Function and Strength after Free Abdominally Based Breast Reconstruction: A 10-year follow-up

Abstract

BACKGROUND:

The long-term impact of abdominally-based free flap breast reconstruction is incompletely understood. The aim of this study is to provide long-term, subjective and objective health data on abdominally based free flap breast reconstruction patients with specific attention to the effects of laterality, flap type and obesity.

METHODS:

Patients enrolled in this prospective study between 2005 and 2010 and completed preoperative, early (<1 year) and long-term(5–10 years) evaluations. Objective examination included an assessment of upper (UA) and lower (LA) abdominal function and a functional independence measure (FIM). Patient-reported outcomes included the Short-Form 36 (SF36) and the Breast-Q abdominal well-being module. Scores were compared by laterality (unilateral vs bilateral), flap type (msfTRAM vs.DIEP), and presence of obesity.

RESULTS:

Fifty-one patients were included with an average of 8.1 years follow-up. Overall, 78.8% of patients had stable or improved scores across the UA, LA and FIM, and minimal objective differences across flap laterality or types were observed. Post-operative scores improved for SF36 Physical health (p<0.001) and Mental health (p<0.001), and did not differ based on laterality or flap type. Obesity negatively impacted Physical Health (p=0.002) and Mental Health (p=0.006).

CONCLUSION:

Abdominally based autologous breast reconstruction is associated with significant improvements in long-term quality of life across key domains of physical and mental health with little functional impairments, and no long-term differences across flap type or laterality. Obese patients, however, may be at risk for subjective physical and mental health impairment, perhaps unrelated to the surgery itself.

Introduction

Donor site morbidity following abdominally based autologous breast reconstruction has been a major focus of concern and a topic of intense clinical research focus since it was first described[1, 2]. Such studies began to emerge in the 1990s, and often focused on differences between flap types, driving the field towards free muscle sparing, fascial sparing, and ultimately perforator based reconstructions, as minimal muscle harvest tended to correlate with improved postoperative function[3, 4] and less risk of hernia and bulge[5–7]. Early functional studies culminated in a systematic review by Atisha and Alderman, which further demonstrated objective differences based on flap type, but minimal subjective differences[8].

In 2009, our group published a prospective study examining abdominal wall function after abdominally-based breast reconstruction and found differences by flap type in early function based closely along the degree of rectus muscle sacrifice. However, the study found that patients with a greater decrease in early postoperative function returned towards their baseline one year after surgery and did not objectively differ from patients with less muscle sacrifice who had advanced early function[9, 10].

Recently, several cross-sectional studies have demonstrated a favorable long-term profile for DIEP flaps, with good quality of life more than 5 years out from surgery and with no defined difference from msfTRAM flaps when examined[11, 12]. Yet, the cross-sectional nature of these studies limits the conclusions which can be reached and highlight a need for further outcomes research using more robust longitudinal and prospective data.

Furthermore, obesity continues to be an epidemic, with multiple adverse health and functional impacts[13–16]. Recent studies have demonstrated that obesity places patients at higher risk for complications after breast reconstruction and can impact early functional recovery, especially in bilateral reconstruction patients[17–21]. The long-term impact of autologous reconstruction on these patients is also undefined.

To date, there is a paucity of objective and subjective data on long-term physical function after autologous breast reconstruction. This information is critical as the risks and benefits of abdominally based breast reconstruction is discussed with the patient preoperatively. The aim of this study was to assess the long-term functional impact of abdominally-based autologous breast reconstruction on the abdominal wall with regards to function and overall health using a prospective, observational cohort. We sought to examine this issue overall and with specific attention to laterality, flap type and obesity.

Methods

This study is a prospective, blinded (to the observer) cohort study of patients undergoing autologous, abdominally based breast reconstruction by the senior author (JMS) at an academic, tertiary care medical center. Patients were enrolled from December 2005- August 2010, and received follow-up in 2015 and 2016. Early findings were published as a two-part series and detailed the early execution of the study[9, 10].

