Scalp Reconstruction after Malignant Tumor Resection: An Analysis and Algorithm

Abstract

Background  An oncologic tumor resection of the scalp can result in complex wounds that result in challenging scalp reconstructions. This study aimed to evaluate the outcomes of microvascular-based scalp reconstructions (MSR) in oncologic patients and to propose an algorithmic treatment approach.

Methods  Within a 5-year period, 38 patients having undergone 41 MSR (15 anterolateral thigh (ALT), 15 gracilis muscle (GM), and 11 latissimus dorsi muscle (LDM) flaps) after extensive scalp tumor resections fulfilled inclusion criteria for this study.

Results  Malignant skin disease included superficial and/or deep infiltration of the calvarium in 26 and combined intracranial infiltration in 12 patients. In case of bone replacement (24 patients), MSR was done concomitant, otherwise MSR was performed after pathological confirmation of tumor-free margins. LDM flaps were used in cases with defect sizes of 400 to 1250cm ² , whereas ALT- and GM flaps were chosen for defects ranging from 40 to 350cm ² . The average length of the pedicle was comparable in ALT- and LDM flaps and longer than in GM flaps. Total flap loss with need for revision surgery and minor donor site morbidity occurred in four and three patients, respectively.

Conclusion  Microsurgical reconstruction of moderate-to-extensive scalp defects remains a reliable method with overall low risks and satisfactory aesthetic results, while, according to our experience, muscle flaps show the best functional and aesthetic results. However, in cases of central scalp defects and in situations when a long vascular pedicle of the flap is important, the ALT flap seems to be the best solution.

Introduction

Following oncologic tumor resection and/or radiation, the extent of scalp tissue loss can range from small partial-thickness- to large full-thickness wounds involving bone and even dura, with potential for subsequent cerebrospinal fluid leakage. ¹ Due to the inelastic galea, the paucity of the adjacent tissue, its limited expandability, and the convexity of its shape, reconstruction of scalp defects is often challenging. ^{2 3 4 5} Several reconstructive options have been discussed in the literature, including primary closure in case of small defects, skin grafting when the calvaria periosteum has been preserved, and local flaps in case of smaller to medium size defects. ^{2 3 4 5} Local flaps from the adjacent regions show the best color- and tissue quality match, which can be beneficial. ^{6 7} However, especially in cases of full-thickness and/or extensive scalp defects, as well as following neoadjuvant radiation, free flap coverage becomes often necessary. ^{2 4 5 6 7} Also, it has been shown that well-vascularized tissue can tolerate adjuvant radiation therapy much better, which minimizes soft-tissue complications. ⁸

However, when faced with the need for oncological procedures and subsequent need for free flap coverage, no ideal approach has yet been found. ^{2 4 5 6 7}

The aim of this retrospective case series was hence to evaluate the functional and aesthetic outcomes of different microvascular-based scalp reconstruction (MSR) approaches for moderate-to-extensive scalp defects following oncologic resections. An algorithmic treatment approach is presented based on the analysis ( Fig. 1 ).

Fig. 1.

Algorithm for oncologic scalp reconstruction in schematic form focusing on free flap reconstructions.

Patients and Methods

Over a 5-year period (2013–2018), 41 free flaps (anterolateral thigh- (ALT), gracilis muscle- (GM), and latissimus dorsi muscle- (LDM) flaps) were performed for scalp defect reconstruction following malignant tumor resection in 38 consecutive patients. There were no distinct exclusion criteria. However, a complete pre-, peri-, and postoperative dataset for every patient to be included was mandatory.

A retrospective chart review was performed to obtain the following data: age, gender, comorbidities such as diabetes, hypertension or peripheral arterial disease, and smoking status. The medical status of each patient, classified according the American Society of Anesthesiologists (ASA) Classification of Physical Status, was obtained in every patient preoperatively by the respective anesthesiologist in charge. ⁹ Oncologic data included histology, tumor extension (with/without intradural involvement), treatment (with/without radiation therapy [RT] prior to MSR), and adjuvant treatment (none vs. RT vs. CTx [chemotherapy]) ( Tables 1 and 2 ).

