Cortiva versus AlloDerm in Prepectoral and Partial Submuscular Implant-Based Breast Reconstruction: A Randomized Clinical Trial

Abstract

Background:

Several acellular dermal matrices (ADMs) are used for soft-tissue support in prosthetic breast reconstruction. Little high-level evidence supports the use of one ADM over another. The authors sought to compare Cortiva 1-mm Allograft Dermis with AlloDerm RTU (ready to use), the most studied ADM in the literature.

Methods:

A single-blinded randomized controlled trial comparing Cortiva with AlloDerm in prepectoral and subpectoral immediate prosthetic breast reconstruction was performed at 2 academic hospitals from March of 2017 to December of 2021. Reconstructions were direct to implant (DTI) or tissue expander (TE). Primary outcome was reconstructive failure, defined as TE explantation before planned further reconstruction, or explantation of DTI reconstructions before 3 months postoperatively. Secondary outcomes were additional complications, patient-reported outcomes (PROs), and cost.

Results:

There were 302 patients included: 151 AlloDerm (280 breasts), 151 Cortiva (277 breasts). The majority of reconstructions in both cohorts consisted of TE (62% versus 38% DTI), smooth device (68% versus 32% textured), and prepectoral (80% versus 20% subpectoral). Reconstructive failure was no different between ADMs (AlloDerm 9.3% versus Cortiva 8.3%; P = 0.68). There were no additional differences in any complications or PROs between ADMs. Seromas occurred in 7.6% of Cortiva but 12% of AlloDerm cases, in which the odds of seroma formation were two-fold higher (odds ratio, 1.93 [95% CI, 1.01 to 3.67]; P = 0.047). AlloDerm variable cost was 10% to 15% more than Cortiva, and there were no additional cost differences.

Conclusion:

When assessing safety, clinical performance, PROs, and cost, Cortiva is noninferior to AlloDerm in immediate prosthetic breast reconstruction, and may be less expensive, with lower risk of seroma formation.

CLINICAL QUESTION/LEVEL OF EVIDENCE:

Therapeutic, I.

In 2022, 287,850 women in the United States were diagnosed with invasive breast cancer, and 51,400 with ductal carcinoma in situ.¹ In 2020, members of the American Society of Plastic Surgeons performed 137,808 breast reconstructions, with 105,665 in the immediate setting. These reconstructions used a tissue expander (TE) in 83,487 cases and an acellular dermal matrix (ADM) 59,247 times, with bilateral reconstructions outpacing unilateral at a ratio of 2:1.² Breast reconstructions are common, and tend to involve a prosthesis in the immediate setting with an ADM.^3,4 Initially, ADMs were used as a sling to extend the subpectoral pocket caudally, limit resistive forces to lower pole expansion of immediate implants, and define infralateral and lateral mammary folds.⁵ Compared with total submuscular coverage, partial submuscular coverage supplemented with ADM is associated with greater initial TE fill volumes, fewer postoperative TE fill visits, and higher likelihood of direct-to-implant (DTI) reconstruction.^6–8 The concepts of improved aesthetic and patient-reported outcomes (PROs) with less postoperative pain when using an ADM has not been proven, however, and some reports suggest that ADMs can increase the risk of seroma, infection, skin necrosis, and reconstructive failure.^6,9–19 Furthermore, ADM used for lower pole support after partial submuscular coverage does not mitigate the risk of animation deformity or pectoralis major muscle fibrosis after radiotherapy. Prepectoral reconstruction with full anterior implant coverage with ADM in TE or DTI reconstruction has gained popularity.^20,21

When considering breast reconstruction, the shared decision-making process should be informed by risk exposure and optimized for preference concordance.²² The decision to use a particular ADM can be considered in this context and then weighed against cost. Risk factors associated with prosthetic breast reconstruction complications include immediate reconstruction, bilateral procedures, radiation, elevated body mass index (BMI), and smoking.^{10,13,23–33} Whether ADM type influences complication rates or PROs remains unclear. In most studies, the particular ADM chosen does not significantly influence complication rates.^18,34–50 A higher infection rate with FlexHD than AlloDerm,⁵¹ more seromas with AlloDerm than Surgimend,⁵² and higher total complications driven by higher seroma rate with AlloDerm than Strattice⁵³ represent the conspicuous exceptions not corroborated by other studies. Moreover, no significant differences in PROs have been shown when comparing ADMs.^37,42,54,55

Selection of approach to postmastectomy breast reconstruction is the culmination of a series of choices, including whether to operate at all, mastectomy type, reconstruction timing, and whether to perform autologous or implant-based reconstruction. Once the decision to proceed with TE or DTI in the subpectoral or prepectoral plane is made, the surgeon must still choose a particular implant and surgical approach for device support. If the best available evidence, however, suggests that one device or ADM is not statistically significantly inferior to another in terms of safety, clinical performance, or PROs, then cost becomes a relevant factor. This study used a noninferiority design to compare the performance of Cortiva with AlloDerm in prepectoral and subpectoral breast reconstructions, focusing on TEs and DTIs, with respect to the primary outcome of reconstructive failure.

