Postoperative complications of hip arthroplasty

Complications following THA have been investigated extensively throughout the history of THA and reported complications have changed the mode of operation many times during the last few decades (Ninomiya et al. 2017). Complications guide clinical practice, with the main goal of THA being to predict and prevent complications which are very unpleasant for patients and are expensive to treat. A typical indication for surgery is primary osteoarthrosis in both THA and HRA, but the difference from HRA is that it cannot be used in more complex and higher complication risk operations. However, HRAs have shown to have overall higher revision rates than THAs in all age and gender groups (Clarke et al. 2015). Typical complications following THA are infection and peri-prosthetic fractures, as well as leg length discrepancy (LLD), dislocation, nerve damage and thromboembolic events (Hofmann et al.

2000, Upadhayay et al. 2007).

Periprosthetic joint infection (PJI) is a leading cause of the failure of THA, with contemporary yearly infection burden varying from 0.76–1.24% (SHAR 2015, AOANJRR 2016, Springer et al. 2017). A large number of studies have been performed with the aim of reducing infection rates, but no improvement has been seen to date in implant registers (SHAR 2015, AOANJRR 2016, NJR 2016). Postoperative THA infections are severe complications and can result in substantial morbidity to patients, multiple procedures, increased costs and lengthy hospitalisations (Ninomiya et al. 2017). One study projects that revisions due to PJI will increase over the next decades compared to other modes of failure, with the anticipation that over 60% of all revisions will be due to infection (Kurtz et al. 2008).

Periprosthetic fractures may occur years or even decades after the index surgery.

Periprosthetic fracture may be caused by a trauma or periprosthetic bone loss may led to implant loosening and further on to fracture (Knutsen et al. 2017). The incidence of femoral

peri-prosthetic fractures after THA is raising worldwide (Streit et al. 2013). It is expected that the incidence of peri-prosthetic fractures overall will increase by a mean of 4.6% every decade for at least the next three decades (Pivec et al. 2015). The risk of femoral periprosthetic fracture after index surgery in 10-year follow-ups has varied between 0.64% – 1.6% (Lindahl 2007, Streit et al. 2014). The operative treatment of femoral periprosthetic fracture is the third most common cause of hip joint revision arthroplasty (Lindahl 2007). Periprosthetic fractures are costly, disabling and also morbid; these types of femoral fractures often require surgical treatment (Lindahl et al. 2005, Moloney et al. 2014). HRA has a unique type of complication, which is femoral neck fracture and femoral component loosening due to osteonecrosis, as well as adverse femoral neck remodelling (Campbell et al. 2000, Amstutz et al. 2004, Amstutz 2012).

The early neck fractures typically occur within 3-4 months after surgery and are attributed to different factors such as uncovered bone, leaving the component proud, notching the neck, osteopenia, cysts, impingement and trauma (Shimmin et al. 2005).

Leg length discrepancy (LLD) following THA is a significant source of patient dissatisfaction and a common reason for litigation (Hofmann et al. 2000, Upadhayay et al.

2007). Patients who are at risk of postoperative limb length discrepancies are those who have undergone previous hip surgery, trauma, infection, growth plate arrest and congenital dysplasia (Sculco et al. 2016). Significant lengthening of the leg may lead to sciatic nerve palsy and lower back pain and limping. Preventing LLD is achieved by evaluating leg length preoperatively by using physical and radiological methods. Physically true leg length is determined by a distance from the anterior superior iliac spine to the medial malleolus. There are few radiological methods to determine leg length and preoperatively template the correct level of femur neck osteotomy, neck length and femoral offset (Maloney et al. 2004). LLD is mainly caused by improper femoral component positioning (Al-Amiry et al. 2017). LLD

>1.5cm is defined as severe LLD because this size of stretch in leg length can affect hip joint function by changing the hip rotation centre and causing abductor dysfunction (Dong et al.

2016). HRAs do not have same problem related to LLD as THAs due to different surgical technique where in HRA femoral neck is preserved and head is resurfaced.

Medical complications are a heterogeneous group of problems related to surgery and anaesthesia and are typically caused by multiple aetiological factors. Of these complications, the most typical is venous thromboembolism (VTE) which is a term including DVT and PE.

