI versus 3.65 for IRCI; P ¼ 0.0453). For patients with clinically benign nodes (cN0), there was no difference in the number of sentinel nodes sampled. For clinically positive nodes (cNþ), there was more sentinel nodes sampled in clinically positive axilla for IRCI (2.25 for PRCI versus 3.63 for IRCI; P ¼ 0.0106; Table 3).

Sentinel node positivity

Of patients with sentinel nodes identified, there were fewer positive SNB with PRCI than IRCI (32.1% for PRCI versus 55.2% for IRCI; P ¼ 0.0432). There was no difference in the positive

Table 3 e Mean number of sentinel nodes removed. Number of sentinel nodes

PRCI

IRCI

P value

Mean number of sentinel nodes pT stage T0 Tis T1 T2 T3 T4 pN stage N0 or Nx N1þ cN stage N0 or Nx N1þ

194/78 (2.49)

97/29 (3.38)

0.0068

9/3 6/2 33/11 21/5 11/5 17/3

(3.00) (3.00) (3.00) (4.20) (2.20) (5.67)

0.5165 1.0000 0.4726 0.0175 0.7571 0.0255

124/51 (2.43) 70/27 (2.59)

35/12 (2.92) 62/17 (3.65)

0.2312 0.0453

131/50 (2.62) 63/28 (2.25)

39/13 (3.00) 58/16 (3.63)

0.3339 0.0106

15/6 12/5 76/29 65/26 22/9 4/3

(2.5) (2.40) (2.62) (2.50) (2.44) (1.33)

Statistically significant P values are set in bold.

PRCI, n (%)

IRCI, n (%)

P value

Positive sentinel biopsy 25/78 (32.1) Pathologic T stage T0 2/6 (33.3) Tis 0/5 (0) T1 8/29 (27.6) T2 10/26 (38.5) T3 2/9 (22.2) T4 3/3 (100) Clinical N stage before neoadjuvant N0 or Nx 14/50 (28.0) N1þ 11/28 (39.3)

16/29 (55.2)

0.0432

0/3 1/2 7/11 3/5 3/5 2/3

(0) (50.0) (63.6) (60.0) (60.0) (66.7)

0.5000 0.2857 0.0654 0.6254 0.2657 1.000

7/13 (53.8) 9/16 (56.3)

0.1030 0.3520

Statistically significant P value is set in bold.

SNB rate when analyzed by cN0 or cN1 before the start of NAC, Table 4.

3.5.

Statistically significant P value is set in bold.

3.4.

Positive sentinel nodes

Axillary dissection

With PRCI, six patients did not have any sentinel node identified and went on to axillary dissection. Of these six, three patients were discovered to have nodal metastasis for a positive axilla rate of 50.0% on dissection. With IRCI, seven patients who did not have any sentinel node identified underwent axillary dissection. Of these seven, five axillae contained nodal metastasis for a positive axilla rate of 71.4% on dissection. These differences were not statistically significant. The total positive axillary metastatic rate overall was higher in the IRCI group (33.3% for PRCI versus 58.3% for IRCI; P ¼ 0.0148). The incidence of nontherapeutic axillary dissection was similar between the two study groups (3.6% for PRCI versus 5.6% for IRCI: P ¼ NS; Table 5).

4.

Discussion

The ability to perform selective axillary dissection for breast cancer through axillary SNB has transformed the surgical care of patients with breast cancer. As a result, the incidence of nontherapeutic axillary dissection and its morbid consequences have been reduced. Radiocolloid detection is considered by most breast surgeons to be essential for SNB. When using a hand-held radiation probe intraoperatively, audio signaling facilitates the detection of the radioactive

Table 5 e Axillary dissection. No sentinel node: positive axilla on dissection Total axillary dissection Nontherapeutic axillary dissection No sentinel node identified: Positive axilla on dissection Total positive axilla

PRCI, n (%)

IRCI, n (%)

P value

33/84 (39.3) 3/84 (3.6)

19/36 (52.8) 2/36 (5.6)