METHODS

Study Design

We performed a randomized controlled trial (Fig. 1) comparing 2 ADMs in immediate postmastectomy prosthetic breast reconstruction: Cortiva 1 mm Allograft Dermis, 1.0±0.2 mm (RTI Surgical) and AlloDerm RTU (ready to use), 1.6±0.4 mm (Allergan Medical). All ADMs had a single-sided, nonpreperforated basement membrane that was fenestrated by the plastic surgeon. Patients were blinded to ADM manufacturer. Reconstructions were performed at 2 sites (Barnes Jewish West County and Barnes Jewish Hospital) by 2 plastic surgeons (T.M.M. and M.M.T.) from March of 2017 to December of 2021. Patients were randomized using SAS PROC PLAN in a block size of 4 to 1 of the 2 ADMs. This study was approved by the Washington University Human Research and Protection Office (approval no. 201608168) and registered at clinicaltrials.gov (NCT02891759).

Fig. 1.

Consolidated Standards of Reporting Trials diagram. Existing data did not inform how prepectoral reconstruction would influence outcomes, so we amended the trial protocol to recruit an additional 180 patients for prepectoral reconstruction beyond those recruited for the subpectoral approach. No additional patients chose subpectoral reconstruction once prepectoral was offered, despite keeping the trial open for an additional year to optimize recruitment. Furthermore, no patient who selected prepectoral reconstruction was converted to a postpectoral approach for clinical reasons. Based on these trends, we continued overall trial enrollment to 302 patients (n = 61 subpectoral, n = 241 prepectoral). The research coordinator screened and enrolled all patients, who were blinded to the intervention; performed block randomization; notified surgeons which ADM would be used for each patient, disclosed in sealed envelopes; coordinated follow-up; administered the BREAST-Q to measure PROs; and ensured study protocol compliance.

TEs were used until planned secondary reconstructive procedure or reconstructive failure. DTI reconstructions were followed up for a minimum of 3 months postoperatively or until reconstructive failure. Time to drain removal was recorded for all breasts, including 2 drains for subpectoral and 1 drain for prepectoral.

Enrollment

All patients undergoing immediate breast reconstruction with subpectoral TE or DTI and lower pole support with ADM were considered for eligibility starting in February of 2017. The option for prepectoral reconstruction with full anterior coverage of the TE or implant was introduced in September of 2018. Patients with a BMI under 36 kg/m², ages 22 to 70 years, undergoing prophylactic or therapeutic, nipple- or skin-sparing, unilateral or bilateral mastectomies were included.

Surgical Procedures

Patient preference drove the decision to proceed with subpectoral versus prepectoral reconstruction, and the decision to use saline versus silicone gel–filled devices where relevant. The decision to proceed with a TE or DTI and initial volumes was based on clinical assessment of perfusion, but never indocyanine green or associated devices. For subpectoral placement, the pectoralis major muscle was caudally detached and sutured with 2-0 polydioxanone suture to the cephalad edge of a 16 × 8-cm sheet of oriented ADM, and anchored to the chest wall to define the lateral and inframammary folds. For prepectoral placement, oriented ADM was secured to the chest wall muscles with 2-0 polydioxanone suture with sufficient redundancy to accommodate an implant or fully inflated TE. Two drains—1 deep (10 Fr) and 1 superficial (15 Fr) to the pectoralis major—were used for subpectoral reconstructions, and 1 subcutaneous drain (15 Fr) accompanied all prepectoral reconstructions. Only saline irrigation was used. Each breast received a 15-cc PEC II nerve block with 0.25% Marcaine and 1:200,000 epinephrine by the plastic surgeon and a standardized enhanced recovery after surgery protocol. All implanted breast prostheses from July of 2019 onward had smooth surfaces due to the U.S. Food and Drug Administration recall of Allergan macrotextured breast TEs and implants.⁵⁶ Only tabbed TEs were used once we switched to smooth devices, before which TEs did not have tabs.