Female patients undergoing autologous breast reconstruction with abdominal tissue (msfTRAM, DIEP or SIEA flaps) at the main hospital in the health system were eligible to enroll. Patients undergoing immediate, delayed, unilateral, and bilateral reconstruction were included. Patients were excluded if reconstruction was planned using non-abdominally based reconstruction modalities. For this long term portion of the study, we chose to exclude SIEA patients given the small number of patients with SIEA flaps in the initial study with follow up (unilateral = 3, bilateral = 10) which made meaningful subgroup analysis and interpretation difficult. The initial flap choice was guided by an intraoperative anatomic algorithm based on size, location and number of perforators[22].

Study Procedures

All patients were invited to continue participation via postal mail and follow-up telephone calls. Respondents completed a study specific strength assessment and two quality of life questionnaires, the Breast-Q and Short-form 36 (SF36) and then received a $25 gift card and validated parking. Patients who did not initially respond were again re-contacted after 1 month.

Strength Testing

The abdominal wall strength assessment focused on upper abdominal (UA) function, lower abdominal (LA) function and functional independence (FIM)(See Video, Supplemental Digital Content 1, which shows the Abdominal Wall Strength Assessment, http://links.lww.com/PRS/D173) (Video Graphic 1). Tests were chosen based on their ability to isolate the upper and lower rectus abdominis muscle [23, 24]. To assess the UA, patients were placed in a supine position and graded on a scale based on the ability to raise the inferior border of the scapula off the bed (Table 1a, scale 0–5). To assess the LA, the supine patient raised her legs to 90 degrees with knees extended, then slowly lowered her legs back to the bed. Function was graded based on the angle at which point posterior pelvic tilt was lost (Table 1b, scale 0–5). The FIM assessed a patient’s ability to independently arise from a supine position. Patients were graded based on the degree of assistance required to perform the task (Table 1c, scale 0–7). Patients were assessed by trained residents or research fellows.

Video Graphic 1.

Supplemental Digital Content 1, See Video, which shows the Abdominal Wall Strength Assessment, INSERT HYPER LINK.

Table 1a.

Grading Scale for the Upper Rectus Abdominis Manual Muscle Function Test

Grade	Result
5	Patient can raise inferior angles of scapulae off table with hands clasped behind head
4	Patient can raise inferior angles of scapulae off table with arms across chest
3	Patient can raise inferior angles of scapulae off table with arms outstretched in full extension above plane of body
2	Patient can raise head off table only
1	Patient cannot raise head off table, but a visible or palpable abdominal muscle contraction occurs
0	Patient cannot raise head off table, and there are no visible or palpable abdominal muscle contractions

Table 1b.

Grading Scale for the Lower Rectus Abdominis Manual Muscle Function Test

Grade	Result
5	The angle between the lower extremities and the table is 0 degrees; the patient is able to lower the legs completely to the table without losing the posterior pelvic tilt
4	The angle between the lower extremities and the table is 30 degrees when the posterior pelvic tilt is lost
3	The angle between the lower extremities and the table is 70 degrees when the posterior pelvic tilt is lost
2	The angle between the lower extremities and the tale is greater than 75 degrees when the posterior pelvic tilt is lost
1	The patient is unable to assume or maintain the position, but a palpable contraction of the rectus abdominis is present
0	No palpable contraction is present

Table 1c.

Grading Scale for the Functional Independence Measure

Grade	Result
7	Patient is able to sit at the edge of the table without assistance in a reasonable amount of time and in a safe manner
6	Patient is able to sit at the edge of the bed using an assistive device, such as a bedrail, or is able to sit at the edge of the bed but requires greater than a reasonable amount of time, or there are safety concerns
5	Patient requires supervision, such as cuing or coaxing, or setup to accomplish the task
4	Patient requires minimal contact assistance, which is defined as no more help than touching, and the patient performs 75% or more of the work of the task
3	Patient require moderate assistance, which is defined as more help than touching, or the patient performs 50–74% of the task
2	Patient requires maximal assistance, in which patient performs 25–49% of the task
1	Patient requires total assistance, in which patient performs <25% of the task
0	Activity does not occur

Patient Reported Outcomes

Patients completed the SF36 form and Breast-Q at the follow-up visit. The SF36 provides an assessment of the patient’s general health with both a physical health component (PHC) and mental health component(MHC)[25], while the Breast Q was developed specifically to examine the subjective quality of life of the breast reconstruction patient[26]. The Breast Q was added to this portion of the study, as the tool was unavailable during the initial 2005–2010 enrollment period. For this study, we assessed only the abdominal well-being module.

