Penetrating Injuries

TRAUMA TO THE GLAND Salivary Fistula

Penetrating injury to the substance of a major gland—for example, the parotid or submandibular gland—will cause direct damage to the gland and possible related structures and may lead to the formation of an external salivary fistula to the skin (Figure 12.1). When the substance of the gland is injured, suture of the parenchyma is recommended (Lewkowicz, Hasson, and Nablieli 2002). In addition to direct closure of the parotid capsule, a pressure dressing for 48 hours is applied to reduce the chances of sialocele formation. In 51 cases of parotid complications following trauma, 15 (29.4%) developed parotid fistula, treated by intravenous fluids and nil by mouth, with faster healing of parenchymal injuries alone than when the ductal system was involved (Parkeh, Glezerson, and Stewart et al. 1989). Similarly, Ananthakrishnan and Parkash (1982) reported that their 3 cases of fistula from the parotid gland parenchyma resolved without treatment, unlike the 14 fistulae related to parotid ductal injury. In a study of 13 patients with traumatic parotid fistulae, 54% resolved with conservative management within 3 weeks and the remainder were cured by internal drainage with a catheter (Cant and Campbell 1991). Landau and Stewart (1985) advocated conservative management of post-traumatic parotid fistulae and sialo-celes and found that parenchymal injuries alone resolved in 5 days, whereas ductal injuries took 14 days. Morestin (1917), in a series of 62 war injuries with parotid fistula, 30 glandular and 32 ductal, reported good success with the creation of an intraoral fistula. In more extensive avulsive injuries with gross scarring, conservative treatment may be less successful (Figure 12.2), and established epi-thelialized fistulae require excision with repair of the parotid capsule and closure.

The submandibular gland is less liable to be involved in the development of traumatic fistulae, perhaps because it is protected by the mandible and its smaller size. Few cases of submandibular gland fistulae have been reported. A case report of submandibular gland fistula secondary to a gunshot wound was reported in 1995, where the authors also reviewed the literature and found only one

Figure 12.1a. Penetrating injury of the parotid after stabbing with a pencil.

Figure 12.1b. Exploration via a modified Blair incision to check for damage to the external carotid artery and facial nerve and to suture the parotid capsule.

other case from 1976 (Singh and Shaha 1995). In their 1995 case the fistula resolved without active treatment in 10 days.

Rarely internal parotid fistulae can occur presenting as rhinorrhea or rhinorrhea related to food,

Figure 12.1c. The length of the recovered pencil illustrates the depth of the wound. Patient treated by Mr. B.S. Avery, Consultant OMS, Middlesborough, England.

usually as a result of maxillary fracture with parotid fistula into the maxillary antrum (Faussat, Ghiassi, and Princ 1993; Scher and Poe 1988). In a recent report of parotid fistula into the maxillary antrum (and a very rare case of a sublingual gland fistula to the skin), excellent results were achieved with botulinum toxin injection (Breuer, Ferrazzini, and Grossenberger 2006). Although the authors state that primary surgical repair should be carefully considered, they found the injection of botulinum toxin to be effective, to shorten fistula closure time, and to be minimally invasive. The current management of fistulae from the parotid gland parenchyma is therefore conservative, as cases that do not involve the duct will resolve. In recalcitrant cases botulinum toxin appears a good option. True fistula post-surgery—for example, superficial parotidectomy—is not common but may occur through the surgical skin incision. Usually management with antisialogogues, nil by mouth, or botulinum toxin will lead to resolution. In a report of 3 cases post-parotidectomy treated with injection of botulinum toxin under electromyographic control, all resolved with no recurrence 14-21 months after therapy (Marchese-Ragona et al. 2006).


A sialocele is formed by the extravasation of saliva into glandular or periglandular tissues due to disruption of the parenchymal or ductal structures of the salivary gland. This is most commonly seen following trauma, and the usual sites are the sublingual gland (ranula) or minor salivary glands (mucocele). Ranulae and mucoceles were discussed in chapter 4, and this section will concentrate largely on parotid sialoceles. Parotid sialo-celes are usually seen after penetrating trauma to the parotid region and will present as painless, cystic swellings that are gradually increasing in size. Aspiration of the sialocele with fluid positive for amylase >10,000 units/liter will confirm the diagnosis. Computerized tomograms will show a cystic mass with smooth margins and a density lower than the surrounding tissues. After 2 weeks there will be enhancing borders due to the development of a capsule (Cholankeril and Scioscia 1993) (Figure 12.3).

