Vancouver B3 periprosthetic fractures of the femur: options for treatment

Abstract

The Vancouver B3 type periprosthetic fracture is a rare but difficult-to-treat complication of hip arthroplasty. Various options for the treatment of Vancouver B3 periprosthetic fractures are described in the medical literature. The problem with many studies is the small number of cases, the fact that the results of Vancouver B2 and B3 fractures are only presented together, the lack of a clear definition of B3 fractures and the resulting subjective categorisation of fractures into the two groups (Vancouver B2 and B3). Therefore, the evidence in the literature for the recommendation of treatment strategies for Vancouver B3 fractures can be classified as poor. This review aims to analyse the various therapeutic options with regard to the current state of knowledge based on reports of case series.

Introduction

Periprosthetic fractures of the femur are relatively rare but difficult to treat complications of total hip replacements. Demographic change, with more and more older patients and the associated increase in hip endoprosthesis implantations, means that the incidence of periprosthetic fractures is increasing too. The incidence of periprosthetic proximal femur fractures is approximately 2.3 ± 0.9 per 100 inhabitants, increased by a factor of 2.3 between 2010 and 2019 and has continued to rise in recent years [1, 2]. Periprosthetic fractures of the proximal femur are now the third most common reason (15.9%) for revision surgery in Germany after loosening (22.7%) and infection (16.4%) [3].

The literature provides a variety of classifications for periprosthetic fractures. The UCS (Unified Classification System) of periprosthetic fractures [4] has included and expanded the generally accepted Vancouver Classification, which was first presented by Duncan et al. [5]. Periprosthetic fractures of the femur with a loose stem and poor bone stock quality are classified as Vancouver B3 and as type IV.3-B3 fracture in the Unified Classification System [4, 5]. Although 75% of postoperative periprosthetic fractures involved loosened prosthetic stems, Vancouver B3 fractures are relatively rare [6]. 43 of 1049 periprosthetic femoral fractures (4.1%) were of type B3 in the Swedish Prosthetic Registry [7]. The one-year mortality rate for these fractures is reported to be around 30% especially in elderly patients over 85 years old [8–11].

The deficient bone status poses an extreme challenge for the surgeon during treatment. However, the bone stock quality for the differentiation between Vancouver type B2 and B3 is not well defined. Most authors classify a fracture as a Vancouver B3 fracture if the proximal femoral bone at the level of the fracture is of poor quality [12, 13]. However, this alone is not decisive for the choice of the revision stem or the fixation method of the new prosthesis. The bone quality of the remaining intact isthmus of the femur and thus the remaining fixation length in the isthmus play a much more decisive role in the choice of the revision stem [14].

Various options for the treatment of Vancouver B3 periprosthetic fractures are reported in the literature. The problem with many studies is the small number of cases, the joint presentation of results of Vancouver B2 and B3 fractures (e.g. Axenhus et al. [15]) even in meta-analyses (e.g. Di Martino et al. [16]), the unclear definition of B3 fractures and the resulting subjective categorisation of fractures into the two groups (Vancouver B2 or B3). Therefore, the evidence for recommending treatment strategies for Vancouver B3 fractures can be classified as weak. The aim of this review is to analyse the various therapeutic options with regard to the current state of knowledge.

Materials and methods

A literature search was conducted in PubMed, Scopus, Embase, and the Cochrane Database of Systematic Reviews databases to identify studies published by June 2025, using the following key terms in association with the Boolean operators “AND”, “OR”: “periprosthetic femoral fracture”, “Vancouver B3”. Clinical studies reporting the patient clinical outcomes, complications, survival rate, and Harris Hip Score (HHS) of Vancouver B3 fractures were included. Studies on animals, biomechanical reports, case reports and reports less than one year of follow-up were excluded. In addition, studies written in non-English languages were excluded.

Results

The initial search produced a result of 3783 articles. All duplicates were removed. The full text was reviewed for the 99 studies included, based on the title and abstract. After evaluating the inclusion and exclusion criteria, 87 studies were included in the final analysis (Fig. 1).

Fig. 1.

