Dynamic, Over-Valgus Correction Without Osteotomy for Nonunion of Subtrochanteric Hip Fractures Using a Dynamic Hip Screw ^†

Featured Application

This work describes in detail a novel technique quite useful for the nonunion of subtrochanteric hip fractures.

Abstract

Subtrochanteric nonunion is especially challenging. Extramedullary fixation using fixed-angle devices with a closing lateral wedge osteotomy is the standard surgical procedure for restoring the hip’s normal alignment and achieving bone union. However, this procedure is technically demanding and relies on devices that facilitate bone union in a non-dynamic manner, along with the limitations that this type of fixation entails, especially in this highly biomechanically stressed area. This paper aims to describe a novel surgical technique to heal subtrochanteric nonunion based on dynamic fixation performed through an over-valgus correction of the hip and fixed with a dynamic hip screw (DHS). Between March 2022 and July 2023, five patients diagnosed with nonunion of a subtrochanteric fracture were operated on by a single surgeon and followed prospectively. The average age of the patients was 64 (range: 34–85). The mean duration of surgery was 112 min (range: 63–153). The femoral neck angle before surgery was 120° (range: 110°–122°) and 147° (range: 142°–150°) after surgery. The mean leg length increased by 8 mm (range: 6–12). The Harris hip score improved from 38 points (range: 30–44) to 90 points (range: 88–96), corresponding to excellent or good results. All patients healed without major complications and were asymptomatic after 12 months of follow-up. In conclusion, over-valgus correction using a DHS is a novel technique that seems helpful for subtrochanteric nonunion. It allows for dynamic fixation, offering the advantages of dynamic fixation, especially in this high-stress area. It is also performed without osteotomy, making it a less demanding technique than the current methods described.

1. Introduction

Subtrochanteric hip fractures account for 5% to 34% of all proximal femoral fractures. [1]. The subtrochanteric region is located between the lesser trochanter of the femur and extends approximately 5 cm below it [2]. It experiences the highest level of mechanical stress in the human skeleton. In this region, the stress distribution is unequal; the medial cortex experiences about 20% more compressive stress (up to 1200 pounds/square inch) than the tensile stress encountered by the lateral cortex, leading to immense stress in this area [3]. The standard treatment for subtrochanteric fractures is intramedullary nailing [2]. Due to the region’s anatomy, fractures are often not appropriately reduced in a significant number of cases, which can result in residual displacement (gaps between the ends of the bone) and leave the proximal fragment of the femur in a varus position. Given the high bending forces present in this area, varus malalignment is considered the most critical factor contributing to nonunion [4].

Poor vascularization in this anatomical zone also contributes to the nonunion of these types of fractures. Nevertheless, the advantages of achieving an anatomic bone reduction, or at least restoring a valgus alignment of the fracture, are likely more significant than the potential risks of disrupting the soft tissue and periosteal blood supply during an open surgical approach. Therefore, an anatomic fracture reduction, or at least restoring hip valgus through an open approach, is preferable to achieving varus malalignment through a percutaneous approach [2,4].

The reported nonunion rate for these types of fractures ranges from 4% to 20% [5,6,7]. The current approach to treating subtrochanteric nonunion is surgical; exchanging nailing is the most widely used technique [2]. It is mandatory to perform nailing in a static mode (which involves both proximal and distal static locking of the new nail) in conjunction with over-reaming of the canal. This technique has reported good outcomes [8,9,10,11]. However, a valgus restoration of the hip angle becomes essential for achieving bone union in subtrochanteric nonunion when residual varus malalignment is present, and achieving proper valgus hip angle correction can be complex (requiring osteotomies of the fracture) or simply impossible with intramedullary fixation [12].

Open reduction using an extramedullary device and a lateral closing osteotomy appears to be the preferable surgical procedure for restoring a physiological valgus alignment of the hip in a subtrochanteric nonunion in varus [13,14,15]. In most cases, anatomical reduction of the bone and rigid fixation (needed to achieve absolute stability and primary bone union) can be quite challenging. Bone defects, technical intraoperative limitations, or the performance of the osteotomy required to restore the physiological valgus of the hip make this optimal surgical fixation impossible, and it is, in fact, poorly documented in the literature [16].

