The total hip arthroplasties popularized by Sir John Charnley utilized components that were fixed to the surrounding bone using a self-curing acrylic bone cement. In these early trials, there was a wide variability in the success of cemented femoral components [125, 126]. This inconsistency can at least partially be explained by the evolution of cementing techniques over the past half century. Originally the cement distribution technique involved finger packing the bone cement into an unplugged femoral canal. Modern cementing techniques involve cleansing the canal with pulsatile lavage, inserting cement in a retrograde fashion, porosity reduction via vacuum mixing, and cement pressurization within the canal. Additionally, the stem is centralized proximally and distally in order to ensure an adequate and symmetric stem mantle. Thus, when critically examining clinical data, it is important to determine which cement technique was used when placing the component.

The mechanisms of failure of cemented femoral components are typically the result of mechanical factors initiating femoral loosening. Debonding, or separation of the cement from the stem, occurs followed by high stresses produced in the cement mantle proximally and at the distal tip of the implant [127–130]. These stresses then initiate crack formation which further destabilizes the implant and produces debris which results in an inflammatory reaction, bone resorption, and a soft tissue membrane which is commonly encountered in aseptic loosening. Interestingly, unlike uncemented components, this fibrous membrane forms late in the loosening process of cemented components and is not thought to play a significant role in the initial loosening process [107, 131]. The factors that led to the initial debonding were examined in numerous studies. Radiolucencies in the cement mantle signifying poor distribution of cement were found to predict later failure [132]. Varus implant position was also found to be associated with a higher risk of aseptic loosening and was thought to result in adverse outcomes due to its propensity to create a poor cement mantle [133, 134]. Thus in an effort to evaluate cement mantles, a grading system was created by Barrack et al. [135] which distinguished complete filling of the proximal diaphysis (A), near complete filling (B), incomplete filling with either greater than 50 % demonstrating radiolucencies (C1) or less than 1 mm of mantle present (C2) and gross deficiencies in the mantle with no cement distal to the tip or multiple large voids (D). Aseptic loosening was associated with C and D mantles with the latter having the greatest amount of implant failures. Other variables that contributed to failure included increased weight of patient, younger age of patient, male sex, and patients with post-traumatic osteoarthritis. Interestingly, the type of implant and surgeon did not correlate with a change in incidence of revision in this study [136].

Clinical outcomes of cemented femoral components have improved with advances in cement techniques. Distal plugging of femoral canal, extensive lavage, and retrograde injection of bone cement resulted in reported mechanical failure rate of 1–5 % in studies performed by Ranawat et al., Madey et al., and Smith et al. at 15 years follow-up [137–139]. Contradictory results were reported by Sanchez-Sotelo et al. who found that the mechanical failure rate was 10 %. Stratification revealed a 4.3 % failure rate in patients older than 50 years and a 27.7 % failure rate in patients younger than 50 years [140]. Further advances in cementing technique including pressurization and vacuum preparation of the cement have resulted in further decreases in failures. A 13.5-year follow-up of 204 THAs placed with this cementing technique showed only 4 revisions: one for osteolysis, two for recurrent dislocations, and one delayed infection [141].

Uncemented implants were originally created to deal with the issue of “cement disease” or the lysis of periprosthetic bone. Cementless femoral components rely on osseointegration, the structural and functional connection between bone and implant, without intervening soft tissue, for their fixation strength. The mechanism of osseointegration seen in uncemented implants is classified as either ongrowth or ingrowth. Traditionally, ingrowth is bone deposition in the interstices of a porous surface and ongrowth is bone attachment to a flat implant surface. As noted earlier, bony ingrowth requires more than just a porous surface to occur successfully. The surface must also be consist of a biocompatible material, have optimal pore size, be in intimate contact with viable bone, and have adequate initial stability for bony ingrowth to occur successfully [103–105, 108, 109]. The ideal values for these parameters are discussed earlier in the chapter when describing the osseointegration of uncemented acetabular components.

While uncemented components have enjoyed great success, it is currently recognized that “cement disease” is a misnomer and is instead referred to as osteolysis. This process is known to cause loosening and eventual failure in both cemented and uncemented implants. Osseointegration of cementless porous coated femoral stems has proven to be a reliable and successful form of fixation but some features of cementless stems such as stem geometry, surface properties of the porous surface, and the extent to which the porous surface is applied to the stem continue to be a source of debate. One way of categorizing femoral components is by stem geometry. The two most common types of geometries are anatomical and tapered. The differences between these include their shape, metallurgy, head-neck design, and ongrowth–ingrowth surface.

