Introduction

Fractures of the proximal phalanx account for 38% of all phalangeal fractures [1]. While stable fractures can be treated conservatively with splinting and early mobilization, unstable or significantly dislocated fractures require surgical intervention [2]. Various surgical techniques have been described for the fixation of proximal phalanx fractures, with Kirschner wires, plates, and screws being commonly used [3].

The percutaneous insertion of K-wires is advantageous due to minimal soft tissue damage, as opposed to open reduction and internal fixation with plates or screws. However, as shown by Eberlin et al. percutaneous K-wires also carry the risk of pin infection, and stiffness with tenolysis required in 7% of their patients [4]. Ultimately, depending on the type of fracture, the results for functional recovery vary little, comparing closed reduction and percutaneous K-wiring to open reduction and lag screw fixation [5].

Trans-articularly inserted cannulated headless screws have emerged as a promising alternative for the fixation of proximal phalanx or metacarpal fractures. Several authors have described the use of this technique [6,7,8,9,10,11]. The technique was initially introduced by Weiss et al. for addressing phalangeal nonunions within the middle third of the phalangeal shaft [12]. Even in that early publication, it was portrayed as a fixation method that provided immediate stability, enabling early rehabilitation.

Maintaining a good range of motion during the early postoperative period is critical in preventing stiffness and adhesions following proximal phalanx fracture surgery. Intramedullary fixation of fractures of the proximal phalanx with a headless screw has proven to fulfil the requirements for early post-operative mobilization and thus lowers the risk of adhesions [11, 13,14,15]. Compared to other fixation techniques, intramedullary screw fixation exhibits a lower rate of implant removal [13, 16, 17]. Despite the satisfactory clinical outcomes and low early post-operative complication rates associated with intramedullary screw fixation, a limited number of studies have been able to determine the incidence of late complications due to the short follow-up periods.

The aim of this study was thus to review the longer term outcome of proximal phalanx fractures treated with an intramedullary headless compression screws, given the promising short-term results.

Patients and methods

Patients and data collection

We conducted a retrospective analysis of eight patients who sustained nine proximal phalanx fractures and were treated with intramedullary compression screw fixation. The patients' medical records were reviewed and analysed. Furthermore, a follow-up examination was conducted three years postoperatively to evaluate both clinical and radiological long-term outcomes.

Ethical approval to this project was granted by the local Ethics Committee.

Outcomes of interest and their relevance

From the patient’s records we retrospectively collected data concerning the type of fracture, the type of surgery and patients’ demographics, as well as total range of motion measurements at 3 months postoperatively.

At the follow-up visits conducted three years postoperatively, the patients were requested to complete the Michigan Hand Questionnaire (MHQ). The MHQ comprises six health domains, including overall hand function, activities of daily living (ADLs), pain, work performance, aesthetics, and patient satisfaction [18].

In addition, grip strength was assessed as a functional outcome using a Jamar dynamometer (Patterson Medical Holding, Warrenville, IL, USA) in both hands at position 2. Pinch grip strength was also measured for the thumb. The total active motion (TAM) of the fractured digit was determined. The TAM results were compared to the TAM at 3 months postoperatively which were obtained from the patients’ medical records.

The medical records of the patients were reviewed for any instances of major or minor complications that occurred during the postoperative period.

A radiological evaluation was performed with plain x-rays and tomosynthesis.

Tomosynthesis is a radiological technique intended to analyse high-contrast structures such as bone. In analyses where bone superimposition is essential or metal structures hide the regions of interest, the tomosynthesis is superior to projection radiography [19].

Conventional radiographs were performed using standard clinical radiography (dorso-palmar, oblique and lateral projections) of the hand/fingers. Standard pre-sets were used (e.g. tube voltage of 50kgV and a current of 40 mA), according to the standard protocol of our department. The tomosynthesis was performed using the same radiography unit (FDR Ascelerate, Fujifilm Europe GmbH) equipped with the vendor-specific tomosynthesis tool. Imaging included dp-projections at a tube voltage of 50 kV and a tube current of 40MA. The tomographic image acquisition resulted in 36 coronal section images with a 2 mm increment.

