Volar Plating versus Headless Compression Screw Fixation of Scaphoid Nonunions: A Meta-analysis of Outcomes

Abstract

Background  Headless compression screw fixation with bone grafting has been the mainstay of treatment for scaphoid nonunion for the past several decades. Recently, locked volar plate fixation has gained popularity as a technique for scaphoid fixation, especially for recalcitrant or secondary nonunions.

Purpose  The purpose of this meta-analysis was to compare union rates and clinical outcomes between locked volar plate fixation and headless compression screw fixation for the treatment of scaphoid nonunions.

Methods  A literature search was performed for studies documenting treatment outcomes for scaphoid nonunions from 2000 to 2020. Inclusion criteria consisted of (1) average age > 18 years, (2) primary study using screw fixation, plate fixation, or both, with discrete data reported for each procedure, and (3) average follow-up of at least 3 months. Exclusion criteria consisted of studies with incomplete or missing data on union rates. Data from each study was weighted, combined within treatment groups, and compared across treatment groups using a generalized linear model or binomial distribution.

Results  Following title and full-text review, 23 articles were included for analysis. Preoperatively, patients treated with plate fixation had significantly longer time from injury to surgery and were more likely to have failed prior surgical intervention. There was no significant difference between union rates at 92 and 94% for screw and plate fixation, respectively. However, plate fixation resulted in longer time to union and lower modified Mayo wrist scores.

Conclusion  Patients treated with locked volar plate fixation were more likely to be used for recalcitrant or secondary nonunions. There was no statistically significant difference in union rates between screw and plate fixation. The results from this meta-analysis support the select use of locked volar plate fixation for scaphoid nonunion, especially recalcitrant nonunions and those that have failed prior surgical repair.

The scaphoid is the most commonly fractured carpal bone accounting for 60 to 70% of all carpal fractures. ¹ Unfortunately, nonunion is a common complication of scaphoid fractures, believed to occur in approximately 5 to 10% of all scaphoid fractures. ² This predisposition toward nonunion is related to several factors including delayed or missed diagnoses, as well as the scaphoid's tenuous blood supply to the proximal pole which flows retrograde and is vulnerable to disruption following an acute fracture. ³ Factors that can increase the risk of nonunion include delayed diagnosis, insufficient immobilization, fracture displacement, and proximity of the fracture to the proximal pole. ^{2 4 5 6} Over the past several decades, many different surgical strategies have been developed to treat scaphoid nonunions. In order to achieve optimal outcomes, there is a general consensus that rigid fixation and fracture compression are necessary to achieve optimal outcomes. However, many different methods of rigid fixation have been described and utilized over the years, with new methods continuing to be developed. ⁷

A common method of fixation in treating scaphoid nonunion is the headless compression screw. This method became widely utilized to treat scaphoid fracture and nonunion after it was first described by Herbert and Fischer in 1984, demonstrating successful results for both acute scaphoid fractures and nonunions. ⁸ Subsequent advancements in compression screw fixation included the development of the cannulated headless compression screw, which resulted in improved union rates by enabling greater accuracy of screw placement relative to noncannulated screws. ⁹ Thus, cannulated headless compression screws have become the mainstay of achieving fixation in scaphoid nonunions for the past two decades. A more recent technique developed for scaphoid fixation is volar locked plating. Although initially described by Ender in 1977, it was not commonly utilized due to the development of the headless compression screw shortly afterwards. ¹⁰ However, recent studies evaluating the efficacy of volar plating for scaphoid nonunions have indicated that this method of fixation can be effective in treating scaphoid nonunions, in particular recalcitrant or multifragmented nonunions. ^{11 12} Plate fixation offers increased rotational stability compared to screw fixation, as well as increased construct stability through the use of locking screws, while minimizing the reliance on cortical or structural bone grafting. ¹³ Despite these advantages of plate fixation, potential disadvantages include symptomatic hardware which may require the need for hardware removal following successful union. ^{14 15} The purpose of this study is to provide a meta-analysis comparing the union rates in patients treated with screw fixation versus plate fixation for scaphoid nonunions, while also evaluating functional outcomes, complications, and the need for reoperation between the two groups.