Table 1. Patient demographics of all assessed patients.

Characteristics	Assessed patients
Number of patients ( n )	38
Male	26 (68.4%)
Female	12 (31.2%)
Mean age (y)	69 (SD: 12.3)
Range of age (y)	40–91
Comorbidities ( n )
Diabetes mellitus	7 (18.4%)
Hypertension	24 (63.2%)
PAD	11 (28.9%)
Smoking status ( n )
Nonsmoker	35 (92.1%)
Smoker	3 (7.9%)
ASA score ( n )
ASA I	4 (10.5%)
ASA II	11 (28.9%)
ASA III	19 (50.0%)
ASA IV	4 (10.5%)

Abbreviation: ASA, American Society of Anesthesiologists Classification of Physical Status; PAD, peripheral arterial disease; SD, standard deviation.

Table 2. Tumor characteristics and infiltrations of all assessed patients.

Characteristics	Assessed patients
Number of patients ( n )	38
Bone infiltration
	26 (68.4%)
Intracranial infiltration
	12 (31.6%)
Bone replacement (Palacos)
	24 (63.2%)
Dural replacement
	8 (21.1%)
Concomitant MSR
Yes	24 (63.2%)
No	14 (36.8%)
Mean time a (d)	11.4 (SD: 8.1)
Range time a (d)	4–36

Abbreviation: MSR, microvascular-based scalp reconstructions.

^aLength of time in cases when MSR was not done concomitant.

The following surgical parameters were included for analysis: type of free flap, size of the defect, size of flap, recipient vessels, technique of arterial anastomosis, number of venous anastomoses, need for additional dural and/or bone replacement, operative times, length of ischemia, and characteristics of the hospitalization ( Table 3 ).

Table 3. Overall flap characteristics in patients needing scalp reconstruction with the use of a free flap (ALT flap; LDM flap; gracilis muscle flap).

Characteristic	Assessed patients
Free flaps
ALT	15 (36.6%)
GM	15 (36.6%)
LDM	11 (26.8%)
Overall defect dimension
Size (cm 2 )	221.9 (SD 176.2)
Range (cm 2 )	42–700
Overall flap dimension
Size (cm 2 )	328.2 (SD: 259.7)
Range (cm 2 )	78–1250
Recipient vessel ( n )
Superficial temporal artery	35 (85.4%)
Facial artery	3 (7.3%)
Superior thyroid artery	3 (7.3%)
Arterial anastomosis
End-to-end	41 (100%)
End-to-side	0 (0.0%)
Number of venous anastomoses ( n )
One	35 (85.4%)
Two	6 (14.6%)
Operative time (min)
Mean (min)	353 (SD: 58.1)
Range (min)	165–663
Length of ischemia (min)
Mean (min)	54.2 (SD: 16.5)
Range (min)	20–82
Hospitalization (d)
Mean (d)	11.2 (SD: 10.4)
Range (d)	4–51
Total flaps ( n )	41

Abbreviations: ALT, anterolateral thigh flap; GM, gracilis muscle flap; LDM, latissimus dorsi muscle flap; SD, standard deviation.

To compare the local conditions and different flap types, a separate analysis was performed ( Table 4 ). Additionally, all encountered complications were analyzed. Complications were divided into major and minor. Total flap loss and wound healing complications at the donor site were declared as major complications. All others complications, including infections, small wound healing disorders at the site of flap inset, and perfusion problems of the free flap with loss of less than 10% of tissue (tip necrosis) were declared as minor complications. Additionally, possible postoperative cerebrospinal fluid leakage was documented ( Table 5 ).

Table 4. Detailed flap characteristics in patients needing free flap scalp reconstruction separated according to the type of free flap (ALT flap, LDM flap, GM flap).

