Internal jugular vein rupture after oncologic resections in the head and neck

Internal jugular vein (IJV) blowout after major oncologic resections in the head and neck is a rare fatal yet preventable complication. The condition is unregistered sufficiently in the literature. The records of patients who underwent oncologic neck surgery were retrospectively reviewed. The study included records between January 2014 and November 2019 at Kasr Al Ainy Educational Hospital. 275 patients underwent cervical ablative procedures. Ten patients developed IJV blowout. Six patients were saved. Four patients had diabetes mellitus with postoperative wound infection and dehiscence. Three patients were given primary radiotherapy; two of them developed flap necrosis. Eight patients acquired pharyngocutaneous fistula (PCF). Regional flap coverage was done in three patients. Sentinel hemorrhage occurred in all patients. IJV blowout is a rare potentially life-threatening complication usually preceded by sentinel hemorrhage. The condition is essentially preventable by the prompt and structured response.


Background
Major vascular blowouts after oncologic resections in the head and neck are rare postoperative complications [1]. The incidence of internal jugular vein (IJV) rupture is escalating with the increased performance of functional and modified radical neck dissections which entail preservation of the IJV [2]. Deep neck space infections following these surgical modalities largely predispose to this life-threatening yet preventable complication [3]. Being an infrequent complication, there was paucity of the reports discussing IJV blowout after head and neck surgery in the medical literature. According to the best of our knowledge, the maximum number obtained by reviewing the medical literature was six patients in a single report [4].
In this article, we are describing our institutional experience with IJV blowout after major oncologic resections in the head and neck. By practicing at a tertiary care center, we were able to retrospectively review the data of 275 patients who had head and neck malignancies treated surgically from January 2014 to November 2019. During this period, we had 10 cases of IJV blowout; representing the highest number of cases describing the condition in the medical literature up till now.

Methods
The medical records of patients who underwent oncologic resections in the neck-at a tertiary care center-were retrospectively reviewed. The study was conducted between January 2014 and November 2019. The incidence of IJV blowout was determined. Associated possible risks and perioperative events were also examined. These included age, associated medical co-morbidities, primary radiotherapy, surgical procedure in detail, postoperative wound infection, postoperative wound dehiscence, regional flap coverage, sentinel bleeding, time of rupture, and the survival.

Results
The total number of patients who had oncologic neck resections was 275. Ten patients had IJV blowout. All patients were men. The median and mean ages were 60.5 years (range: 56 and 67). The median interval between the ablative procedure and the IJV blowout was 15 days (range: 12 and 30) while the mean was 18 days. The associated possible risks for IJV blowout and the perioperative events are shown in Table 1. All patients developed at least an attack of sentinel bleeding, but not all patients underwent urgent neck exploration after this attack. The first patient (Fig. 1) had T4a laryngeal carcinoma (LC) extending to the upper 3 tracheal rings. They were resected en-block with the laryngectomy together with another 2 rings as a safety margin. The second patient had T2 LC but he had restrictive lung disease and partial laryngectomy was not feasible. He refused primary radiotherapy as a treatment modality and preferred surgery. He developed wound infection followed by postoperative pharyngocutaneous fistula (PCF) on a postoperative day 10 (D10). He had IJV blowout on D13 which was managed by urgent neck exploration, ligation of the IJV, and wound debridement. After that, he had repeated wound dressings and antibiotics with a successful cure of infection. The residual wound dehiscence was well managed by pectoralis major myocutaneous flap (PMMCF) coverage. The third patient ( Fig. 2) had recurrent papillary thyroid carcinoma (rPTC) for which he had two operations-with no available details-followed by external beam radiotherapy and radioactive iodine (RAI) ablation outside our institution. After the failure of all the previously mentioned measures, he was referred to our hospital and he had total laryngectomy (TL), total thyroidectomy (TT), bilateral modified radical neck dissection (BMRND) with preservation of the IJV on both sides, and PMMCF augmentation of the wound. The fourth patient had recurrent laryngeal carcinoma (rLC) after primary chemoradiation (CRTH), and he had solitary single pulmonary metastasis (PM). The medical oncologist strongly recommended salvage surgery to be followed by postoperative chemotherapy to decrease the morbidity resulting from the metastatic nodule. The patient had TL, TT, and bilateral selective neck dissection (BSND). He developed a rupture of the IJV with successful ligation. That was followed by ICU admission at which he developed CCA rupture 24 h after the previous event. The 5th patient had T4a LC which was managed surgically. Despite the fact that he had PCF, he developed neither wound infection nor dehiscence. When he developed brisk IJV bleeding, he was urgently transferred to the theater and the vein was ligated. He was then dispatched to the ICU but he died early after. Patients 6, 7, and 8 had PCF after urgent neck exploration to control the bleeding (iatrogenic pharyngeal wall injury). The sixth patient had LC which was surgically treated 3 years ago. Twelve days after the initial surgery, he developed brisk bleeding from the stoma. Urgent wound exploration revealed IJV rupture and it was successfully ligated. The patient survived but later developed tumor recurrence. After 3 years, he had massive bleeding urging immediate exploration which revealed invasion of the common carotid artery (CCA) by the tumor recurrence. It was successfully ligated but the patient died early after. Indeed, understanding the risk factor-that might have been the cause for the IJV rupture in this patient-is very challenging. This patient did not have any associated medical comorbidities. He had neither postoperative wound infection nor dehiscence. Also, he was not given primary radiotherapy; however, he developed IJV blowout on D12. Probably, the absence of any risk factor was one of the leading causes that this patient was a survivor after the initial bleeding attack. The 7th patient had rLC after primary radiotherapy. He had supraclavicular (SCF) flap augmentation of the pharyngeal closure together with the ablative procedure. He started to have sentinel bleeding on postoperative day 11. He had another event on day 13. On day 16, he developed brisk bleeding from the stoma warranting urgent exploration during which the pharynx was injured resulting in postoperative PCF. The IJV was found to be the source of bleeding, and it was successfully ligated and controlled. He was then referred to the ICU and later to the ward with repeated wound dressing and fistula care. Finally, the fistula was closed and he was discharged home to become the last survivor in the study. The 8th patient had postoperative subclavian vein thrombosis which was managed by low molecular weight heparin (LMWH) therapeutic dose injection. After the first attack of sentinel bleeding, the patient underwent urgent neck exploration during which the pharynx was injured and repaired but resulted in PCF. Surprisingly, neck exploration was negative for a definite bleeder but revealed diffuse ooze from the operative bed probably due to (LMWH) toxicity. On the next day, the drain had 600 cc of fresh blood warranting another neck exploration which was negative for a definite bleeder as well. A few days later, he had severe bleeding from IJV blowout which was successfully ligated but the patient passed away shortly after the operation. The 9th patient (Fig. 3) had T3 LC for which total laryngectomy and bilateral lateral neck dissection was done. He developed wound infection followed by postoperative pharyngocutaneous fistula (PCF) on a postoperative day 10 (D10). He had IJV blowout on day 12 which was managed by urgent neck exploration, ligation of the IJV, and wound dehiscence. After that, he had repeated wound dressing and antibiotics with a successful cure of infection. The 10th patient (Fig. 4) had T4 LC for which total laryngectomy and bilateral lateral neck dissection was done. He developed wound infection followed by postoperative pharyngocutaneous fistula (PCF) on a postoperative day 9 (D 9). He had IJV blowout on day 12 which was managed by