Methods and Materials

Literature Review

A literature search of studies documenting outcomes for scaphoid nonunion treatment with either screw fixation, plate fixation, or both using PubMed, SCOPUS, Cochrane Library, and Google Scholar from 2000 to 2020 was performed. Search terms included “(scaphoid nonunion) AND (screw fixation OR headless compression screw OR plate fixation OR volar plate).” Two of the authors (D.V. and M.R.) reviewed all identified article titles for appropriateness and relevancy. Afterward, we applied specific inclusion and exclusion criteria for full-text review and final article selection. The Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines were used to conduct the review. ¹⁶ Inclusion criteria consisted of (1) adult patients with average age > 18 years, (2) primary study using either screw fixation, plate fixation, or both, with discrete data reported for each procedure, (3) average follow-up of at least 3 months, and (4) a full English language manuscript available for review. Exclusion criteria consisted of studies with incomplete or missing data on union rates.

Data Extraction and Statistical Analysis

For the studies meeting final inclusion criteria, relevant data was identified and collected for the specific cohort of patients from each study including the method of fixation, the number of patients undergoing each method of fixation, gender, age, mean time from injury to surgery, rate of avascular necrosis (AVN), union rate, mean time to union, modified Mayo wrist scores (MMWS), complication rate, and reoperation rate. Two reviewers were utilized in this process to ensure accuracy and completeness. Patients from studies that included cohorts undergoing Kirschner wire fixation were excluded from the analysis. When reported, individual complication rates were summarized quantitatively and qualitatively for screw and plate fixation. Reoperation rates were collected for revision surgeries performed for persistent nonunions, symptomatic hardware, or salvage operation for scaphoid nonunion advanced collapse (SNAC).

Data collected from the studies were weighted and combined within treatment groups, and then compared across treatment groups. For the continuous variables, including age, time from injury to surgery, time to union, and MMWS, weighted means were calculated and compared across treatment groups using a generalized linear model. For categorical variables, including gender, rate of AVN, union rates, complication rates, and reoperation rates, the number of events was compared across treatment groups using a binomial distribution. All statistical analyses were performed using R Studio (Version 3.6.3, Vienna, Austria).

Results

The initial literature search identified 1,320 articles based on keywords. A total of 348 articles were excluded because they were published prior to the year 2000. Of the remaining 972 articles, 925 were excluded based on title, leaving 47 studies for full-text review. Twenty-four studies were excluded because they failed to meet final inclusion criteria. Ultimately, 23 studies were included for final data extraction and analysis. ^{11 12 15 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36} Fig. 1 outlines the results of the literature search and article selection.

Fig. 1.

Flowchart of literature review and article selection process according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines.

Of the 23 studies meeting inclusion criteria, 3 studies analyzed screw fixation versus plate fixation, 11 studies reported results for screw fixation alone, and 9 studies reported on the results of plate fixation alone. Three studies using screw fixation also used vascularized bone grafting compared to two studies with plate fixation. The rest of the studies used nonvascularized bone grafting techniques. In total, there were 395 patients treated with headless compression screw fixation and 205 patients treated with volar plate fixation. Demographic data from each of the study cohorts are presented in Table 1 . Patients undergoing plate fixation trended toward being older than patients undergoing screw fixation at 30.4 years for plate fixation versus 27.3 years for screw fixation ( p  = 0.056). Preoperative patient characteristics and functional outcomes are presented in Table 2 . Patients undergoing plate fixation had a significantly longer time from injury to surgery at 22.2 months compared to 13.4 months for screw fixation ( p  = 0.001). Plate fixation patients were also significantly more likely to have undergone prior scaphoid surgery at 15.1% compared to 0.60% for the screw fixation group ( p  = 0.011). Overall, reported rates of preoperative AVN in these cohorts were low and were not significantly different between plate and screw fixation at 4.10 and 0.53%, respectively ( p  = 0.383).

