|Year : 2015 | Volume
| Issue : 1 | Page : 14-19
Surgical intervention in haemangiomas of the head and neck Bradford Cannon closure: results reassessment
Ahmed M Hegazy1, Atef A Allam1, Abeer A Hodeib2, Hassan M Hegazy3
1 Department of Plastic and Reconstructive Surgery, Faculty of Medicine, Tanta University, Tanta, Egypt
2 Department of Dermatology and Venereology, Faculty of Medicine, Tanta University, Tanta, Egypt
3 Department of Otolaryngology-Head and Neck Surgery, Faculty of Medicine, Tanta University, Tanta, Egypt
|Date of Submission||26-Nov-2014|
|Date of Acceptance||07-Apr-2015|
|Date of Web Publication||7-Aug-2015|
Abeer A Hodeib
Assistant Professor of Dermatology and Venereology, Faculty of Medicine, Tanta University, Tanta
Source of Support: None, Conflict of Interest: None
Haemangioma is the most common vascular anomaly that affects infants. Although it is self-regressing, it can block visual field and airway. It might also cause mutilation. In surgically indicated cases, the residual defect is often large enough to present as a true reconstructive problem.
The aim of this study is to evaluate the purse-string closure method for circular and ovoid defects after excision of haemangioma in the face and scalp.
Patients and methods
We operated 25 targeted lesions (25 patients) with head and neck infantile haemangiomas. Ten lesions were in the involution stage, eight lesions were still in the proliferation stage and seven lesions were ulcerated and/or bleeding.
There were reductions in the defects' dimensions by 53.8% for the longest dimension, 75% for the shortest dimension and 88% of the defect surface area. The results were very encouraging in areas with mobile skin, but we discourage any use of this method in the areas around the mouth and the eyelid because of initial distortion of the treated areas. One scalp case developed infection and the purse string had to be removed. The number of lesions in the scalp was markedly lower than the least number to enable a statistical analysis.
The purse-string closure (Bradford Cannon closure) is a powerful tool for the closure of skin defects after haemangioma excision in the face. It should be considered a first option whenever it does not distort a vital structure such as the mouth and the eyelids. For the scalp, the value of this technique needs further study considering that there are other good options as alternatives for this technique, particularly for the scalp.
Keywords: Bradford Cannon closure, haemangioma, purse-string closure
|How to cite this article:|
Hegazy AM, Allam AA, Hodeib AA, Hegazy HM. Surgical intervention in haemangiomas of the head and neck Bradford Cannon closure: results reassessment. Egypt J Dermatol Venerol 2015;35:14-9
|How to cite this URL:|
Hegazy AM, Allam AA, Hodeib AA, Hegazy HM. Surgical intervention in haemangiomas of the head and neck Bradford Cannon closure: results reassessment. Egypt J Dermatol Venerol [serial online] 2015 [cited 2020 May 31];35:14-9. Available from: http://www.ejdv.eg.net/text.asp?2015/35/1/14/162457
| Introduction|| |
Haemangiomas are the most common vascular anomaly in infants. It affects up to 12% of White infants, with a lower incidence in other races by the first year of age. The female to male incidence ratio is 3: 1. Eighty percent of the cases present as a single lesion, 60% occurs in the head and neck, 25% in the trunk and 15% in the extremities ,. Haemangiomas are classified as vascular tumours among a broad classification of vascular lesions. Other members of this classification and other highly vascular tumours of infancy, for example, sarcoma represent the differential diagnosis of haemangioma .
Vascular lesions can be classified as follows :
(1) Vascular tumours:
- Haemangioma of infancy (appear after birth): superficial, deep and mixed. It is the most common type of haemangioma and the most common tumour of infancy. They have a short progressive phase for 1 year, followed by regression in 50% of cases by the fifth year and 100% by the 10th year of life. Recently, a glucose transporter 1was recognized as a universal marker exclusive for this type.
- Congenital haemangioma (present at birth): rapidly involuting (within 10 months) and noninvoluting. It is the least common type of haemangioma.
