Outcomes of Local Bone Graft and Fixation of Proximal Pole Scaphoid Nascent Nonunions and Nonunions

Abstract

Background  Fractures of the proximal pole of the scaphoid have an increased risk of nonunion due to its tenuous blood supply. The optimal treatment of proximal pole scaphoid nonunions remains controversial.

Objectives  To review a single surgeon's experience with proximal pole scaphoid nascent nonunions (delayed unions) and nonunions that underwent surgical fixation with a cannulated headless compression screw and local autologous bone graft from the distal radius.

Patients and Methods  After obtaining Institutional Review Board approval, the electronic medical record of one tertiary care center was queried for patients with the diagnosis of “proximal pole scaphoid fractures” who underwent surgical fixation by a single surgeon over an 11-year period (2006–2017). Fifteen patients met initial query criteria; upon review of records, four patients were excluded due to the acute nature of the fracture, and one was excluded as surgical fixation included a vascularized bone graft.

Results  The final study cohort consisted of 10 patients with a total of 10 proximal pole scaphoid nonunions. Almost all of the patients in this study were male (9/10 [90%]), and sporting activities were the most common mechanism of injury (8/10 [80%]). Volumetric measurements of the scaphoid fractures on computed tomography (CT) revealed that the mean total volume of the scaphoid was 2.4 ± 0.48 cm ³ and the mean volume of the proximal pole fragment was 0.38 ± 0.15 cm ³ . Postoperative CT scans were performed at a mean of 12.4 weeks (range: 8–16 weeks), with seven (7/10 [70%]) showing signs of complete union and three (3/10 [30%]) demonstrating partial union. None of the patients required additional procedures and there were no complications.

Conclusions  Our results suggest that proximal pole scaphoid fractures with delayed union and nonunion treated with surgical fixation and autologous local bone graft heal without the need for more complex vascularized procedures. The volume of the proximal pole fragment did not correlate with increased risk of ongoing nonunion after the index procedure.

Level of Evidence  This is a Level IV, case series study.

Scaphoid fractures often present in delayed fashion as they initially may be perceived as minor wrist injuries. ^{1 2} Fractures of the proximal pole of the scaphoid are believed to be at an increased risk of delayed union and nonunion as compared with those at the waist or distal pole due to the more tenuous blood supply about this region. ^{3 4 5 6 7 8 9 10} The optimal treatment of proximal pole scaphoid fractures demonstrating delayed union and nonunion remains controversial. ^{3 4 5 6 7 8 9 10} Traditional treatment methods include screw fixation with local autologous bone graft or excision of the proximal pole fragment. ^{3 4 5 6 7 8 9 10 11 12 13 14} Some studies have reported unfavorable results with the use of conventional autologous bone graft, ⁴ and, as such, various alternative vascularized bone grafting techniques, including pedicled and free vascularized bone grafts, have been developed and more frequently studied over the past decade. ^{1 2 4 5 6 7 8 9 10 15 16 17 18 19 20 21} The purpose of this study was to review a single surgeon's experience with proximal pole scaphoid fractures with delayed union or nonunion that underwent surgical fixation with local autologous bone graft and screw fixation.

Patients and Methods

After Institutional Review Board approval, the electronic medical record was queried to identify all skeletally mature patients who underwent surgical fixation of a scaphoid fracture by a single surgeon at a level I trauma center from 2006 to 2017. Using Current Procedural Terminology codes (25440 and 25628), 72 patients with a total of 80 scaphoid fractures who underwent surgical intervention were identified. After initial review of medical records, 65 patients with fractures that did not involve the proximal pole of the scaphoid, 4 patients with acute proximal pole scaphoid fractures, and 1 patient who did not have a preoperative computed tomography (CT) scan were excluded. While Krimmer et al proposed a CT-based classification of scaphoid fractures based on fracture location and stability of the fracture pattern, a specific definition of the boundaries of the proximal pole is not stated; throughout the literature, the proximal pole has ranged from the proximal one-fifth to the proximal one-third of the scaphoid. ^{3 22 23 24 25 26} Therefore, for the purpose of this study, a proximal pole scaphoid fracture was defined as the majority of the fracture line located proximal to the head of the capitate. Patient demographics, mechanism of injury, pre- and postoperative imaging, and operative reports were reviewed and recorded ( Table 1 ). There is no current consensus regarding the classification of a fracture as a nonunion with regard to the specific amount of time from injury with no or minimal bony healing. For the purpose of this study, nonunion was defined as greater than 3 months of symptoms, and patients seen within 3 months of injury were considered to have “nascent nonunions” or “delayed union.” Volumetric parameters of the proximal scaphoid fracture fragments were determined by a fellowship-trained musculoskeletal radiologist from the preoperative CT scan ( Fig. 1 , Table 2 ).