Characteristic	Assessed patients
Free flaps	ALT	GM	LDM
Defect dimension
Size (cm 2 )	140.9 (SD: 66.5)	183.5 (SD: 73.9)	602.7 (SD: 104.5)
Range (cm 2 )	42–240	45–300	460–760
Flap dimension
Size (cm 2 )	189.4 (SD: 73.8)	262.3 (SD: 89.5)	751.8 (SD: 206.3)
Range (cm 2 )	78–320	126–400	520–1250
Pedicle length (cm)
Size (cm 2 )	11.2 (SD: 2.5)	7.0 (SD: 0.92)	9.3 (SD: 1.9)
Range (cm 2 )	7–15	5–9	7–13
Recipient vessel ( n )
Superficial temporal artery	12 (80.0%)	15 (100%)	8 (72.7%)
Facial artery	1 (6.7%)	0 (0.0%)	2 (18.2%)
Superior thyroid artery	2 (13.3%)	0 (0.0%)	1 (9.1%)
Operative time (min)
Mean (min)	344.7 (SD: 93.7)	307.1 (SD: 86.5)	340.6 (SD: 124.6)
Range (min)	205–603	145–421	182–634
Ischemia (min)
Mean (min)	52.1 (SD: 19.1)	49.8 (SD: 19.2)	51.8 (SD: 12.9)
Range (min)	28–82	20–78	35–75
Total flaps ( n )	15	15	11

Abbreviations: ALT, anterolateral thigh flap; GM, gracilis muscle flap; LDM, latissimus dorsi muscle flap; SD, standard deviation.

Table 5. Detailed postoperative complications in patients needing free flap scalp reconstruction separated according to the type of free flap (ALT flap; GM flap; LDM flap).

Complications	ALT	GM	LDM	Sum
Total flap loss	1 (6.7%)	1 (6.7%)	2 (18.2%)	4 (9.8%)
Partial flap loss > 10%	2 (13.3%)	1 (6.7%)	0 (0.0%)	3 (7.3%)
Revision surgery	3 (20.0%)	2 (13.3%)	2 (18.2%)	7 (17.1%)
Arterial thrombosis	2 (13.3%)	2 (13.3%)	1 (9.1%)	5 (12.2%)
Venous thrombosis	1 (2.4%)	0 (0.0%)	0 (0.0%)	1 (2.4%)
Hematoma	0 (0.0%)	0 (0.0%)	1 (9.1%)	1 (2.4%)
Minor a	1 (6.7%)	1 (6.7%)	1 (9.1%)	3 (11.5%)
CFL	2 (13.3%)	1 (6.7%)	1 (9.1%)	4 (9.8%)
Donor site morbidity	1 (6.7%)	0 (0.0%)	2 (18.2%)	3 (11.5%)
Total flaps ( n )	15	15	11	41

Abbreviations: ALT, anterolateral thigh flap; CFL, cerebrospinal fluid leakage; GM, gracilis muscle flap; LDM, latissimus dorsi muscle flap.

^aWound dehiscence, wound infection, partial flap loss < 10%.

The operative techniques of the employed free flaps have been previously described in detail. ^{10 11 12} After careful dissection, the flap is transferred to the defect site and microvascular anastomosis is performed, usually with the superficial temporal vessels as the recipient vessels. In all patients, arterial as well as venous anastomosis was done by the use of interrupted, nonresorbable, monofilament sutures (size: 9–0). Other recipient vessels were only used in case of insufficient vessels or primary flap loss.

Statistical Analysis

Descriptive statistics, Pearson chi-square analysis with Fisher's exact test, nonparametric Wilcoxon signed-rank test for paired continuous data, and nonparametric Wilcoxon rank sum (Mann–Whitney U) test for comparison between treated and nontreated side were applied for statistical analysis. All analyses were performed using SPSS version 24. Statistical significance was defined as a p values of < 0.05.