Therefore, secondary bone healing using extramedullary fixation with fixed-angle devices (such as condylar blade plates, condylar locking plates, and 95° dynamic condylar screw plates) following a valgus lateral closed osteotomy is actually the standard surgical procedure aimed at correcting varus deformities and achieving bone union [8,9,14,15,16,17]. None of these allow for dynamic fixation. Condylar blade plates and condylar locking plates are static devices that do not enable interfragmentary dynamic compression of the bone. In contrast, the 95° dynamic condylar screw plate is a dynamic device; however, in subtrochanteric nonunions, the proximal fragment is fixed in a manner that does not follow the anatomical direction of the neck of the hip, making sliding between the fragments impossible as well.

In short, to date, surgical procedures for subtrochanteric nonunion rely on methods that achieve bone union based on secondary bone healing through static fixation, either intramedullary nails or extramedullary devices. Non-dynamic surgical procedures in the subtrochanteric area, the most stressed zone of the human skeleton, entail specific issues. The stability of the nonunion site is primarily achieved through the device (not the bone). Because of this, in addition to the intrinsic characteristics of nonunion (bone defects and previous surgical procedures that can reduce the contact area between fragments), only a minimal amount of beneficial loading forces (compression forces) are directly conducted to the bone. These forces are essential for increasing stability at the nonunion site (decreasing the strain at the fracture site) to achieve bone union and reduce biomechanical stress on the device [18]. They are especially relevant in this area because high bending varus shear forces can cause significant strain in the interfragmentary tissue, promoting nonunion. In addition, the device is under high stress and can break or fail if the bone union is not achieved in a timely manner. Although the overall union rate is reported to be between 71% and 98% [1,8,17], recent studies indicate a failure rate of up to 29% [10]. Furthermore, these procedures are challenging. Full weight bearing is not immediately allowed postoperatively (a devastating circumstance for elderly patients) [7], and bone grafting is often necessary to promote bone union [6,7,9,13].

Dynamic fixation outperforms nondynamic fixation and will overcome these problems in this highly stressed zone. Dynamic fixation progressively results in greater compression forces, reinforcing the stabilization of the nonunion directly through the bone contact areas. A dynamic device exerts continuous compression forces, enhancing the stability of the bone fragments at the nonunion site and promoting quicker healing without the need for bone grafting. Additionally, it decreases stress on the device (increasing the time to achieve bone union) and allows for full weight-bearing from the first postoperative day.

The DHS is the only device that can achieve dynamic compression at the nonunion site. However, due to its design, it cannot properly conduct the vertical physiological vector of the loading forces predominant in this area and produces excessive medialization of the shaft (especially if there is no intact lateral cortex), increasing the possibility of loss of reduction and, consequently, nonunion [19]. For the sliding hip screw to function according to its biomechanical principle, an intact lateral cortex is mandatory [19]. The lateral cortex is always damaged in subtrochanteric nonunions, so the use of the DHS in subtrochanteric fractures has reported a high rate of complications, making it unsuitable for these injuries. Furthermore, in subtrochanteric fractures or nonunions, the screw insertion site is mostly at or just near the fracture site. It cannot maintain a proper hold in the lateral cortex. Due to this, this principle has not been successfully applied to subtrochanteric injuries.

We propose that achieving some degree of dynamic fixation in the subtrochanteric zone using a DHS would be feasible, avoiding excessive medialization of the shaft. This may be accomplished through a valgus overcorrection of the hip angle applied to the proximal fragment. In addition, it can be performed without the need for a lateral closed osteotomy by simply decompressing the nonunion site. This paper aims to describe a new surgical technique for addressing dynamic fixation in subtrochanteric nonunion in varus through over-valgus correction fixed with a DHS and to evaluate its radiological and clinical outcomes in actual patients.

2. Material and Patients

Between March 2022 and July 2023, five patients diagnosed with the nonunion of a subtrochanteric fracture were operated on by a single surgeon [ADDM] and followed prospectively. These patients presented with pain at the fracture site, a varus angle in the hip, and established nonunion of the fracture. All patients had previously undergone surgery using a cephalo-medullary nail and were severely limited in their activities of daily living. The youngest patient was 34 years old and had previously undergone several surgeries involving nail exchanges.

The exclusion criteria included (1) deep infection, (2) previous non-ambulatory patients, and (3) a concurrent ipsilateral limb injury.

All subjects involved in the study gave informed written consent. The study was conducted following the Declaration of Helsinki and was approved by the Institutional Review Board (or Ethics Committee) of the Hospital Universitario de Jaén (protocol code INV-2022-023, approved on 4 February 2022).