Outcomes

Clinical, patient-reported, and cost outcomes were prospectively recorded in REDCap version 7.^57,58 The primary outcome was reconstructive failure, defined as unplanned, premature TE or implant explantation for any reason. Infection requiring reinstitution of oral or intravenous antibiotics or reoperation, seroma or hematoma identified by imaging or requiring intervention, necrosis or incisional dehiscence requiring intervention, TE or implant exposure, and time to final drain removal represented secondary outcomes. Analgesia use 2 and 4 weeks postoperatively was also collected. PROs were reported as Q scores generated by BREAST-Q version 1.1.

Variable costs, excluding the cost of the ADM, were compiled for all patients. These are patient-specific costs that the hospital incurs (eg, operative staff salaries, supplies, drugs, laboratory tests, imaging, and so on). Surgeon reimbursement and fixed costs, such as overhead and administrative salaries, were excluded. Our institution is contractually prohibited from disclosing exact costs for discrete products, such as ADMs. Rather, we can present ranges of $28 to $31/cm² for AlloDerm RTU (ie, 128 cm²: $3584 to $3968; 320 cm²: $8960 to $9920) and $23 to $26/cm² for Cortiva (ie, 128 cm²: $2944 to $3328; 320 cm²: $7360 to $8320), which was 10% to 15% less expensive during the study period.

Sample Size Calculation

Sample size determination assumed a reconstructive failure rate of 20% in 1 cohort and 9% in another group based on outcomes from more than 2000 of our previous reconstructions.^26,35,59 A total of 180 eligible patients were determined to be required to achieve 80% power based on the 1-sided Z test with pooled variance to detect a difference between the incidence rates of the 2 ADMs at 0.1 significance. Interim analysis did not alter enrollment strategy.⁶⁰ The trial was initially designed for subpectoral reconstruction, but our practices reflected a strong national switch to prepectoral reconstruction 1 year into the trial (Fig. 1).

Statistical Analyses

Chi-square and Fisher exact tests were used for categorical variables to detect differences in reconstructive failure rate between ADM types. The nonparametric Wilcoxon rank sum test was used for ranked BREAST-Q data and continuous variables that were not normally distributed. Multivariable logistic regression models were performed to assess the effect of ADM type, controlling for prepectoral versus subpectoral plane, and significant factors found in univariate analysis. Stepwise selection was applied with a P value threshold of 0.2. The 95% confidence interval for the area under the receiver operating characteristic curve was 0.773 (0.695 to 0.851) for reconstructive failure and 0.784 (0.696 to 0.872) for the logistic regression model for seroma formation, indicating acceptable model fits. A significance of α = 0.05 was used, and all tests were 2-sided. Missing data were considered missing at random and excluded. SAS version 9.4 (SAS Institute) was used for analyses.

RESULTS

Patient Demographic and Cancer Characteristics

A total of 302 patients (557 breasts) were enrolled and completed the study: 151 patients (280 breasts) were randomized to AlloDerm and 151 patients (277 breasts) to Cortiva. There were no significant differences in age, race, BMI, diabetes, or smoking status (Table 1). Hypertension was more prevalent in the AlloDerm cohort (25% versus 15% in the Cortiva cohort; P = 0.04).

Table 1.

Patient-Level Demographic Characteristics and Comorbidities

Variable	AlloDerm RTU (n = 151)	Cortiva 1 mm (n = 151)	P
No. of breasts, total	280	277	—
Age, median (IQR), yr	47.2 (38.7, 53.5)	48.2 (40.6, 55.4)	0.23
Race, no. (%)			0.43
White	139 (92)	142 (95)
Black	9 (6.0)	7 (4.7)
Asian	2 (1.3)	0 (0.0)
Other	1 (0.7)	0 (0.0)
BMI, median (IQR)	26.7 (23.0, 30.1)	26.5 (23.3, 30.6)	>0.99
Diabetes, no. (%)	9 (6.0)	7 (4.6)	0.61
Hypertension, no. (%)	37 (25)	23 (15)	0.04a
Autoimmune disease, no. (%)	10 (6.6)	12 (8.0)	0.66
Smoking, no. (%)			0.16
Never smoker	107 (71)	101 (67)
Former smoker	29 (19)	41 (27)
Current smoker	15 (9.9)	9 (6.0)
Neoadjuvant chemotherapy, no. (%)	36 (24)	43 (28)	0.36
Adjuvant chemotherapy, no. (%)	35 (23)	36 (24)	0.89
Immunotherapy, no. (%)	25 (17)	22 (15)	0.63

BMI, body mass index; RTU, ready to use.

^aStatistically significant.

Cancer characteristics and treatment variables are displayed in Table 2. Most mastectomies in both cohorts were prophylactic (54% in AlloDerm versus 52% in Cortiva), and there were no significant differences in cancer stage in the treated breasts (P = 0.07). Use of radiation, chemotherapy, immunotherapy, and sentinel lymph node biopsy did not differ across groups, and resected mastectomy specimen weights were similar (median: 561 [interquartile range {IQR}: 386, 738] grams AlloDerm versus median: 582 [IQR: 370, 813] grams Cortiva; P = 0.27).