Results

A total of 26 men (29 reconstructions) and 12 women (12 reconstructions) with mean age of 69 (±12.3; range: 40–91) years underwent 41 scalp (15 ALT-, 15 GM-, and 11 LDM-flaps) reconstructions following malignant tumor resection ( Table 1 ).

The original presenting pathological features included isolated malignant skin disease with or without superficial and/or deep infiltration of the calvarium in 26 patients and combined malignant skin disease with intracranial infiltration in 12 patients. Eight patients had dural exposure and required repair of the dura (6 artificial dural patches and 2 pericardium patch). In 63.2% of cases, the calvarium bone had to be resected and the defect reconstructed with artificial bone substitute (Palacos, Heraeus Medical GmbH, Wehrheim, Germany). In all patients that needed bone reconstruction using artificial bone substitute, concomitant MSR was performed. In case of isolated malignant skin tumors, prior to MSR a tumor resection was performed. Following histological approved complete resection of the malignant skin tumor, MSR was done. In these cases, mean time between tumor resection was 11.4 (standard deviation [SD]: 81, range: 4–36) days ( Table 2 ).

In 85.4%, the superficial temporal artery and vein were used as recipient vessels. Due to neoadjuvant radiation ( n  = 3) or primary flap loss ( n  = 3), in six patients the flow of the superficial temporal vessels was not sufficient, and thus the facial-( n  = 3) or superior thyroid artery and vein ( n  = 3) were used as recipient vessels. In all patients, vascular anastomosis was performed end-to-end. The mean operative time, including resection, was 353 (range: 165–663; SD: 58.1) minutes. Hospitalization after the reconstructions was 11.2 (range: 4–51, SD: 10.4) days on average ( Table 3 ).

A detailed analysis of the different flap types is demonstrated in Table 4 . LDM flaps were used for significantly larger defects (450–760 cm ² ) ( p  < 0.05), while ALT- (40–250) and GM flaps (45–350 cm ² ) were used for similar sized defects ( p  > 0.05). This correlation was also noted regarding the flap sizes (LDM: 520–1250 cm ² vs. ALT: 78–320 cm ² , p  < 0.05; LDM: 520–1250 cm ² vs. GM: 126–400 cm ² , p  < 0.05; GM: 126–400 cm ² vs. ALT: 78–320 cm ² , p  > 0.05), while length of the pedicle was significantly longer in ALT- (mean: 11.2, SD: 2.5, range: 7–15) and LDM flaps (mean: 9.3, SD: 1.9, range: 7–13) compared with GM-flaps (mean: 7.0, SD: 0.92, range: 5–9) ( p  < 0.05). Comparison of operative times and length of ischemia was similar among all free flaps ( p  > 0.05).

Complications

A detailed summery of postoperative complications is presented in Table 5 . Except for one ALT- and two LDM flaps, wound healing at the donor site was uneventful in all cases. In the three patients, no additional surgery was needed and the wounds healed under conservative therapy including daily wound irrigation by the use of saline solution and changing wound dressings. In these patients, additional antibiotic therapy was not necessary.

Regarding the recipient sites, total flap loss occurred in four patients (one ALT-, one GM- and two LDM flaps). In all of these lost flaps, recipient vessels were the superficial temporal arteria and vein. In three of these patients, an additional free flap was performed 5.3 (SD: 5.5; range: 3–8) days on the average after initial free flap. In one of these patients, again the superficial temporal vessels and in two patients the superior thyroid vessels were used as recipient vessels. Thereafter, the further postoperative course was uneventful. Given a palliative situation, in one patient a split-thickness skin graft was transferred to the patients' dura, which provided coverage for the time being. Three patients (one GM and two ALT flaps) suffered tip necrosis of the flap and needed secondary procedures. One of these patients had artificial dural and skull reconstruction. In two patients, the necrotic part of the flap was resected and the free flap could be mobilized and advanced for coverage of the resulting defect. However, in these patients no artificial dural and/or skull reconstruction was done. In one patient that received artificial dural and skull reconstruction due to cerebrospinal fluid leakage, a contralateral cerebral shunt was set and due to possible contamination of the artificial bone this was removed. After 12 days and suspending of the cerebrospinal fluid, the necrotic part of the flap was resected and the free flap was mobilized; however, the resulting defect needed to be covered with a split-thickness skin graft ( Fig. 2 ).