Flow-chart of the analysis of the literature

Osteosynthesis

Osteosynthesis of periprosthetic fractures with loosened prostheses has the advantage of generally shorter operating times and less blood loss [3, 16]. Pure osteosynthesis has been reported in a small number of Vancouver B3 fractures cases. Although the fracture may heal, the stem only becomes firm again in exceptional cases, especially if it is a cemented stem with a loosened cement mantle [17]. Kim et al. [18] described a successful osteosynthesis with additional use of proximal strut grafts in one patient, but the cemented stem did not reattach and the patient was only able to bear partial weight on the leg. Lampert et al. [19] reported on 6 cases without detailed information on the results for this group. In a review by Kahn et al. [20], only 28 of 167 B3 fractures underwent osteosynthesis alone. The revision rate of 28.6% was almost twice as high as the 14.4% in the group with revision arthroplasty. Holley et al. [21] saw a significantly higher union rate with stem replacement than with osteosynthesis in 9 patients with Vancouver B3 fractures (once with an uncemented stem and 8 times with a cemented stem). In 12 Vancouver B3 fractures with osteosynthesis alone, De Maio et al. [22] achieved an excellent result only once and a good result 11 times according to Beals and Tower’s [23] (with stable subsidence of the prosthesis or healed fracture with moderate deformity or shortening). Spina et al. [24] even achieved a poor result 3 times with a pure osteosynthesis in 4 Vancouver B3 fractures. However, in a systemic review article by Stoffel et al. [25] comparing osteosynthesis and revision arthroplasty in Vancouver B2 and B3 fractures, it was shown that the common lack of rigorous statistical analyses and significant methodological weaknesses made identification of outcome predictors impossible. Like Powell-Bowns et al. [17], they concluded that ORIF may be a viable option if bone stock is adequate around uncemented or tapered polished stems with an intact cement mantle and the fracture geometry allows stable anatomic reconstruction [17, 25]. Since adequate bone stock is not present in Vancouver B3 fractures and stable anatomic reconstruction of the fragments is extremely difficult to achieve, pure osteosynthesis is therefore only an exception as a surgical option for bedridden patients with a Vancouver B3 fracture in individual cases.

Cemented revision stems

Because of the thin, sclerotic bone bed and associated poor cementing qualities, cemented prosthesis revision stems are not recommended for type B3 fractures by most authors. For adequate cementing, the proximal fracture fragments must be repositioned and the proximal femoral medulla reconstructed. In addition, more or less correct reduction and retention of the fracture fragments is also necessary for the fracture to heal so that as little cement as possible can penetrate into the fracture gap. However, due to the poor bone quality of Vancouver B3 fractures, this is very difficult to achieve and difficult to reproduce [14].

In cemented revision stems, healing of the fracture is mostly impeded by intrusion of cement into the fracture gap [23, 26]. Non-union was found in 31% of revisions with cemented stems by Beals et al. [23] and persistent bone defects in 15%, whereas only 7% of revisions using cementless stems revealed new fractures. Only 60% of 42 patients with type B2 and B3 fractures treated with a cemented long stem had a stable implant with a healed fracture after an average of 68 months, according to Springer et al. [26]. Two patients had revisions because of infection, and six patients had revisions because of loosening, non-union, or dislocation (Fig. 2). There was radiographic evidence of loosening in 18% of the unrevised stems. Tsiridis et al. [27] showed that additional impaction grafting in Vancouver B2 and B3 fractures with cemented stems tends to improve fracture healing. However, Corten et al. [28] report a one-year mortality rate of 43% for cemented revision stems in Vancouver B2 fractures in elderly patients (mean age 82 years), meaning that this risk must also be taken into account for Vancouver B3 fractures, which occur more frequently in older patients. Axenhus et al. [15] found a higher mortality rate for cemented stems in Vancouver B2 and B3 fractures than for cementless stems. In contrast, they found a lower loosening rate and higher fracture healing rate for cemented stems. However, in relation to the topic of the present study (Vancouver B3 fractures), it must be said that only 15% (n = 17) of the cemented stems in this study had a Vancouver B3 fracture. In an isolated study of 54 Vancouver B3 fractures, 33 restorations with cemented stems and impaction grafting showed significantly higher failure and complication rates compared to 21 patients with cementless modular distally fixated revision stems [29]. It can therefore be concluded, that the use of cemented stems in Vancouver B3 fractures should only be performed by surgeons familiar with this technique due to the difficulty of adequate reduction of the fragments and cementing technique, and that additional impaction grafting in these cases is helpful for fracture healing [15].

Fig. 2.

