Introduction
Rupture of the common carotid artery (CCA) or internal carotid artery (ICA) is known as carotid blowout. In 1996, a classification of carotid blowout syndrome (CBS) was proposed, based on clinical severity, which included three types.Reference Chaloupka, Putman, Citardi, Ross and Sasaki1
Type 1 is a “threatened artery”, described as carotid artery exposure found on examination or imaging (i.e., air surrounding the vessel, adjacent abscess or tumour associated with a fistula), or by areas of arterial wall disruption found on vascular imaging studies.Reference Suárez, Fernández-Alvarez, Hamoir, Mendenhall, Strojan and Quer2 Type 2 is an “impending blow out” characterised by sentinel bleeds which may stop spontaneously or after treatment. Type 3 is the “acute” massive hemorrhage.
CBS can be caused by variety of aetiologies; it is most commonly described in the head and neck cancer literature where it is caused either directly by the head and neck cancer or is a complication of its treatment. Incidence of CBS in the newly diagnosed head and neck cancer patient population ranges from 2.9 to 3.9 per cent.Reference Lu, Chen, Chen, Chu, Tai and Wang3, Reference Chen, Wang, Wang, Jiang, Lin and Liu4 However, CBS risk after major oncological surgery post-radiotherapy is reported to be between 4.5 and 21.1 per cent.Reference Suárez, Fernández-Alvarez, Hamoir, Mendenhall, Strojan and Quer2 Arteries with tumour encasement either by the primary tumour or a cervical lymph node metastasis with greater than 180° are thought to be particularly at risk.Reference Yazici, Sanlı, Cengiz, Yuce, Gultekin and Hurmuz5, Reference Yamazaki, Ogita, Himei, Nakamura, Kotsuma and Yoshida6 This CBS risk increases after radiotherapy. Radiation obliterates the vasa vasorum, causing fibrosis of the adventitia and weakening of the arterial wall.Reference Cohen and Rad7
Even if the initial surgery and/or radiotherapy does not put the artery at risk, problems with wound healing can lead to CBS. Surgical site complications, such as infection, dehiscence and necrosis can involve the underlying vascular structures. In patients with oral or pharyngeal fistula, the adjacent carotid wall can be directly subjected to the digesting effect of salivary enzymes.Reference Powitzky, Vasan, Krempl and Medina8 Either the damage of the surrounding adventitia of the carotid artery can make it vulnerable to CBS, or the surgical site complication directly leads to carotid exposure.Reference Chen, Wang, Wang, Jiang, Lin and Liu4 In combination with the aforementioned influencing factors, the risk of disruption of surgical wound healing is significantly increased after radiotherapy and becomes even worse post re-irradiation with cumulative doses over 70 Gray (Gy).Reference Hoebers, Heemsbergen, Moor, Lopez, Klop and Tesselaar9, Reference Ward, Riaz, Caudell, Dunlap, Isrow and Zakem10
In terms of managing CBS, the literature tends to focus on CBS type 2 and 3,Reference Suárez, Fernández-Alvarez, Hamoir, Mendenhall, Strojan and Quer2, Reference Cohen and Rad7 with less discussion of the early identification of CBS type 1, though this remains a real post-operative risk in daily practice for both irradiated and non-irradiated patients.
In our institution, we aggressively manage CBS type 1 with a pectoralis major myofascial flap (PMMF) in order to bring well-vascularised tissue into the at-risk area in order to protect the carotid. Whilst vessel coverage by vascularised tissue is a well-accepted principle in head and neck oncology, particularly in the setting of salvage surgery,Reference Guimarães, Aires, Dedivitis, Kulcsar, Ramos and Cernea11 there is scarce literature data on the treatment of CBS type 1.
For this reason, we conducted a case note review on the outcome of timely recognition of CBS type 1 and management by a PMMF.
Methods
A retrospective case note review was performed of all patients suffering CBS in the Netherlands Cancer Institute/Antoni van Leeuwenhoek in the period between 2010 and 2020. The registry of surgical procedures was searched for the codes for hemorrhage and exploration of the neck, as well for pectoralis major muscle flap as sole procedure. All entries were thoroughly reviewed to select all patients with a threatened blowout (type 1) of the CCA or the ICA as defined by Chaloupka et al. Reference Chaloupka, Putman, Citardi, Ross and Sasaki1 A compromised external carotid artery (ECA) was excluded as were CBS type 2 and 3. Medical records were thoroughly examined for relevant patient, tumour and therapy characteristics (Table 1).
