2017-18 HSC Section 3 Green Book

Table 1 Continued

Study

Patient Age, y

Foreign Body

Path of the Foreign Body

Sharif et al., 9 2000

44

Pen and pencil

Transnasal then transcranial between the hemispheres of the brain Lower eyelid through the lamina into the ethmoid Upper eyelid through the ethmoid, sphenoid, then intracranial Right upper eyelid just above the medial canthus inferomedially through the lamina and ethmoid to the inferior sphenoid Through the submentum, tongue, maxillary, and ethmoid sinuses into the orbit and transcranial Transnasal, transcribriform into the frontal lobe Transnasal transcribriform through most of the frontal lobe Upper outer left orbit through the left orbital floor, maxillary sinus through the pterygopalatine fossa into the NP and ending just anterior to the spinous ligament Transnasal to the sphenoid through the sella and into the mammary bodies

Udwadia et al., 31 1994 Verret et al., 18 2005

30 27

Toothbrush

Branch

Windle-Taylor et al., 32 1978

15

Harpoon

Zweckberger et al., 22 2011

40

Branch

No surgery

Bhatjiwale et al., 3 2001

7

Wire fragment

Chan et al., 7 2014

49

Chopstick

Hickman, 25 1984

41

Steak knife

Rawlinson et al., 27 1988

27

Pool cue

OCR = Optico-carotid recess; ICA = internal carotid artery; NP = nasopharynx.

the Netherlands) was performed with the keywords: “penetrat* and intracranial and foreign,” with a result of 215 publications. Articles were noted from 1949 to 2016. Exclusion criteria were the following: no sinus involvement, foreign language, forensic articles that did not describe management, nontraumatic injury ( i.e., iatrogenic), and high- velocity projectiles. It was deemed that articles with no sinus involve- ment were unlikely to have otolaryngology involvement. In addition, high-velocity projectiles were excluded because the pathophysiology of injury is dramatically different in high-velocity compared with low-velocity penetrating injury. After a title review, 98 articles were excluded, which left 117 publications that underwent abstract review and, if needed, imaging review to assess for the above criteria. Eighty- one articles were excluded for no sinus involvement, which left 36 articles for full review. Two articles were excluded for high-velocity injury and one for nontraumatic injury; the final count included 33 articles that describe 35 cases of transorbital or transnasal penetrating trauma with sinus involvement. A schematic of the systematic review is presented in Fig. 5. RESULTS A total of 35 cases in 33 publications were systematically reviewed (Table 1). The average age of the patient described was 30.1 years (median age, 36 years). The most common FB found were plant branches in eight cases (23%), chopsticks in five cases (14.2%), and pen or pencil in four cases (11.4%) (Fig. 6). Overall, wooden material was identified in 14 cases (40%). The mechanism of injury was acci- dental (fall or MVC primarily) in 22 cases (63%), assault in 9 cases (26%), and self-inflicted in 4 cases (11%). Seven patients (20%) had delayed presentation, which ranged from 36 hours after injury to two patients who presented 3 years after the original injury. All the patients received CT imaging initially for identification of the trajec- tory of the FB. In 24 cases (69%), a trajectory was shown that penetrated intracra- nially, all except 1 patient with intracranial penetration presented with CSF leak or CSF leak noted after FB removal. 3 In 20 cases (57%), a transorbital trajectory was shown. A transnasal trajectory was noted in 15 cases (43%). The ethmoid sinuses were the most commonly involved sinuses in 30 patients (86%), followed by sphenoid in 13

patients (37%), maxillary in 7 patients (20%), and frontal in 2 patients (6%). Multiple sinuses were involved in 16 patients (46%). Nine patients (26%) with suspected arterial injury received a CT angiogram or formal angiogram. Antibiotic administration was reported in 31 patients (89%). The majority of patients, 31 (89%), were treated sur- gically. Of these, 16 patients (46%) were treated with endoscopic surgery, which allowed for visualization of the FB during removal, removal of any debris, and CSF leak repair after removal. Reported complications ranged from CSF leak in 23 patients (66%); meningitis in 6 patients (17%), particularly in cases with delayed presentation; and brain abscess in 1 patient (3%). Orbital complica- tions included transient or permanent ophthalmoplegia in five pa- tients (14%). Permanent blindness was noted in only eight patients (23%): in two patients due to globe rupture and in six due to optic nerve injury. Decreased visual acuity occurred in three patients (9%). Vascular injury was noted in two patients (6%): one patient with right internal carotid artery and right posterior cerebral artery occlusion, and one patient with carotid-cavernous fistula. 1,4 A summary of find- ings from the systematic review is shown in Fig. 6. DISCUSSION Although low-velocity transorbital and transnasal penetrating in- tracranial injures overall are uncommon, they are usually caused by falls, MVCs, or assaults. The most important determinants of the severity of injury from these penetrating FBs are the velocity and trajectory. Purely transorbital injuries have two relatively distinct pathways of intracranial penetration due to anatomic factors. 1,2 These pathways occur either through the thin orbital roof or are directed posteriorly through the optic nerve and superior orbital due to the “cone” shape of the orbit. 1,2 Transnasal injuries can traverse the sinuses or can be directed up the narrow nasal cavity to the cribriform plate and gain access to the intracranial cavity. 2,3,5–10 However, when the penetrating injury takes a more oblique path transorbitally (as in our case) or transnasally, then the pattern of injury is more unpre- dictable. In some cases, the object can be lodged intracranially and the patient’s presentation can be delayed for years. 8,11 The workup of these patients includes a history and physical ex- amination, especially a full neurologic examination, an ophthalmo-

March–April 2017, Vol. 31, No. 2

44