Radiograph of the right hip of a 76 year old female 6 months after stem revision with a cemented long revision stem due to a periprosthetic fracture Vancouver B3. Loosening of the stem is seen by radiolucent lines and cement has intruded into the fracture

Another technique is the so-called hybrid fixation. Here, a revision stem is cemented distally to the fractures in the distal femur if the bone quality is not suitable for cementless fixation. The fracture fragments are then osteosynthesised proximally around the proximal part of the revision stem, which is left cementless, using cerclage. Cementless revision stems (monoblock or modular) can be used for this purpose as an off-label application or cemented distal components can be combined with proximal cementless components (e.g. MP stem, Waldemar Link, Norderstedt, Germany). The advantages of both revision stems (cemented and cementless) are combined here. In the fracture area, cementation is avoided with the disadvantage of poor cementation and, above all, poor fracture healing due to cement leakage into the fracture zone. Distally, the difficulty of cementless fixation in the short residual isthmus and above all thin cortical thickness is avoided by cementing. Only a few individual case descriptions with low case numbers can be found in the literature to date, meaning that this method cannot yet be evaluated. Javar et al. [30] describe 2 cases with good results using a cementless monoblock revision stem.

Cementless revision stems

Cementless revision stems can be divided into proximal and distal fixation stems according to their fixation principle. When using uncemented stems with proximal fixation, the fixation point lies within the fracture zone, which leads to reduced quality of primary stem fixation. Invaco et al. [31] reported stem loosening in a quarter of their cases with this technique, whereas Sledge et al. [32] observed stem subsidence in more than a quarter of their cases.

Cementless stems with distal fixation in the isthmus offer the advantage that their anchoring zone spans the fracture site. As a result, stem stability is independent of the quality of the osteosynthesis, and the fracture can heal without unwanted loading effects. Spina et al. [24] observed an excellent result in 6 of 7 Vancouver B3 fractures with the distal fixating monoblock Wagner revision stem according to Beals and Towers [23].

However, even in periprosthetic fractures, modular revision stems with distal fixation also provide the benefit of enabling the two surgical goals to be addressed separately during revision: secure distal fixation of the implant in the femoral isthmus with fracture bridging by the distal stem component, followed by precise adjustment of leg length, offset, and anteversion in a second step using the proximal component [3, 14, 33].

When using cementless stems for Vancouver B3 fractures, additional impaction grafting for Paprosky II defects [34] or an additional strut graft for Paprosky III and IV defects can help to support fracture healing [35, 36]. Other authors found 100% healing rates of periprosthetic Vancouver B3 fractures with cementless revision stems without strut grafts [37, 38]. As there are no comparative studies with and without strut grafts in comparable surgical techniques and initial situations, it is not possible to evaluate the necessity of the strut grafts.

Taper-fluted stems

If the isthmus remains intact (Paprosky Type IIIA defects), tapered, fluted revision stems achieve good outcomes even in Vancouver B3 fractures, since the isthmus still provides adequate support for distal fixation of these cementless revision stems [12, 13, 34]. The new implant must span the entire fracture with a sufficiently long bridge. Earlier recommendations suggested that the fracture should be bridged by at least twice the stem diameter [5, 39–41]. As a general principle, the necessary bridge length depends on the type of distal stem fixation and the bone quality below the fracture. A cylindrical stem implanted with scratch-fit fixation requires at least 4 cm of fixation length [42]. Because this fixation zone may be located at the stem tip, the stem must be long enough to span the fracture zone and still achieve a minimum of 4 cm of fixation in the isthmus. When using a 2 degree tapered stem with distal cone-in-cone fixation, a minimum fixation length of 3 cm is required [33]. Because this fixation may also occur at the stem tip, the same principle applies as above, but with a 3 cm requirement instead of 4 cm. If a tapered stem with a greater taper angle is used and the femoral isthmus has an intact cortex, the fixation zone may be shorter (at least 2 cm) according to the findings of Tangsataporn et al. [43]. However, the taper angle of these stems means that the fixation zone is not at the tip but proximal to the tip, so the position of the fixation zone together with the more distal portion of the stem determines the minimum bridging length of the fracture with such a stem. An essential prerequisite for using a shorter bridging length is adequate bone stability and quality below the fracture. If this is not the case, bridging of the fracture with a longer stem is required. This longer stem is therefore thinner than the configurations mentioned above, but it is anchored in the same region of the femur as the shorter, thicker prosthetic stems [14]. Particularly in these Vancouver B3 fractures, where only a short or thin cortical bone remains - equivalent to a Paprosky type IV defect [34] - this approach is no longer effective, as adequate cone-in-cone fixation in the isthmus can no longer be achieved. In such cases, alternative strategies must be used. Some authors recommend additional augmentation using a combination of impaction grafting and strut grafting, or a proximal allograft with or without supplementary strut grafts or a long revision stem with additional distal interlocking (see next chapter) [42, 45–48]. However, results for these techniques specifically in Vancouver B3 fractures are rare or even lacking.