The radiological images were used for the evaluation of screw position, fracture healing and joint damage, with the tomosynthesis imaging accounting for superposition-free, focused depiction of the interphalangeal joints and respective cartilage.

Surgical technique

The type of anaesthesia varied according to the patient’s wish. Six patients had general anaesthesia or regional anaesthesia under tourniquet. Two patients underwent the procedure under local anaesthesia administered by the surgeon without tourniquet (WALANT) [20].

Fractures were reduced using a traction technique and manipulation of the affected finger, under the guidance of fluoroscopy. A guide wire, sized to match the screw, was subsequently inserted through a small incision in the skin, either in an anterograde fashion via the MCP joint, or retrograde via the PIP joint. Once the guide wire was in place, the screw was inserted over it. The screw size was chosen based on the size of the medullary canal as measured in the lateral x-ray view. Either 2.2 mm or 3.0 mm diameter self-tapping cannulated headless screws (SpeedTip CCS Screws, Medartis, Basel Switzerland) were used. In the two multi-fragmentary fractures 2 screws were used, in one case two times a 2.2 mm diameter screw, in the other one a 2.2 mm and a 1.8 mm screw. For one finger a screw by another manufacturer was inserted (HBS mini screw, KLS martin, Tuttlingen, Germany).

Post-operative follow-up

The initial clinical and radiological follow-up took place at 6 and 12 weeks postoperatively. Early active mobilisation was started within one week after surgery. The long-term follow-up took place in an average of 3,4 years (range 37–52 months) after the last surgery.

Statistical analysis

Descriptive statistics included frequencies for categorical variables. Means and ranges were reported for continuously coded variables. R software environment for statistical computing and graphics (version 3.4.3) was used for all statistical analyses.

Results

Patient characteristics

Three women and five men with a mean age of 45 years were included (range 18–64 years). Five fractures were on the non-dominant hand and four on the dominant. One fracture was on the thumb, one on the index finger, one on the middle finger, three on the ring fingers and three on the little fingers. There were three open fractures with concomitant extensor tendon lesions in zone 4. In one of these open fractures the two digital arteries were damaged as well, and the finger needed to be revascularized during the surgery. Two fractures were multi-fragmentary, and one of these two intraarticular. All other fractures were simple, extraarticular dislocated fractures. One patient with a fracture on the ring finger had additional fractures on the adjacent metacarpal IV and V as well as metacarpal fractures on the other hand. The patient with the index finger which needed to be re-vascularized had another open proximal phalanx fracture on the adjacent middle finger. The remaining 6 patients had isolated proximal phalanx fractures.

Complications

Among the patients included in the study, one patient with a proximal phalanx fracture of the ring finger and adjacent metacarpal IV and V fracture experienced a flexion contracture of the PIP IV joint and malrotation, which required revision surgery two years after the initial operation. Another patient with a multi-fragmentary fracture suffered an early dislocation of the fracture and underwent revision surgery after 4 weeks. No other major or minor complications were observed.

Michigan hand questionnaire

The Michigan Hand Score average was 83% (range 43%-100%). The score average for activities of daily living (ADLs) was 83% (range 40%-100%), for work performance 83% (25%-100%), for pain 81% (range 70%-100%), and for aesthetics 75% (range 12.5%-100%). Regarding the overall performance of the hand the score average was 83% (range 40%-100%) with a satisfaction score average of 86% (range 29%-100%).

Total active range of motion and grip strength

Clinical examination at the 3-year follow-up showed a mean total active range of motion (TAM) of 216.4º (range 95º-285º), compared to 185.1° 3 months postoperatively. In the thumb a TAM of 95º was observed 3 years postoperatively and 75° 3 months postoperatively. The detailed results of the TAM results can be seen in (Tables 1 and 2).

Table 1 TAM results 3 months postoperatively
Full size table
Table 2 TAM results 3 years postoperatively
Full size table

Statistical analysis revealed a significant improvement in TAM values over the course of three years post-surgery, with a p-value < 0.05. (Figs. 1, 2, 3, 4 and 5).