- Kaposiform haemangioendothelioma.
- Tufted angioma.
- Pyogenic granuloma.
(2) Vascular malformations:
- Simple malformations: capillary (portwine stain), venous, lymphatic and arteriovenous malformations.
- Combined malformations: capillary-venous, capillary-venous with anterovenous shunting and/or fistula, capillary-lymphatic-venous and cutis marmorato telangectatic congenital.
Because of the benign nature of most haemangiomas and the fact that most haemangiomas are nonsyndromic, little was done to determine its aetiology . Several placental antigens were isolated besides glucose transporter 1 and a single cell with a somatic chromosome alteration is responsible for cloning . Histopathology of proliferating haemangiomas shows active proliferating endothelial cells and immunohistochemical staining shows the presence of endothelial progenitor cells compared with flat endothelial cells in vascular malformations .
The management of haemangiomas includes the following ,,:
- Medical treatment options: intralesional steroids, systemic steroids, b-blockers, calcium channel blockers, interferon-α and occasionally vinchristine.
- Laser treatment.
- Surgical treatment, which is the main topic of this work.
Surgery as a first option in the following cases ,,:
- Cases including lesions blocking airway, lesions with massive bleeding and lesions potentially cause handicapping e.g. blocking the visual field as their medical treatment takes time.
- Incomplete resolution leaving fibrofatty tissues and/or excess skin.
- Lesions with mutilating potential even after resolution, i.e. leaving ugly and/or large scars in apparent cosmetic sites such as the nasal tip, eyelid, scalp, etc.
Thus, surgery can be considered at any stage of the natural history of haemangioma according to the above-mentioned criteria. Because of the fact that surgery always raises the questions of both scarring and blood loss, which can be considerable from a vascular tumour excision in an infant, there are contentious efforts to minimize both.
Standard elliptical excision will result in flattening in curvy areas, for example, the cheek. This flattening will also be aggravated by dog ears. To avoid dog ears, the ellipse has to be a (1: 3) width: length ratio, that is, the operation will result in a much longer final scar than the longest dimension of the original lesion . Mulliken et al.  published a predictive formula that estimates the final scar length expected if the lesion is excised in a circular manner and a minimum skin ellipse is excised to result in a linear scar without dog ears as follows:
Final scar length = initial defect widest dimension × 3 × (shortest defect dimension/longest defect dimension).
According to this formula, a circular defect of 2.4 cm reconstructed in a lenticular manner will result in a 7.2 cm scar. According to Mulliken et al. , the only advantage of this scar is to be thin and linear.
Circular excision and purse-string closure is not new in surgery. It has been used to treat umbilical hernias and appendicular stump closure. The first recorded use in plastic surgery was by Bradford Cannon, in 1964, to treat a cheek defect after the removal of a sebaceous cyst . Further studies appeared in the following decades following the method of Bradford Cannon ,. According to the results of Mulliken et al. , Cannon closure enables the maximum likelihood for reduction of scar size.
The aim of this work is to assess the value of Bradford Cannon closure for reconstruction of posthaemangioma defects in the face and scalp.
| Patients and methods|| |
A total of 25 patients were included in this study (25 patients) in the period between March 2011 and March 2013. Patients were recruited from Plastic Surgery, Dermatology and Venereology, and ENT clinics. Consents of the institutional review board and ethics committee were obtained. Surgery was carried out, and then follow up, prescription of scar-modulating treatment and final scar measurement and assessment. We operated targeted patients with head and neck infantile haemangiomas with mutilating potentials and those with ulcers and bleeding. All patients were diagnosed by a careful assessment of clinical history and clinical examination. All of them were subjected to routine preoperative laboratory tests (complete blood film and coagulation profile). A computed tomographic scan was ordered for patients with a suspected underlying skull defect and/or lesions near midline or skull suture lines to exclude intracranial pathology . A detailed health education meeting was offered to parents about the child's medical condition and the planned surgical procedure including the goals of surgery, and the fact that the wound will appear distorted for a few weeks. Parents of all patients were informed about the purpose and method of the operation. Informed consent was taken after discussion of the procedure.