Overview patients with carotid blowout syndrome (CBS) type 1 of common carotid artery (CCA) or internal carotid artery (ICA) treated by pectoralis major muscle flap. RT = radiotherapy | CCRT = concurrent chemoradiotherapy | Gy = Gray | RND = radical neck dissection | MRND = modified radical neck dissection | SND = selective neck dissection | TL = total laryngectomy | COMMANDO = combined mandibulectomy and neck dissection operation |PM = pectoralis major myofascial flap | FRFF = free radial forearm flap | FFF = free fibula flap | TLP = total laryngopharyngectomy | SSC = surgical site complication | mf = myofascial | mc = myocutaneous | CDC = Clavien-Dindo Classification
1 Palliative procedure.
2 Oral defect reconstruction.
In cases of primary radiotherapy or primary chemoradiotherapy, the total dose of Gray delivered to the carotid artery was noted. For cases that went on to salvage surgery, the time interval to surgery was recorded. Details of surgery, especially of neck dissection were noted, together with residual tumour activity, if any. The post-operative course until CBS type 1 was recorded, including the delay between both identification and treatment of CBS type 1. Treatment was always with a PMMF. Details of whether this was used for mucosal closure, cutaneous closure, or both, as well as details of skin paddles and or skin grafting, together with possible surgical donor site complications, were noted.
Approval of the Institutional Review Board was acquired, locally registered as d19-098.
Statistical analysis
Descriptively frequential statistical analyses were performed using Statistical Package for the Social Science (SPSS) version 27.0 (SPSS Inc., Chicago, IL, USA).
Results
In total, 15 patients with threatened CBS (CBS type 1) were identified. Four females and 11 males, with age ranging from 51 to 75 years old with a median of 63 years old. Fourteen patients (14/15 = 93 per cent) had undergone primary radiotherapy or primary concurrent chemoradiotherapy in the neck region of the following CBS type 1 (range 46 - 136 Gray). Two of these patients (2/14) had received two courses of radiotherapy, with a total dose of 98 and 136 Gray, respectively.
Surgery that led to CBS type 1 was because of recurrent disease (N = 6), residual disease (N = 4), second primary disease (N = 3) and radionecrosis (N = 2) (see Table 1).
Both the initial ablative surgeries as well as the secondary PMMF reconstructions took place between 2010 and 2020. Initial surgery involved neck dissection (often combined with locally recurrent or residual tumour resection) in all but one patient. This one patient suffered repetitive free flap failure for oral reconstruction, leading to wound problems in the neck (case 12).
Seven patients were still actively using tobacco, while 13 consumed alcohol on regular basis.
CBS 1 was diagnosed by clinical examination in 11 patients and by computed tomography (CT) scan in four patients. Digital subtraction angiography is not used for the identification of CBS type 1 at our institute.
In 13 patients, a complicated post-operative course led to CBS type 1. Two patients suffered from progressive tumour ulceration (cases 6 and 7). In the first, the tumour was found to be invading the adventitia of the ICA during initial surgery, and in the second, a progressive ulcerating metastasis threatened the CCA.
Our institutional policy is to treat threatened CBS with a PMMF,and this was done in all 15 patients. A split skin graft was used to cover the muscle in nine patients, one patient received a full thickness graft, while primary closure of the cutaneous defect was possible in five patients.
One patient (1/15 = 7 per cent) suffered a superficial surgical donor site infection (Clavien-Dindo grade 2) after PMMF surgery, which was treated with antibiotics.
CBS treatment was performed with curative intent in 12 of 15 patients. In the three palliative cases, cases 6 and 7 suffered from tumour ulceration leading to CBS type 1, whilst in case 13, a neck dissection was complicated by abscess formation around the carotid.
Time in weeks from initial radiotherapy or chemoradiotherapy until surgery ranged from 14 to 671 weeks (median 35 weeks). Diagnosis of surgical site complications leading to CBS type 1 ranged from 5 days post-operatively to 35 days (median 11.5 days). Time from diagnosis of CBS type 1 until PMMF surgery ranged from the same day till 4 days (median 1 day). One of the palliative patients (case 6) needed a 4-day period to decide whether or not she wanted surgery.
Of all patients, both date and cause of death was known. None of the 15 patients progressed to CBS type 2 or 3, and no one deceased of CBS. Median survival after PMMF for the 12 patients treated with curative intent was 160 weeks (range 2–693 weeks). For the three palliative patients, survival range was between 18 and 43 weeks. It should be noted that one patient treated with curative intent suffered such severe wound healing problems that subsequently interventional radiological (IR) embolisation of the ICA was advised to prevent long-term CBS risk. This particular patient could not bear the chance of a cerebrovascular accident – as one of the side effects of embolisation – and chose euthanasia, performed 2 weeks after the PMMF.