0.2278 0.6355

3/6 (50.0)

5/7 (71.4)

0.5921

28/84 (33.3)

21/36 (58.3)

0.0148

Statistically significant P value is set in bold.

j o u r n a l o f s u r g i c a l r e s e a r c h 1 9 8 ( 2 0 1 5 ) 1 4 9 e1 5 4

sentinel node in the axilla. Although breast cancer can metastasize to the internal mammary nodes, the overall incidence is low and the instances of internal mammary node metastasis without axillary metastasis is even lower [11]. Routine sampling of the internal mammary nodes has largely been abandoned. The IRCI, however, has introduced a new morbidity: the pain associated with the preoperative retroareolar and intradermal radiocolloid injection for identification of the sentinel node. Chandler [12] reports 35% of patients attested to significant pain associated with the IRCI, which occurred regardless of the use or timing of anesthetic cream. In a randomized study of 39 patients, topical lidocaine and/or prilocaine cream versus placebo before radiocolloid injection fails to reduce the perceived pain associated with the injection [13]. The addition of 1% lidocaine to the radiocolloid solution decreases the perceived pain in a prospective randomized study of 140 patients; however, this method is associated with a decreased number of sentinel lymph nodes detected, even if the overall sentinel node identification did not change [14]. Radowsky [15] also notes providers tended to underestimate the level of pain patients experienced when the patient reported the highest pain rating. Clearly, if the IRCI can occur after anesthesia (IRCI), the pain of injection could be completely avoided. IRCI for SNB has been validated in published studies. Layeeque’s experience with 96 IRCI resulted in an SNIDR of 97% [1]. With 122 IRCI, Zogakis [2] reported a 99.2% SNIDR. Thompson [3] reported an SNIDR of 96% in 252 patients. Dauphine [4] prospectively randomized 100 patients each to PRCI or IRCI; both groups had equivalent SNIDR with 96% and 100%, respectively. Dixon [5] reported an SNIDR of 98% with 163 IRCI. Johnson [6] reported on 775 IRCI; the SNIDR remained excellent at 98.6%. Stell [7] compared the incidence of positive sentinel lymph nodes between 102 patients with PRCI and that of 112 patients with IRCI. Both groups had equivalent tumor characteristics and incidence of positive sentinel lymph nodes: 20.6% for PRCI versus 19.6% for IRCI. The VCUHS experience with PRCI and IRCI was similar to published reports. Of the 739 SNBs analyzed, 647 had PRCI and 92 had IRCI. The overall SNIDR were similar for the two groups: 93.5% for PRCI and 94.6% for IRCI [8]. The overall rates of positive sentinel nodes were comparable for the two groups: 25.6% for the PRCI group and 26.4% for the IRCI group. However, none of these studies included SNB after NAC. The greatest concern for SNB after NAC is the false negative SNB. During the conduct of NSABP B-27 study of NAC for breast cancer, 428 patients underwent SNB before the required axillary dissection; the SNIDR was 84.8% [9]. The VCUHS identification rates for PRCI (92.9%) and IRCI (80.6%) approximate the NSABP finding. Of the 218 B-27 patients with negative SNB, the false negative rate was 10.7%. The ACoSOG Z1071 trial contained 649 evaluable patients with cN1 disease who underwent chemotherapy followed by both SNB and axillary lymph node dissection. In 39 patients, cancer was not identified in the sentinel nodes but was found on axillary dissection, resulting in a false negative rate of 12.6% [10]. However, most current regimens of NAC do not save any treatment cycles postsurgical resection. The need for adjuvant radiation is predicated on breast preservation; the need for postmastectomy radiation is predicated on clinical tumor size