Table 1. Demographic data from included articles according to surgical intervention cohort.

Research article	Year	Surgical technique	Study design	No. of subjects	Mean age, y	Gender, % male
Screw fixation cohorts
Talia et al	2019	Screw vs. plate fixation	Retrospective	15	30.1	100
Quadlbauer et al	2019	Screw vs. plate fixation	Retrospective	22	33	95.4
Sander et al	2018	Screw vs. plate fixation	Retrospective	18	39.6	89.5
Mahmoud and Koptan	2011	Screw fixation	Retrospective	27	28.3	96.3
Gurger et al	2018	Screw fixation	Retrospective	12	27	83.3
Morris et al	2018	Screw fixation	Retrospective	12	20	66.6
Somerson et al	2015	Screw fixation	Retrospective	17	21.1	76.5
Garcia et al	2014	Screw fixation	Retrospective	19	21	94.7
Cohen et al	2013	Screw fixation	Retrospective	12	22	100
Altay et al	2014	Screw fixation	Retrospective	33	29	93.9
Gereli et al	2011	Screw fixation	Retrospective	29	28.9	96.5
Ritter and Giachino	2000	Screw fixation	Retrospective	10	26.6	92.8
Poggetti et al	2015	Screw fixation	Retrospective	80	35.2	76.25
Papatheodorou and Sotereanos	2019	Screw fixation	Retrospective	89	28	79.8
				Total = 395
Weighted mean					27.3	91.2
Plate fixation cohorts
Talia et al	2019	Screw vs. plate fixation	Retrospective	15	29.9	100
Quadlbauer et al	2019	Screw vs. plate fixation	Retrospective	20	32	95
Sander et al	2018	Screw vs. plate fixation	Retrospective	20	39.6	89.5
Putnam et al	2019	Plate fixation	Retrospective	13	32	76.9
Esteban-Feliu et al	2018	Plate fixation	Retrospective	15	32	93.3
Dodds et al	2018	Plate fixation	Retrospective	20	24	90
Dodds and Halim	2016	Plate fixation	Retrospective	9	26.67	88.9
Ghoneim	2011	Plate fixation	Retrospective	14	26	100
Mehling et al	2019	Plate fixation	Retrospective	17	33.1	88.2
Leixnering et al	2011	Plate fixation	Retrospective	11	37	63.6
Schormans et al	2018	Plate fixation	Prospective	21	31	95.2
Putnam et al	2019	Plate fixation	Retrospective	34	31	94.1
				Total = 209
Weighted mean					30.3	90.3

Table 2. Continuous and categorical outcomes for demographic and preoperative data, as well as postoperative and functional outcome data for screw fixation versus plate fixation.

Variables	Screw fixation	Plate fixation	p -Value
Demographics and preoperative data	Mean (95% confidence interval)
Age	27.34 (25.37–29.30)	30.36 (27.96–32.75)	0.056
Gender (male)	91.20% (84.10%–95.31%)	90.33% (82.23%–94.97%)	0.834
Injury to surgery time (mo)	13.39 (10.17–16.62)	22.18 (18.37–25.99)	0.001
Prior scaphoid surgery	0.60% (0.05%–6.77%)	15.05% (7.46%–28.02%)	0.011
Avascular necrosis	0.53% (0.02%–13.08%)	4.09% (0.16%–52.86%)	0.383
Postoperative outcomes
Union rate	91.98% (82.35%–96.58%)	93.93% (86.48%–97.40%)	0.642
Time to union (mo)	2.70 (2.28–3.13)	4.45 (3.44–5.45)	0.002
Modified Mayo wrist score	90.54 (88.58–92.49)	71.64 (63.52–79.77)	< 0.001
Complication rate	1.53% (0.14%–14.45%)	6.12% (1.51%–21.76%)	0.314
Reoperations rate	3.62% (0.98%–12.48%)	18.38% (4.99%–49.15%)	0.076