Discussion
Our study showed that ten patients developed IJV blowout after having neck surgery for oncologic resection with some sort of neck dissection. Although selective surgical techniques for addressing the cervical lymph nodes during the procedure of neck dissection has led to more preservation of vital structures, those preserved structures have become liable to postoperative morbidity and complications. Primary radiotherapy and salivary fistulae remained as major contributing factors to catastrophic vascular insults. Breaching the outer adventitial layer of the vessel has always been a culprit in the pathogenesis of vessel rupture [5]. In the present study, five patients had PCF before the IJV blowout while another three acquired it while managing the bleeding. Three patients were given primary radiotherapy. One did not develop PCF, the other developed it before IJV rupture and the last one acquired the fistula as a traumatic complication during urgent neck exploration to control IJV bleeding. Owing to the paucity of this condition, statistical significance could not be determined for specific factors; however, it seems that there is a strong association between primary radiotherapy, the development of PCF, and the blowout of the IJV if the operative procedure entailed the preservation of the IJV with extensive dissection around its lower part. Radiotherapy has double effects (early and late) on the process of wound healing. Connective tissue cells have an intermediate response to the radiation. The irradiated tissues always exhibit hypoxia owing to small blood vessel obliteration, and there is also a reduction in the production of collagen by fibroblasts. The net result of the previously mentioned effects is inverse affection of both the proliferative and maturation phases of wound healing [6]. In our study, we had two patients who had PMMCF coverage of the wound and one patient with SCF coverage. A patient had PMMCF in the same session with the ablative procedure because he was previously given radiotherapy as an adjunctive therapy to recurrent papillary thyroid carcinoma. Unfortunately, this patient died of IJV blowout despite the fact that his neck vessels were covered with the flap. Another patient had PMMCF coverage of the wound in a separate procedure after the event of IJV ligation. Fortunately, this patient survived. One other survivor was given primary radiotherapy and had SCF in the same session with the  laryngectomy and neck dissection operation. So, out of four survivors, two patients had regional flap coverage. Sentinel bleeding is described as an accidental and short gush of the blood from the wound. A history of such an event warrants comprehensive and urgent evaluation. If no further bleeding is observed, the patient is admitted for follow-up and imaging of the neck vessels is mandatory. Computerized tomography (CT) is usually preferred. Air bubbles on imaging found in the vicinity of a major vascular structure is a risk sign of impending rupture. The required intervention is always adjusted according to the patient condition, and it may vary from instituting aggressive intravenous antibiotic therapy to elective reconstructive procedures [7]. In the present study, sentinel bleeding occurred in all the patients. Indeed, not all patients had neck exploration after the first attack which eventually leads to brisk bleeding later with the need for urgent IJV ligation.
Sulaiman and charters reported the shortest interval between neck surgery and IJV rupture which was 9 days [8]. The longest duration before IJV blowout was 23 days as given by Hudorović and Vucetic [9]. In their review, Genden and colleagues reported that the incidence of IJV hemorrhage occurred 14 days after the operation [3,10]. In the present study, our research team found that the median period between the operation and the incidence of IJV burst was 15 days ranging between 12 and 30 days postoperatively. We had 5 events in the 2nd postoperative week, two in the 3rd week, and the two on postoperative day 30.
In summary, the research team for the present study concludes that IJV blowout is a catastrophic postoperative complication which ended up with mortality in half of the patients. The use of primary radiotherapy and organ preservation protocols for managing locally advanced head and neck carcinoma has led to an increase in the number of cases. The dangerous gap for the blowout seems to be present between the second and the fourth weeks postoperatively. Prompt response with urgent ligation of the vein may save the patient life. Sentinel bleeding usually precedes all cases of IJV rupture. The condition could be prevented with the proper identification of the sentinel bleeding which warrantsas a minimum-hospital admission, close observation, aggressive parenteral antibiotics, thorough wound care, and anatomical imaging of the neck vasculature. Clinical suspicion may justify emergency neck exploration and dealing accordingly. Such an approach may lead to prevention or early recognition of the condition which will decrease both patient morbidity and mortality.