Table 2.

Breast-Level Cancer Characteristics and Treatment

Variable	AlloDerm RTU (n = 280)	Cortiva 1 mm (n = 277)	P
Cancer stage, no. (%)			0.07
Prophylactic	152 (54)	145 (52)
DCIS	24 (8.6)	34 (12)
I	52 (19)	37 (13)
II	47 (17)	47 (17)
III	5 (1.8)	14 (5.1)
Previous radiation therapy, no. (%)	6 (2.1)	2 (0.7)	0.29
Adjuvant radiation therapy, no. (%)	40 (14)	43 (16)	0.68
Mastectomy type, no. (%)			0.07
Nipple-sparing	142 (51)	119 (43)
Skin-sparing	138 (49)	158 (57)
SLNB, no. (%)	140 (50)	128 (46)	0.37
Mastectomy specimen weight, median (IQR), g	561 (386, 738)	582 (370, 813)	0.27

DCIS, ductal carcinoma in situ; RTU, ready to use; SLNB, sentinel lymph node biopsy.

Breast Reconstruction Characteristics

Most breasts were reconstructed using TEs (Table 3), although a large minority of patients underwent DTI reconstruction in both groups (38% AlloDerm versus 37% Cortiva; P = 0.65). Of patients who initially received TEs, the most common second-stage reconstruction was with implants alone (59% of TEs in AlloDerm versus 62% of TEs in Cortiva; P = 0.50), with a smooth surface (68% in AlloDerm and Cortiva; P = 0.97) and in the prepectoral plane (81% in AlloDerm versus 80% in Cortiva; P = 0.78). Median time to final drain removal was 13 days in both cohorts (IQR, 9, 16 for both; P = 0.38).

Table 3.

Breast-Level Reconstruction Characteristics^a

Variable	AlloDerm RTU (n = 280)	Cortiva 1 mm (n = 277)	P
Reconstruction plane, no. (%)			0.78
Prepectoral	227 (81)	222 (80)
Subpectoral	53 (19)	55 (20)
Shell type, no. (%)			0.97
Smooth	190 (68)	189 (68)
Textured	89 (32)	88 (32)
Reconstruction type, no. (%)			0.65
Direct to implant	107 (38)	102 (37)
TE	38 (14)	35 (13)
TE to implant	102 (36)	108 (39)
TE to flap	32 (11)	28 (10)
TE to flap + implant	1 (0.4)	4 (1.4)
Ratio of initial fill volume to mastectomy specimen weight, median (IQR), mL/g	0.53 (0.28, 0.98)	0.50 (0.27, 0.95)	0.38
Time to final drain removal, median (IQR), days	13 (9, 16)	13 (9, 16)	0.38

RTU, ready to use; TE, tissue expander.

Complications

The incidence of reconstructive failure was 9.3% in patients with AlloDerm versus 8.3% in patients with Cortiva (P = 0.68). The most common complications were seroma (12% in patients with AlloDerm versus 7.6% in patients with Cortiva; P = 0.09) and infection (9.3% versus 8.7%; P = 0.80), which did not differ by cohort. Incidences of hematoma, necrosis, incisional dehiscence, implant exposure, or surgical exploration without explantation, or analgesic use 2 and 4 weeks postoperatively, did not differ across groups (Table 4).

Table 4.

Breast-Level Postoperative Complications and Analgesia Use

Variable	AlloDerm RTU, No. (%) (n = 280)	Cortiva 1 mm, No. (%) (n = 277)	P
Reconstructive failure	26 (9.3)	23 (8.3)	0.68
Hematoma	10 (3.6)	4 (1.4)	0.17
Seroma	33 (12)	21 (7.6)	0.09
Infection	26 (9.3)	24 (8.7)	0.80
Necrosis	22 (7.9)	17 (6.1)	0.43
Incisional dehiscence	16 (5.7)	14 (5.1)	0.73
Exposure of implant or TE	4 (1.4)	6 (2.2)	0.51
Surgical exploration	14 (5.0)	8 (2.9)	0.20
Analgesia use at 2 weeksa	112 (91)	126 (95)	0.25
Analgesia use at 4 weeksa	46 (48)	63 (58)	0.18

RTU, ready to use; TE, tissue expander.

^aPatient-level analgesia use.

The incidence of reconstructive failure was not different between ADMs in either plane, and the most common complications remained infection and seroma (Table 5). A subgroup analysis by reconstruction type revealed no significant differences in complication rates within patients who underwent TE reconstruction; however, among patients who underwent DTI reconstruction, surgical exploration was more common with AlloDerm than Cortiva (5.6% versus 0%; P = 0.03; Table 6).