Traditional management has been conservative, the same as for parotid parenchymal fistulae (Cant and Campbell 1991; Landau and Stewart 1985; Parkeh, Glezerson, and Stewart et al. 1989), with resolution reported in approximately the same time period as for fistulae. Most sialoceles develop 8-14 days post-injury, and the development of a late capsulated sialocele is more difficult to treat. Literature reviews show that treatments proposed include multiple aspirations, pressure dressings, secondary duct repair if this is the etiology, creation of an intraoral fistula, sectioning the auriculotemporal nerve, the use of antisialogogues (atropine, probanthine, glyco-pyrolate), duct ligation, and even radiation or parotidectomy (Canosa and Cohen 1999; Lewkow-icz, Hasson, and Nablieli 2002) (Figures 12.4 and 12.5).

In recent years the use of botulinum toxin has caused a paradigm shift in the way these injuries can be managed. In 1999 Ragona, Blotta, and Pastore et al. reported a case of post-traumatic parotid sialocele resistant to conservative therapy that was successfully cured using botulinum injection. These authors used botulinum F due to its earlier and shorter efficacy compared to botulinum A, and injected the gland with electromyographic control. Botulinum toxin works by causing a chemical denervation of the gland by blocking the cho-linergic neurotransmitter. Following this paper a

Figures 12.3a and 12.3b. Frontal and three-quarter views of patient with right parotid sialocele post-surgery.

report of 4 cases of recurrent post-parotidectomy sialoceles treated with botulinum A toxin injected subcutaneously with 100% success was published (Vargas, Galati, and Parnes 2000), as well as other

Figures 12.3c, 12.3d, and 12.3e. MR views of sialocele with enhancing capsule on Figure 12.3d.
Figure 12.4a. A 41-year-old male post-gunshot wound that entered in the left parotid region and traversed to the right parotid region with fracture of both left and right condyles. He developed an increasing sialocele in the right parotid gland.

Figure 12.4b. Panoramic film shows the retained bullet in the right parotid gland (arrow).

Figure 12.4c. Modified Blair incision and partial parotidectomy, with mosquito forceps indicating the bullet. The bullet was situated between the superior and inferior branches of the facial nerve, which was intact with no weakness. The capsule of the parotid gland was repaired and the sialocele resolved. Reprinted with permission from Blanchaert R, Ord RA. 1997. Management of late complications of penetrating injuries to the parotid gland. Pan American Journal of Trauma 6(1):52-57.

Figure 12.5a. Plain film of a 20-year-old man following a gunshot wound shows the bullet "floating" in parotid sialocele.

case reports (Chow and Kwok 2003). There is a single case of a submandibular sialocele treated with resolution using botulinum toxin A (Capac-cio, Cuccarini, and Benicchio et al. 2006).

Nerve Injury

The facial nerve is at risk from penetrating injury to the facial region both in the parotid and in the distribution of its peripheral branches to the facial musculature. It is stated that damage to branches distal to a line drawn from the lateral canthus to the commissure does not require repair and may be managed expectantly. All patients with facial wounds should have a careful clinical examination of facial nerve function. Where this is not possible—for example, in the unconscious patient or the uncooperative infant (Figure 12.6)—the wound should be carefully explored at the time of surgery to exclude transaction of the branches of

Figure 12.5a. Plain film of a 20-year-old man following a gunshot wound shows the bullet "floating" in parotid sialocele.

Figure 12.5b. Surgical exploration to remove the bullet revealed an abscess cavity where the bullet had lodged. Note draining pus. Reprinted with permission from Blanchaert R, Ord RA. 1997. Management of late complications of penetrating injuries to the parotid gland. Pan American Journal of Trauma 6(1):52-57.