Table 1. Description of cohort ( N  = 10) .

Gender	1 female
	9 male
Age at surgery	24 y old (median)
	26.2 y old (mean)
Side of injury	3 left
	7 right
Medical comorbidities	7: none
	3: ≥1
Mechanism of injury	9: sports
	1: fall
Time from injury to clinic visit	3.5 mo (median)
	10.7 mo (mean)
Time from injury to surgery	6 mo (median)
	9.5 mo (mean)
MRI	6 yes
	4 no
Location of bone graft insertion	7: fracture site
	1: screw track
	2: none
Last follow-up	4 mo (median)
	5.85 mo (mean)

Abbreviation: MRI, magnetic resonance imaging.

Fig. 1.

Computed tomography (CT) volumetric scan of the scaphoid with proximal pole fragment.

Table 2. Scaphoid fracture characteristics.

Patient	Total volume (cm 3 )	Proximal pole fragment volume (cm 3 )	Percentage
1	1.46	0.107	7.3
2	2.93	0.579	19.8
3	2.71	0.392	14.5
4	2.81	0.486	17.3
5	1.82	0.351	19.3
6	2.09	0.215	10.3
7	2.40	0.342	14.3
8	2.44	0.380	15.6
9	2.38	0.357	15.0
10	2.82	0.609	21.6
Mean	2.39	0.382	15.5

Operative Procedure

All procedures were performed under infraclavicular regional anesthetic and laryngeal mask anesthesia. The wrist was flexed over a bolster, and a longitudinal dorsal approach was used to expose the scaphoid. A longitudinal dorsal capsulotomy was performed, and care was taken to maintain the integrity of the scapholunate interosseous ligament. The fracture site was examined clinically and fluoroscopically. Some patients had a visible cartilage defect of the scaphoid; this was pried open gently dorsally and radially, and the fracture site was thoroughly curetted. In the absence of an obvious cartilage defect, the bone graft was placed through the screw track after drilling and prior to screw placement. Local bone graft was harvested from the distal radius: a 0.045-inch Kirschner wire was used to make multiple drill holes around Lister's tubercle to create a bone window. Through this bone window, a large amount of cancellous bone graft was harvested and carefully morselized. The morselized pieces were then placed directly into the nonunion site in the scaphoid, and the window created in the distal radius was replaced. The scaphoid fracture was then stabilized with a headless compression screw. A thumb spica splint was placed following soft tissue and skin closure.

Results

The final study cohort consisted of 10 patients with a total of 10 proximal pole scaphoid nascent nonunions (delayed union) and nonunions ( Table 1 ). At the time of surgery, the median age was 24 years (mean: 26; range: 17–42 years). Almost all of the patients in this study were male (9/10 [90%]), and sporting activities were the most common mechanism of injury (9/10 [90%]). None of the patients were active smokers. Six (6/10 [60%]) patients underwent preoperative magnetic resonance imaging (MRI), four of which were obtained by the surgeon. Two (2/10, 20%) patients were found to have signs of dysvascularity of the proximal pole on MRI as identified by a fellowship-trained musculoskeletal radiologist; signs of dysvascularity were described as T1 signal hypointensity as well as sclerosis of the proximal pole. Volumetric measurements of the scaphoid fractures were calculated using Aquarius TeraRecon (Foster City, CA). The mean total volume of the scaphoid was 2.4 ± 0.48 cm ³ (range: 1.46–2.82 cm ³ ), and the mean volume of the proximal pole fragment was 0.38 ± 0.15 cm ³ (range: 0.107–0.609 cm ³ ) ( Fig. 1 , Table 2 ). The mean proportion of the proximal pole fragment volume relative to the total scaphoid volume was 0.15 ± 0.043 (range: 0.073–0.216).