The average age of the patients was 64 (range: 34–85). Three patients had the right lower limb involved, while two were affected on the left side. Only one patient sustained a fracture from a high-velocity motor vehicle accident, whereas the others experienced fractures due to simple falls.

Preoperative radiographic examinations included anteroposterior and lateral view X-rays of the affected hip and telemetry of the lower limbs. Measurements included residual hip varus, limb length discrepancy, the presurgical offset of the hip, and the contralateral cervical–diaphyseal angle (which was assessed to determine the normal angle in the affected limb), among others. The preoperative dysfunction of the affected hip was evaluated using the Harris Hip Score (HHS), which was documented accordingly. A summary of the patients’ baseline data can be found in

Table 1. Baseline demographics of the patients and preoperative radiologic and functional parameters.

PATIENTS	1	2	3	4	5	Mean
Age *	34	54	85	69	78	64 (34–85)
Sex	Male	Female	Female	Female	Female	--
Time from fracture to nonunion diagnosis (months) *	40	12	12	10	13	17.4 (40–10)
Affected hip	Left	Left	Right	Right	Right	--
Injury mechanism	Vehicle accident	Simple fall	Simple fall	Simple fall	Simple fall	--
Previous implant	Long intramedullary nail $	Short intramedullary nail	Short intramedullary nail	Short intramedullary nail	Long intramedullary nail &	--
Femoral neck angle pre-surgery *	122°	128°	125°	110°	115°	120° (110–122)
Pre-surgery offset (mm) *	59	42	50	54	47	50.4 (42–59)
Vertical distance nonunion-head pre-surgery (mm) *	57	43	32	46	40	43.6 (32–57)
Limb length discrepancy (mm) *	12	8	9	4	7	8 (4–12)
Harris Hip Score pre-surgery *	30	42	40	44	34	38 (30–44)

*: Parameters were measured and presented with the mean and range in parenthesis. ^$: First-generation nail. ^&: Third-generation nail (cephalo-medullary nail).

.

3. Key Points of the Surgical Technique

Before describing the surgical procedure, the philosophy of our new technique must be understood. Because of the eccentric mechanical axis of the femur, high bending shear forces in subtrochanteric injuries cause the proximal fragment to deviate in varus. Thus, high bending stress is created medially and tensile forces laterally, increasing the strain at the fracture nonunion (Figure 1A). After the decompaction of the fragments (without removing the fibrous tissue from the nonunion site), the head screw is introduced into the undamaged lateral trochanteric cortex of the proximal fragment (Figure 1B). Posteriorly, the plate is applied to the head screw and secured to the shaft to achieve valgization of the proximal fragment (Figure 1C). Over-valgization of the proximal fragment is the key point of our novel technique. It implies a verticalization of the proximal fragment. This way, the lateral cortex is positioned against the medial area of the distal fragment, producing a lateralization of the shaft (Figure 1C). The DHS will bear strong weight-bearing loads, and due to its dynamic behavior, a progressive medialization of the previously lateralized shaft will occur, resulting in progressive compaction between fragments. Thus, the main drawback of the DHS, which is the medialization of the diaphysis, is now one of the keys to the technique’s success (Figure 1D). This zone of bone contact is subjected to strong loads, creating firm, progressive contact between the fragments. It will behave as a true bone stop that prevents excessive medialization of the diaphysis. Due to the verticalization of the proximal fragment, load transmission occurs more vertically, which is the dominant form of transmission in this subtrochanteric zone. Moreover, due to the overcorrection of the hip valgus, which shifts the mechanical axis of the femur laterally, shear forces are transformed into compressive forces. Thus, even through this small area, a powerful bony bridge is created that adequately conducts the loads to the diaphysis and marks the beginning of an early and significant bony callus.

4. Preoperative Planning

The goal of the surgical technique was to overcorrect the valgus hip angle to approximately 150° by valgizing the proximal fragment and achieving dynamic fixation at the nonunion site using a DHS.

If possible, a 135° DHS was the preferred surgical option. Due to its acute angle, the head screw was inserted low and through intact bone at the femoral head, considering the position of the previous implant device (assuming it had been appropriately placed at the center of the femoral head). When a 135° dynamic hip screw (DHS) was utilized, achieving an overcorrection of 150° at the hip presented more technical challenges, necessitating comprehensive radiological preoperative planning. The wire for the head screw needed to be inserted at a 15° varus angle to the neck angle. This additional 15°, combined with the 135° of the device, resulted in a final valgus angle of 150° at the hip (Figure 2).