Study Variables

Pre-operative physical assessment scores and SF36 scores were carried forward for each patient. Additional variables examined included baseline patient characteristics (age, body mass index [BMI], hypertension, chronic obstructive pulmonary disease, hyperlipidemia, active smoking, coronary artery disease, and peripheral arterial disease)(reported as a Comorbid condition sum), reconstructive details (immediate vs. delayed, unilateral vs. bilateral, flap type), intraoperative complications (venous or arterial thrombosis), and postoperative surgical complications (flap loss [partial and total], delayed breast or donor site wound complications, infection, seroma, and hematoma). Delayed wound healing at the abdominal donor site and mastectomy site was defined as skin necrosis or wound breakdown necessitating topical care or dressing changes for more than three weeks. Fat necrosis was defined as a palpable firmness greater than 1 cm in diameter. Partial flap loss was defined as flap loss or atrophy up to 50 percent but not requiring immediate return to operating room. Superficial infection was defined as incisional cellulitis at either the donor site or breast that occurred within 30 days. We characterized complications as minor (wound healing, seroma, infection, partial flap loss, and fat necrosis) or major (flap loss and hernia).

Statistical Analysis

Long term respondents were compared to those in the initial enrollment cohort with respect to demographics, surgical characteristics and complications. We then examined the cohort overall to determine any changes which occurred over time. Patients were then compared by laterality (comparing unilateral reconstructions to bilateral reconstructions), flap types (within unilateral (msfTRAM vs. DIEP) and bilateral reconstructions (msfTRAM/msfTRAM vs. DIEP/DIEP vs. msfTRAM/DIEP)) and by obesity (obese vs. non-obese).

Data was entered into a REDCap database, and exported to Microsoft Excel. Statistical analysis included Chi-squared and Fisher’s exact tests (where appropriate) for categorical variables and the Wilcoxon-Mann-Whitney test, Wilcoxon Signed Rank test, and Kruskall Wallis test (where appropriate) for continuous variables. Linear regression analyses were performed on the cohort overall then by laterality to control for confounding factors. All tests were two-tailed and statistical significance was defined as p<0.05. All analyses were performed using STATA IC 13.0 (StataCorp, College Station, TX).

Results

One-hundred-fifty-nine patients who enrolled and completed early follow-up were eligible for participation. Of these, 51 (32%) completed long-term follow-up, with the average follow-up time being 8.2 years (range 6–10)(Figure 1). Participants did not significantly differ from non-responders with regards to baseline demographics, past medical history, or surgical characteristics/flap types (table 2).

Figure 1.

Study Attrition Diagram

Table 2 –

participant characteristic analysis

Variables	No Response	Participants	p
N	108 (68%)	51 (32%)	-
Age	50.7 (8.1)	52.3 (8.1)	0.4
Years from Reconstruction	8.3 (1.5)	8.1 (1.3)	0.4
Distance from center (miles)	69.4 (165)	31.6 (26)	0.2
BMI	29.1 (5.8)	27.8 (4.9)	0.2
Obesity	40 (37%)	15 (29%)	0.4
Any Smoking History	53 (49%)	27 (53%)	0.7
Comorbidity Sum ≥2	39 (36%)	17 (33%)	0.9
Hernia Complication	5 (4.6%)	2 (3.9%)	1.0
Minor Surgical Complication	64 (59%)	25 (49%)	0.2
Major Surgical Complication	3 (2.7%)	1 (1.9%)	1.0
Bilateral	60 (56%)	27 (53%)	0.9
TRAM/TRAM	37 (62%)	20 (74%)	0.4
DIEP/DIEP	6 (10%)	3 (11%)
TRAM/DIEP	17 (28%)	4 (15%)
Unilateral	48 (44%)	24 (47%)	0.9
TRAM	35 (73%)	15 (63%)	0.4
DIEP	13 (27%)	9 (38%)