The design rationale behind tapered components is geared towards promoting long-lasting osseointegration between the component and the surrounding bone. This requires rigid initial fixation until full osseointegration can be achieved. This initial stability is provided by the taper when it is forced into the medullary canal thereby producing circumferential hoop stresses that do not allow an axially loaded tapered stem to advance any further [142]. Rotational stability is provided by a rectangular cross section of the tapered stem although circular cross sections also exist and have shown highly satisfactory results as well [143]. An advantage of the circular cross section over the rectangular one is the ability to correct femoral anteversion if necessary. The tapered design has been shown to reduce stress shielding and lead to a decreased prevalence of thigh pain [144–147]. Femoral fixation requires osseointegration of the femoral component, typically in the proximal stem because of the maximal contact between bone and implant afforded by this area. Once osseointegrated, the stem is rigidly and lastingly fixed to the surrounding bone.

Clinical studies have favored the use of tapered designs over anatomic ones. In one study of 311 Porous Coated Anatomic (PCA) stems the overall survival rate of the femoral component at 14 years was 95 %. Unfortunately this design also showed a 36 % prevalence of thigh pain and 42 % of individuals had significant amount of femoral osteolysis as evidenced by radiography [148]. One study with 10- to 13-year follow-up of tapered total hip replacements in 283 patients found a 99 % survival rate of the femoral component with osteolysis seen in only 6.2 % of cases [149]. Activity related thigh pain was also low, noted in only 3 % of patients [149]. Very long-term studies of greater than 20 years have also shown positive results for tapered designs. The 20- to 25-year survival rate for tapered total hip replacements with follow-up ranging from 20 to 25 years was found to be between 86 and 95 % [150–152]. A long-term study of 47 obese patients with 18- to 27-year follow-up found a 94 % survival [153]. Risks for failure in a study of 326 patients receiving tapered THA with a follow-up of 22 years were found to be undersized stems, and hips where the cup was already revised. This study had a high (38 %) rate of cup revision due to the use of smooth-threaded cementless sockets [151].

Extensively porous coated prostheses are defined as those with porous coating of more than 80 % of the surface area of the stem. The most common of these is the Anatomic Medullary Locking (AML) stem produced by DePuy. A critical design feature of the AML is its straight, cylindrical, non-tapered distal stem geometry. Thus, the stem does not wedge in place and fixation depends on a “scratch fit” between the rough external surface of the implant and a similar shaped bone canal. The theoretical advantage of the extensive coating is that it allows for osseointegration over the entire length of the stem. In this design, distal porosity is particularly important because the distal part of the stem most consistently contacts with the cortical bone. Interestingly, even in cases where successful osseointegration does not occur and a fibrous tissue membrane forms between the bone and stem, adequate fixation is still achieved and there is sufficient radiographic stability as well as patient satisfaction owing to the extensively coated surface [154]. Conversely, proximally porous coated femoral implants can be either tapered or cylindrical and were intended to achieve biologic fixation solely in the femoral metaphysis. This property was intended to reduce proximal stress shielding and preserve bone stock.

Both designs have proponents that cite multiple publications with reproducible long-term follow-up. The reported incidence of thigh pain in patients receiving THA with a fully coated femoral component has varied between 3 and 20 % [155–157]. Today, most surgeons presume that proximally coated stems and cemented stems are less predisposed to causing thigh pain than fully coated stems although this idea remains controversial [158–161]. Similarly, femoral bone loss as a result of stress shielding around well-fixed femoral components is thought to occur more frequently in extensively porous coated stems. In theory, bone loss in the femur could lead to greater and lesser trochanteric avulsion fractures and make future revisions more difficult. However, currently, it is important to note that stress shielding is more of a radiographic finding than a diagnosis and the clinical ramifications of this process have not yet been demonstrated. A prospective randomized blinded clinical trial followed 388 patients receiving either proximally porous coated and fully porous coated femoral components (Table 11.2). A minimum follow-up of 2 years with a mean of 6 years was used and found that post-operative clinical outcome scores were similar at all follow-up intervals. There were no differences in incidence of thigh pain at any time although bone density reduction was greater in the fully coated stem as compared with the proximally coated [162].