The true incidence of VTE is not well-defined because symptoms of DVT might be mild and can remain undiagnosed. In different parts of the world, there are many guidelines for antithrombotic prophylaxis after THA which makes a comparison of the true incidence of VTE challenging (AAOS 2011, Flack-Ytter et al. 2012, NICE 2015). Fujita et al. showed that the overall incidence of VTE after THA is 4.4% in patients receiving recommended antithrombotic prophylaxis (Fujita et al. 2015). In a cohort study of over 58,000 patients, 0.9% had pulmonary embolism after primary THA (Phillips et al. 2003). In the revision THA cohort of over 12,000 patients, complication rates were higher than in primary THA, and PE was diagnosed in 0.8%

of the patients during the first six months after THA (Phillips et al. 2003). Optimal prophylaxis of prevention of DVE after THA is debated. Ideal prophylaxis would have a high efficacy in preventing DVE and PE, with minimal bleeding, cost-effectiveness and ease of administration (Agaba et al. 2017). Commonly used prophylactic pharmacological agents are antiplatelet agents (aspirin), low molecular weight heparin (LMWH), vitamin K antagonists (warfarin), synthetic indirect inhibitors of activated Factor Xa (fondaparinux) and selective and reversible direct Factor Xa inhibitors (apixaban and rivaroxaban) (Agaba et al. 2017). Mechanical compression devices can also be used for VTE prevention together with pharmacological prophylactic agents.

2.4.2.1 Aseptic loosening

The cementless component requires good initial fixation to allow bony ingrowth to occur (Pilliar et al. 1986, Haddad et al. 1987). The initial press-fit is typically achieved by under-reaming the acetabulum or by using non-hemispherical (flanged) components. Cementless femoral component is force impacted in femur to achieve press-fit stability. In cementless MoM THAs, revision typically occurs during early follow-up. In a recent study of 80 MoM hips that underwent acetabular revision, 92.5% of revisions were performed within 3 years of the index surgery (Fadi et al. 2012).

Various factors such as bone quality, female gender, elderly age and geometry of the acetabulum have been postulated as factors influencing primary fixation of the cementless acetabular component (Dorr et al. 2000, Piarulli 2013). In addition, underlying systemic diseases like RA and osteoporosis affect the properties of bone and also influence implant osseointegration (Aro et al. 2012, Zwartelé et al. 2012). From a mechanical point of view, acetabular component malpositioning or failure to restore the centre of rotation of the hip joint may lead to impingement and increased edge-loading, which may lead to a lack of bony ingrowth into the porous surface and subsequent early loosening of the acetabular component (Brown et al. 2014).

Compared to conventional modular acetabular components, the suggested benefits of the monoblock components used in LDH MoM THA and HRA are the reduced risk of dislocation and better biomechanics (Stroh et al. 2013). On the other hand, monoblock components do not allow supplementary screw fixation. Another drawback to monoblock components is the difficulty in assessing whether the component is fully seated to the acetabulum as the bone bed cannot be visualised through holes in the component. This may lead to a complication where the acetabulum component is not steady enough in index surgery and components may spin-off over a short period of time (Figure 7). In addition, aseptic loosening of the hemispherical acetabular components was recently shown to be one of the leading causes of early failure of primary THA (Carcia-Rey et al. 2012, Fadi et al. 2012).

Figure 7. Loosening of the acetabulum component due to a failure in the initial press-fit and insufficient fixation during the index surgery.

2.4.2.2 Dislocation

Dislocation is an unfortunately frequent and serious complication following conventional THA, but is rare with LDH MOM THAs (Bozic et al. 2009, Jameson et al. 2011) (Figure 8). The prevalence of dislocation has been reported to vary between 0.3% and 10% in primary MoP THAs and < 1% in LDH MoM THAs or even non-existent in HRAs (Parvizi et al. 2008, Jameson et al. 2011, Leclercq et al. 2013, Girard et al. 2017, Tao 2018). In a large registry-based study, the revision rate due to dislocation of different head size THAs was ≤2.5% at the 12-year follow-up (Kostensalo et al. 2013). In another study of 58,000 patients, the incidence of dislocation was 3.9% during the first six months after primary THA (Phillips et al. 2003).

Furthermore, the incidence of dislocation increases with time. In a Mayo Clinic series of 19,680 hips, the incidence of dislocation was 1.8% at one year and 7% at five years, increasing by 1%

per year for the subsequent 5-year period (von Knoch et al. 2002). The median time to the occurrence of dislocation was 11.3 years after primary THA (von Knoch et al. 2002).

Figure 8. Dislocation of large diameter head metal-on-metal total hip arthroplasty in plain radiograph.