Examining objective overall results, 78.8% of patients had stable or improved scores across the UA, LA and FIM. No significant differences were noted across time comparing preoperative to long-term scores for UA function (p=0.91) or FIM (p=0.19) (table 4). LA function demonstrated a slight improvement in score over time (p=0.002). Subjective patient reported outcomes demonstrated improvement in the SF36 PHC (p<0.001) and MHC (p<0.001) comparing preoperative to long-term follow-up (table 5), with 90% of patients overall demonstrating improvement in SF36 PHC compared to preoperative values. Overall Breast-Q abdominal well-being at long-term follow-up was 82.5 (table 6).

Table 4:

Abdominal Function Analysis

	N	Upper rectus function					Lower rectus function					Functional independence measure
		pre	early	med	Long	Δ	pre	early	med	Long	Δ	pre	early	med	Long	Δ
Overall	51	4.6	3.7	4.2	4.6*	0.04	4.4	3.6	4.0	4.8^	0.4	6.9	6.4	6.8	6.9*	0.03
Unilateral	24	4.5	3.9	4.4	4.7*	0.3	4.3	3.6	4.1	4.8^	0.5	6.8	6.4	6.8	7.0^	0.2
Bilateral	27	4.7	3.6	4.1	4.4*	−0.3	4.6	3.7	4.0	4.8*	0.2	7.0	6.3	6.7	6.9*	−0.1
P value		0.3			0.2	0.1	0.3			0.6	0.2	0.03			0.2	0.02
Unilateral	24
TRAM	15	4.5	3.7	4.4	4.8*	0.3	4.5	3.5	4.1	4.9*	0.4	6.9	6.1	6.9	7.0*	0.13
DIEP	9	4.1	4.1	4.3	4.6*	0.5	3.9	3.6	4.1	4.7*	0.7	6.8	6.8	6.6	7.0*	0.22
P value		0.8			0.2	1	0.1			0.03	0.4	0.6			1.0	0.6
Bilateral	27
TRAM/TRAM	20	4.7	3.4	3.9	4.4*	−0.3	4.7	3.6	3.7	4.8*	0.1	7.0	6.3	6.7	6.9*	−0.2
DIEP/DIEP	4	4.68	4.7	5.0	5.0*	0.3	4.7	4	5	5.0*	0.3	7.0	6.5	6.8	7.0*	0.0
TRAM/DIEP	3	4.8	3.0	4.3	4.8*	0.0	4.8	3.5	5	5.0*	0.3	7.0	6.0	7.0	7.0*	0.0
P value		0.9			0.3	0.4	0.9			0.7	0.9	1.0			0.9	0.9
Obese	15	4.3	2.9	3.6	4.7*	0.5	4.5	3.5	4.1	4.7*	0.1	6.8	6.2	6.7	6.8*	0.0
Non-obese	36	4.7	4.0	4.4	4.5*	−0.2	4.4	3.7	4.0	4.9^	0.5	6.9	6.4	6.8	7.0*	0.1
P value		0.3			0.2	0.04	0.4			0.07	0.1	0.04			0.1	0.2

^*pre vs. long – p>0.05,

^{^}pre vs. long – p<0.05,

⁺pre vs. long – p<0.001

Table 5:

SF36 Analysis

	N	SF36 Physical Health Component					SF36 Mental Health Component
		pre	early	Med	Long	Δ	pre	early	med	Long	Δ
Overall	51	65.3	56.9	68.0	77.5+	12.2	77.0	73.2	82.5	84.6+	7.6
Unilateral	24	66.3	57.6	68.1	76.5^	10.1	79.4	73.9	81.3	85.3^	5.9
Bilateral	27	64.4	56.1	68.0	78.8+	14.3	75.2	72.4	83.6	84.1^	9.0
P		0.9			0.9	0.43	0.7			0.9	0.6
TRAM	15	65.1	56.0	65.5	76.8^	12.5	76.0	69.6	78.8	83.9^	7.9
DIEP	9	68.5	60.0	72.1	75.9^	7.4	85.2	80.4	85.2	87.6*	2.5
P		0.7			0.6	0.9	0.7			0.9	0.7
TRAM/TRAM	20	62.4	54.6	68.3	77.5+	18.1	73.7	75.7	84.1	85.3^	13.7
DIEP/DIEP	4	65.0	53.8	68.3	91.7^	26.7	74.2	64.2	77.5	80.3*	6.12
TRAM/DIEP	3	72.3	70.0	65.1	74.0*	1.7	79.6	61.8	85.4	80.3*	0.7
P		0.2			0.2	0.3	0.9			0.6	0.6
Obese	15	66.6	55.8	67.2	63.1*	−3.5	78.6	72.9	82.9	75.2*	−3.4
Non-obese	36	64.9	57.3	68.2	83.6+	18.7	76.6	73.3	82.4	88.5+	11.9
P		0.4			0.001	0.003	0.4			0.005	0.02

^*pre vs. long – p>0.05,

^{^}pre vs. long – p<0.05,

⁺pre vs. long – p<0.001

Table 6:

BREAST Q – abdominal well-being analysis

	Overall	Unilateral	Bilateral	p	TRAM	DIEP	p	TRAM/TRAM	DIEP/DIEP	TRAM/DIEP	p	Obese	Non-Obese	p
N	51	24	27		15	9		20	4	3		15	36
Abdominal Well Being	82.5 (18.8)	86.2 (16.6)	79.3 (20.2)	0.1	83.6 (19.9)	90.4 (8.2)	0.8	75.7 (21.5)	96.3 (6.4)	84.5 (12.9)	0.1	80.7 (23.0)	83.3 (17.0)	0.9

Laterality

The analysis included 24 patients who underwent unilateral reconstruction and 27 who underwent bilateral reconstruction. Comparison of unilateral and bilateral reconstructions for baseline characteristics yielded no significant differences (table 3), though bilateral reconstructions trended towards being younger (50 vs. 54, p=0.08).

Table 3:

subject demographics and complication analysis

	Unilateral			Bilateral				Obesity
	msTRAM	DIEP	P value	msfTRAM	msfTRAM/DIEP	DIEP	P value	Obese	Non-obese	P value
N	15	9		20	3	4		15	36
Age	54.8 (8.3)	53.8 (8.1)	0.9	51.7 (7.9)	47.9 (9)	44.9 (4)	0.3	52.5 (8.9)	52.2 (7.9)	0.9
Years from Reconstruction	8.2 (1.5)	8.6 (1.01)	0.5	8.0 (1.3)	7.8 (1.7)	7.7 (1.2)	0.9	8.1 (1.1)	8.1 (1.5)	0.9
Miles from center	25.8 (24.3)	16.7 (9.6)	0.4	27.1 (15.9)	45.3 (75.1)	45.2 (14.2)	0.1	25.8 (25)	27 (4.7)	0.9
BMI	27.8 (5.3)	28.8 (6.03)	0.7	27.9 (6.0)	26.8 (5.5)	25.5 (3.2)	0.6	32.3 (5.3)	25 (4.9)	<0.001
Obesity	5 (33.3)	4 (44.4)	0.7	5 (25)	1 (25)	0 (0)	0.6	15 (100)	0 (0)	<0.001
Any Smoking History	7 (46.7)	5 (55.6)	1	12 (60)	2 (50)	1 (33.3)	0.7	11 (73.3)	17 (45.9)	0.1
Comorbidity Sum ≥2	6 (40)	3 (33.3)	1	6 (30)	2 (50)	0 (0)	0.4	12 (80)	5 (13.5)	<0.001
Hernia Complication	1 (6.7)	0 (0)	1	1 (5)	0 (0)	0 (0)	0.8	0 (0)	2 (5.4)	1
Minor Surgical Complication	8 (53.3)	3 (33.3)	0.4	9 (45)	3 (75)	2 (66.7)	0.5	10 (66.7)	15 (40.5)	0.1
Major Surgical Complication	0 (0)	0 (0)	0	1 (5)	0 (0)	0 (0)	0.8	0 (0)	1 (2.7)	1