153

and nodal status before NAC. Because most do not view therapeutic axillary dissection as having overall or even disease-free survival benefit, the actual clinical impact of a false negative SNB after NAC is unclear. A rationale to pursue SNB after NAC would be to identify patients with complete axillary pathologic response and avert a nontherapeutic axillary dissection. Of 525 Z1071 patients, no residual cancer was identified in the axillary lymph nodes of 215 patients, yielding a pathologic complete nodal response of 41.0% [10]. In a study of 274 biopsy-proven, node-positive breast cancers before NAC, Straver [16] found the complete pathologic axillary response rate was 20.0%. Here, at VCUHS, the complete pathologic axillary response rate by SNB only is 53.1% for PRCI and 35.0% for IRCI. The axillary pathologic response for IRCI is within the published range (20%e41%). The overall incidence of pathologic positive axilla was 33.3% for PRCI and 58.3% for IRCI (P ¼ 0.0148) and is likely related to the higher incidence of advance cT stages with IRCI. However, a lower axillary response rate and a higher pathologic axilla rate may suggest a lower likelihood of a false negative SNB. Both the PRCI and IRCI had similar and low incidence of nontherapeutic axillary dissection at 3.6% and 5.6%, respectively. There are several limitations associated with this nonrandomized study. First, there is no quantification of pain associated with injection of the radiocolloid. However, those undergoing IRCI after anesthesia cannot report any pain with injection. Quantification of pain would be limited to those undergoing PRCI, and any measured pain would certainly be different when compared to none for those undergoing IRCI. Second, the two study groups may not be comparable with differences in cT stages. The IRCI had lower proportion of cT2, higher cT3, and a higher incidence of positive axillary nodes. Third, the overall numbers analyzed, 84 and 36, are small and may limit statistical significance. Fourth, there is no standardization of injection site for SNB due to surgeon’s preference; some patients underwent peritumoral or intradermal injection as well as retroareolar injection. Randomized studies, however, have endorsed the retroareolar injection site, used in all studied patients, as reliable [17e20]. Fifth, the time interval from radiocolloid injection to surgical incision is not quantified for any of the studied patients. Empirically, there is about 1 h from injection to incision for the PRCI group and 15 min for the IRCI group. Because the analysis compares injection before or after the induction of operative anesthesia, the actual minutes count was not deemed important. Finally, and most importantly, there is no routine axillary dissection thus no true negative control for analysis. However, the rate of nontherapeutic axillary dissection was similar.

5.

Conclusions

For breast cancer patients who have undergone NAC, IRCI for SNB results in lower sentinel lymph node identification as compared with PRCI but the rate of nontherapeutic axillary dissection was similar. Though empirically evident, this study does not provide data quantifying the differences in pain associated with PRCI as compared with IRCI, the primary advantage for IRCI for SNB. Breast centers adopting IRCI should be cautious in its application in patients post-NAC.

154

j o u r n a l o f s u r g i c a l r e s e a r c h 1 9 8 ( 2 0 1 5 ) 1 4 9 e1 5 4

Acknowledgment The authors thank Jenni Brumelle PhD. Authors’ contributions: H.N.V. contributed to the conception, design, and article preparation. H.N.V. and W.W. did the data analysis. R.R.S. and P.F.O. collected the data. M.J.F. performed the data and article review.

Disclosure H.N.V., R.R.S., P.F.O., W.W., and M.J.F. report neither any disclosure nor any financial support.

