A Comparison of Perioperative Blood Transfusions and Other Complications Following Transverse Rectus Abdominis Myocutaneous Flaps versus Deep Inferior Epigastric Perforator Flaps

Objective: Abdominal based free flaps such as the Transverse Rectus Abdominis Myocutaneous (TRAM) flap and Deep Inferior Epigastric Perforator (DIEP) flap are essential tools in the reconstructive armamentarium post mastectomy. These reconstructions are often prolonged, complex and associated with complications. This study aims to compare the complication rate, particularly focusing on perioperative transfusions, between TRAM and DIEP flaps performed by the senior author in our tertiary referral centre. Methods: A retrospective review was conducted of one hundred and seven consecutive TRAM and DIEP flaps from 2000 to 2014. The two groups were analysed for demographics, preoperative risk factors and post-operative complications including blood transfusions, haematomas, flap losses, redo-anastomoses, flap infections, abdominal wound sequelae and non-surgical complications. Results: Sixty-three patients underwent 67 free/muscle sparing TRAM flaps and 35 patients underwent 40 DIEP flaps. There were no statistically significant differences in patient demographics or preoperative risk factors between the two groups. Five TRAM flap cases (7.9%) required transfusion whilst no DIEP flap cases required transfusion. This difference was not found to be statistically significant (p = 0.16). However, free/muscle-sparing TRAM flaps were found to have a significantly higher overall complication rate compared to DIEP flaps (23.8% vs. 5.7%, p = 0.02). Conclusion: The current study demonstrated no difference in perioperative transfusion requirement between TRAM and DIEP cases. There was however a significantly higher rate of overall complications associated with TRAM flaps warranting the authors to conclude that care be taken when opting for this reconstructive method.


Introduction
The use of autologous free tissue transfer for breast reconstruction post mastectomy has become a main stay of management throughout the past decade [1] [2] [3] [4] [5]. Abdominal based free flaps such as the Transverse Rectus Abdominis Myocutaneous (TRAM) flap and Deep Inferior Epigastric Perforator (DIEP) flap are essential tools in the reconstructive armamentarium post mastectomy. These are often prolonged, complex procedures which can result in more complications than other reconstructive options [6]. Research comparing the two approaches seems to indicate that each method has its strengths and weaknesses. More specifically TRAM flaps have been associated with more abdominal hernias and fat necrosis [7], longer hospital stays and worse postoperative pain [8], whilst DIEP flaps have been reported to have higher numbers of flap losses [9]. Despite an adequate level of research comparing most complication rates there has been a relative paucity of attention given to comparing perioperative transfusion rates between these two reconstructions. Individually TRAM cases have been found to have transfusion rates ranging from 6% to 95% [10]. Similarly, DIEP cases have been reported to have transfusion rates ranging from 9.1% to as high as 80.3% [11] [12].
Perioperative blood loss is a significant issue as it directly impacts flap outcomes [13]. Equally important was the finding of a recent review which highlighted the tendency for increased medical and surgical complications in free flap reconstructions that underwent blood transfusion post operatively [14]. Thus the decision to transfuse blood should not only take into consideration the scarcity of blood products and the cost to the health care system but also the potential harm to the patient. This study aims to compare the rate of complications, particularly focusing on blood transfusion rates, between free/musclesparing TRAM and DIEP flap breast reconstructions at a tertiary referral center in Sydney, Australia.

Method
A retrospective review was conducted of consecutive free/muscle-sparing TRAM and DIEP breast reconstructions performed by the senior author [TL] and his team between January 2000 and January 2014 at Westmead Hospital, a tertiary referral center in Sydney, Australia. One hundred and two cases were identified of which 4 patients were excluded as they underwent a combination of a free/ muscle-sparing TRAM and a DIEP flap for bilateral reconstructions in the single procedure leaving 98 cases for analysis.
Patient files and electronic records were reviewed for each case. Records were analysed for preoperative risk factors, post-operative complications and transfu-sion requirement. Risk factors analysed included age, body mass index (BMI), smoking status, prior adjuvant therapy, pre-operative haemoglobin concentration, timing and extent of reconstruction. Complications assessed included haematoma, flap loss (partial or complete), flap infections, re-do anastomoses, abdominal wound sequelae and non-surgical complications. The protocol for transfusion opted by the unit was a haemoglobin concentration less than 7 g/dl or if patients were clinically symptomatic. A combination of unpaired t tests and Fisher's exact tests were performed to assess for a difference between TRAM and DIEP groups in relation to demographics, risk factors, transfusion rate and other complication rates. A value of p < 0.05 was considered significant in all statistical analyses.

Flap Choice
Since transitioning to DIEP reconstructions in 2003, the senior author discovered zone 4 to be unreliable and with varying venous insufficiency making the choice to do DIEP more selective. TRAM flap reconstructions were opted for when patients needed large flaps especially if zone 4 was included. TRAM flaps were also chosen for patients who had previously received radiotherapy.

Demographics
A total of 107 consecutive free/muscle-sparing TRAM and DIEP flaps were performed on 98 patients. Sixty-three patients underwent 67 free/muscle-sparing TRAM flaps and 35 patients underwent 40 DIEP flaps. The mean age of both groups was 50 years. Mean BMI was 30.0 for TRAM patients and 32.1 for DIEP patients. Thirty (47.6%) TRAM patients and 17 (48.6%) DIEP patients underwent preoperative adjuvant therapy. Tobacco use was reported in 22 (37.3%) TRAM cases and 16 (47.1%) DIEP cases. Four (6.3%) TRAM cases and 5 (14.3%) DIEP cases underwent bilateral reconstructions. Immediate reconstruction was performed on 14 (22.2%) TRAM cases and 6 (17.1%) DIEP cases with the remaining cases being delayed reconstructions. Mean preoperative haemoglobin concentration was 13.6 g/dL in both TRAM and DIEP cases.
Using an unpaired t test, there was no difference identified in age, BMI or preoperative haemoglobin levels between the groups. There was also no significant difference identified between the groups in exposure to adjuvant therapy, tobacco use, number of bilateral cases and immediate reconstructions using Fisher's exact tests (Table 1).