Examining intra-cohort objective outcomes, unilateral reconstructions were noted to have slight improvements over time in LA function as well as FIM (table 4). However, no differences were noted in bilateral patients comparing scores from preoperative time points to long-term follow-up. Comparing long-term scores, no significant differences were noted in raw objective scores. Change over time differed statistically in FIM, with slight improvements in scores were for unilateral reconstructions with slight decreases in bilateral reconstructions.

Patient reported data also demonstrated significant findings. SF36 PHC improved significantly over time for unilateral and bilateral reconstructions (table 5). SF36 MHC analysis mirrored these findings. However, no significant differences were noted at the long-term follow-up comparing unilateral and bilateral reconstructions for either the SF36 PHC or MHC score, or the change. No differences were noted with the Breast-Q abdominal wellbeing (table 6).

Unilateral Flap Type

The study included 15 msfTRAM flaps and 9 DIEP flaps, with no differences in baseline characteristics (table 3). No differences were noted over time comparing UA, LA or FIM in either flap type (table 4). Comparing long-term scores between flap types demonstrated no differences in UA or FIM scores at long-term. LA scores for msfTRAM were slightly higher than the DIEP cohort (p=0.03). However, scores at the preoperative timepoint trended towards being higher in the msfTRAM cohort, with no differences were noted in change between flaps. There was a significant increase in the SF36 PHC and MHC scores for msfTRAM patients over time as well as the PHC score for DIEP patients (table 5). No differences were noted in Breast-Q abdominal wellbeing scores (table 6).

Bilateral Flap Types

The study included 20 bilateral msfTRAM patients, 4 bilateral DIEP patients, and 3 combination msfTRAM/DIEP patients, without differences in baseline characteristics (table 3). No statistical differences were noted in UA, LA or FIM scores (table 4). There was a significant increase in the SF36 PHC and MHC scores for msfTRAM/smfTRAM patients over time as well as the PHC score for DIEP/DIEP patients (table 5). Additionally, no significant differences were noted in the Breast-Q abdominal wellbeing scores (table 6).

Obesity

Overall, 29.4% (n=15) of patients were obese. Comparing these patients to non-obese patients, differences were noted in BMI (32.3(5.3)kg/m² vs. 25.0(4.9)kg/m², p<0.001) and in the presence of multiple comorbidities (80.0% vs. 13.5%, p<0.001) (table 3). No differences were noted in laterality or in flap types utilized for reconstruction.

Non-obese patients were noted to have an improvement in LA function over time (p=0.001). No objective differences in raw score at long-term follow-up were noted in LA or FIM. Change in UA was significantly greater higher for obese patients (p=0.03).

Non-obese patients experienced significant improvements in SF36 PHC over time (p<0.001), which obese patients did not experience (table 5). A similar finding was noted in SF36 MHC scores, with an improvement in non-obese patients (p=0.006) but no significant change in obese patients over time (p=0.18). As such, obese patients had significantly lower long-term follow-up scores compared to non-obese patients for both SF36 PHC (p=0.0016) and SF36 MHC (p=0.006), with no differences in pre-operative values. Change also significantly differed for both SF36 PHC (p=0.003) and SF36 MHC (p=0.02) by obesity, with scores demonstrating an improvement in health for non-obese patients but a worsening score for obese patients. The Breast-Q abdominal well-being scores however did not differ (table 6).