Table 5.

Postoperative Complications and Analgesia Use with Subgroup Analysis by Plane of Reconstruction (Prepectoral versus Subpectoral)

Variable	Prepectoral			Subpectoral
	AlloDerm RTU, No. (%) (n = 227)	Cortiva 1 mm, No. (%) (n = 222)	P	AlloDerm RTU, No. (%) (n = 53)	Cortiva 1 mm, No. (%) (n = 55)	P
Reconstructive failure	22 (9.7)	18 (8.1)	0.55	4 (7.6)	5 (9.1)	0.77
Hematoma	7 (3.1)	4 (1.8)	0.54	3 (5.7)	0 (0)	0.12
Seroma	25 (11)	18 (8.1)	0.30	8 (15)	3 (5.5)	0.10
Infection	22 (9.7)	20 (9.0)	0.80	4 (7.6)	4 (7.3)	>0.99
Necrosis	19 (8.4)	15 (6.8)	0.51	3 (5.7)	2 (3.6)	0.68
Incisional dehiscence	13 (5.7)	13 (5.9)	0.95	3 (5.7)	1 (1.8)	0.36
Exposure of implant or TE	3 (1.3)	5 (2.3)	0.50	1 (1.9)	1 (1.8)	>0.99
Surgical exploration	10 (4.4)	5 (2.3)	0.20	4 (7.6)	3 (5.5)	0.71
Analgesia use at 2 weeksa	90 (89)	103 (95)	0.15	22 (100)	23 (96)	>0.99
Analgesia use at 4 weeksa	34 (48)	50 (59)	0.17	12 (50)	13 (54)	0.77

RTU, ready to use; TE, tissue expander.

^aPatient-level analgesia use.

Table 6.

Postoperative Complications and Analgesia Use with Subgroup Analysis by Reconstruction Type (Direct to Implant versus Tissue Expander)

Variable	Direct to Implant			Tissue Expander
	AlloDerm RTU, No. (%) (n = 107)	Cortiva 1 mm, No. (%) (n = 102)	P	AlloDerm RTU, No. (%) (n = 173)	Cortiva 1 mm, No. (%) (n = 175)	P
Reconstructive failure	5 (4.7)	3 (2.9)	0.72	21 (12)	20 (11)	0.84
Hematoma	4 (3.7)	0 (0)	0.12	6 (3.5)	4 (2.3)	0.54
Seroma	5 (4.7)	4 (3.9)	>0.99	28 (16)	17 (9.7)	0.07
Infection	5 (4.7)	7 (6.9)	0.50	21 (12)	17 (9.7)	0.47
Necrosis	7 (6.5)	3 (2.9)	0.33	15 (8.7)	14 (8.0)	0.82
Incisional dehiscence	5 (4.7)	3 (2.9)	0.72	11 (6.4)	11 (6.3)	0.98
Exposure of implant or TE	0 (0)	0 (0)	>0.99	4 (2.3)	6 (3.4)	0.75
Surgical exploration	6 (5.6)	0 (0)	0.03a	8 (4.6)	8 (4.6)	0.98
Analgesia use at 2 weeksb	42 (89)	47 (96)	0.16	69 (93)	78 (94)	>0.99
Analgesia use at 4 weeksb	14 (45)	21 (59)	0.33	32 (50)	41 (57)	0.42

RTU, ready to use; TE, tissue expander.

^aStatistically significant.

^bPatient-level analgesia use.

Multivariable stepwise logistic regression identified associations between nipple-sparing mastectomy, prepectoral plane, textured shell, other breast surgeon (ie, breast surgeon contributing fewer than 10 cases to the study), and mastectomy specimen weight with increased rates of reconstructive failure (Table 7). ADM type did not achieve the 0.2 P value threshold for entry into the model. Stepwise regression for seroma identified other breast surgeon, former smoking, and mastectomy specimen weight as significant risk factors (Table 7). However, ADM type was also significantly associated with seroma formation while accounting for these risk factors: use of AlloDerm was associated with an almost two-fold higher odds of seroma than use of Cortiva (OR, 1.93 [95% CI, 1.01–3.67]; P = 0.047).

Table 7.