Figure 12.6. Infant with laceration from broken glass. No facial nerve damage was present.

the facial nerve. Primary repair soon after the injury with end to end anastomosis is the ideal scenario, as paralysis of the facial nerve is a devastating injury for the patient, and even when "successful" nerve repair has been carried out with satisfying results (based on House-Brackmann, Stennert, and May grading), patients experience a reduced quality of life (Guntinas-Lichius, Straesser, and Streppel 2007) (Figures 12.7 and 12.8). This section will discuss the management of the primary nerve injury and will not discuss the techniques for facial reanimation or static slings, which are beyond the scope of this text. The interested reader will find many recent review articles addressing these topics (Guntinas-Lichius, Streppel, and Stennert 2006; Malik, Kelly, and Ahmed et al. 2005).

Classically the nerve is sutured under the microscope using 9-0 or 10-0 nylon sutures attempting to coapt the nerve ends without tension (Figure

Figure 12.7a. Entrance wound for bullet below ear lobe.

12.9). The suturing can be epineural or fasicular. In epineural suture less damage is caused to the neural bundles with less foreign body reaction in the fasicles due to the suture materials; however, fasicular suturing should allow better adaptation of the fasicles and trimming back the epineurium to prevent fibrous tissue in-growth. However, anatomic studies have shown the fasicular and connective tissue anatomy of the facial nerve to be complex, with the number of fasicles increasing in a proximo-distal way from the geniculate ganglion

Figure 12.7c. Frontal and buccal facial nerve paralysis.

Figures 12.8a and 12.8b. A 72-year-old man diagnosed with a stroke in the emergency room after falling through a plate glass window. The diagnosis was made due to the dense facial nerve palsy.

Figure 12.8c. After removing the dressings the deep penetrating wound in the region of the facial nerve trunk is appreciated.

Figures 12.8d and 12.8e. One year post-repair of nerve trunk and microneural suture the patient still has some weakness of the upper lid and cannot "blow out" his cheek on the left side, but has good facial symmetry and function.

with diminishing diameter (Figure 12.10). This variability in number of fasicles and structure along the extratemporal facial nerve constitutes a difficulty in facial nerve repair (Captier, Canovas, Bonnel, Seignarbieux 2005). The use of tubes (e.g., collagen) to support the anastomosis and prevent connective tissue in-growth and also tissue glues (e.g., fibrin) to replace sutures and their foreign body reactions have been advocated. However, the tubes can themselves cause foreign body reactions and possible compression. Regarding the tissue glues, animal experiments appear to show sutures to be superior. In the rabbit model axonal growth was faster and greater with epineu-ral suture than fibrin adhesive (Junior, Valmaseda-Castellon, and Gay-Escoda 2004). Although the rate and amount of reduction in conduction velocity was equivalent between the two methods, the authors concluded that epineural suture appears to be the method of choice. In another study in the rat model looking at suture and the effects of platelet rich plasma (PRP) and fibrin sealant, the best return of function for the facial nerve was again with suture (Farrag, Lehar, and Verhaegen et al. 2007). The authors did note a favorable neuro-trophic effect for the PRP but no benefit for the fibrin sealant. As it is vital that the anastomosis be tension free, difficulty is encountered when a gap between the nerve ends exists. In a cadaver study, Gardetto, Kovacs, and Piegger et al. (2002) showed that removal of the superficial part of the parotid gland could allow overlap of the cut branches of the facial nerve. They found it possible to bridge gaps of 15 mm in the temporo-zygomatic branches, 23 mm in the buccal-mandibular branches, and 17 mm in the nerve trunk. Following this experimental work the authors have reported successful clinical results on 3 patients, recommending the

Figure 12.9b. Exploration of wound reveals transection of the superior branch of the facial nerve. (Ruler from surgical marker makes a good background for microsuture if custom microsurgical background material is unavailable.)

Figure 12.9c. Post-microsurgical repair.