Fig. 2.

Case illustration. A 72 years old male patient suffering from a (radiation-induced) meningiomatosis of the right hemisphere. ( A ) Multiple recurrent meningiomas WHO grade II with intracranial and transcutaneous extension after multimodal treatment including craniectomy. ( B ) Surgical result after complete tumor resection, dura replacement, scull reconstruction (Palacos, arrow) and anterolateral thigh flap (A. temporalis). ( C ) Secondary tip-necrosis with exposure of the Palacos plastic and cerebrospinal fluid leakage. ( D ) Revision surgery with removal of the Palacos plastic, dura sealing, repositioning of the flap plus consecutive skin graft and secondary ventriculoperitoneal shunt implantation (arrows) for chronic hydrocephalus malresorptivus. ( E ) Bird‘s eye and side view at 8 weeks follow-up control.

Minor complications occurred in three patients (one ALT-, one GM-, and one LDM flap). In one of these patients (GM flap), at 7 days postoperative despite adequate underlying flap perfusion there was a wound healing disturbance on the meshed skin graft. This was likely due to bacterial colonization. The mesh graft was removed and replaced, which finally resulted in good take. In four patients, postoperative cerebrospinal fluid leakage occurred that resulted in minor-to-major flap complications. In these patients, lumbar drain or cerebroabdominal shunt was set. Among these measures secretion suspended within 6.4 postoperative weeks on the average ( Table 5 ).

Discussion

The presented patient series illustrates the advantages and disadvantages of different reconstructive techniques of large scalp defects in oncological patients. A variety of different reconstructive techniques have been described before in the literature. ^{2 13 14} These include cases of acquired scalp defects, ¹³ free flap reconstructions of the scalp and calvaria, ¹⁴ and MSR in tumor resection. ² To our knowledge, a specialized algorithm for oncologic scalp reconstruction with a focus regarding the optimal type of free flap coverage has not been described so far ( Fig. 1 ).

Several therapeutic options can be offered to the patient depending on the underlying disease and possible need for neo- and/or adjuvant radiation, the size and location of the defect, preservation of the periosteum, and the overall condition of the patient. Given the exposed location of the scalp, an additional challenge for the reconstructive surgeon is to achieve an adequate functional coverage. ²

In case of smaller defects (up to 3–4 cm in diameter), primary closure with wide undermining of the galea enables excellent functional and aesthetic outcomes. ^{2 3 13 14 15 16 17}

In cases of preserved periosteum/remaining galea or even small circumscribed bone defects with intact dura coverage after tumor resection skin grafts can be used for coverage. ^{2 13 14} However, especially after radical oncological tumor resection, the periosteum is normally gone. The technique of stimulating granulation tissue by removing the outer table of the calvaria bone should only be done as a last resort, given the unacceptable relation of risks and benefits. ² Although skin grafts can provide satisfactory aesthetic results, their use within the hair-bearing scalp often results in unstable and unaesthetic scars, because they do not perfectly match with the color, structure and thickness of the scalp. Furthermore, postoperative radiation, when necessary, is not well tolerated. ^{18 19} For these reasons, skin grafts are not the first choice for reconstruction.