Multivariable Logistic Regressions for Reconstructive Failure and Seroma Formation^a

Variable	Reconstructive Failure			Seroma Formation
	Odds Ratio	95% CI	P	Odds Ratio	95% CI	P
AlloDerm RTU (ref = Cortiva 1 mm)	—	—	—	1.93	(1.01, 3.67)	0.047b
Nipple-sparing mastectomy (ref = skin-sparing)	2.63	(1.14, 6.05)	0.02b	—	—	—
Prepectoral plane (ref = subpectoral)	2.92	(1.08, 7.89)	0.03b	—	—	—
Textured shell (ref = smooth)	4.32	(1.93, 9.67)	<0.001b	1.98	(0.97, 4.05)	0.06
Breast surgeon (ref = surgeon A)
Surgeon B	0.75	(0.08, 6.77)	0.80	0.50	(0.06, 4.39)	0.53
Surgeon C	0.92	(0.24, 3.52)	0.90	1.48	(0.52, 4.21)	0.47
Surgeon D	2.88	(0.76, 10.97)	0.12	1.48	(0.31, 7.16)	0.62
Other	4.53	(1.45, 14.15)	0.009b	5.87	(2.11, 16.35)	<0.001b
Hypertension	1.79	(0.84, 3.79)	0.13	—	—	—
Smoking status (ref = never smoker)
Former smoking	1.94	(0.94, 4.00)	0.07	2.78	(1.41, 5.48)	0.003b
Current smoking	2.10	(0.79, 5.58)	0.14	1.38	(0.48, 3.95)	0.55
Chemotherapy	0.61	(0.30, 1.27)	0.19	0.51	(0.25, 1.05)	0.07
Specimen weight, g	1.002	(1.001, 1.003)	<0.001b	1.002	(1.001, 1.003)	<0.001b

RTU, ready to use.

^aCandidate variables that did not meet threshold for inclusion in the final model for reconstructive failure were acellular dermal matrix type, reconstruction type (direct to implant versus tissue expander), plastic surgeon, age, body mass index, diabetes, and radiation. Candidate variables that did not meet threshold for inclusion in the final model for seroma formation were mastectomy type, plane of reconstruction, reconstruction type (direct to implant versus tissue expander), plastic surgeon, age, body mass index, diabetes, hypertension, and radiation.

^bStatistically significant.

Patient-Reported Outcomes

Preoperative Q scores assessed with the BREAST-Q were not significantly different between the AlloDerm and Cortiva cohorts (Table 8). Postoperative well-being scores did not differ by ADM type. Satisfaction with Breasts, Psychosocial Well-being, and Sexual Well-being Q scores were similar postoperatively compared with preoperatively; however, in both groups, Physical Well-being tended to be lower postoperatively, with no difference by ADM type (−9 [IQR −21, 0] in AlloDerm versus −10 [IQR −20, 0] in Cortiva; P = 0.74).

Table 8.

Preoperative, Postoperative, and Change in BREAST-Q Scores in Patients with AlloDerm RTU and Cortiva 1 mm ADM

Variable	AlloDerm RTU, Median (IQR) (n = 151)	Cortiva 1 mm, Median (IQR) (n = 151)	P
Satisfaction with Breasts
Preoperative	58 (43, 63)	53 (43, 63)	0.38
Postoperative	58 (50, 71)	58 (45.5, 71)	0.51
Change	2 (−8, 19)	2 (−12, 15)	0.66
Psychosocial Well-being
Preoperative	63 (55, 76)	63 (57, 76)	0.45
Postoperative	67 (57, 92)	64 (57, 92)	0.98
Change	0 (−12, 16)	0 (−6, 11)	0.77
Physical Well-being
Preoperative	77 (71, 91)	81 (74, 91)	0.10
Postoperative	68 (60, 81)	71 (63, 81)	0.10
Change	−9 (−21, 0)	−10 (−20, 0)	0.74
Sexual Well-being
Preoperative	54 (45, 63)	54 (45, 63)	0.90
Postoperative	57 (41, 67)	54 (43, 63)	0.64
Change	0 (−14, 13)	−2 (−15, 8)	0.28
Postoperative satisfaction
Satisfaction with outcome	75 (61, 86)	75 (61, 86)	0.59
Satisfaction with information	80 (62, 100)	77 (60, 91)	0.28
Satisfaction with plastic surgeon	100 (100, 100)	100 (100, 100)	0.83
Satisfaction with medical team other than plastic surgeon	100 (100, 100)	100 (100, 100)	0.46
Satisfaction with office staff	100 (100, 100)	100 (100, 100)	0.16

RTU, ready to use.

There were no significant differences in Q scores across ADM cohorts when subgroup analysis was performed by reconstructive plane or type. (See Table, Supplemental Digital Content 1, which shows a comparison of preoperative, postoperative, and change in BREAST-Q scores in patients with AlloDerm RTU and Cortiva 1 mm ADM, with subgroup analyses by plane of reconstruction. All scores are reported as median [IQR], http://links.lww.com/PRS/H47. See Table, Supplemental Digital Content 2, which shows comparison of preoperative, postoperative, and change in BREAST-Q scores in patients with AlloDerm RTU and Cortiva 1 mm ADM, with subgroup analyses by reconstruction type. All scores are reported as median [IQR], http://links.lww.com/PRS/H48.)