Regression analyses

Linear regression analyses demonstrated significant findings within the SF36-PHC and SF36-MHC models (F-test <0.0001, R2=0.68 and F-test = 0.0002, R2=0.55 respectively). Examining the SF36-PHC, long term scores were negatively impacted by obesity and preoperative SF36-PHC score (table 7), while laterality had no significant impact. These findings were mirrored in the SF36-MHC analysis (table 7). No significant regression models were found examining long term function (UA, LA, FIM). Subgroup regression by laterality revealed no association with flap type. No significant associations were noted for breast-Q abdominal well-being.

Table 7 –

Linear Regression SF36-PHC and SF36-MHC

	SF36-PHC long term				SF36-MHC long term
	Coef.	P-Value	[95% Conf. Interval]		Coef.	P-Value	[95% Conf. Interval]
Age	−0.69	0.046	−1.37	−0.01	−0.24	0.50	−0.97	0.48
Obesity	−20.08	<0.0001	−30.09	−10.07	−15.00	0.007	−25.59	−4.41
Follow up period (years)	−2.38	0.13	−5.46	0.69	−0.89	0.59	−4.15	2.37
Diabetes	8.26	0.44	−12.91	29.43	16.08	0.16	−6.53	38.68
Hypertension	2.58	0.63	−8.13	13.28	−3.31	0.56	−14.74	8.11
Coronary Artery Disease	−10.91	0.26	−30.14	8.31	−6.30	0.53	−26.60	14.00
Dyslipidemia	−0.12	0.98	−10.52	10.28	−4.14	0.46	−15.22	6.95
Active Smoking	−3.31	0.61	−16.22	9.60	0.36	0.96	−13.31	14.03
Preoperative chemotherapy	1.24	0.75	−6.64	9.12	1.93	0.65	−6.47	10.32
SF36 component score preoperative	0.80	<0.0001	0.51	1.09	0.48	<0.0001	0.25	0.71
laterality (unilateral/bilateral)	−2.91	0.47	−11.01	5.18	−5.29	0.22	−13.93	3.35
Constant	89.56	0.004	31.01	148.11	79.81	0.011	18.95	140.67
	Prob>F <0.0001, R2 = 0.6817				Prob>F = 0.0002, R2 = 0.5526

Discussion:

Over the past two decades, numerous studies have shed light on the impact of abdominal wall function and donor site morbidity after abdominally based free flap breast reconstruction[8]. A desire to improve function and minimize morbidity has driven significant progress away from full rectus muscle sacrifice towards fascial and muscle preservation. Yet, while surgeons have continued to hone techniques, making the procedures faster, more efficient and more precise, the understanding of postoperative function has been limited to early follow-up, with nearly all studies presenting data with less than 3 years of follow-up. As rates of mastectomies increase, it is imperative that we come to a better understanding of the long-term impact of autologous reconstruction on the abdominal wall function and subsequently health overall. This information should be integrated into the decision-making process for women. Additionally, this information would help us determine if the refined techniques truly improve functional outcomes for our patients in the long term.

In an attempt to answer some of these critical questions and fill some of the existing knowledge gaps, we present a 10-year follow-up on 51 patients who underwent breast reconstruction with free abdominal tissue(average 8.1 years). Patients overall functioned extremely well with minimal to no functional decline over time in the objective study specific measurements which included an examination of upper and lower abdominal function as well as a functional independence measure. Furthermore, patients noted improvement in subjective health over time as demonstrated by the SF36 PHC and MHC. Examinations focusing on unilateral compared to bilateral reconstructions yielded minimal differences, as did a direct comparison of msfTRAM flaps to DIEP flaps. Interestingly, a sub-group analysis of obese patients demonstrated that, while maintaining objective function, these patients may have lower long-term scores in both subjective physical and mental health long after the surgery compared to non-obese patients. These findings continue to support autologous breast reconstruction as a reconstructive option, with excellent functional results and subjective experiences.