Understanding the mechanisms and causes of dislocation are essential in clinical decision-making before and during THA operations. The main risk factors for dislocations have been found to be female gender, advanced age, neuromuscular or cognitive disorders, substance abuse, previous hip surgery and soft-tissue or bone deficits of the hip (Werner et al. 2012). Hip dysplasia or technical errors in surgery (e.g. acetabulum component malposition or failure in restoration of hip rotation centre) are known risk factors for dislocation (Lecerf et al. 2009, Brown et al. 2014). Head size is a known risk factor for THA dislocation, with head size of < 36 mm the risk of dislocation have found to be 1.25% while in larger head size risk is < 1%, respectively (Jameson et al. 2011, Allen et al. 2014). Articulation material (MoM, MoP or ceramic-on-ceramic) have not been noticed to be a risk factor for dislocation in head sizes of 28 mm and 32 mm (Shah et al. 2017). Surgical opening is a risk factor for dislocation while in posterior approach the risk for dislocation have been found to be 2.0 – 5.8%, in anterolateral approach 2.2 – 2.-3% and in direct lateral approach 0.6%, respectively (Patel et al. 2007).

LDH MoM THA and HRA were widely used in the early 2000s; one of the main reasons for its popularity was the low risk of dislocation due to large head size implant and ease of retain natural jumping distance of hip

(Amstutz et al. 2004, Sariali et al. 2009) The jumping distance is the degree of lateral translation of the femoral head centre required before dislocation occurs (Sariali et al. 2009). Other implant options to decrease the risk of dislocation are constrained acetabular liners and dual mobility articulation (Mäkinen et al. 2016, Plummer et al.

2016).

2.4.2.3 Problems related to metal-on-metal bearings and trunnion

MoM bearings produce less wear debris than MoP bearing surfaces. However, in the case of alloys, soluble ions and compounds are released by corrosion. This phenomenon accelerates metal wear and the hip joint becomes filled with high concentrations of chromium and cobalt.

Subsequently, this debris accumulates in the body and causes exposure to metal ions for prolonged periods of time. In particular, large diameter head implants have a greater tendency to rub against the edges of the cup and subsequently accelerate corrosion. LDH MoM THAs have been shown to have greater systemic exposure to cobalt and chromium ions than smaller head sizes (Clarke et al. 2005).

Modularity in THA has been studied with great interest throughout the history of hip implants; the most significant player in modularity is the femoral head-stem taper junction which is called a trunnion. Modularity of the hip implant design has certain advantages over monoblock implants. The use of modularity allows the surgeon to select the correct combination of implants in order to restore the biomechanics of the hip joint. Modularity provides the ability to perform a modular component change with the retention of other implant parts in some revision surgery cases. Disadvantages of modularity are the increased risk of mechanical failure and the production of metal debris that may cause ARMD (Dangles et al. 2010). Trunnion corrosion was first described nearly three decades ago and is nowdays claimed to account for approximately 2% of the revision burden (Collier et al. 1991, Whitehouse et al. 2015). Trunnion wear and corrosion occurs most often with mixed metal couples and may further lead to the development of galvanic corrosion (Goldberg et al. 2002).

This process is exacerbated by mechanical load which is further increased by large head size and high offset (Del Balso et al. 2015). It is proposed that corrosion at the trunnion is an important source of metal ion release and may cause ARMD in the same way as MoM bearings do (Gill et al. 2012).

There are concerns about metal hypersensitivity, osteolysis, pseudotumours, chromosomal mutation, foetal exposure to high metal ion levels and carcinogenicity (Beaule et al. 2000, MacDonald 2004, MacDonald 2004, Milosev et al. 2005). There are few studies where post-mortem subjects with stainless steel and cobalt-chromium implants demonstrated the dissemination of metal debris to local and also distant lymph nodes, bone marrow, the spleen and the liver (Black 1988, Case et al. 1994, Urban et al. 2000). Metal wear particles derived from the articular surface can invade the bone-implant interface where they are phagocytosed by macrophages which further induce the release of various cytokines and cause a cascade of reactions leading to ARMD (Case et al. 1994, Urban et al. 2000).

Carcinogenesis and malignant tumours due to local exposure to metal debris at the implant site are extremely rare when comparing the number of individuals with THA. However, the effect of the systemic release of metal ions from THA on carcinogenesis is uncertain. There are concerns about the possible association between THA and malignancies of the lymphatic and haematopoietic systems (Wagner et al. 2012). Due to metal ion release from MoM bearings and trunnions with the addition of further progression of pseudotumours and early loosening of implants, serious problems emerged in the early 2010s and many implant systems were recalled; therefore, the use of these implants should be discontinued (MHRA 2010, SHAR 2015).