After skin disinfection and draping, a series of transverse mattress sutures using (0) prolene were performed on the normal skin 1 cm away from the nearest skin pathology as a custom-made tourniquet to minimize bleeding [Figure 1].
The lesion was completely excised along with any skin with abnormalities such as scar, hyperpigmentation, etc., but any normal skin was preserved. This type of excision usually resulted in a circular, ovoid or sometimes irregular defect. Proper haemostasis was performed and the longest dimension of the defect was measured and recorded [Figure 2].
|Figure 2: The residual defect after circular excision and the potential length of the scar in case of linear closure.|
Click here to view
No or minimal undermining was performed to maximize blood supply to the defect edges ,,.
A purse-string suture using a nonabsorbable suture (3/0 prolene) putting equal bites. All were at the same level in the dermis. The point of inlet of the needle was the exit point of the previous stitch , [Figure 3]. The stitches were tightened and a few 5/0 prolene simple stitches were applied to approximate the skin in the direction of the relaxed skin tension lines ,,.
The simple stitches were removed 5 days after surgery and the purse-string stitch was removed 8 weeks after surgery .
Then, the patient was followed up by the dermatologist. A silicon-based scar-modulating cream was prescribed and finally the scar was measured and assessed after 6 months.
Statistical analysis 
Statistical analysis was carried out using the SPS Statistical Software (SPSS Inc., Chicago, Illinois, USA). The differences in the measurements of the defect and the final scars were analyzed using the t-test to determine the significance of the correlated means. The same test was used to compare haemangiomas of different stages in the natural history of haemangioma. P-value of 0.001 for all statistical analyses carried out in this work.
| Results|| |
A total of 25 patients were included in this study. Five were males and 20 were females (ratio 1: 4). All of them had head and neck infantile haemangiomas. One patient had two lesions: one in the lateral forehead and one in the lower lip (lip lesion was excluded: see later) [Figure 4]. Ten of the lesions were in the involution stage, eight lesions were still in the proliferation stage and seven lesions were ulcerated and/or bleeding. The lesions were located on the cheek (12), forehead (five), glabella (three) and nose (five). Of these 25 lesions, two were blocking the visual field: one in the cheek and the other in the glabella. Actually, there were five additional lesions (one in the scalp, two in the lip and two in the lower eyelid), but the lesions were excluded from this study during surgery (see the Discussion section). Therefore, the total number of lesions was reduced from 30 to 25 [Figure 5].
The mean longest dimension of the defect was 2.6 cm whereas that of the shortest dimension was 2.4 cm. The mean longest dimension of the final scar was 1.2 cm whereas that of the shortest dimension was 0.6 cm. Therefore, a 53.8% reduction in the mean longest dimension and a 75% reduction in the mean shortest dimension were achieved using the single purse-string procedure.
There was a statistically significant reduction in both dimensions after a single purse string. The mean percentage reduction of the longest dimension was less than that of the shortest dimension even in cases with almost circular defects with the longest axis parallel to relaxed skin tension lines.
The mean surface area of the defect after excision was = 22/7 (2.6 × 2.4) = 19.6 cm 2 . The mean final scar surface area was = 22/7 (1.2 × 0.6) = 2.26 cm 2 . This yielded an ~88.4% reduction (i.e. significant reduction) of the total surface area after single purse-string closure. In terms of blood loss, we did not use more than two 4 × 4 gauzes in any of our cases. All the cases were treated as daycare surgery. Examples of the results of our work are shown in [Figure 6].