Figure 12.9d. Post-suturing of wounds. This case treated by Dr. J. Caccamese, Dept. OMS, University of Maryland.

technique for gaps up to 15 mm but cautioning against its use in the presence of infection or nerve defects (Piza-Katzer et al. 2004). In another approach the use of a rapid nerve expander (2 cm/30 minutes) was used to bridge gaps up to 3 cm (Ya, Gao, and Wang 2007). In 9 patients 5 achieved good results with EMG peak value of mimetic muscles 82-95% of the normal side, 3 cases were fair with EMG 60-90%, and 1 case poor with EMG 55%. Other surgical options in this situation are the use of a tube as a conduit and grafting. In 7 patients with post-traumatic defects up to 3 cm the use of a bioabsorbable polyglycolic acid tube was reported as giving very good results in 1 case, good in 4, and fair in 2 (Navissano et al. 2005). The two commonest sites for donor nerves for the facial nerve are the greater auricular, which is adjacent to the surgical field, and the sural nerve. Surgical principles for repair follow those for direct anastomosis. As expected, facial nerve function from cable nerve graft interposition is not as good as end to end anastomosis (Malik, Kelly, and Ahmed et al. 2005). If there is widespread destruction of multiple branches from an injury such as a gunshot the entire superficial cervical plexus can be used to supply multiple grafts.

Frey's Syndrome

Although Frey's syndrome is now most commonly seen in relation to parotid surgery, it was originally described after a shotgun injury to the parotid gland (Frey 1923); however, despite Frey's landmark paper, gustatory sweating was probably first described by Baillarger in 1853 (Dulguerov, Marchal, and Gusin 1999). In the trauma situation obviously the problem is management and treatment of the condition, whereas in the post-parotidectomy cases much work has been done on prevention.

Previously reported treatments of Frey's syndrome have included topical and systemic anticho-linergics, tympanic neurectomy, sectioning of the auriculotemporal or glossopharyngeal nerves, or interposition of fascia lata between the parotid bed and overlying skin. Currently the use of botulinum toxin is the most frequently reported therapeutic modality for Frey's syndrome. A report of 33 patients with Frey's syndrome treated by intra-cutaneous injection of 16-80 IU of botulinum toxin A showed all symptoms to resolve within a week (Eckardt and Kuettner 2003). In a prospective

Figure 12.10a. Extensive penetrating wound anterior to the parotid gland, which involves the peripheral branches of the facial nerve.

Figure 12.10b. Nerve branches identified for microneural repair.

Figure 12.10c. High-power view of completed repair.

Figures 12.10d, 12.10e, 12.10f, and 12.10g. Post-repair. Note upper branch weakness persists. This case treated by Dr. J. Caccamese, Dept. OMS, University of Maryland.

nonrandomized, nonblinded study of 11 patients treated with botulinum toxin A with follow-up 623 months, only 1 patient recurred and was successfully retreated (Kyrmizakis, Pangalos, and Papadakis et al. 2004). A prospective randomized trial to establish the ideal dosage and length of effect of botulinum toxin A was carried out on 20 patients divided into two groups receiving either

2 MU/cm2 or 3 MU/cm2. In the 3 MU/cm2 group a single injection resulted in nearly complete absence of gustatory sweating during the 12 month follow-up period. In the 2 MU/cm2 group 44% of the total skin areas were still sweating and required a second injection, and the authors concluded that

3 MU/cm2 is the recommended dose (Nolte, Goll-mitzer, and Loeffelbein et al. 2004). Some authors have cautioned that the effect of botulinum toxin

Figure 12.11a. Lateral facial view of patient complaining of gustatory sweating (Frey's syndrome).

in Frey's syndrome is often temporary and further injections may be necessary depending upon the initial dose and the length of time followed (Ferraro et al. 2005).

In regards to parotidectomy patients, probably 100% will have gustatory sweating if tested with starch and iodine (Laage-Hellman 1957) (Figure 12.11). However, few patients clinically notice this problem and most do not wish for treatment, so that this condition will be underestimated in clinical reports. Frey's syndrome is rarely reported after submandibular gland removal. Berini-Aytes and Gay-Escoda (1992) reviewed 206 submandibular gland excisions and found only 1 case of Frey's syndrome, while Teague, Akhtar, and Phillips (1998) reviewed the literature and could find 7 reported cases since 1934.

Figure 12.11b. Bottle of iodine solution that is painted on the patient's face and then covered in corn starch.
Figure 12.11c. While the patient eats an apple the cornstarch is discolored blue-black, indicative of gustatory sweating.