references

[1] Laveeque R, Kepple J, Henry-Tillman RS, et al. Intraoperative subareolar radioisotope injection for immediate sentinel lymph node biopsy. Ann Surg 2004;239:841. discussion 845e8. [2] Zogakis TG, Wetherille RE, Christensen RD, et al. Intraoperative subareolar injection of 99mTc-labeled sulfur colloid results in consistent sentinel lymph node identification. Ann Surg Oncol 2005;12:167. [3] Thompson M, Korourian S, Henry-Tillman R, et al. Intraoperative radioisotope injection for sentinel lymph node biopsy. Ann Surg Oncol 2008;15:3216. [4] Dauphine CE, Khalkhali I, Vargas MP, Isaac NM, Haukoos J, Vargas HI. Intraoperative injection of technetium-99m sulfur colloid is effective in the detection of sentinel lymph nodes in breast cancer. Am J Surg 2006;192:423. [5] Dixon JM, Mak C, Radhakrishna S, et al. Effectiveness of immediate preoperative injection of radiopharmaceutical and blue dye for sentinel node biopsy in patients with breast cancer. Eur J Cancer 2009;45:795. [6] Johnson CB, Boneti C, Korourian S, Adkins L, Klimberg VS. Intraoperative injection of subareolar or dermal radioisotope results in predictable identification of sentinel lymph nodes in breast cancer. Ann Surg 2011;254:612. [7] Stell VH, Flippo-Morton TS, Norton HJ, White RL Jr. Effect of intraoperative radiocolloid injection on sentinel lymph node biopsy in patients with breast cancer. Ann Surg Oncol 2009; 16:2300. [8] Vu HN, O’Connor PF, Shoemaker RR, Wan W, Fratkin MJ, Bear HD. Intraoperative injection of radiocolloid for sentinel node biopsy in breast cancer. J Nucl Med Technol 2013; 41:263.

[9] Mamounas EP, Brown A, Anderson S, et al. Sentinel node biopsy after neoadjuvant chemotherapy in breast Cancer: results from National Surgical Adjuvant Breast and Bowel Project Protocol B-27. J Clin Oncol 2005;23:2694. [10] Boughey JC, Suman VJ, Mittendorf EA, et al. Alliance for Clinical Trials in Oncology. Sentinel lymph node surgery after neoadjuvant chemotherapy in patients with nodepositive breast cancer: the ACOSOG Z1071 (Alliance) clinical trial. JAMA 2013;310:1455. [11] Mansel RE, Goyal A, Newcombe RG. ALMANAC Trialists Group. Internal mammary node drainage and its role in sentinel lymph node biopsy: the initial ALMANAC experience. Clin Breast Cancer 2004;5:279. discussion 285e6. [12] Chandler KJ, Hunt CH, Morreale R, Johnson GB, Peller PJ. Effect of patient age, breast density, and topical anesthetic cream on perceived pain with sentinel lymph node scintigraphy. J Nucl Med Technol 2012;40:44. [13] O’Connor JM, Helmer SD, Osland JS, Cusick TE, Tenofsky PL. Do topical anesthetics reduce periareolar injectional pain before sentinel lymph node biopsy? Am J Surg 2011;202:707. discussion 711e2. [14] Arciero CA, Henry LR, Howard RS, et al. Technical effects of adding 1 % lidocaine to technetium sulfur colloid for sentinel lymphatic mapping in early breast cancer: analysis of data from a double-blind randomized controlled trial. Ann Surg Oncol 2013;20:2548. [15] Radowsky JS, Baines L, Howard RS, Shriver CD, Buckenmaier CC 3rd, Stojadinovic A. Pain ratings by patients and their providers of radionucleotide injection for breast cancer lymphatic mapping. Pain Med 2012; 13:670. [16] Straver ME, Rutgers EJ, Russell NS, et al. Towards rational axillary treatment in relation to neoadjuvant therapy in breast cancer. Eur J Cancer 2009;45:2284. [17] Smith LF, Cross MJ, Klimberg VS. Subareolar injection is a better technique for sentinel lymph node biopsy. Am J Surg 2000;180:434. discussion 437e8. [18] Mateos JJ, Vidal-Sicart S, Zano´n G, et al. Sentinel lymph node biopsy in breast cancer patients: subdermal versus peritumoural radiocolloid injection. Nucl Med Commun 2001;22:17. [19] Povoski SP, Olsen JO, Young DC, et al. Prospective randomized clinical trial comparing intradermal, intraparenchymal, and subareolar injection routes for sentinel lymph node mapping and biopsy in breast cancer. Ann Surg Oncol 2006;13:1412. [20] Rodier JF, Velten M, Wilt M, et al. Prospective multicentric randomized study comparing periareolar and peritumoral injection of radiotracer and blue dye for the detection of sentinel lymph node in breast sparing procedures: FRANSENODE trial. J Clin Oncol 2007;25:3664.