Cost Analysis

Overall variable costs were obtained for all patients. Three patients were excluded (2 AlloDerm, 1 Cortiva) because of an additional pharmacy cost for an unrelated pharmacy charge of $13,920 to $13,990, far in excess of the $19 to $1931 range exhibited in the remaining 299 patients (Fig. 2). Excluding ADM, variable costs for unilateral reconstructions ranged from $4354 ± $786 for subpectoral to $4677 ± $1216 for prepectoral TEs, and for bilateral reconstructions from $5509 ± $776 for subpectoral to $6606 ± $1274 for prepectoral TEs (Fig. 2). (See Table, Supplemental Digital Content 3, which shows variable costs by treatment cohort, with and without effect of ADM, http://links.lww.com/PRS/H49.) ADMs increased variable costs of reconstruction from 71.4% for a unilateral subpectoral TE reconstruction with Cortiva at its lower cost range of $23/cm² to 323.8% for a bilateral prepectoral DTI reconstruction with AlloDerm at its higher cost range of $31/cm².

Fig. 2.

Cost of subpectoral TE or prepectoral TE or DTI breast reconstructions using AlloDerm or Cortiva in unilateral or bilateral settings. Mean variable costs generated from actual patients are presented in bar graphs and exclude the cost of ADM. Costs with ADM assuming $28 to $31/cm² for AlloDerm (blue) and $23 to $26/cm² for Cortiva (red) are depicted as circles. The use of an ADM increases variable cost from 1.7 times (unilateral, subpectoral TE with Cortiva) to 4.2 times (bilateral, prepectoral DTI with AlloDerm) at the time of primary immediate reconstruction.

DISCUSSION

An evidence-based approach accounting for safety, clinical performance, and PROs should drive the breast reconstruction shared decision-making process. However, there are few comprehensive randomized studies comparing the various ADMs.^{34,38,42,47,55,61} Current meta-analyses are heavily biased to retrospective studies of subpectoral implant placement.^40,41 AlloDerm is the most studied and used noncrosslinked human ADM for breast reconstruction.^{18,34,35,37–43,45–53,55,61–71} Cortiva, another noncrosslinked human ADM, has been safely used in a series of 183 prepectoral breast reconstructions,⁷² and demonstrated no significant differences compared with AlloDerm after retrospective review of 298 reconstructions.⁴⁴ AlloDerm and Cortiva share similar elastin, collagen, and vascular remodeling characteristics.⁷³ Herein we present the first randomized controlled trial comprehensively comparing Cortiva with AlloDerm in immediate prepectoral and subpectoral prosthetic breast reconstruction.

The complications reported in our study are consistent with those reported in the literature surrounding implant-based breast reconstruction with ADM, and are no different between AlloDerm and Cortiva: overall seroma rate of 9.7% (literature range 0% to 30%); infection, 9% (0% to 29%); mastectomy flap necrosis, 7% (0% to 25%); and reconstructive failure, 8.8% (0% to 30%).^{6–9,13–15,17–19,27,28,31,32,34,35,37–39,43–45,47–49,51,52,61,64,66,69,70,74–82} However, multivariable logistic regression found that reconstruction with AlloDerm increased odds of seroma formation two-fold. We also found that prepectoral implant placement increased odds of reconstructive failure three-fold. One hypothesis is that the risk of the known complications associated with ADM use would be increased with prepectoral reconstruction, as more ADM is used for full anterior prosthesis coverage. Why odds of seroma formation were higher with AlloDerm is less clear, but also a potentially less important finding, as ADM type was not associated with increased odds of reconstructive failure. Only 2 studies comparing ADMs have found seromas to be more common with AlloDerm when compared with another ADM—Surgimend⁵² and Strattice.⁵³ However, the AlloDerm used in both studies was different from ours as it underwent aseptic versus sterile processing. In addition, Butterfield⁵² noted that drain practices changed during the study, which could confound results. Drain duration in our study was no different between ADMs and on average 13 days, well within the ranges reported by Glasberg and Light⁵³ and Keifer et al.⁴⁴