If a 150° dynamic hip screw was used, the guide for the wire of the head screw was inserted at the center of the femoral head, following the same varus neck angle of the neglected hip. This approach eliminated the need to calculate valgus correction beforehand, simplifying the surgical procedure. As a result, a 150° hip angle was restored successfully (Figure 3).

5. Surgical Procedure

Surgery was performed on a traction table, and preoperative cefazolin (2 g) was used, with standard skin preparation done as usual. All patients underwent open reduction through a standard lateral approach (subvastus approach). Using standard methods, the first step was to withdraw the previous material (e.g., broken nails). Second, three samples of the nonunion site were taken for microbiological study to rule out infection. Third, the void in the head caused by previous fixation material was filled with an allogenic bone graft (40 g of cancellous bone from the tissue bank), which was impacted into the head to secure a better fixation device. Fourth, the wire for the head screw was inserted at the planned degrees of varus relative to the neck, regarding the degrees of the DHS used. Subsequently, the head screw was introduced.

Fifth, decompaction of the nonunion site was performed. It is paramount not to damage vascularization at this site while trying to be as minimally invasive as possible. For this purpose, a chisel was inserted through the nonunion site. Opening the medial site is critical for later valgus correction. No bone graft was added to the nonunion site.

Sixth, reduction was achieved by gathering the plate to the femoral shaft. Care should be taken to do this slowly and progressively to avoid intraoperative cut-out. The wound was routinely closed. No bone graft was added to the nonunion site. Finally, the plate was distally fixed. The whole surgical procedure is shown in Figure 4.

All patients were permitted to bear full weight the day after the correction was verified. Functional outcome measures and the same preoperative radiological assessments were conducted at the final follow-up. Patients were assessed at 3, 6, and 12 months post-surgery. The primary outcome measured was the Harris Hip Score (HHS) at the 12-month follow-up.

Complications were categorized as major when unplanned surgical intervention was required or substantial functional problems developed, such as implant failure, deep infection, and nonunion. Minor complications did not require surgical intervention or lead to significant final functional problems (temporary nerve palsies, superficial infections, etc.).

6. Results

All patients were prospectively followed for at least 12 months. The mean duration of surgery was 112.8 min (range 63–153). The mean femoral neck angle before surgery was 120° (range 110–122) and 146.8° (range 142–150) postoperatively. The DHS of 150° was the most widely used device (3/5). The median offset was 50.4 mm (range 42–59) before surgery and 19.6 mm (range 17–24) after surgery. Leg lengthening increased to 8 mm (range 6–12). The Harris Hip Score improved from a preoperative mean of 38 (range 30–44) to 90 (range 88–96). There were no major complications. In one patient, there was a suspicion of a cut-out in the X-rays, but the CT scan showed that the head screw was just subchondral and showed no invasion of the joint cavity. All patients were asymptomatic at the last follow-up and had a mean bone union healing time of 5 months (range 4–6) (Figure 5 and Figure 6). The results are summarized in

Table 2. Intraoperative parameters and postoperative radiological and clinical results.

Patients	1	2	3	4	5	Mean/Median
DHS degrees used $	135°	150°	150°	135°	150°	150°
Duration of surgery (min) *	120	63	123	102	156	112.8 (63–153)
Femoral neck angle post-surgery *	142°	150°	150°	147°	145°	146.8 (142–150)
Post-surgery offset (mm) *	24	19	18	20	17	19.6 (17–24)
Vertical distance nonunion–head post-surgery (mm) *	65	55	38	52	48	51.6 (38–65)
Postoperative leg lengthening (mm) *	8	12	6	6	8	8 (6–12)
Time to union (months)	4	6	6	5	4	5 (4–6)
Harris Hip Score post-surgery *	96	96	88	90	88	90 (88–96)

*: Results are presented as a mean and range in parenthesis. ^$: Results are presented as median.

.

7. Discussion

Varus malalignment in subtrochanteric fractures is the leading factor in nonunion, often resulting from technical errors during osteosynthesis. The fracture site endures substantial shear forces in both the medial and lateral zones, which impede the healing process [20]. In fact, residual varus is frequently observed at the nonunion site, particularly if the fixation devices fail.

