Preserving Nipple Sensitivity after Breast Cancer Surgery: A Systematic Review and Meta-Analysis

Purpose As breast-conserving procedures become increasingly safe and viable options for surgical management of breast cancer, efforts have focused on assessing and optimizing patient-reported outcome measures (PROMs), such as nipple sensation. This study aims to evaluate the current understanding of nipple-areolar complex (NAC) sensation outcomes in breast cancer patients undergoing breast cancer surgeries, namely, nipple-sparing mastectomies (NSM), skin-sparing mastectomies (SSM), and lumpectomies. Methods Articles including terms related to “nipple,” “mastectomy,” “sensation,” and “patient-reported outcome” were queried from three databases according to PRISMA guidelines. Study characteristics, patient demographics, and surgical details were recorded. Outcomes of interest included objective nipple sensitivity testing and PROMs. Results Of 888 manuscripts identified, 28 articles met the inclusion criteria. Twelve studies (n = 578 patients) used objective measures to evaluate sensitivity, such as monofilament testing. Sixteen studies (n = 1785 patients) assessed PROMs through validated or investigator-generated surveys. Three of the included studies reported NAC sensitivity in patients who received NSM with neurotization (n = 203 patients) through a variety of techniques that used various grafts to coapt a lateral intercostal nerve to the NAC nerve stumps. Results of investigator surveys showed that of 1565 patients without neurotization, nipple sensation was maintained in 29.0% (n = 453) of patients. Of 138 NSM patients without NAC neurotization, SWM testing showed an average loss of protective sensation in the nipple (average SWM score: 4.7) compared to normal or diminished sensation to light touch in nonoperated controls (average SWM score: 2.9, n = 195). Of patients who underwent NSM with neurotization, one study (n = 78) reported maintenance of NAC sensation in 100% of patients, while another study (n = 7) reported average diminished protective sensation in the nipple (average SWM score: 3.9). Conclusion Our study has shown that objective and patient-reported results of nipple sensitivity support nipple-sparing techniques as a viable option for preserving NAC sensation, although patients can expect a decrease in sensation overall. Neurotization of the NAC during NSM shows promising results of improved postoperative nipple sensitivity, though additional studies are warranted to confirm this finding. Variations between study methodologies highlight the lack of standardization in sensory testing techniques when evaluating NAC sensation.


Introduction
Breast cancer is the most common type of cancer in women globally, the majority of which require surgery to treat their disease [1]. Although radical mastectomy remains an option, surgical therapy for breast cancer has evolved to include breast-conserving procedures, such as lumpectomies and nipple-sparing mastectomies (NSM), as viable and safe alternatives [2][3][4]. According to 2016 National Comprehensive Cancer Network (NCCN) guidelines, NSMs are oncologically safe, given specifc indications such as earlystage disease, clear nipple margin, and no nipple involvement on imaging [5].
As such, studies have increasingly begun to focus on postmastectomy NAC sensation and how it may afect patient satisfaction after surgery due to improved techniques and patient-reported surgical expectations [6,7]. NAC sensation is an important factor in maintaining "normality" of the postsurgical breast and plays a major role in women's psychological and sexual health [8]. Several studies report that patients prefer to preserve nipple sensation to achieve a more normal NAC [9,10]. Although patients can undergo nipple reconstruction, they report lower satisfaction with a reconstructed nipple [11]. Terefore, a greater emphasis has been made to improve NAC sensation after nipple-sparing procedures.
Normal sensation of the breast arises from cutaneous innervation by the intercostal nerves [12]. Medial innervation of the breast is from anterior cutaneous branches of the 1 st through 6 th intercostal nerves, which produce a medial and lateral branch after passing through the deep fascia at the lateral margin of the sternum [12]. Lateral innervation of the breast arises from lateral cutaneous branches which originate from the 2 nd through 7 th intercostal nerves at the midaxillary line between the transversus thoracis and internal intercostal muscles [12]. Tese nerves travel through the external intercostal and serratus anterior muscles, giving of an anterior branch that runs over pectoralis major into the mammary gland where fne branches travel toward the NAC [12].
While current research demonstrates the importance of preserving both the structure and sensory function of the NAC, a succinct review of the current literature on nipple sensitivity after oncologic surgery does not exist. Tis systematic review and meta-analysis summarizes the current literature on NAC sensation outcomes after nipple-sparing surgeries.

Methods
For inclusion in this study, all papers included women receiving nipple-sparing surgeries and objective and/or subjective measures of NAC sensitivity. Primary study outcome was the degree of NAC sensation preservation, either by objective sensation exams or patient-reported outcomes (PROs).