Despite the success of monoblock implants, these designs were limited by their inability to fine-tune two properties which can differ greatly between patients: offset and leg length. Since dislocations remain one of the most frequent post-operative complications after THA, designers sought to reduce the dislocation rate by allowing surgeons to make adjustments in the offset and neck length of implants. Thus modular femoral stems were intended to allow a more accurate reproduction of patient anatomy. The use of a modular head-neck junction allowed the surgeon greater freedom in adjusting for leg length discrepancies as well as optimizing the function of the abductors [163]. Modular head-neck components also allowed for easier revision of femoral components when the femoral stem is clearly fixed and requires no further fixation. Unfortunately, increased customizability came at a cost and these implants suffer from an increased risk of fatigue failure, fretting, and crevice corrosion.

There is a vast array of modular femoral stems that allow for retention and sacrifice of various components of hip anatomy. While most modular femoral components are used in revision total hip arthroplasty, some complex primary total hip arthroplasties can benefit from the use of these devices as well. Modularity can occur at various points throughout these femoral stems and great emphasis has been placed on the association between modularity positioning and implant failure. Systems such as the Zimmer ZMR utilize a mid stem modularity allowing independent selection of the sizing and positioning of the proximal and distal components. The DePuy S-ROM is a proximal modular femoral implant that utilizes a titanium stem with distal spines to achieve initial fit and rotational stability. Standard and calcar replacement options are available with variable offset options. A separate sleeve that incorporates a “step-like” geometry to convert shear to compressive forces is available in both porous and hydroxyapatite coated designs. Between the proximal and distal stem segments is a tapered section that engages this enveloping sleeve. The sleeve is positioned first and the stem is placed through the sleeve and engaged. This allows the S-ROM stem to separate hip biomechanics and component fixation. The sleeve achieves fixation whilst the stem allows for adjustments in length and offset. In all, the S-ROM design allows for 10,398 different reconstructive possibilities [164]. Unfortunately, stems with distal modularity have underperformed other systems and most have been withdrawn from the market.

Modular stems have produced good clinical outcomes in certain situations. Restrepo et al. examined initial distal fixation, femoral offset restoration, leg length equalization, and hip stability in 118 patients who underwent revision with Stryker Restoration Stem [165]. This system consists of a fluted, titanium conical distal stem, which attaches to a proximal body. Adequate bone ingrowth and fixation was obtained in 100 % of patients and the offset was corrected in 66 % while leg length discrepancy was corrected in 78 %. Ultimately, stability was achieved in 97 % of patients who received this implant. Initial concerns were raised regarding failure of the modular junction but in this study with a 4–7 years follow-up, no failure/fracture was observed in a total of 118 patients [165].

When two components lock into each other, wear and corrosion are an inevitable consequence. The production of wear particles can lead to osteolysis and corrosion can lead to implant failure. More alarmingly, metallosis and adverse local tissue responses (ALTR) similar to that seen in metal-on-metal bearing couples are being reported in patients with modular hip arthroplasties [166]. One subtype of these adverse responses is termed aseptic lymphocyte-dominated vasculitis-associated lesions (ALVAL). This lymphocyte-dominated reaction is very severe and can result in not only implant loosening but also severe soft tissue necrosis and pseudotumor formation [53]. Modular stems that incorporate a double taper (head-neck and neck-stem) add additional sites for failure and corrosion, with the added risk of increased wear particles.

Kop et al. examined 57 retrieved modular stems of 7 different designs. Of these, three were cobalt-chromium-molybdenum based and four titanium based [167]. The aim of this retrieval study was to assess whether the same degradation mechanism was present at the head-neck and neck-stem junction, whether or not the additional junction contributed to the revision, if the implant alloy affected the extent of degradation, and if the trunion machine finish affected the degradation mechanisms. Corrosion and fretting were both lowest in the titanium components with 62 % of the Co-Cr-Mo components having corrosion of the trunion, and 90 % fretting. In contrast, 30 % of Ti-based components showed corrosion, with 50 % exhibiting fretting. However cold welding of the titanium components did occur. It was concluded that titanium modular components may reduce the amount of degradation, but at the expense of an increased risk of cold welding.

Kop et al. also used a retrieval study of 16 modular components to examine the relationship between corrosion, material, and implant time. Of the retrieved implants 6 tapers showed fretting corrosion, with the average implant time being 39 months. No corrosion was shown in the remaining 10 tapers, which had an average implantation time of 2.7 months. It was concluded that even with modern materials and taper designs corrosion is still a concern with added modularity [168].

Gilbert et al. examined 148 retrieved modular hip prostheses. Significant corrosion was noted on 16 % of necks and 35 % of heads [169]. Concerns have also been raised regarding the elevated serum metal ions levels produced due to corrosion of modular stems. Once implanted, a protective surface oxide layer forms on the implant (modular implants are typically composed of titanium or Cobalt-chrome). While this film provides added corrosion resistance to the implant, it is subjected to repeated disruption as a result of stresses applied to the prosthesis. As the film reforms it reduces oxygen in the surrounding soft tissue. This process repeats itself and results in a reduced ability of the film to protect the implant [29].