The increase in mastectomy rates has been fueled in large part by the increase in contralateral prophylactic procedures[27, 28]. As such, many women are electing bilateral mastectomies and reconstruction. As many of the women who elect bilateral procedures are younger, being able to counsel on the impact of unilateral compared to bilateral procedures is important. In the first part of our study, we did not directly compare unilateral to bilateral reconstructions[9, 10]. However, after considering the recent trends, we believed examining this would be appropriate for such a comparison in the current study. Interestingly, few recent studies have compared unilateral and bilateral function, even though this data could factor directly into the preoperative decision-making process for the patient and her surgeon. Data from the early 1990’s does however suggest decreased ability to do sit-ups in bilateral reconstructions compared with unilateral reconstructions [29], though this does not specifically look at msfTRAM and DIEP reconstructions. We demonstrate minimal objective or subjective differences between these cohorts, with the small objective differences having little clinical significance, as patients on average were all able to raise the inferior aspect of the scapula off the exam table with arms across chest or out-stretched. Subjective findings were similar, with improvements noted in both unilateral and bilateral reconstructions over time, but no difference in raw score at long-term follow-up.

Multiple studies have compared flap types, and the evidence suggests that muscle sacrifice (of varying degrees) leads to objective early differences on comparison[3, 7–10, 29]. However, many of these tests utilize isokinetic dynamometers, which are able to differentiate true, small differences in muscle strength but which may not be clinically meaningful. Yet, even more simplistic objective studies, such as our own or those examining sit-up ability, have also demonstrated early differences or favorable profiles in DIEP flaps[30, 31]. But subjectively, regardless of flap type, patients appear to do well, even at early follow-up following surgery(1–3 years), with recent cross-sectional, multicenter data using the Breast-Q also supporting this conclusion[12]. The current study adds to this literature, presenting up to 10 years of prospective data demonstrating sustained objective function with nearly full return or maintenance at baseline levels. Subjectively, although slight variations exist in SF36 scores, no significant differences exist based on flap type and no difference in the Breast-Q abdominal wellbeing was noted.

Obesity is a known risk factor for complications after autologous breast reconstruction[18], which have been shown to impact early subjective health[17, 32]. In the early follow-up, UA function is impacted in obese patients, however no difference is demonstrated in subjective outcomes[17]. Interestingly, the current findings show the converse at long term follow up – no significant difference in objective findings, but a decrease in general patient reported outcomes (SF36 PHC and MHC). As many autologous breast reconstruction patients are obese, this is a key finding for preoperative counseling which plays out in regression analyses. Obesity is understood to adversely impact both physical and psychological function over time[15, 16, 33]. Given the lack of objective deficit and no difference in Breast-Q outcomes, it is unlikely that the subjective decrease in health is directly related to the surgery, and may be related to the disease state itself. This is an important issue which surgeons should be cognizant of in long-term follow up of obese patients, and should consider appropriate referral if a mental health concern is uncovered.

This study has a number of significant limitations. First, it is prone to observer bias, as over the 10 year study, several examiners (residents or research fellows) helped collect data used for analysis. While each observer was trained for the study, the inter-reliability of their measurements was not assessed. Additionally, our response rate was low, at 32%. However, comparisons of responders to non-responders demonstrated no significant differences between these cohorts, and as such, we view this current cohort as representative of the initial study. Even with multimodal attempted contact[34, 35], we were unable to encourage further participation. Finally, the findings presented here should be viewed with a degree of caution, as subgroups examined in the analyses quickly reduced in size, making truly meaningful analysis challenging. The analysis was also limited to msfTRAM and DIEP flaps, therefore we cannot comment on the impact of other flaps on function. However, even with these limitations, this study helps bridge an important gap in our literature, providing new data on the long-term impact of autologous breast reconstruction, gathered through prospective methodology.

Conclusions:

Abdominally based autologous breast reconstruction causes minimal to no objective or subjective functional impairment 5–10 years after surgery, regardless of laterality or flap type. However, obese patients may be at risk for long-term subjective physical and mental health impairment which may be unrelated to the surgery itself.