With the exception of very few studies [29, 36, 37, 45], all studies in the literature involve mixed cohorts of Vancouver B2 and B3 fractures, which differ in terms of stem subsidence, clinical outcomes, complications, and revision rates [for example 20,43,46–50]. Consequently, it cannot be excluded that some fractures classified as Vancouver B3 in reports using 2° tapered revision stems did not actually present with poor bone in the area of the isthmus of the femur and thereby in the area of fixation. This would, of course, influence the reported results for these stems [14, 20, 49, 50]. This is consistent with the observations of Parry et al. [44], who found a correlation between the occurrence of subsidence when using modular fluted tapered stems with a taper of 2 or 2.5 degrees and Dorr-type bone morphology (significantly more frequent in Dorr type C).

Taper-fluted stems with distal interlocking

In Vancouver B3 fractures with a short sufficient isthmus (Paprosky typ IIIB defect) or a thin cortical bone at the isthmus (Paprosky type IV defect), there is no intact isthmus region of adequate length to achieve fixation, and reliable cone-in-cone anchorage of a taper-fluted cementless revision stem in the isthmus is no longer possible. In this case, improvement of the short cone-in-cone fixation (e.g. Revitan Curved, ZimmerBiomet GmbH, Winterthur, Switzerland) or cone-in-cylinder fixation (e.g. MRP stem, Peter Brehm, Weisendorf, Germany) can be achieved by additional distal interlocking (Fig. 3a-c). As a result, these stems are longer and extended due to the added distal interlocking, and the fracture is bridged over a correspondingly greater length (Fig. 3b, c).

Fig. 3.

a Periprosthetic fracture Vancouver B3 of a 90 year old female. b Radiograph 6 days after stem revision via a transfemoral approach using a curved tapered revision stem with distal interlocking (Revitan, ZimmerBiomed, Winterthur, Switzerland). c Radiograph 2 years postoperative showing osseous consolidation of the fracture and incorporation of the revision stem without subsidence, radiolucent lines or even breakage of the interlocking screws

The 77.3% rate of subsidence observed by Mulay et al. [51] in this type of fracture using the MP stem (Waldemar Link, Hamburg, Germany) may be due to the fact that the MP stem does not allow for additional distal interlocking and therefore requires a minimum fixation zone of 3 cm in the isthmus [14]. Nevertheless, even with distal interlocking, stem anchorage still depends on cone-in-cone fixation within the remaining isthmus of less than 3 cm, supported by the additional stabilization provided by distal interlocking screws below the isthmus. Thus, additional stability is provided until callus formation occurs - resulting from the transfemoral approach (extended trochanteric osteotomy) – and allows bone to grow onto the proximal parts of the newly implanted stem. This type of locking mechanism serves as a means of supporting the short cone-in-cone fixation in the compromised isthmus [14]. In biomechanical studies, Mohamed et al. [50] showed that distal locking screws can significantly enhance both axial and rotational stability of distally fixed stems. However, the fixation of the revision stem cannot depend solely on the distal interlocking screws; a short cone-in-cone fixation within the remaining isthmus is still required [14, 52]. Sole fixation with locking screws alone is insufficient to consistently achieve good outcomes, which was shown by Eingärnter et al. [53], who reported system failure in more than 12% (5 cases) of 41 periprosthetic fractures (2 Vancouver type A, 14 B1, 8 B2, and 13 B3 Vancouver B3 fractures) treated with the low-tapered Bicontact Revision stem (Aesculap, Tuttlingen, Germany) (0.6 degrees of taper) and additional distal locking. Due to the stem’s low taper of 0.6°, fixation depended predominantly on the locking screws. Additionally, poorer outcomes have also been reported for multiple locking with the cylindrical Huckstep stem (with no cone-in-cone fixation). Aspinall et al. [54] presented short-term results for the Huckstep stem in 57 patients with difficult primary implantations and revision procedures with femoral defects. Revision surgery was required in 19.3% of cases, 7% experienced screw fractures indicating stem instability, 3.5% sustained periprosthetic fractures and 8.8% developed periprosthetic infections.