The median time from injury to surgery was 6 months (mean: 9.5; range: 2–25 months). All patients underwent open reduction internal fixation (ORIF) of the proximal pole nonunion. Eight (8/10 [80%]) patients underwent local bone grafting. Of those who underwent local bone grafting, bone graft was inserted at the site of fracture in seven patients (7/8 [87.5%]) and in the screw track in one patient (1/8 [12.5%]). In the remaining two patients, bone graft was not used. The median duration of follow-up as calculated from the date of surgery to the final clinic visit was 4 months (mean: 5.9; range: 3–16 months).

Postoperatively, all patients were immobilized for 8 to 10 weeks in a thumb spica cast. Patients were then transitioned to an Ispoprene thumb spica splint (Rolyan San-Splint - Performance Health, Warrennik, IL), and a rehabilitation program was instituted. To confirm union, a CT scan was performed at 12 to 16 weeks postoperatively (mean: 13 weeks). Postoperatively, seven (7/10, 70%) fractures showed signs of complete union at the time CT was obtained ( Fig. 2 ) and three (3/10, 30%) patients demonstrated partial union, with healing of at least 50% of the fracture surface area ( Fig. 3 ). After confirmation of scaphoid union on CT, the splint was discontinued. None of the patients reported to have pain at their final visit ( Fig. 4 ). None of the patients have had any subsequent procedures related to their scaphoid nonunion.

Fig. 2.

Case example. Preoperative images: ( A ) anteroposterior radiograph, ( B,D ) coronal computed tomography (CT) scan, and ( C ) sagittal CT scan.

Fig. 3.

Case example. Postoperative images: ( A ) anteroposterior (AP) radiograph at 6 weeks, ( B ) AP radiograph at 10 weeks, ( C ) coronal computed tomography (CT) scan at 12 weeks, and ( D ) sagittal CT scan at 12 weeks.

Fig. 4.

Clinical images demonstrating range of motion at the final appointment.

Discussion

The scaphoid is the most commonly fractured carpal bone, with reported nonunion rates as high as 10 to 15%. ^{1 2 4 5} Proximal pole fracture fragments are thought to be more susceptible to nonunion due to the distal insertion of the vascular pedicles, with the incidence of dysvascularity reported to be between 16 and 42%. ⁸ Some reports suggest that proximal pole nonunions treated with conventional autograft bone graft result in unfavorable outcomes. ^{5 6 27} Currently, there is no consensus for optimal treatment. The goals of treatment for proximal pole scaphoid nonunions include providing stability through fixation, and bone grafting, as needed, in an effort to achieve fracture union.

The exact definition and classification of a fracture as a nonunion is controversial. ^{28 29} Currently, there is no universally accepted time of continued lack of bony healing that defines a nonunion as compared with a delayed union or nascent nonunion. Several authors have defined scaphoid nonunions as fractures that have not demonstrated bony healing at least 6 months postinjury. ³⁰ Some also define scaphoid fractures that have no evidence of healing at 3 months from injury as nonunions, ^{28 31} whereas others consider these to be nascent nonunions. ^{10 31}

Conventional treatment of scaphoid nonunions includes screw fixation with nonvascularized bone graft or excision of the proximal pole fragment. ^{3 4 5 6 7 8 9 10 11 12 13 14 15 32 33 34} Luchetti et al performed a retrospective study of 20 patients with proximal third scaphoid nonunions who underwent curettage and cancellous bone autograft from the distal radius and screw fixation, with 18 (90%) of 20 patients achieving union. ³ In their study, the vascularity of the proximal pole was not specifically assessed; MRI was not obtained and punctate bleeding of the proximal pole was not assessed during surgery. Their findings suggest that proximal pole nonunions have the propensity to heal following thorough curettage across the fracture site, augmentation with local bone graft, and stable fixation. ³