The parents of 16 patients were satisfied with the single purse-string procedure, five decided on another purse-string later and four opted for a linear excision later. None of the parents reported dissatisfaction.
| Discussion|| |
Because of the fact that 85-90% of haemangiomas are benign in nature, follow-up is emphasized by most doctors. However, still, 10 to 15% of cases develop complications such as ulceration, bleeding, visual field block and/or airway block . The negative psychological impact on the patient and his/her relatives either from the lesion or its residues should also sometimes be strongly considered . Medical treatment by propranolol and captopril was prescribed ,,. Propranolol was described as the first line of treatment for noncomplicated cases, and was reported to be successful in some ulcerating haemangiomas by Naouri et al. .
Medical treatment is not a wise choice in painful ulcerated haemangiomas, or a mutilating lesion in surgically fit patients because it usually exposes these patients to a risk of infection and bleeding, and later mutilation and psychological distress for the child and his family is obviously not a wise choice .
Mulliken et al.  studied purse-string closure for haemangiomas all over the body, but in this work, we chose haemangiomas of the face and scalp not only because of the functional and aesthetic importance of these areas but also because of the variability in skin characteristics in these areas and the aesthetic and anatomical landmarks that should be considered, besides the fact that in these areas, haemangiomas are prevalent with the highest incidence.
In our series, a nonabsorbable monofilament stitch was used and this was removed 8 weeks after surgery. Mulliken et al.  and Yuen  used slowly absorbed stitches in their purse string, whereas Tremolada et al.  used a nonabsorbable stitch that was removed 6-8 weeks later. An 8-week period is sufficient for mechanical support until there is a proper increase in the tensile strength of the wound.
In this series, the mean longest dimension of the defect was 2.6 cm whereas that of the shortest dimension was 2.4 cm. In the study of Mulliken et al. , the mean longest dimension was 2.4 cm, whereas the mean shortest dimension was 2.3 cm.
The mean longest dimension of the final scar in this work was 1.2 cm (53.8% reduction) whereas that of the shortest dimension was 0.6 cm (75%), with a reduction in the mean surface area reduction of about 88.4%. In the study of Mulliken et al. , the longest dimension of the final scar was 1.3 cm (45% reduction) and the shortest dimension of the final scar was 0.6 cm (74% reduction), with a mean surface area reduction of 85%.
Both ours and Mulliken et al.'s  studies found a statistically significant reduction in postexcision defect dimensions and surface area. The better outcome of this work: In our opinion, the better outcome of this work is due to the usage of non absorbable stitches while the absorbable stitches lose their tensile support to the wound long time before their disappearance. It is noteworthy that the study of Mulliken et al.  included cases of haemangiomas in the limbs and the nature of the skin of the limbs in terms of thickness and anchorage to deeper structures is different from that of the face, and this definitely affected the outcome of their study.
We agree with the statement that haemangiomas act as tissue expanders to normal skin around them . In addition, the phenomenon of creep that begins with tightening of the purse string also helps to mobilize the skin and close these defects . The minimal blood loss in our series is because of the use of a custom-made tourniquet whenever possible.
Five lesions were excluded intraoperatively in our work. Two of the excluded lesions were in the lower eyelid. During surgery, we found unacceptable lower lid ectropion. Therefore, purse string was cut and linear closure was performed parallel to skin tension lines. The other two excluded lesions were on the lip. We found that purse string caused severe distortion of the lip because the skin is naturally adherent to muscles of the lip. Again, the purse string suture was removed and linear closure was performed keeping in consideration skin tension lines. In our opinion, in both lip and eyelid cases, the scars were acceptable or at least amenable to scar revision later [Figure 4].