Discussion
This study describes the efficacy of swift intervention with a PMMF transfer (median 1 day, range 0–4 days) in preventing carotid bleeding in patients with CBS type 1 of the CCA or ICA, after head and neck cancer treatment. In their valuable and comprehensive review, Suarez et al. published a practical management algorithm for CBS types 2 and 3.Reference Suárez, Fernández-Alvarez, Hamoir, Mendenhall, Strojan and Quer2 However, they did not address the optimal intervention for type CBS 1 of the CCA/ICA. In our opinion, the PMMF is superior in this setting to IR procedures of the CCA/ICA, which have potential neurological complications. The PMMF has the additional benefit of bringing a large quantity of well-vascularised tissue into the at risk area, helping solve the causal wound healing problem. With our case note review, we aimed to demonstrate that it is advisable to seriously consider the first step of PMMF in the clinical decision-making process and advocate for a modified management algorithm as shown in Figure 1. Even for CBS type 2 cases where stenting has been performed, we recommended to additionally perform a PMMF transfer to improve wound healing because the presence of a wound infection at the stent site poses an increased risk for recurrent CBS.Reference Pyun, Lee, Yoo, Lee, Choi and Kim12
Modified management algorithm for Carotid Blowout Syndrome (CBS) of the common carotid artery (CCA) or internal carotid artery (ICA). (2) Cancer Management and Research 2018:10 5617-5628 (Adapted) - originally published by and used with permission from Dove Medical Press Ltd.
All but one of our patients (14/15) had undergone prior radiotherapy or chemoradiotherapy, with a total dose ranging from 46 to 136 Gray, before their salvage surgery to the neck that led to CBS type 1.
This is in line with the data from literature indicating that prior radiotherapy is a main predisposing factor for CBS.Reference Suárez, Fernández-Alvarez, Hamoir, Mendenhall, Strojan and Quer2, Reference Zhu, Huang, Lu, Yang, Huang and Zhou13 For patients irradiated (or re-irradiated) to a dose of 70 Gy or higher, this risk is nearly 14-fold increasedReference Chen, Wang, Wang, Jiang, Lin and Liu4 due to the damaging influence of radiotherapy to feeding vessels in the adventitia layer. Patients who underwent salvage surgery of the primary tumour with en bloc neck dissections after radiotherapy or chemoradiotherapy have a significantly higher risks to develop post-surgical fistulas. Bacterial infection combined with the enzymatic effects of saliva to the mucosal wound, exacerbates the negative effect of radiotherapy on wound healing, leading to significantly increased risk of erosion of the vascular wall of the major vessels in the neck. In the literature, there is ample evidence that the prophylactic use of (free) muscle flaps leads to improved wound healing with less risk of fistulas.Reference Guimarães, Aires, Dedivitis, Kulcsar, Ramos and Cernea11 In our series, a PMMF is also well suited for its protective role with prevention of progression to a higher type of carotid blow out. The morbidity of harvesting a PMMF is limited, while anastomosing a free vascularised flap pose a serious challenge in this subacute situation.Reference Schneider, Wu and Wax14 In exceptional cases where the pectoralis major muscle has been involved in previous surgery, a latissimus dorsi myofascial flap can be used as an alternative.
The fascia has a complex vasculature, which can easily adhere to underlying tissues.Reference Stecco, Tiengo, Stecco, Porzionato, Macchi and Stern15 It is hypothesised that hyaluronan-rich deep fascia plays an important role in early phases of wound healing by establishing wound extracellular matrix.Reference Turley, Noble and Bourguignon16 Moreover, contraction of the fascia helps wound closure and improves alignment with the contours of the neck.Reference Ye and Rinkevich17 The PMMF can subsequently be covered with a split skin graft, either directly or after 24 hours without the need of pressure dressing. The use of split skin provides also better colour match than a myocutaneous flap.Reference Bos, Balm, Schouwenburg, Hilgers and de Boer18
Although in cases of CBS types 2 and 3 actual bleeding can effectively be stopped by endovascular procedures, the risk of nonviability and further breakdown of both the arterial wall and surrounding tissues in the neck remains a problem. So, in our view also in these cases protection of the CCA or ICA by well vascularised tissue remains of vital importance.