Surgical techniques of revision prostheses in periprosthetic fractures

An analysis of the literature on the treatment of periprosthetic fractures Vancouver types B2 and B3 reveals differences in the surgical techniques. The distally fixed prosthetic stem can be implanted in two different ways.

Technique1

In the first technique, the fracture is initially reducted and the fragments are held in position using cerclage wires or reduction forceps. The prosthetic stem is then implanted endofemorally.

Technique 2

In the second technique, the proximal fragment (usually spiral-shaped) is first exposed up to its tip using a transfemoral approach (extended trochanteric osteotomy). Next, the fixation bed for the revision prosthesis is prepared in the distal fragment under direct visualization. The revision stem is then implanted into the distal fragment (with distal locking if necessary), bridging the fracture site. Finally, the proximal fragments are repositioned around the prosthetic stem and secured with cerclages. For modular stems, the in-situ assembly and osteosynthesis of the proximal fracture fragments around the implanted stem are performed only after a successful trial fitting has been completed.

The disadvantage of the first technique is that the position of the prosthetic stem’s fixation zone cannot be reliably assessed. Consequently, it is not possible to verify whether the fixation zone adequately bridges the fracture. As a result, the fixation zone of the endofemorally implanted stem is likely to lie at least partially within the fracture area, which makes achieving consistent outcomes—particularly in cases with poor bone quality, such as Vancouver B3 fractures—considerably more difficult [41, 55, 56]. The main advantages of the second implantation technique using the transfemoral approach are that in cemented stems the removal of the rest of the cement is easier and the fixation bed of the new prosthesis can be prepared directly under visual control. This is particularly important in Vancouver B3 fractures, where fixation relies on the remaining isthmus and on additional distal interlocking (Paprosky IIIB and IV). This is confirmed by the study of Han et al. [57], who found a significant higher subsidence rate after revision arthroplasty of 21 cemented stems with periprosthetic fractures (compared to 31 cementless stems) using a transfemoral approach only in 2 cases and explained this with the incomplete cement removal in the distal fixation zone of cementless revision stems (because of avoidance of the extended trochanteric osteotomy). Technique two also reduces the risk of additional uncontrolled fractures or perforations [58]. The disadvantage of this technique is that the osteotomy performed during the transfemoral approach enlarges the surgical exposure, creating three fragments from the two originally caused by the fracture. However, studies of the literature show, that this enlargement of the approach was not regarded as a disadvantage and the healing rate of the fracture was not negatively affected by it [59, 60] (Table 1).

Table 1.

Overview of the different surgical techniques and stems for stem replacement in Vancouver B2 and B3 fractures