Search Strategy.
Te systematic review adhered to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines and 2009 checklist as adopted from the Cochrane Collaboration. A systematic search of databases as summarized in Figure 1 was performed using Medical Subject Heading (MeSH) terms and keywords including but not limited to "nipple," "mastectomy," "sensation," and "patient-reported outcome" [13]. Full-text manuscripts available in English and published in peer-reviewed journals were included. No limitations were placed on year of publication or country of origin.

Study Selection.
Two independent reviewers screened each citation (V.H. and J.B.) using Rayyan (Qatar Computing Research Institute, Doha, Qatar) systematic review web application. First, studies were screened for relevance based on title and abstract. If a screening decision was not unanimous, a third reviewer (Z.H.) was consulted to discuss their reasoning until consensus was reached. Te remaining studies then underwent full-text review. Papers were screened for duplicate patient populations and excluded based on commonalities between author list, study period, and cancer center location.

Data Collection and Analysis.
Studies were reviewed to collect primary outcomes and factors that may have impacted these results, such as patient demographics, comorbidities, cancer characteristics, and surgical techniques. PRO measures were evaluated based on total number of patients, and objective sensory testing was evaluated based on total number of operated breasts, nonoperated breasts, and control groups.
Results of investigator-generated surveys were grouped together by reported outcomes. Outcomes such as "excellent" or "good" were recategorized as normal sensation, while "fair" and "poor" were pooled with "decreased" sensation results. We defned "overall maintained" sensation as any sensation above reported absent sensation. Reported outcomes of sensory testing using Semmes−Weinstein monoflaments (SWM) were combined using a modifed classifcation system for assessing quality of sensation as described by Imai et al. (Supplementary Table 1), which correlates higher scores with increased loss of sensation [14]. Te continuous variables were analyzed by a random efects model with statistical signifcance defned as p < 0.05.

Study Selection Process.
Te initial literature search identifed 888 nonduplicate articles. Of these, 81 abstracts were deemed relevant and underwent a full-text review. Fifty-four studies were excluded based on eligibility criteria. One additional study was found incidentally and was added to our analysis. Te remaining 28 articles were included for the systematic review, which identifed 2915 study subjects ( Figure 1).

Study
Characteristics. Study characteristics are described in Table1. Of 28 total included studies, twelve (n = 578 patients) used objective measures to evaluate sensitivity, such as monoflament testing which included SWM or von Frey hairs [7][8][9][10][15][16][17][18][19][20][21][22]. Areas of the NAC tested with these flaments are displayed in Figure 2. Sixteen studies (n = 1,785 patients) assessed patient-reported sensitivity through either validated surveys, such as Breast-Q or investigator-generated surveys [9,[21][22][23][24][25][26][27][28][29][30][31][32][33][34][35].    though age was comparable between groups (p � 0.999). 1433 patients received immediate reconstruction compared to 9 patients who received delayed reconstruction. Reconstruction timing was not reported for the remaining 1473 patients. Upon review of reported surgical techniques, three of the included studies measured NAC sensitivity in patients who received NSM with neurotization procedures (n � 203 patients). Tevlin et al. preserved the lateral intercostal nerves during the mastectomy, coapted the nerves to a nerve graft, tunneled them through a free fap, and coapted those nerves to the nerve stumps of the NAC [19]. A similar procedure was performed by Peled and Peled, though the authors specifcally reported it using 1 -2 × 70 mm nerve allografts from Avance (Axogen, Jacksonville, FL) [18]. Peled and Peled additionally used silicone implants and anterior implant coverage with acellular dermal matrices over which the nerve allograft was laid [18]. Djohan et al. also used a technique similar to that of Peled et al. regarding nerve coaptation and the use of cadaveric nerve allografts, but included patients receiving reconstruction with tissue expanders in their study population [36]. Tese techniques involved dissection of third, fourth, or ffth lateral intercostal nerves at the lateral border of the pectoralis major and coaptation to nerve allografts and remaining nerve stumps of the preserved NAC [18,19,36].