Jacobs et al. describe corrosion is also affecting the structural integrity of the implant. This has been implicated in isolated incidents of fracture. Wright et al. report a case of fracture of the modular neck in a 49-year-old man. The fracture occurred while bending forward to tie his shoes, in the context of a fall onto his hip 2 months previously. On examination with light microscopy marked fretting and corrosion were noted along with debris [170].

Despite the intra-operative advantages offered by modular stems, the added junctions with the implant can lead to increased corrosion and wear particle production. While implants such as the S-ROM and modular taper stems have revolutionized revision surgery the use of modularity in primary THA should be avoided due to the risk of fatigue fracture and corrosion.

While THA is the treatment of choice for osteoarthritis of the hip there is an increasing cohort of patients who are requiring replacements earlier in life in order to remain physically active. While THA offers good symptomatic relief for these patients, they are likely to require revision in the future as a result of the increased physical demands placed on the implants, their increased life expectancy, and the insufficient longevity of traditional total hip arthroplasty. Hip resurfacing was developed as one solution to this problem and involves the preservation of the femoral neck through the use of a cap over the femoral head. This approach conserves femoral bone and theoretically allows for easier revision in the future, should it be needed.

The results of early resurfacing procedures in the 1970s and 1980s were poor. Accelerated wear and large volumes of biologically active wear debris resulted from the combination of a large articulating surface with a thin polyethylene liner and led to high rates of implant loosening and bone loss. With the development of new, more wear resistant materials, resurfacing has been reintroduced, this time incorporating a metal-on-metal (MoM) articulating surface [171]. As such, the downsides of MoM bearing couples which have been discussed earlier in this chapter must be weighed when deciding whether to pursue this reconstruction option.

As the stability of the femoral head is a key component in the effectiveness of resurfacing, severe bone loss, cysts or osteonecrosis of the femoral head or neck are contraindications for resurfacing. Retention of the femoral neck also means that unlike THA, femoral neck fractures may occur. Marker et al. used a prospective cohort study to identify the incidence of femoral neck fracture. 550 resurfacings performed by a single surgeon were studied. It was shown that 14 (2.5 %) had resulted in fracture of the femoral neck. Of these, 12 had occurred in the first 69 resurfacings performed, with women and obese patients shown to have a higher cumulative incidence of fracture. It was concluded that the risk of femoral neck fracture is multifactorial, associated with both the surgical learning curve and patient selection [172].

In vitro studies have suggested more limitations to hip resurfacing. Bengs et al. and Kluess et al. have both demonstrated reduced range of motion (ROM) in resurfacing compared to THA. Kluess examined the ROM of eight resurfacing prosthetics using 3D CAD models. The ROM of the resurfacing systems was found to be substantially less than that of total hip prosthetics with the large diameter of the femoral neck leading to impingements in all maneuvers analyzed [173]. Bengs examined the ROM of eight different hip replacement designs implanted into composite femurs and pelvises. It was found that compared to the THA prosthetics, resurfacings showed reduced ranges of motion, with, once again, early impingement of the femoral neck [174]. This is important since hip resurfacing is largely being indicated for younger and more active patients who require an increased ROM for various activities. Fortunately, despite these studies demonstrating reduced ROM in resurfacing systems, these reductions have not been seen when applied to the clinical setting. Clinical studies have shown similar ROM for both resurfacing and THA systems. Le Duff et al. examined 35 patients who had undergone bilateral surgery receiving THA on one side and resurfacing on the other, with a mean follow-up of 88 months. They found no difference in ROM between the two systems [175]. Shimmin et al. suggested that while THA shows greater ROM in laboratory studies, this cannot be recreated in patients with normal flexibility leading to similar ROM in both THA and resurfacing [176]. Shimmin also reports that of the nine papers comparing functional outcomes between THA and resurfacing, eight showed consistently similar outcomes [176].

Springer et al. used a large meta-analysis to compare the results of 3269 resurfacings with 6408 cementless THAs. Femoral revision for mechanical failure was used as an endpoint and was found to be 1.3 % in the THA group with a mean follow-up of 8.4 years compared to 2.6 % in the resurfacing group with a mean follow-up of 3.9 years [177]. Johanson et al. used the Nordic Arthroplasty Register Association database to examine the non-septic 2-year revision risk of 1638 resurfacings and compared to 172,554 THAs. By 2 years the revision rate for resurfacings was 2.4 % compared to 1.1 % for THA [178].