O’Shea ´05 [55]	10/12	+	2	2	33.7	91%	≈ 75	9.1%	0%	0%	4.5%	0%
Mulay ´05 [51]	10/12	+	2	4	24	91%	69	77.3%	22.7%	4.5%	4.5%	0%
Zaki ‘07 [61]	0/37	-	1	5	168	100%	n.a.	0%	5.4%	0%	0%	0%
Levine ´08 [84]	12/5	+	2	2,4 (in B3)	44.5	100%	n.a.	17.6%	5.8%	5.8%	5.8%	0%
Park ´09 [56]	16/11	in 2	1	4	57.6	92.6%	84.7	7.4%	0%	7.4%	3.7%	7.4%
Fink ´12 [59]	22/10	+	2	4	32.2	100%	81.6	0%	3.4%	0%	0%	0%
Marx ‘12 [85]	15/14	-	1	3	74	100%	69 B2 62 B3	n. a.	6.9%	0%	0%	3.4%
Neumann ‘12 [38]	35/20	in 4	1	4	67	100%	72	3.8%	3.8%	0%	1.9%	4%
Munro ´13 a. ´14 [47]	30 /17	+	2*	4	54	98%	76 W	8.5%	4%	4%	2%	2%
Abdel ‘14 [63]	25/19	(+)	2(*)	4	24	98%	83	4.5%	11%	2.3%	4.5%	2.3%
Amenabar ´15 [64]	66/10	+	2	4	74.4	93.4%	55.1’ 37.4’’	n.a.	5.3%	3.9%	1.3%	6.6%
Da Assuncao ‘15 [65]	31/6	(+)	1	4	35	100%	35 OHS	2.7%	10.8%	2.7%	2.7%	0%
Hernandez-Vaquero ‘15 [66]	7/5	+	2	4	43	100%	78	n.a.	16.7%	8.3%	0%	0%
Rodriguez ‘17 [62]	12/2	+	2	4	40	100%	83.9	7.1%	14.3%	0%	0%	0%
Parry ‘18 [44]	57/6	n.a.	n.a.	4	54	93.6%	84	13%	6.5%	0%	0%	0%
Lizaur-Utrilla ‘19 [68]	25/21 HA 17/14 THA	(+)	2	4	39.6 HA 46.8 THA	97.8% HA 100% THA	MD 6.2 HA 7.3 THA	4.3% HA 3.2% THA	0% HA 9.6% THA	0%	0%	0%
Moreta ‘19 [67]	17/7	-	n.a.	4	60	87.5%	73.6	0%	16.7%	0%	0%	0%
Van Laarhoven ´20 [13]	70/12	n.a.	n.a.	4	34.8	94.3%	n.a.	8.1%	18%	n.a.	4.6%	0%
Munegato ‘20 [69]	19/4	+	2	4	29	100%	76.1	8.7%	17.4%	0%	4.3%	0%
Sun ‘20 [86]	69/14	in 7	2(*)	4	94.5	97.6%	48.3°	n.a.	15.6%	n.a.	2.4%	9.6%
Santiago ‘21 [71]	16/19	n.a.	n.a.	4	42	100%	78.1	11.4%	5.7%	5%	2.8%	0%
Schreiner ‘22 [70]	12/6	-	1	4	18.5	94.4%	72.5	0%	5.5%	5.5%	11%	5.5%
Fink `24 [59]	0/46	+	2	4	48.8	100%	76.1	2.2%	4.3%	0%	0%	0%
Di Matteo ‘24 [87]	44/5	(+)	2	3,4,5	63.4	100%	85.5	n.a.	6.1%	2.0%	4.1%	0%

ETO = extended trochanteric osteotomy, Surgical technique 1 = reduction and retention of the fragments before implantation of the stem, Surgical technique 2 = transfemoral, implantation of the stem and then cerclage of the fragments around the stem; * = additional strut grafts for Vancouver B3, Stem 1 = with proximal fixation; Stem 2 = non-modular, distal fixation, fully porous-coated; Stem 3 = monoblock, distal fixation, conical; Stem 4 = modular, distal fixation, conical; Stem 5 = long cemented stems; n. a. = no answer; ( ) = partially; ° = SF-36 score; OHS = Oxford Hip Score; MD = Merle D’Aubigné Score; ‘ = SF-12 mental; ‘’ = SF-12 physical; HA = Hemiarthroplasty; THA = Total hip arthroplasty

In the study of Fink et al. [60], 46 stems in Vancouver B3 fractures (23 with Paprosky type IIIA, 19 with type IIIB, and 4 with type IV defects) were revised using a transfemoral approach to the modular tapered cementless revision stem Revitan curved (ZimmerBiomet, Winterthur, Switzerland) with distal cone-in-cone fixation. The patients were prospectively followed for 48.8 ± 23.9 months (range: 24–112 months). Additional distal interlocking was used in 23 cases in which distal cone-in-cone fixation within the isthmus measured less than 3 cm. After 6 months, 43 patients showed osseous and 3 patients fibrous consolidation of both the fracture and the bony flap. According to the Engh classification [72] for biological stem fixation, bony ingrowth fixation was observed in 42 patients and 4 patients showed stable fibrous fixation of the stem. Breakage or radiolucency of the interlocking screws was not seen. There was one case of non-progressive subsidence and two hip dislocations. The Harris Hip Score increased from 57.8 ± 7.9 points at three months and to 76.1 ± 10.7 points 24 months postoperatively. One patient died during the first postoperative year (2.2%).