Small-to-moderate-sized defects can be closed by local scalp flaps. A variety of different reconstructive techniques have been described. ^{13 20 21 22} Functional and aesthetic results are often excellent, with minimal donor site defects and good wound healing. The maximal defect size to use local flaps depends on the location of the defect. Although Orticochea reported on the use of his local flap technique in defects up to 16 × 13 cm, ^{20 21} in our opinion defects within the hair-bearing scalp of up to 6 to 8 cm in diameter are best suited for this method. According to the reconstructive ladder, prior to free flap surgeries, local flap reconstructions could be an option. ²³ Especially elderly patients are good candidates for local scalp flaps, as laxity of the skin prevents distortion of the surrounding tissues and the hairline. ²⁴ In previously irradiated patients, the use of local flaps can be limited, as neighboring scalp tissue is often fibrotic resulting from the radiation therapy. ^{25 26}

Should local- or transposition flaps not be possible for any of the aforementioned reasons, free flap reconstruction usually becomes the next best options. The major advantage of free tissue transfer is that it provides well-vascularized, healthy tissue from a remote area. ¹⁴ This becomes especially important in patients who have been previously treated with radiation therapy ⁸ and whose tissue is consequently damaged. Patients with previous history of osteomyelitis, osteoradionecrosis, prior local flap failure, heavy trauma, or high-grade malignancy usually also benefit from free flap reconstructions. ³ The well-vascularized tissue allows adjuvant radiation therapy and reduces the risk of soft-tissue complications. ⁸ Additionally, the robust vascularization of free flaps enables better antimicrobial defense and wound healing, which leads to shorter hospital stays. ^{4 27}

In our study, all patients had defects from oncological tumor resection requiring free flap reconstruction. Regarding the question which free flap is the best choice, various criteria such as the defect size and depth, location, and surgeon preference have to be taken into account. We thus compared the characteristics of three different standard workhorse-free flaps: the ALT flap, the LDM flap, and the GM flap.

In our patient collective, the mean defect size was 221.9 (range: 42–700, SD: 176.2) cm ² . In the current literature, defect sizes up to 150 cm ² are usually defined to be small to moderate. ^{28 29} We used the ALT- and the GM flaps for defects ranging from 45 to 350 cm ² . As demonstrated by our group before, by spreading of the epimysium, the free GM flap can be expanded and then cover defects of up to 400 cm ² . ^{3 29} However, the GM flap is limited by a significantly shorter vascular pedicle compared with the LDM- and ALT flap. Therefore, in central defects (i.e. involving the vertex) or insufficient temporal vessels, the ALT flap presents the best option. Because in ALT flaps spreading is not possible, in our experience only defects of up to 350 cm ² can be covered. In cases of defects greater than 350 cm ² , an LDM free flap is the best solution, even though the LDM flap is associated with higher donor site morbidity. ^{3 29 30 31 32}

The main advantage of the free ALT flap, besides its low donor site morbidity, lies in its versatility regarding design and composition, the possibility to harvest it simultaneous with the resection, and its long vascular pedicle, ³³ which was 11.2 (range: 7–15, SD: 2.5) cm on average in our series. A long pedicle becomes especially important in cases of parietal or central defects, and when nearby recipient vessels are not sufficient. ³³ However, care has to be taken not to twist the pedicle to avoid arterial and/or venous congestions. The ALT flap can be harvested as a cutaneous-, fasciocutaneous, or myocutaneous flap, and it can be split into multiple skin islands or into two different free flaps from a single donor site based on multiple perforators. ^{31 33 34} As the ALT flap can provide an extremely large area of skin, muscle, and fascia, it is suitable to cover also extensive defects of the scalp. ³⁵ It can also be harvested to include the vastus lateralis muscle if dead space obliteration is required. ³⁶ However, excessive bulkiness and thickness of subcutaneous tissues present the main disadvantage of the flap, as it does not match with the thickness of the native scalp. Postoperative debulking by liposuction can become necessary in these cases. ^{37 38 39} Another disadvantage of the ALT flap lies in its tendency to migrate caudally over time, which may lead to poor aesthetic results, especially if defects are located close to the forehead. Finally, the flap might be quite hairy in male patients, which can become problematic if defects include the forehead. ^{37 38 39}