Sensory Testing.
To measure sensitivity, studies reported both objective and subjective PROMs. Objective sensitivity measures included SWM, von Frey hairs, pinprick sensation tests, and a pressure-specifed sensory device. Locations of monoflament testing are shown in Figure 2.   6 Te Breast Journal Studies varied with regards to reporting objective sensation in the nipple, areola, or both as part of the NAC. SWM testing (n � 138 patients) demonstrated an average loss of protective sensation in the nipple (average SWM score: 4.7) compared to normal or diminished sensation to light touch in nonoperated controls (average SWM score: 2.9, n � 195) ( Table 3). In the areola, 71 patients reported an overall loss of protective sensation (average SWM score: 5.5) compared to normal or diminished average sensation in nonoperated controls (average SWM score: 3.1, n � 57). Tree papers utilizing monoflament testing (n � 196) demonstrated that 50.7% of patients had preserved NAC sensation when compared to the contralateral breast in unilateral cases or nonoperated controls; however, these studies did not specify the location of sensory testing within the NAC nor provided specifc monoflament scores in their results [15,17,22].
Other studies reported sensitivity outcomes using the pinprick sensation test, two-point discrimination test, and pressure-specifed sensory devices. Chirappapha et al. used the pinprick sensation test and found that 70% (n � 7) of patients followed for a year experienced partial sensation recovery [16]. Te NAC two-point discrimination test and the pressure-specifed sensory device were used by Peled and Djohan, respectively, both of which also performed neurotization and are discussed later in this section. Due to the varied methods of objectively evaluating sensation, their results were not incorporated in our pooled analysis reported above. It was unclear if Stanec et al. (n � 288) used patient-reported outcomes or objective sensation measures, but they reported 22% of patients described normal postoperative NAC sensation, 62% reported decreased sensation, and 16% reported no sensation [38].
Of patients who underwent NSM with neurotization, Tevlin et al. (n � 7) reported average diminished protective sensation in the nipple (average SWM score: 3.9) and loss of protective sensation in the areola (average SWM score: 4.8) [19]. Peled and Peled (n � 16) reported that 87% of patients with minimum 3 months follow-up had intact 2-point discrimination [18]. Djohan et al. assessed sensory preservation using the pressure-specifed sensory device and found that patients who underwent NAC neurotization had better sensation in six of eight areas compared with nonneurotized breasts [36].

Patient-Reported Outcomes of NAC Sensation.
Patient-reported subjective methods included validated surveys such as the Breast-Q survey (n � 102 patients), the Michigan Breast Reconstruction Outcome Study Survey (n � 10 patients), and several investigator-generated surveys (n � 1643 patients). Investigator-generated surveys assessed patient-reported NAC sensitivity through various scoring systems, most commonly Likert scales. Te results showed that of 1565 patients, NAC sensation was "overall maintained" in 29.0% (n � 453) of patients, of whom 13.4% (n � 61) reported normal sensation and 35.2% (n � 132) reported decreased sensation (Table 4). 70.9% of the total patients (n � 1110) reported absent sensation (Table 4).