The approach for healing subtrochanteric nonunion has been based on static fixation. The most commonly used surgical procedure to achieve bone union is the exchange of a nail for a newly over-reamed nail [8,9,10]. Nailing has shown good outcomes, especially without residual varus [8,9,21]. A subtrochanteric nonunion in varus is challenging to manage using a nail and may even be impossible. Barquet et al. [12] have reported that, in some instances, an osteotomy through the fracture site was necessary to correct varus malalignment. This is a challenging surgical procedure to perform, and despite this, they report a failure rate of up to 13% when using a long intramedullary nail [12]. Open reduction using an extramedullary device and a lateral closing osteotomy appears to be the preferable surgical procedure for restoring a physiological valgus alignment of the hip in a subtrochanteric nonunion in varus [13,14,15]. However, Loztien et al. [13] recently reported an even higher rate of complications concerning this type of static fixation [13]. A total of 11 of the 40 (27.5%) patients required secondary revision surgery; one patient had a persistent nonunion, nine patients had persistent nonunions leading to hardware failure, and one patient had a peri-implant fracture due to low-energy trauma four days after the index surgery. Although static fixation using the relative stability philosophy has shown good results in other anatomical areas, the results remain suboptimal in the subtrochanteric area. Due to the difficulty of achieving primary bone healing (through rigid fixation) in a previously operated area and such a biomechanically stressed area, dynamic fixation would be desirable. However, to date, no surgical procedures have been described as able to achieve this.

Our preliminary results show that dynamic compression can be successfully applied to treat subtrochanteric nonunion, offering several relevant advantages in comparison to static fixation in this specific area, such as the following:

(a)

The stability of the nonunion site is primarily achieved through the bone, avoiding stress on the device. If the bone does not achieve union in a timely manner, there is more time before its breakage. Conventional techniques have reported up to 13% implant failures [13]. Also, using additional plates is a widespread practice and aims to increase the stability of the nonunion site [13,19]. Although our number of patients is small, no implant failures have occurred in our study. Dynamic fixation secures robust stability between the fragments. In addition, it is implemented by improving load transmission (over valgus correction), making the implementation of the DHS with other devices unnecessary.

(b)

Immediate full weight bearing is allowed from the first postoperative day. In previous procedures, full weight bearing is often limited to weeks, along with the limitations this entails, especially for the elderly population.

(c)

Grafting of the nonunion site is never required. Bone grafting is necessary in most procedures to promote bone union, which adds comorbidity for the patient. A bone union rate of up to 100% using extramedullary devices that involve bone grafting has been reported [15]. The same 100% bone union rate was obtained in our series. Even expensive, non-risk-free therapies, such as teriparatide or recombinant human bone morphogenetic protein (BMP), have also been used to aid fixation, with promising results [22,23]. Dynamic compression offers an excellent biomechanical environment that promotes quick and significant callus formation, making them unnecessary.

Our technique is friendlier to perform than the current surgical procedures described to date using extramedullary devices. Achieving the desired intraoperative close bone contact to achieve intraoperative compression using non-dynamic fixation is challenging. In our technique, close bone contact does not necessarily need to be achieved intraoperatively. Even if no bone contact is achieved intraoperatively (due to nonunion patterns like bone defects, comminution, or technical aspects), strong areas of bone contact will appear when walking due to the dynamic construct created. As described above, this results from the shaft’s medialization conducted through the DHS (Figure 7).

Only a tiny area of bone contact between fragments is necessary to develop a large bone callus (Figure 8 and Figure 9). Dynamic fixation reinforces a progressive and strong (due to this highly stressed area) contact between the ends of the bone fragments. In other cases, a wide bone contact area was achieved intraoperatively (Figure 10).

Whatever the case, only a tiny medialization of the shaft occurred due to the strong bony contact at the nonunion site.

The literature reports that a closing lateral wedge osteotomy must be performed to restore the hip valgus in varus subtrochanteric nonunion. This surgical procedure is laborious. Our technique is more straightforward to perform, and restoring hip valgus can be achieved through a simple decompaction of the nonunion site. In previous studies, the mean valgus angle obtained is 128° in some studies [13] and 127° in others [24]. A mean of 147° is reported in our study. This is because our technique is different and is performed with the objective of over-valgizing the hip to allow successful dynamic fixation. Over-valgizing the hip angle has a theoretical drawback; the hip offset notably decreases. Decreasing the hip offset is a concern. A reduced offset is a predisposing factor for developing osteoarthritis. Changes in hip offset have not been reported in studies we have revised regarding the current techniques. Still, theoretically, this parameter should not be altered; if it is, it should only be very subtle. As the collected postoperative parameters demonstrated, our surgical technique shows a notable decrease in this parameter. The median offset was 50.4 mm (range 42–59) before surgery and 19.6 mm (range 17–24) after surgery. This is especially relevant for young patients; therefore, this technique is not desired as the first option for this population. Despite this, the main objective when operating on a subtrochanteric nonunion is to achieve bone union, and we believe it provides more benefits than drawbacks, as was the case with our youngest patient, who underwent surgery up to three times prior.