One of the excluded lesions was in the scalp. Purse string led to wound infection 4 weeks after the operation. The stitch was removed. Antibiotics were administered and wound dressing was performed. The resultant small area of alopecia was treated by elliptical excision and linear closure and the case was excluded from the series [Figure 7]. Infection may be due to the non absorbable stitch that acts as a foreign body and difficulty in maintaining good hygiene in the scalp hair of the child.
|Figure 7: Left, scalp haemangioma; middle, obvious alopecia after control of infection; right, after elliptical excision and closure.|
Click here to view
The number of scalp cases was too small in our study to obtain statistical data on purse-string stitching of the scalp. However, there are plenty of other good options for closure of a scalp defect, for example, up to a 3 cm defect of the scalp can be treated in one stage by simple closure and up to a 9 cm defect of the scalp can be treated with a local flap . Therefore, the surgeon is not obligated to perform the purse-string stitches on the scalp. Because of the excluded cases, we disagree with Mulliken et al.'s  statement that purse-string should be the absolute first choice for haemangioma excision.
| Conclusion|| |
Purse-string closure (Bradford Cannon closure) is a powerful tool in the closure of skin defects after haemangioma excision in the face. It should be considered a first option whenever it does not distort an important structure such as the mouth or the eyelids. For the scalp, the value of this technique should be studied further taking into consideration that there are other good options as alternatives for this technique, particularly for the scalp.
| Acknowledgements|| |
Conflicts of interest
| References|| |
Kim KH, Rohrer TE, Geronemus RJ. Vascular Lesions. In: Goldberg DJ ed. Laser and lights
. Philadelphia, PA: Elsevier; 2005:14.
Galiano RD, Gurtner GC. Vascular Anomalies. In: Kryger ZB, Sisco M, eds Practical plastic surgery
. Austin, TX: Landes Bioscience; 2007:139.
Chang MW. Updated classification of hemangiomas and other vascular anomalies. Lymphat Res Biol 2003; 1
North PE, Waner M, Mizeracki A, Mrak RE, Nicholas R, Kincannon J, et al
. A unique microvascular phenotype shared by juvenile hemangiomas and human placenta. Arch Dermatol 2001; 137
Mulliken JB. Vascular Anomalies. In: Thorne CH, Beasley RW, Aston SJ, Bartlett SP, Gurtner GC, Spear SL. Grabb and Smith′s plastic surgery
. Philadelphia, PA: Lippincott Williams and Wilkins; 2007:191-200.
Marler JJ, Mulliken JB. Current management of haemangiomas and vascular malformations. Clin Plas Surg 2005; 32
Tan ST, Itinteang T, Day DJ, O′Donnell C, Mathy JA, Leadbitter P. Treatment of infantile haemangioma with captopril. Br J Dermatol 2012; 167
Naouri M, Schill T, Maruani A, Bross F, Lorette G, Rossler J. Successful treatment of ulcerated haemangioma with propranolol. J Eur Acad Dermatol Venereol 2010; 24
Thorne CH. Techniques and Principles in Plastic Surgery. In: Thorne CH, Beasley RW, Aston SJ, Bartlett SP, Gurtner GC, Spear SL. eds Grabb and Smith′s plastic surgery
. Philadelphia, PA: Lippincott Williams and Wilkins; 2007: 6.
Mulliken JB, Rogers GF, Marler JJ. Circular excision of hemangioma and purse-string closure: the smallest possible scar., Plast Reconstr Surg 2002; 109
Yuen JC. Versatility of the subcuticular purse-string suture in wound closure. Plast Reconstr Surg 1996; 98
Tremolada C, Blandini D, Beretta M, Mascetti M. The ′round block′ purse-string suture in wound closure: a simple method to close skin defects with minimal scarring. Plast Reconst Surg 1997; 100
Leaute-Labreze C, Prey S, Ezzedine K. Infantile haemangioma: part II. Risks, complications and treatment. J Eur Acad Dermatol Venereol 2011; 25
Zweegers J, Jeffery I, van der Vleuten CJ, Carine JM. The psychological impact of infantile haemangioma on children and their parents. Archiv Dis Child 2012; 79
Yap LH, Langstein HN. ??. In: Thorne CH, Beasley RW, Aston SJ, Bartlett SP, Gurtner GC, Spear SL. eds Grabb and Smith′s plastic surgery
. Philadelphia, PA: Lippincott Williams and Wilkins; 2007:358.
[Figure 1], [Figure 2], [Figure 3], [Figure 4], [Figure 5], [Figure 6], [Figure 7]