Fig. 1
figure 1

Statistical analysis of the TAM results

Full size image
Fig. 2
figure 2

A postoperative X-ray control of the right V finger, B-C X-ray and tomography control in 2019 demonstrating light arthrosis due to screw protrusion into the joint

Full size image
Fig. 3
figure 3

A postoperative X-ray control of the right V finger. B-C X-ray and tomography control in 2019 demonstrating no arthrosis and bone union

Full size image
Fig. 4
figure 4

A postoperative X-ray control of the left IV finger. B-C X-ray and tomography control in 2019 demonstrating no arthrosis and bone union

Full size image
Fig. 5
figure 5

A postoperative X-ray control of the left II and III finger. B-C X-ray and tomography control in 2019 demonstrating no arthrosis and bone union

Full size image

We found no extension lag in the MCP joints at either time point. The average extension lag for the PIP joint reduced from 17° 3 months postoperatively to 6° at the 3-year follow-up.

The mean Jamar grip strength in the injured hand was 29.9 kg (range 19 kg-48 kg). In the non-injured hand, the mean grip strength was 35.8 kg (range 29 kg-40 kg). The injured non-dominant thumb pinch strength was 3 kg. The dominant thumb had pinch strength of 5 kg.

Fracture consolidation

The follow-up x-rays took place on average 3.4 years after the last surgery (range 37–52 months). Clinically all the nine fractures were healed in the follow-up consultation. Radiologically the fracture gap was still slightly visible in the tomosynthesis picture in 2 patients, but not in the normal x-ray. There was no secondary screw dislocation or break of any screws. In one case, the screw protruded into the joint, due to incorrect placement from the beginning (Figure S2). In all the other cases there was no sign of, arthritis in the joints where the screws were inserted. Figures S3-5 show examples of the x-rays and tomosynthesis pictures.

Discussion

The treatment of unstable proximal phalanx fractures requiring surgery can be difficult [2]. The main purpose of surgery in these cases is to reduce the fracture and render the result stable enough to allow for early active mobilization in order to protect from stiffness in the adjacent joints [4]. The most commonly used techniques for fracture fixation are the percutaneous insertion of K-wires or open reduction and internal fixation with screws or plates [3]. Open reduction and internal fixation (ORIF) is associated with extensive soft tissue dissection, which may lead to a restricted range of motion. Moreover, in many cases, the osteosynthetic material used in ORIF needs to be removed at a later stage [21]. The overall complication rate remains high with up to 52% [22].

Percutaneous pinning with K-wires on the other hand causes less tissue damage but often times requires some sort of immobilization which again can bears the risk of causing stiffness, and pin infection is a frequent complication [4, 5].

Intramedullary fixation with compression screws offers another fracture fixation method. Giesen et al. showed that the short-term results for proximal phalanx fracture fixation with this method are very promising [16]. Based on those results we decided look at the long-term outcome after such treatment.

Our study yielded several noteworthy findings:

We found fracture consolidation in all of our cases, whereas the literature indicates that nonunion remains a common adverse event after k-wire pinning or lag-screw fixation of proximal phalanx fractures [4, 5, 23]. The literature suggests excellent rates of radiographic union in proximal phalanx fractures treated with intramedullary screws [11, 14, 24, 25]. In our group full radiological consolidation was shown in 7 cases, in 2 cases the fracture line was partially visible on digital tomosynthesis. This does not imply that the fractures were not healed, but rather highlights the superiority of digital tomosynthesis over digital radiography in visualizing fracture healing in the hand, especially in the presence of orthopaedic hardware [26].

We found a satisfying average score of 83 for the Michigan Hand questionnaire. The MHQ is a well-established instrument for measuring patient reported outcomes related to symptoms, function, aesthetics and satisfaction [18, 27]. Nolte et al. described normative values for the MHQ in patients with and without hand conditions, and found an average value of 95.2 in a population of 247 healthy individuals and 88.2 in a group of 266 patients with a hand or wrist condition [28]. To our knowledge there is no study describing normative values for the MHQ for finger fractures. Most studies on the operative techniques for proximal phalanx fractures fail to report patient reported outcome measures or use other tools than the MHQ. Gao et al. reported a mean MHQ of 91.0 ± 9.44 one year after surgery, for patients with open proximal or middle phalangeal shaft fractures treated with K-wire fixation [29]. Our much lower average value might be partly explained by a particularly low mean MHQ of 43 in the patient who suffered multiple fractures on both hands in the context of a polytraumatic injury after a motorbike accident, and another very low value of 63 in the patient with the fracture on the thumb.