APC (Allograft prosthesis Composite)

Another option for treating periprosthetic Vancouver B3 fractures is the allograft prosthesis composite (APC). Here, a conventional prosthesis is cemented into the allograft and the remaining proximal bone fragments are fixed with cerclages around the allograft [73, 74]. An APC may be indicated if, in addition to a Vancouver B3 periprosthetic fracture, there are additional larger segmental bone defects extending at least 8 cm into the diaphysis [76]. Since bone regeneration is necessary for integration of the allograft, APC is only recommended for younger patients [73, 74, 76]. Wong et al. [74] report 5-year results in 15 type B3 fractures treated with an allograft. The rate of bony incorporation of the allografts was 93.3% and loosening of the implanted prostheses was 6.6%. The most common problems associated with this procedure are dislocations and infection. Maury et al. [77] reported 25 Vancouver B3 fractures in cases with an average age of 70.2 years (60–82 years). In 12 cases the allograft was united with the host bone, 17 of the 25 greater trochanters were united, 15 patients required walking aids and in 4 cases (16%) a revision was necessary within the first 5 years postoperatively. Due to the disadvantages of allografts in terms of procurement, possible disease transmission and irregular healing, this technique is only performed in very few centres.

Proximal femoral replacement (Megaprostheses)

A further option for the treatment of Vancouver B3-fractures is the use of megaprostheses (proximal femoral replacement) with cemented fixation. They may be considered for Vancouver type B3 fractures with substantial bone loss in elderly patients because, compared with allografts or cementless revision stems, they enable earlier weight bearing of the limb [76, 78]. The remaining bone fragments are attached to the megaprosthesis along with the corresponding soft tissue structures [76, 78]. Proximal femoral replacement using megaprostheses has the disadvantages of limited functional outcomes due to impairment of the abductor muscles caused by replacement of the proximal femur as well as higher rates of dislocation and periprosthetic infection [76–79].

Generally, the most common complications of this procedure reported in recent reviews for proximal femur replacement in non-oncolgogical conditions (periprosthetic fractures in 27% of cases) involve dislocation with rates between 10% and 15%, infection with rates between 7% and 10% as well as reoperation rates between 13% and 40% [80–82]. Reports of megaprostheses for the single indication of periprosthetic fracture are rare and have only small case numbers. Klein et al. [79] reported a Harris Hip Score of 71 points (range: 56–90 points) in 21 cases of proximal femoral replacement for B3 fractures with a mean patient age of 78.3 years (range 52–90 years). After an average follow-up of 3.2 years, they observed two dislocations and two infections. At the final follow-up, only 8 patients (38%) were able to walk without assistive walking devices. Apprich et al. [83] reported on 33 patients with megaprostheses for periprosthetic fractures (12 proximal femur replacements and 7 total femur replacements in 13 Vancouver B3 fractures). The one-year mortality rate was 18.2% and the revision rate 39%.

However, the use of megaprostheses requires that the bone distal to the fracture remains sufficiently stable to allow implantation of a proximal femoral replacement. If this is not the case, or if the distal femoral region is also fractured, implantation of an intramedullary femoral replacement (push-through prosthesis) combined with replacement of the knee joint is the only remaining option. Here, as well, the bone with its attached musculature should be preserved rather than using an extramedullary femoral replacement (total femoral replacement).

In general, B3 fractures carry a slightly higher risk of revision than B2 fractures [20]. However, Munro et al. [47, 48] found no differences in complication rates, survival rates or clinical scores between B2 and B3 fractures.

Conclusion

The choice of one of the different treatment options for Vancouver B3 periprosthetic fractures remains a case-by-case decision. However, based on the current knowledge of the literature, stem revision to distal tapered fluted cementless revision stems is the solution with the most predictable results in Vancouver B3 fractures with an isthmus according to Paprosky type II and IIIA-defect. In insufficient isthmus (Paprosky IIIB and IV) tapered fluted stems with a higher degree of conicity (in sufficient cortical thickness of the rest of the isthmus), distal interlocking of these stems or distal cementing may be the treatment of choice. It would be helpful for the comparison of results in future studies to present the results for Vancouver B2 and B3 fractures separately and to clearly describe the bone quality in the remaining isthmus of the femur, for example, by additionally using the Paprosky classification for bone defects.

Author contributions

All authors contributed to the study conception and design. The idea for the article came from Bernd Fink with additional literature search. Data collection and analysis were performed by Bernd Fink and Marius Hoyka. The first draft of the manuscript was written by Bernd Fink and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.

Data availability

No datasets were generated or analysed during the current study.

Declarations

Conflict of interest

The authors declare no competing interests.