While previous work suggests that seromas increase implant loss and infection four-fold,⁸³ AlloDerm was not associated with increased odds of reconstructive failure, or a significantly greater incidence of infection compared with Cortiva. Despite the differences in manufacturing processes (sterilization assurance level of 10⁻³ for AlloDerm RTU versus 10⁻⁶ for Cortiva 1 mm), infection may not be solely predicted by ADM type. The breast is known to be a clean-contaminated surgical site populated with bacteria throughout and a microbiome more diverse than skin.⁸⁴ Mastectomy with immediate reconstruction undoubtedly leads to some level of bacterial contamination despite surgeons’ best efforts to maintain sterility, and even imperceptible defects in a healing incision can provide portals for bacterial translocation.⁸⁵ Drain placement and home drain care, which controls seromas, provides a conduit for cutaneous flora. In addition, there are multiple additional nonsurgical factors that increase risk for infection, including BMI, radiation, and smoking.^{10,13,23–33} Given these multiple contributing factors, it is unlikely that the 1000 times greater sterilization assurance level of Cortiva makes a clinically significant difference in infection and reconstructive failure outcomes.

Because complications and PROs did not differ significantly between AlloDerm and Cortiva, cost becomes a differentiating factor with ADM selection. Cortiva was 10% to 15% less expensive than AlloDerm at our institution. With ADM excluded, variable costs did not differ significantly between cohorts. ADM use increased variable costs dramatically in our series—a minimum of 1.7 times in unilateral subpectoral reconstructions with Cortiva that used a relatively small, single 128 cm² sheet. In bilateral prepectoral reconstructions, with two 320 cm² sheets, ADM increased variable costs at least 4 times. Indeed, this finding supports the need to identify more economical approaches to prepectoral breast reconstruction with implants or autologous tissue. Furthermore, the cost of ADM has a considerable effect on the overall cost of primary prosthetic breast reconstruction and its economic sustainability, particularly when there are no detectable differences in outcome. Although breast reconstruction is covered by insurance, up to a third of patients undergoing breast reconstruction still pay over $5000 of out-of-pocket costs.⁸⁶ Moreover, patients are frequently forced to make tradeoffs (with decreased spending on basic items) or leverage savings to pay for cancer care.⁸⁶ This financial toxicity is associated with diminished quality of life, treatment nonadherence, and increased mortality rates.^86–89

This study has limitations. It was a 2-center study involving 2 plastic surgeons, with the majority of mastectomies performed by a single breast oncologic surgeon. Therefore, there was little heterogeneity in technique, and our findings may not be generalizable to all practices. There were several changes that occurred during the study period, including the transition from subpectoral to prepectoral reconstruction, conversion to only smooth devices owing to the Food and Drug Administration macrotextured implant recall, and interruption in enrollment due to COVID-19. These changes were likely mitigated in part by our randomized design. In addition, although complications such as reconstructive failure and seroma typically arise in the first 3 months after prosthetic reconstruction,^26,90,91 we did not examine long-term outcomes, which may have influenced reoperation rates and overall cost.

CONCLUSIONS

This randomized controlled trial establishes that Cortiva is noninferior to AlloDerm in immediate prosthetic breast reconstruction when assessing safety, complications, and PROs, regardless of reconstructive type (DTI or TE) or plane (subpectoral or prepectoral). AlloDerm may also increase risk of seroma formation. Surgeons should use an evidence-based approach informed by safety, clinical performance, PROs, and cost to aid the shared decision-making process in breast reconstruction.

DISCLOSURE

Dr. Myckatyn receives investigator-initiated grant funding, royalties, and advisory board renumeration from RTI Surgical, and investigator-initiated grant funding from Sientra. Royalties are not derived from the product studied in this article, and Dr. Myckatyn has not ever used the product for which he has received royalties. Dr. Tenenbaum receives consulting fees from NC8, Allergan, and RTI Surgical, and grant funding from Mentor. The remaining authors have no financial interests to report.

ACKNOWLEDGMENTS

This study was funded by an unencumbered investigator-initiated research award to Dr. Myckatyn from RTI Surgical, Inc., and by a Barnes Jewish Foundation Grant to Dr. Myckatyn earmarked for studies related to postmastectomy breast reconstruction. Use of REDCap at Washington University was supported by a Clinical and Translational Science Award grant (UL1 TR000448), Siteman Comprehensive Cancer Center, and a National Cancer Institute Cancer Center support grant (P30 CA091842). The funders had no involvement in research study design or conduct of the study; collection, management, analysis, or interpretation of data; preparation, review, or approval of the manuscript; or the decision to submit the manuscript for publication.

The authors thank Colleen Kilbourne Glynn, Kelly Koogler, Corrine Merrill, and Tracey Guthrie for study administration. This study was funded by an unencumbered, investigator-initiated research grant to Dr. Myckatyn from RTI Surgical as well as funds from a Barnes Jewish Foundation grant to Dr. Myckatyn.