Blood Transfusions
A total of five cases (5.1%) required transfusion. All five cases were TRAM flaps (7.9% of TRAM cases). Two of these cases were transfused due to intraoperative blood loss. A further two cases were transfused after returning to theatre for evacuation of chest wall haematomas. One case required transfusion for significant post-operative blood loss following revision of a thrombosed arterial anas- Three of 78 delayed cases (3.8%) required transfusion compared to two of 20 (10%) immediate reconstructions. Also four out of 89 (4.5%) unilateral reconstructions and one out of 9 (11.1%) bilateral reconstructions underwent transfusion. These differences were also found not to be statistically significant using Fisher's exact test (p = 0.27 and p = 0.39).

Other Complications
Overall Fisher's exact test, the higher rate of complications associated with TRAM flaps was found to be significant (p = 0.02). There was no statistically significant dif-ference found for any individual complication between the two groups ( Table   2).

Discussion
Autologous  This is markedly higher than our rate of 7.9%. The disparity maybe due to the larger scale of their study or their indication to transfuse being a haemoglobin value less than 8 g/dL, as opposed to less than 7 g/dL in our study.
Similarly the DIEP flap transfusion rate of 0% in our study was lower than that reported in the literature. Appleton et al. (2011) after reviewing 144 DIEP cases reported 18.8% of patients required blood transfusion. The decision to transfuse was similar to our study in that cases were transfused if there were clinical signs of hypovolaemia or the haemoglobin was less than 7 g/dl [18].
Likewise, Lymperopoulos et al. (2013) studied 131 patients who underwent DIEP reconstructions and found 9.1% of patients required transfusion post operatively [11]. However a study performed in Melbourne, Australia noted that up to 80.3% of DIEP cases in their study required perioperative transfusion [12]. This high number of cases may be due to the fact that some patients included in the study were preoperatively anaemic. Though the authors have not mentioned how many cases were anaemic the mean preoperative haemoglobin for the study was 11.8 g/dL, with a value less than 12.0 g/dL deemed anaemic. The authors found a significant relationship between preoperative anemia and post-operative transfusion requirement. The current study had a mean preoperative haemoglobin of 13.6 g/dL for both TRAM and DIEP cases with only two patients having a haemoglobin level less than 12.0 g/dL.
In the current study the difference in transfusion rates between free/musclesparing TRAM and DIEP flaps was not found to be significant. This may be a result of the small sample size having inadequate power to detect a significance.
Perhaps with a larger sample size the increased rate of transfusion in TRAM flaps may have been deemed significant. If this were the case then a possible explanation may be that there is greater blood loss when harvesting TRAM flaps due to the harvesting of muscle. It is however difficult to conclude that the reason for transfusion could be ascribed to the type of flap performed in this series. In relation to overall complication rate this study found TRAM flaps to have significantly more complications than DIEP flaps. This finding is in contention with the literature which has mainly found no difference in overall complication rates between the two methods [7] [19]. In their meta-analysis Wang et al (2014) found TRAM and DIEP flaps to have complication rates of 25.6% and 27.9% respectively [7]. The TRAM complication rate in the current study was similar at 23.8% however our DIEP complication rate was lower at only 5.7%. Furthermore they noted that TRAM flaps were associated with significantly higher rates of hernias whereas DIEP flaps were associated with more incidences of fat necrosis.
These findings were not shared in the current study with no statistically significant difference found in any individual complication type between the two groups. This finding may be due to the initial learning curve experienced when TRAM reconstructions were first performed. It is likely that experience with complications arising from TRAM flap reconstructions would have aided in preventing similar events in DIEP cases. Alternately, the higher complication rate of TRAM flaps may be due to the procedure being more invasive and harvesting a larger amount of tissue than DIEP reconstructions. Also, the resultant defect is larger making primary closure more difficult and possibly leading to abdominal wound issues. Furthermore, the subsequent pain and reduced mobility may explain the increased incidence of atelectasis and pneumonia in this group compared to DIEP flaps.
The current study also analysed the data to assess for a possible operator dependency given our centre is a teaching hospital and surgeons in training often perform procedures. The majority of cases were performed by the senior author who is a Senior Consultant Surgeon in Plastic and Reconstructive Surgery. He transitioned to DIEP flap reconstructions from 2003. There was no evidence of operator dependency in regards to complications or perioperative transfusion requirement.

Conclusion
Given the association of blood transfusion with increased post-operative complications in free flap surgery it is important to determine whether one form of abdominal based reconstruction is more likely to result in blood loss requiring transfusion than the other. The current study found no difference in transfusion rates between TRAM and DIEP flap reconstructions. However TRAM flaps were found to have a higher overall complication rate than DIEP flaps. Consequently the authors advocate for greater care and awareness when opting for free/muscle-sparing TRAM reconstructions. Future research in the form of large prospective randomised trials may be better able to elucidate a difference in transfu-sion rates if one exists between these two work horses of the reconstruction ladder.