As the understanding of carpal anatomy and wrist kinematics has improved, novel treatment options have been developed with the intent to maximize healing of the proximal pole of the scaphoid. Several types of vascularized bone grafts have been used for the management of proximal pole nonunions, including those from dorsal and volar radius, metacarpal artery, and medial femoral condyle. ^{6 17 35 36} Several studies have examined the use of these various types of pedicled and free vascularized bone graft for proximal pole nonunions; however, many have small cohorts and limited periods of follow-up. Janowski et al reviewed current methods of treatment for scaphoid nonunions and discussed various options of bone graft and vascularized bone graft. ⁴ A literature review found that the union rates following utilization of 1,2 intercompartmental supraretinacular artery (ICSRA) as a vascularized pedicle are between 60 and 100%. ⁴ They concluded that the treatment of scaphoid nonunion remains a difficult problem and that recent studies including meta-analyses do not reveal a substantial difference in union rates between sources of bone graft and types of fixation. ⁴ Steinmann et al concluded that the management of scaphoid nonunion with bone grafting with or without internal fixation may be considered in the absence of degenerative carpal arthritis. ¹⁹ They noted that the relative risks and benefits of nonvascularized and vascularized bone grafting must be considered and that the literature on vascularized bone grafting for scaphoid nonunion shows variable healing rates, from 27 to 100%. The authors suggested that nonvascularized bone graft is sufficient for scaphoid waist nonunions and proximal pole nonunions with preserved vascularity. Steinmann et al also reviewed a series of 49 patients with 50 scaphoid nonunions treated with ICSRA reverse flow vascularized bone graft. ¹⁹ They determined that the overall factors associated with an increased rate of nonunion include female gender, tobacco use, proximal pole avascularity, humpback deformity, comminution of the proximal pole, and use of Kirschner wire or no internal fixation. They also noted that the type of internal fixation affected union rate, as there was an increased rate of nonunion with Kirschner-wire fixation as compared with screw fixation. ¹⁹ Burger et al performed a retrospective review of 16 cases of medial femoral trochlea (MFT) osteocartilaginous flap for the reconstruction of proximal pole scaphoid nonunion in the setting of inadequate proximal fragment bone quality. ³⁷ During the procedure, the proximal pole fragment was excised prior to insertion of the osteocartilaginous flap that was followed by fixation; the authors noted that in the majority of cases, fixation was accomplished using cannulated screws. Fifteen (93.8%) of 16 patients healed without the persistence of nonunion. The authors suggest that the size of the vascularized osteocartilaginous MFT graft allows for both substantial resection of the dysvascular proximal pole and screw fixation, neither of which are able to be accomplished with smaller vascularized or free bone graft, and, therefore, both vascularity and stability are more adequately addressed with this technique. ³⁷

In our study, there was very delicate handling of the proximal pole fracture fragment with thorough preparation of fracture surfaces, careful management of the local bone graft, and then meticulous fixation of the small proximal pole fragments with detailed attention to avoiding fragmentation on screw insertion. Ten of 10 patients went on to achieve union without any additional procedures. To avoid fragmentation of the proximal pole, in all cases the screw size used was chosen such that the amount of bone available around the proposed insertion site of the screw was at least two to three times the diameter of the trailing end of the screw.

The volumetric analysis performed in this study demonstrated no difference with regard to the volume of the proximal pole fragment and healing. The limitations of this study include the small number of patients from a single surgeon, as well as its retrospective nature. Vascularized pedicle bone graft and free vascular grafts are technically challenging and fragile; while they promote vascularity, they concomitantly can preclude screw fixation and therefore may result in decreased overall stability of the nonunion. Our findings suggest that local bone graft with screw fixation should be considered as an option for the treatment of proximal pole scaphoid nonunion.

The optimal treatment of proximal pole scaphoid nonunions or nascent nonunions (delayed unions) remains controversial. Traditional treatment methods include screw fixation with local autologous bone graft or excision of the proximal pole fragment, whereas novel treatment methods include the use of pedicled and free vascularized bone grafts. Even though new treatments may be alluring, based on the results of this study, traditional treatment of screw fixation with local autologous bone graft remains effective with respect to promoting union of proximal pole scaphoid nascent nonunions (delayed union) and nonunions.