Clinical and functional outcomes are also summarized in Table 2 . Overall, there was no significant difference in union rates between the two treatment groups at 93.9% for plate fixation and 92.0% for screw fixation ( p  = 0.642). However, time to union was significantly longer for patients undergoing plate fixation at 4.45 months compared to 2.70 months for screw fixation ( p  = 0.002). In addition, the mean MMWS was significantly lower for plate fixation versus screw fixation at 71.64 and 90.54, respectively ( p  < 0.001). The reported complication rate for plate fixation overall was 6.12 and 1.53% for screw fixation ( p  = 0.314). The rate of reoperation was higher for plate fixation at 18.38 versus 3.62%, but this difference did not achieve statistical significance ( p  = 0.076).

Among the screw fixation group, the most commonly reported complication was persistent radial-sided wrist pain in a total of 27 patients. Other reported complications in the screw fixation group were screw migration in 12 patients, temporary neuropraxia of sensory branch of the radial nerve in two patients, and superficial wound complications in two patients. The most commonly reported complication in the plate fixation group was symptomatic hardware impingement in a total of 19 patients. Other reported complications in the plate fixation group included breakage and/or failure of hardware in six patients, and superficial wound complications in three patients. In the screw fixation group, there were a total of 38 reported reoperations. Seven of these reoperations were for persistent nonunion, whereas the remaining 31 were for removal of hardware. In the plate fixation group, there were a total of 57 reported reoperations. Seven of these reoperations were for persistent nonunion and the remaining 50 reoperations were for removal of hardware.

Discussion

The establishment of an optimal surgical strategy for scaphoid nonunions continues to prove a difficult task due to the variety of surgical options implemented by hand surgeons. While there is a consensus among hand surgeons that rigid fixation is necessary for optimal healing, there lacks a consensus for which method of fixation is optimal. Currently, there are three main methods of rigid internal fixation used to treat scaphoid nonunions, which are Kirschner wires, headless compression screws, and volar plates. While Kirschner wires are still used as a surgical treatment in some scaphoid nonunion cases, this technique has been used with less frequency as screw fixation has demonstrated superior union rates while also allowing for earlier postoperative mobilization. ^{37 38 39} More recently, volar plating has been used for fixation of scaphoid nonunions, and has been increasingly reported in the literature. While initial indications for volar plating were described as a salvage procedure for failed screw fixation, subsequent studies have expanded indications to include treatment of certain primary scaphoid nonunions, such as those with humpback deformity or multifragmentation. ^{11 12 24} Therefore, considering the variability in clinical presentation for patients with scaphoid nonunion, as well as the expanding indications for volar plate fixation, the aim of this study was to categorize clinical presentations and analyze outcomes for patients treated with screw versus plate fixation using meta-analytical techniques.

This meta-analysis found that both methods of rigid fixation achieved similarly high rates of union. However, there was a significantly longer mean time to union for the plate fixation group. Plate fixation also resulted in significantly decreased MMWS, and there is a trend toward increased rates of complications and reoperations for plate fixation. While these clinical and functional outcome measures appear to favor the use of screw fixation over plate fixation, it is important to note the differences in clinical presentation between these two groups. Based on the results of this analysis, patients treated with plate fixation had a significantly longer mean time from injury to surgery, and they were also more likely to represent revision surgery for patients having failed prior surgical intervention for acute scaphoid fracture or scaphoid nonunion in the past. Because clinical presentations varied significantly between the two treatment groups, the more chronic nature of the nonunions in the plate fixation group serves as a confounding factor when comparing differences in outcomes such as mean time to union and MMWS.