The use of a dynamic hip screw for trochanteric nonunion was first reported in 2011 [24] and recently reproduced with promising results [25,26]. The trochanteric area is quite different from the subtrochanteric area. The physiological bone distribution forces are different; it is a less stressed and well-vascularized zone compared to the subtrochanteric area. Additionally, dynamic compression fixation is more straightforward to perform with a DHS.

Over-valgus corrections of the hip have also been reported in recent papers. El Alfy et al. [27] performed a valgus osteotomy, overcorrecting the valgus of the hip and fixing it with a molded 95° dynamic condylar screw. This technique does not allow for dynamic fixation; only static compression between fragments at the fracture site is achieved intraoperatively, and iliac autograft was needed in eight cases. Our technique is quite different and less demanding. Our valgus correction is gross, and the bone union is promoted through dynamic fixation using DHS. In contrast, the fibrous tissues at the nonunion site are unnecessary to remove, avoiding extra surgical procedures that can worsen the vascular deficit in this area, and no bone grafting is needed. They report that the preoperative mean HHS was 40, which improved to 85 postoperatively, with most patients reporting good or excellent results. They achieved a mean of 1.5 cm of postoperative lengthening. Though our series is much smaller, similar functional outcomes are reported with our technique. The Harris Hip Score improved from a preoperative mean of 38 (range 30–44) to 90 (range 88–96) postoperatively; all corresponded to good or excellent results. Additionally, our patients had a similar mean leg lengthening of 8 mm (range 6–12).

In our series, the mean time for union time was five months, and no major complications were reported. They are similar results to the studies reported in the literature [12,27,28]. Our results are compared with some in the literature, where the surgical procedure has been performed like ours using an extramedullary device (

Table 3. Comparison of our study with those in the literature.

Study	N° of Patients	Design	Device Used	Follow Up	Results
Vaishya et al. [17]	12	Retrospective	A reversed DF-LCP with bone grafting	13 months	100% bone union rate at a mean of 9.5 months. No major complications.
Lotzien et al. [13]	40	Retrospective	Dynamic condylar screw plating along with bone grafting	26 months	93% bone union rate at 26 months. Thirteen patients required a second surgical procedure due to persistent nonunion (n: 1), persistent nonunion leading to implant failure (n: 9), deep infection (n: 2), and periprosthetic fracture (n: 1). Ultimately, 93% (n: 37) achieved union at a mean of 12 months.
De vries et al. [6]	33	Retrospective	A blade plate, along with cancellous bone grafting	31 months	97% bone union rate at a mean of 5 months. Complications included osteonecrosis of the femoral head (3%), refracture (3%), blade plate protrusion (3%), deep infection (3%), femoral neck fracture with fracture of blade plate (3%).
Giannoudis et al. [9]	11	Retrospective	A blade plate, along with cancellous bone grafting	26 months	100% bone union rate at 7 months (5 to 12). The blade plate failure was 7%.
Alfy et al. [27]	26	Prospective	95° dynamic condylar screw	36 months	There was only one nonunion (1/26). The mean HHS was 40 (range 25 to 65), which improved to 85 (range 55 to 95). All the results were excellent or good, and only one was poor.
Present Study	5	Prospective	135°/150° dynamic hip screw. No bone grafting.	12 months	100% bone union rate (5/5). All excellent and good results. There were no major complications (it was a very small series).

).

The main limitation is the small number of patients treated to date. This nonunion is somewhat unusual, so it is difficult to recruit a large number of patients. Even so, to our knowledge, this is the only prospective study performed on the pseudoarthrosis of subtrochanteric fractures fixed dynamically.

8. Conclusions

A DHS can safely ensure dynamic fixation of subtrochanteric nonunion in varus if an over-valgus correction of the proximal fragment is performed. It can also be performed by avoiding demanding procedures that require a close lateral osteotomy. Based on our promising initial clinical results, although it is a recently developed surgical technique, it seems helpful for this type of injury.