Complications requiring secondary surgery were seen in two of our cases. In one case of a multi-fragmentary fracture early screw dislocation occured, necessitating revision surgery after 2 weeks. In another case, a polytrauma patient with an adjacent metacarpal fracture developed malrotation and stiffness, requiring corrective osteotomy surgery after 2 years. No other complications were observed in our series, and no implants needed to be removed. Compared to an overall complication rate after plate and screw fixation reaching up to 52%, intramedullary screw fixation thus still shows satisfying results, if applied correctly [22, 30].

We could observe an average TAM of 216.4° at the follow-up visit 3 years post surgery. Slightly better results were seen by Giesen et al. and Gaspar et al. with 222° and 257° respectively [14, 16]. It should be noted that one of the fingers was a thumb, which naturally has a lower TAM. Moreover a significant improvement in the TAM could be found comparing the short-term and long-term results in our patient population.

Faruqui et al. found TAM values of 201° for extra-articular proximal phalanx fractures treated with trans-articular or extra-articular cross-pinning and an average of 27° of flexion loss at the PIP joint [23]. In comparison Brei-Thoma et al. reported a TAM of 205° for fractures fixated with plates, and 227° after plate removal [21]. For fractures treated with intramedullary screw fixation a TAM of 246° was found. The average extension lag at the PIP joint was 18.9° for plate fixation and 5.9° for intramedullary screw fixation.

Whilst the average extension lag in our population at 3 months postoperatively was 17.4° in the PIP joints, it reduced to 6.3° at 3 years postoperatively. This result suggests a favourable outcome for this fracture fixation method compared to plate fixation with PIP extension lag in up to 67% of the fractured fingers [21].

Regarding grip strength, we found a mean of 29.9 kg in the injured hand, which we considered a fairly good outcome. Brei-Thoma et al. reported similar results, with 33.2 kg for fractures treated with intramedullary screw fixation and 33.1 kg for plate fixation [21]. Horton et al. similarly found a mean grip strength of 30 kg in patients treated with Kirschner-wire fixation and 35 kg in fractures treated with lag screw fixation [5]. Given that the gender, age and profession of the patients were not accounted for in this analysis, direct comparisons are difficult. Nonetheless, it can be stated that intramedullary screw fixation yields similar grip strength results to the commonly used fixation methods.

Although the use of headless intramedullary screw fixation has become more popular recently, a frequent point of debate is the cartilage damage at the intraarticular entry point. The damaged to the articular surface of the MCP and PIP joints in proximal phalanx fractures appears to be minimal with the intramedullary fixation technique [13]. In our series, there was one case where a screw protruded into the joint, resulting in joint arthrosis. This complication, caused by incorrect screw placement, can be easily avoided intraoperatively by ensuring the screw head is fully countersunk.

Pogetti et al. reported that metacarpal fractures treated with intramedullary headless screw fixation showed no signs of osteoarthritis postoperatively [31]. Similarly, Jovanovic et al. found no features of osteoarthritis in phalanx fractures treated with intramedullary fixation after a 4.1 year follow-up although the study did not specify the nature or location of the fractures [15]. In our series, we observed no signs of arthritis in the joints adjacent to the fractures, except for the previously mentioned case.

Our results are further supported by Martini et al., who demonstrated that digital tomosynthesis is superior to conventional radiography in detecting osteoarthritic changes in small joints of the hand [32].

This study has several limitations. Primarily the cohort size is too small to draw general conclusions. Aditionally, inherent biases are associated with the retrospective design of the study. We chose a follow-up period of 3 years for our long-term result. While this period seems reasonable, there is still the possibility of complications occurring later.

We conclude that the use of intramedullary headless screw fixation for proximal phalanx fractures shows good clinical outcomes at 3 years postoperatively. The fixation is stable enough to allow for early mobilization, and minimizes the risk of adhesions. Patient satisfaction is high, with an early return to work, a low implant removal rate and excellent results regarding cartilage damage and osteoarthritis.