Another option for reconstruction of small-to-moderate defects is the GM flap. This flap also has a low donor site morbidity and it can be harvested simultaneously to the resection of the tumor. It can be used as a sole muscle- or musculocutaneous flap. ^{30 40} Unlike the ALT flap, the GM flap is not as bulky and it tends to adhere better to the underlying surface. When comparing the donor site morbidity between the ALT and GM flap, there is a slight tendency toward lower complication rates in GM flaps than ALT flaps. ^{41 42} However, in our opinion based on donor site morbidities, no recommendation which flap is favorable can be done. In case of ALT flaps, donor site complications range from 11 to 21.3% and include wound dehiscence (3.8%), seroma (2%), as well as hematoma (0.9%). ⁴² Especially when harvesting wide ALT flaps, these donor site complications rise. ⁴³ In case of GM flaps, wound healing disorders including dehiscence (0.9%) and seroma (1.3%) are the most common complications. ⁴² However, compared with the ALT- and GM flap harvesting of the LMD flap results in higher donor site morbidity. ³² These results are also confirmed by our findings.

Another important advantage of the GM flap is its great variability in shape, as it can be used to fill deep defects or can be flattened out to cover large areas. ^{30 40} This enlargement of the flap can be managed by microscopically aided intramuscular dissection of the perimysium. ^{3 29 30} When using the GM flap as a muscle flap, a split-thickness skin graft for coverage is necessary. To achieve the best results in quality and color, the skin grafts should be taken close to the recipient side. ³⁰ However, in our experience, it is advisable to use the skin island located directly over the GM as a donor site for the split-thickness skin graft. ^{30 40}

In case of larger or whole scalp defects with an average size of 602.7 (range: 460–760, SD: 104.5) cm ² , we used the latissimus dorsi flap, because it supplies the largest flap dimension with 751.8 (range: 520–1250, SD: 206.3) cm ² on average. Further, it can be draped very nicely over the convex scalp. ^{44 45} It has a long, constant, and reliable pedicle, which can be extended using the subscapular system, and can also be used as an osteocutaneous free flap in combination with the scapula and/or a rib to reconstruct cranial bone as described by Lee et al. ^{46 47} A major benefit of the LMD flap is its excellent vascularity, particularly when being used over bone. ¹ This is very important in cases of previously irradiated patients, ⁸ or if adjuvant radiation therapy is anticipated. ¹ However, similar to the GM flap, the LMD flap results the good match in the color of the scalp, due to the reddish coloring of the muscle becomes visible through the graft. ¹ This color advantage also remains after the muscle has atrophied over time.

To prevent bacterial colonization and infection, in the authors' opinion, in all patients that need bone reconstruction by the use of artificial bone substitute, concomitant MSR should be performed. In case of isolated malignant skin tumors, prior to MSR an oncological tumor resection should be done, and following histological approved complete resection of the malignant skin tumor, MSR is necessary. This is also underlined by different other authors. During the time period of primary tumor resection and finally MSR, wound closure was done by the use of topical negative wound pressure therapy or alloplastic artificial skin analog. Additionally, during that time period all patients receive antibiotic prophylaxis by intravenous cephalosporins.

Limiting factors of the presented study include its retrospective character and small number of patients. In the future, prospective studies including more precise acquisition of intraoperative details as well as pooling data of large microsurgical centers could help to further elucidate risks and prognostic factors associated with complex microsurgical reconstructions.

Conclusion

Following oncologic tumor resection, microsurgical reconstruction of moderate-to-extended scalp defects remains a reliable method with overall low risk and good aesthetic results. In our experience, fasciocutaneous flaps like the ALT flap may show poor color match, whereas muscle flaps (GM and LDM) tend to adhere better down to where originally placed. However, in case of parietal or central scalp defects and when nearby recipient vessels are not sufficient, a long vascular pedicle of the flap is important. In these situations, the ALT flap seems to be the best solution. In case of even more extensive or complete scalp defects, the free LDM flap should be the first choice for reconstruction.

Funding Statement

Funding Statement This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.