Given its simplicity and effectiveness, it could be argued that PMMF should be used prophylacticly during initial surgery, but one should bear in mind that overtreatment is lurking in those cases with minimal risk of developing CBS.Reference McDonald, Moore and Johnstone19 The final decision to use a prophylactic flap insertion in isolated salvage (modified) radical neck dissections after radiotherapy is at the discretion of the individual surgeon, with prior chemoradiation treatment carrying the greatest weight. In other cases, intra-operative observations such as the condition of the tissues and their perfusion and preservation of the sternocleidomastoid muscle in modified radical neck dissections may play an important role in the decision-making process.Reference Henneman, Rouwenhorst, Karakullukcu, Smeele, Lohuis and van den Brekel MWm20
The definition of CBS type 1 remains limited to “threatened” with “exposure” of the carotid artery in the neck.Reference Chaloupka, Putman, Citardi, Ross and Sasaki1 In our opinion, however, the threat begins already in case of a limited wound dehiscence in an operated and irradiated patient without viable tissue. Not just when the carotid artery is recognised with palpable or visible pulsations in the neck. We would argue that these patients, with an early wound defect directly over the course of the carotid artery, should be classified as a high-risk patient and that a PMMF transfer is justified. Awaiting further wound healing with intensive though conservative wound treatment carries the risk of a progression to CBS type 2 and should be avoided.
Also in our two palliative patients with CBS type 1 due to regional ulcerating tumour progression, PMMF proved to be an adequate palliative measure that successfully prevented progression to CBS type 2 or 3. It is questionable whether one should recommend to extend preventive flap transfers to patients with incomplete dissection of recurrent disease from the CCA or ICA. In one of the two recurrent patients, there was a delay of more than two years before tumour progression led to CBS type 1. In the second patient, a second course of palliative radiotherapy 5 weeks prior led to CBS type 1. So, in case of additional re-irradiation one could consider pre-emptive transfer of a PMMF.
It should be emphasised that our study only includes CBS of the CCA and ICA, while in several studies no difference is made between the two categories: CCA/ICA and ECA.Reference Lu, Chen, Chen, Chu, Tai and Wang3, Reference Chen, Wang, Wang, Jiang, Lin and Liu4, Reference Citardi, Chaloupka, Son, Ariyan and Sasaki21–Reference Liang, Guedes, Duvvuri, Singh, Chaer and Makaroun24 IR procedures of the ECA have significantly less risks for ischemic cerebrovascular events as compared to stenting of embolisation of the CCA or ICA. In this respect, clinical distinction between CBS 1 CCA/ICA and ECA is of great importance to minimise the risk of potential neurological complications.
As our research indicates, in none of the cases did we employ digital subtraction angiography to inform us about which carotid branch is involved and about the status of the carotid artery wall. In contrast to statements by Kozin et al. for a CBS type 1,Reference Kozin, Kapo, Straton and Rosielle25 it is not necessary in our opinion to perform a routine digital subtraction angiography, when it is possible to responsibly provide external coverage of the CCA or ICA after careful dissection of the defect surrounding tissue. The presence of healthy muscle tissue provides even sufficient assurance for the recovery of the vascular wall in the event of potential loss of vital tissue.
• Carotid blowout syndrome (CBS) is a rare but devastating complication of head and neck cancer, especially of salvage surgery to the neck, after definitive radiotherapy or chemoradiotherapy
• “Impending” bleeds or “acute massive” bleeds (respectively, CBS type 2 and 3) are preferably treated by intervention radiology with stenting or embolisation of the carotid artery
• This case note review demonstrates that early detection of the “threatened” common or internal carotid artery (CBS type 1) with subsequent timely transfer of a pedicled myofascial flap (PMFF) is a safe approach to prevent development to CBS type 2 or 3.
Limitations
This case note review has its limitations and does not rise above level IV evidence.
First, patients in this series are collected out of our surgical registry. Theoretically we could have missed cases of CBS type 1 that recovered by non-surgical measurements, although our institutional protocol dictates surgery.
Secondly, also without intervention possibly not all CBS type 1 cases would have progressed to a type 2 or 3. However, for this potentially life-threatening condition, it is medically unethical and therefore impossible to conduct a prospective trial because prompt medical intervention is indicated.
Only future publications of prospective studies on a protocol-based treatment, as described in our study for a well-defined indication of CBS type 1 of the ICA/CCA, can help raise the level of scientific evidence.
Conclusion
Our case note review seems to demonstrate that timely transfer of a PMMF in case of CBS type 1 CCA/ ICA, post salvage neck dissection, seems to be a safe approach and might prevent further development to CBS type 2 or 3. In our view, CBS type 1 should be included in the CBS management algorithm. And it should then be regarded as an emergency case that needs to be treated within an interval of 24 hours, without the need of pre-operative digital subtraction angiography.
Acknowledgements
We thank professor MWM van den Brekel, professor LE Smeele and dr TF Pézier for their valuable comments.
Competing interests
The authors report no conflicts of interest.
Funding
None.