It is clear that not all resurfacing systems yield similar results. Seppanen et al. examined the Nordic Joint registry between 2001 and 2009. During this time 4401 hip resurfacings were performed. When comparing the Articulating Surface Replacement (ASR) Hip resurfacing system (DePuy Orthopedics, Warsaw, Indiana), to the Birmingham resurfacing system (BHR) it was found that the ASR had inferior outcomes to the BHR with a relative revision risk of 1.8 (CI: 1.2–2.7) [179]. The ASR Hip resurfacing system was recalled by DePuy voluntarily in August 2010, based on unpublished data from the National Joint Registry of England and Wales that showed a 12 % revision rate at 5 years [180]. In 2007, the Australian Orthopedic Association National Joint Replacement Registry (AOANJRR) reported that the ASR revision system had higher than expected revision rates (3.0 revisions per 100 observed component years), that were twice that of other resurfacing systems with a cumulative percent revision at 2 years of 5.16 % [181]. De Steiger et al. reviewed the AOANJRR between 2003 and December 2009 identifying 1167 ASR resurfacing procedures. It was found that the cumulative revision rate at 5 years for the ASR resurfacing system was 10.9 % compared to 4.0 % with all other hip resurfacing prosthesis [180].

While the literature shows that resurfacing is a viable alternative to THA, especially in younger patients, it is clear that adequate surgical experience must be combined with careful selection of the patient and prosthesis in order to achieve optimal results. Additionally, it is important to note that hip resurfacings utilize a MoM bearing couple which carries inherent issues as discussed earlier in the chapter.

Numerous surgical approaches to the hip exist for use in hip arthroplasty. Although evidence in the literature can be found for the use of certain approaches over others for given indications, realistically most surgeons will use the approach that they are most comfortable with for the vast majority of their cases. In certain situations, however, there are definite advantages to either modifying the surgeon-preferred approach or using a different one altogether. The need to access a particular anatomic region either because of bony deficiency or in order to remove hardware from a previous surgery are two such examples. Additionally, modifications of existing approaches must be utilized in certain populations. Patients with difficult anatomy, such as obese or muscular patients, require larger incisions, longer retractors, and more surgical assistants in order to have a satisfactory clinical outcome.

Recently there has been a surge in interest within the orthopedic community to perform THAs using minimally invasive surgery (MIS). Mini-incisions are either smaller versions of conventional approaches or novel incisions which are used to gain access to the acetabulum and femur. Ultimately, the invasiveness of a procedure is more dependent on the amount of soft tissue damage that it causes rather than the size of the incision. MIS appears to be less disruptive to soft tissues and advocates point to the potential for reduced intra-operative blood loss, decreased muscle damage, shorter length of hospitalization, reduced post-operative pain, and improved cosmesis that these methods offer. Conversely, some believe that the current practices in THA already produce excellent results with low complication rate, significant improvement in patient function, and excellent long-term prognosis. They argue that smaller incisions impair intra-operative visualization and lead to implant malposition, increased risk of intra-operative fractures and the potential for increased muscle damage. Concerns have also been expressed over the risk of neurovascular injury and poor implant fixation [182].

Hip reconstruction is the subject of ongoing efforts to improve clinical outcomes. It is especially challenging to improve on a treatment that has already produced excellent results. THA is successful in 85–95 % of cases. Given this finding, many will ask: why fix what isn’t broken? One reason is that the surgical volume of hip replacements is staggering. Over 500,000 THAs are performed annually in the United States [241]. Even if only 5 % of these fail, the result is a significant amount of burden on the healthcare system but more importantly on those patients who are unlucky enough to have a poor clinical result. Secondly, implant failure that results in revision is costly, technically difficult, and more likely to fail than a primary procedure. Lastly, while THR is used often, there are many debilitated patients who are currently not candidates for this procedure due to their young age. Significant improvements in implant longevity can have a tremendous impact on the lives of these individuals by returning function to their joints earlier and allowing them to resume their normal way of life. Still, the multitude of new technological options that exist for total hip arthroplasty greatly exceed the evidence supporting their use. New designs should be tried but all should be tested rigorously in order to come to find the optimal combination of principal components. Lastly, despite the emphasis on technology, one of the main contributors to the success of THA is surgical ability. Component wear, soft tissue damage, and implant stability have all been shown to be affected by surgical technique. Thus, improvement in total hip arthroplasty must come from advances in implant design, biomechanics, and surgical technique.