Discussion
As more women elect to undergo nipple-sparing methods as surgical treatment for their breast cancer, preservation of the NAC and its sensitivity has become more emphasized as a patient-centered outcome [2,39]. Our study shows that overall NAC sensation was preserved, even though there was average loss of protective sensation when evaluated using objective measures. Results from patient-reported sensitivity measures support these fndings and show maintained sensation in almost one-third of patients. Terefore, these results support the increasing success of nipple-sparing procedures as viable options for maintaining nipple sensitivity in surgical treatment of breast cancer. A 2016 literature review by Sisco and Yao reported similar results regarding sensory outcomes in NSM, particularly that 10-43% of NSM patients self-reported normal sensation [40]. Notably, 14 out of the 28 papers included in our systematic review were published in 2016 or later. Advancements in NAC sensation preservation are expected to have occurred during this time, and continued eforts should be focused on improving neurotization techniques. However, regardless of recent advancements, the fndings reported by Sisco and Yao that normal sensation is preserved to varying degrees are supported by our analysis [40].
Tree studies included neurotization of the NAC and reported preserved sensation; Peled and Peled demonstrated similar preoperative and postoperative sensation, while Djohan et al. reported decreased sensation in 83% of patients [18,19,36]. Tis diference is likely attributed to variation in the surgical technique, particularly relating to NAC reinnervation. Sensory results reported in Djohan et al. suggest that neurotization of the NAC with cadaveric nerve allografts yields lower-than-expected PROs and satisfaction compared to objective sensory outcomes. In fact, patients in this study reported similar outcomes as other studies without neurotization. Given that Djohan et al. reported a similar surgical technique as Peled, one can postulate that use of tissue expanders, placement of nerve allografts, or type of allograft used may have afected these results, the degree to which each of these technical diferences afected the observed outcomes remains unclear. Te lower-than-expected sensation preservation with allografts is further supported by Rochlin et al., which performed female-tomale nipple-sparing mastectomy with neurotization. Tis study did not use allografts and reported no signifcant diference in NAC sensation between preoperative and postoperative groups compared to a signifcant decrease in sensation in the nonneurotized control group [41]. However, Ducic et al. report that allografts may in fact be necessary to allow for tensionless repair [42]. Further studies are warranted to overcome the shortcomings with the various neurotization techniques and to assess measures of NAC sensitivity using objective monoflament testing and validated PROMs to determine if these observed outcomes are truly similar. In addition, Benediktsson et al.mentioned that the increasing difculty of NAC reinnervation as peripheral nerves is severed [15]. Te NAC is innervated by a plexus under the areola formed by variations of the second, third, fourth, and ffth intercostal nerves [43]. Tese nerves course through the gland to the posterior surface, increasing the likelihood of injury during resection of retroareolar tissue and making the preservation of the anterior branches more important [8,16]. Te frst report of sensory repair in autologous breast reconstruction used the anterior ramus of the lateral branch of the fourth intercostal nerve, which emerges at the midaxillary line after traveling through the serratus anterior muscle and later reports used the third anterior intercostal nerve, most likely due to the decreased likelihood of injury [12,44,45]. Khan et al. also found that preservation of the anterior intercostal neurovascular bundles resulted in very few reports of severe loss of light touch sensation [17]. Novel techniques, such as the use of endoscopic NSM, may be potentially successful in achieving this preservation [8].
While our study had aimed to analyze data from validated surveys such as Breast-Q and the Michigan Breast Reconstruction Outcome Study Survey, only fve included studies utilized one of these surveys [21,27,[32][33][34]. Studies using these validated measures generally reported overall patient satisfaction after surgery rather than satisfaction regarding nipple sensation. In contrast, seventeen papers implemented investigator-generated surveys, likely in order to inquire specifcally about NAC sensation. Tis highlights the need for development of a validated breast reconstruction survey that addresses NAC sensation.
Complications such as nipple asymmetry and NAC necrosis were more likely to arise in larger breasts [18,33]. Although there remains signifcant potential to improve nipple sensation, many patients reported overall satisfaction with the surgery, despite generally lower satisfaction with nipple sensation. One reason for this may be related to higher patient satisfaction with the aesthetic outcome of preserving the NAC and with the preoccupation of postmastectomy women with disease-free survival as opposed to sensation and arousal [9,22]. Djohan et al. hypothesized that decreased satisfaction with the procedure may be related to the development of complications, including NAC necrosis, nipple malposition, and delayed wound healing [9,26]. However, further studies are warranted to assess whether this correlation truly exists.
Several aspects of the included studies may additionally limit the results of this study. Te reliance on nonvalidated, investigator-generated surveys may have introduced a reporting bias within our study results. Each investigatorgenerated survey used diferent terms to categorize residual NAC sensation after surgery, which would allow the possibility for patients to interpret questions diferently, thus afecting the ability to pool results. To mitigate any error due to ambiguity of these study results, we created broadlydefned categories to include the various investigator-generated terms. Diferences between individual study questionnaires also refected the heterogeneity of our included papers, which is another limitation of our study. Of note, variation between study methodologies highlights the lack of standardization in sensory testing techniques when evaluating NAC sensation. Rodriguez−Unda et al. also described the limitations of monoflaments, specifcally the need for recalibration with repeated use [7]. Another confounding factor may be the use of diferent nonoperated control groups in our analysis, which consisted of contralateral nonoperated breasts, preoperative control testing, or patients from a nonoperated control group. It is unclear how inclusion of these control groups may have afected our results. In addition, our fndings may be limited due to the inclusion of studies on only women rather than other patient populations such as trans men undergoing female-to-male mastectomies. Finally, the number of studies reporting sensation specifc to the NAC following nipple-sparing procedures was limited and demonstrates the need for continued research in this area.

Conclusion
Te literature demonstrates that NAC sensation is preserved in nipple-sparing surgeries alongside overall satisfaction after surgery. Neurotization of the NAC may provide better sensation outcomes with limited improvement in PROs. However, studies on these reinnervation techniques were limited, and additional studies are warranted to confrm this fnding. Additionally, future studies should consider creating and utilizing validated patient surveys to allow for more standardized, patient-reported assessments of NAC outcomes. As oncological safety of nipple-sparing procedures has become widely accepted, advancements in NAC sensation preservation have improved patient satisfaction. Despite increasing success in NAC sensation preservation, however, further eforts in this area are needed to improve postoperative NAC sensation and increase patients' quality of life.

Data Availability
Te data supporting this systematic review and metaanalysis are from previously reported studies and datasets, which have been cited. Te processed data are available from the corresponding author upon request.

Conflicts of Interest
Te authors declare that they have no conficts of interest.

Supplementary Materials
Imai et al. correlated the quality of sensation with the monoflament markings and their calculated forces. Tis study adapted those classifcations when assessing objective sensory outcomes in order to evaluate the results in a standardized manner. (Supplementary Materials)