The trend toward an increase in complications and reoperations in the plate fixation group is also important when considering the appropriate indications for its use. Many authors have reported on the tendency for symptomatic hardware as a disadvantage of plate fixation. ^{14 15} Because of the presence of the hardware on the articular surface and the potential to cause wrist impingement, some authors have recommended that all plates be removed once radiographic union is achieved. ⁴⁰ However, as described by Dodds et al the identification of particular landmarks on the scaphoid can help aid in optimal plate placement to reduce the risk of hardware impingement and need for removal. ³³ If the proximal end of the volar plate is fixed distal to the equatorial line of the proximal scaphoid, it can help prevent impingement and reduce the need for subsequent removal of hardware. It is possible that some studies with particular low MMWS, high rates of impingement, and hardware removal, may have had tendency for excessively proximal plate placement. ¹⁵ Some fracture patterns may necessitate proximal plate placement, however, and wrist impingement might be unavoidable in these cases, and thus not necessarily considered a complication. ³²

Based on the data presented in this analysis, in addition to these perceived disadvantages of plate fixation, it stands to reason that indications for plate fixation as an initial treatment for scaphoid nonunion should be limited. Because of the documented reduction in functional wrist scores for plate fixation versus screw fixation, and the lower rates of reoperation, and the similarly excellent rates of union, headless compression screws should be considered the primary option for internal fixation for scaphoid nonunions. Because of the excellent union rates reported with plate fixation, this meta-analysis supports the use of plate fixation for secondary nonunions that have failed prior surgical intervention with headless compression screws. While authors have also reported multifragmented scaphoid nonunions as well as humpback deformity as appropriate indications for primary locked volar plating, fragmentation and degree of deformity were not commonly reported in the articles included in this analysis. Because of the high predisposition for comminuted scaphoid fractures and nonunions to result in persistent nonunion and SNAC wrist, as well as the inability of screw fixation to capture comminuted fragments, plate fixation seems to be an appropriate surgical strategy for these cases. ⁴¹ Recent use of two screw fixation has been proposed as a safe and effective alternative treatment for scaphoid nonunion. ^{21 24} This method provides increased rotational stability compared to single screws with a reduced risk of symptomatic hardware compared to plate fixation. However, further studies are required to determine outcomes and appropriate indications for use.

There are several limitations to this study due to the inherent nature of meta-analyses and the quality of data across the studies included in this study. Scaphoid nonunions are relatively rare in the general population, and thus, most of the individual studies included in this analysis have low statistical power, and all but one study were retrospective reviews. This meta-analysis was also limited by the heterogeneity of the data that was reported by each study. In some instances, standard deviations were not reported for certain variables, and thus variance was calculated using a conservative estimate based on reported ranges. Similarly, several studies did not report data on all variables of interest, such as complications or MMWS, thus limiting statistical power. Additionally, because of the diversity of treatment options utilized for scaphoid nonunion, particularly the multiple options for bone grafting used in conjunction with rigid fixation, there exists limitation due to the heterogeneity of surgical interventions used in these studies. There were similar numbers of studies between treatment groups that used vascularized bone grafting and there was no difference in preoperative rates of AVN. While this indicates treatment groups were likely similar in terms of number of patients receiving vascularized versus nonvascularized bone grafts, this still exists a possible confounding variable that our meta-analysis cannot fully account for.

Conclusions

With ongoing improvements and introductions of new surgical techniques to address scaphoid nonunions, the establishment of optimal treatment strategies remains a difficult task. With an increase in the past several years of studies reporting outcomes following volar plate fixation, it is important to critically analyze these results and compare to the current methods of rigid fixation with headless compression screws. The results of this systematic review and meta-analysis demonstrated no difference in union rates between the two methods of fixation but a longer time to union for plate fixation. Additionally, there were decreased functional wrist outcomes and trend toward increased reoperations for the plate fixation group. However, the plate fixation group had a significantly longer time from injury to surgery and was more likely to have failed previous surgical intervention. Therefore, headless compression screws appear to be the preferred method of fixation for primary scaphoid nonunions, but volar plate fixation serves as an appropriate alternative for fixation in secondary scaphoid nonunions that have failed prior surgical intervention with union rates equivalent to primary scaphoid nonunion repair.