The techniques described for prevention of Frey's syndrome depend upon placing a barrier between the parotid bed and skin to prevent the growth of the secretory parasympathetic nerves from the parotid into the sweat glands, causing a paradoxical innervation. Acellular dermis has been used with success but with a higher complication rate (Govindaraj, Cohen, and Genden et al. 2001). Temporoparietal flaps and the superficial musculo-aponeurotic system (SMAS) have been used with good outcomes in 146 parotidectomy patients (Cesteleyn et al. 2002). In reviewing 160 patients followed from 5 to 22 years treated with an inter-positional SMAS layer at the time of parotidectomy and tested with starch/iodine during follow-up, no cases of Frey's were encountered (Bonanno et al. 2000). Other reported barriers have been the use of parotid gland fascia (Zumeng, Zhi, Gang et al. 2006) and sternocleidomastoid muscle flaps, with mixed results (Filho et al. 2004; Kerawala, McAlo-ney, and Stassen 2002).

It would appear that Frey's syndrome is preventable in most cases, and perhaps the use of the SMAS layer is the most convenient for the surgeon.


This is a complication seen in patients following parotidectomy rather than trauma and can be managed in a variety of ways. Reconstruction at the time of surgery to prevent the defect occurring is preferable to secondary reconstruction, when scarring and the superficial position of the facial nerve post-parotidectomy increase the risk of nerve damage. Various techniques have been used, some of which have been discussed in this chapter in relation to Frey's syndrome and in chapter 8. Techniques include the use of layered acellular dermis, free fat grafts (Figure 8.4), use of the SMAS layer, temporalis, and sternocleidomastoid flaps, as well as microvascular free flaps including fascial forearm flaps for larger defects. Choice of technique will depend on the size of the defect, the surgeon's own experience, and the wishes of the patient.


As has already been discussed above in the sections on fistulae and sialoceles related to parenchy-mal trauma, conservative management is usually satisfactory except in those cases where the injury involves partial or complete transection of the duct. Under these circumstances most papers have indicated that resolution is less certain and takes longer, with active management frequently required. There are studies that support conservative measures in duct injuries and in one report, of 19 patients with duct injury confirmed by methylene blue dye injection who were treated non-operatively, 9 patients (47%) healed without complications. Although 7 patients (36.8%) developed salivary fistulae and 4 (21%) sialoceles, these were described as short-term and resolved without the need for surgery (Lewis and Knottenbelt 1991). However, in most cases current management is

Figure 12.12. Surface markings of the parotid duct are shown by a line drawn from the tragus of the ear to bisect a line drawn from the alar base to the commissure. The middle third of this line (arrow) is surface marking of the parotid duct.

directed toward primary repair and the clinician must therefore have a high level of suspicion for injuries involving the region of the parotid duct. The anatomic surface markings of the duct are illustrated in Figure 12.12. Confirmatory evidence for transaction is obtained by cannulating the distal portion of the duct through Stenson's papilla and observing the catheter in the wound (Figure 12.13), or by injecting saline or a small (1 cc) amount of methylene blue through Stenson's papilla. Identification of the proximal end may be difficult, as it can retract into the gland substance. Milking the gland to obtain salivary flow is helpful in these circumstances, and the anesthesiologist must be cautioned preoperatively against the use of anti-parasympathetic agents. If the proximal and distal ends of the duct are identified and can be coapted, then microsurgical repair can be carried out (Hallock 1992) (Figure 12.14).

The use of stents (usually indwelling catheters) for 10-14 days to prevent stenosis is advocated by some and appears a reasonable hypothesis, although no long-term studies of these injuries with and without stenting have been published. A technique of using a 4 F Foley embolectomy catheter for identification of the transection and then leaving it in place as a stent has been described (Etoz, Tuncel, and Ozcan 2006). When the proximal and distal ends of the duct cannot be coapted due to tissue loss, repair using a vein graft has been reported (Heymans et al. 1999). Steinberg and Herrera (2005) recommended the use of sia-lography postoperatively to assess the result of duct repair, stating that this technique may not always be practical or possible in the acute setting. However, we have used sialography intraopera-tively (Figure 12.13).

When the injury is too proximal, the wound is avulsive, or the duct cannot be identified, then the clinician can either create an intraoral fistula or ligate the proximal duct. A controlled fistula can be created by suturing the proximal duct through the buccinator into the oral cavity if enough length is present, or by placing a catheter or drain from the area of the wound into the mouth and leaving it to fistulize (Figure 12.15). Although tying off the proximal duct to cause eventual atrophy has been proposed (Van Sickels 1981), in our experience this is unpredictable and even with the use of pressure and antisialogogues these patients can have considerable swelling and pain. Chemical denerva-tion using botulinum toxin A may help to achieve a good outcome in these circumstances (Arnaud et al. 2006).

In the case of the submandibular duct, tran-section is usually iatrogenic as a result of surgery on the sublingual gland, sialolithotomy from Wharton's duct, or resection of floor of the mouth cancer. In this case sialodochoplasty with repositioning of the duct posteriorly is all that is required. One of the authors formerly used a catheter as a stent (Ord and Lee 1996) following reposition of Wharton's duct in floor of the mouth cancer; however, now the duct lumen is identified and one blade of a sharp iris scissors is inserted and a vertical cut through one wall of the duct carried out. The duct is now "fish-tailed" and sutured to a newly created hole in the oral mucosa with 6-0 nylon sutures. Stenosis and stricture have not been a problem with this technique.

Figure 12.13a. Patient with cheek laceration that was primarily sutured and that now has developed sialocele due to missed duct injury.

Figure 12.13b. Wound reopened for re-exploration; the duct is discovered to be transected. Vessel loop around distal end of duct. Lacrimal probe passed from intraoral through Stenson's duct into the wound (short arrow).

Figure 12.13c. Stenson's duct is approximated after finding the proximal end of the duct by milking the gland. The duct is cannulated and contrast dye injected for intraoperative sialogram.

Figure 12.13d. Fluoroscopic image of intraoperative sialogram with repaired duct (arrow). Case treated by Dr. D. Coletti, Dept. OMS, University of Maryland.
Figure 12.14c. Final skin repair. This case treated by Dr. J. Caccamese, Dept. OMS, University of Maryland.
Figure 12.15a. Gunshot wound entering at right parotid region with exit at left infraorbital region as indicated by suction tubing.

Figure 12.15b. The parotid duct is identified and dissected from the wound to be cannulated as shown.

Stenosis of the Duct

When ductal injuries are not surgically repaired immediately, complications such as fistulae and sialoceles may arise and their management has been discussed. If the duct has not been surgically repaired by 72 hours, conservative or medical therapy is recommended (Arnaud et al. 2006). In the long term, stricture of the duct may occur (Figure 12.16), although most strictures are sec-

Figure 12.15c. The duct is diverted intraorally via the cannula. Note powder burn at entrance wound.

Figure 12.15d. Final repair. This case treated by Dr. J. Caccamese, Dept. OMS, University of Maryland.

ondary to inflammatory or infective conditions. In cases of intraductal salivary gland obstruction, 22.6% of 642 cases were due to strictures, which were more common in females (Ngu, Brown, and Whaites et al. 2007). When this occurs at the distal end of Stenson's duct, excision and diversion of the duct into the oral cavity may be feasible. When the main duct is involved with strictures, sialoen-doscopy may be useful to dilate the strictures using i &

Figure 12.16a. Patient with soft tissue scarring from penetrating wounds caused by a road traffic accident 10 years previously; now has a 7-year history of parotid and cheek swelling.

saline pressure, balloon dilatation, or the midforceps grasper and even the insertion of a stent to the duct lumen (Nahliel, Bar, and Shacham et al. 2004). Simple balloon angioplasty was successful in 7 of 9 patients, and 5 of these patients remained asymptomatic on follow-up (Salerno, Lo Casto, and Comparetto et al. 2007). If this is unsuccessful and the patient continues to have recurrent swelling and sialadenitis, denervation with botulinum toxin or even parotidectomy may be required.

In the submandibular gland the stricture can be excised and sialodochoplasty performed as described above.

Was this article helpful?

0 0
10 Ways To Fight Off Cancer

10 Ways To Fight Off Cancer

Learning About 10 Ways Fight Off Cancer Can Have Amazing Benefits For Your Life The Best Tips On How To Keep This Killer At Bay Discovering that you or a loved one has cancer can be utterly terrifying. All the same, once you comprehend the causes of cancer and learn how to reverse those causes, you or your loved one may have more than a fighting chance of beating out cancer.

Get My Free Ebook

Post a comment