Review: Trunnionosis leading to modular femoral head dissociation

Abstract

Component dissociation secondary to trunnionosis is rare, and its causes are multifactorial. It is a major complication of total hip replacement, in most cases requiring revision arthroplasty.

In this paper, we present a case of taper/head modular interface dissociation in a metal-on-metal total hip replacement. We review the literature of both trunnionosis, as well as a systematic review of modular dissociation of the femoral component in total hip arthroplasty, identifying commonalities with our own case.

1. Introduction

The use of modular femoral components has allowed greater flexibility in total hip arthroplasty (THA): they provide options to aid restoration of leg length and hip offset, choice of bearing surface, and revision procedures. Their use, however, brings a theoretical risk of component dissociation, either as an early or late complication, secondary to trunnion corrosion. Dissociation of a modular femoral component secondary to trunnionosis is rare, but reasonably well reported in literature.

1.1. Corrosion

Corrosion is the dissolution of metal in a solution. There are multiple mechanisms by which corrosion occurs (galvanic, fretting, crevice, pitting, stress, intergranular, intragranular, inclusion). This is through both mechanical and chemical mechanisms. The loss of metal ions through corrosion results in reduced structural integrity of the modular components, as well as the possibility of eliciting a local inflammatory response.

The most common metals used in THA, such as cobalt-chrome, and titanium, form a surface oxide layer, known as the passivation layer. Corrosive mechanisms disrupt this layer, resulting in crevices. Re-oxidation of the underlying metal substrate by direct contact with oxygen-rich fluid then occurs, and by repetition, the fluid becomes acidic. Metal ions are released in order to neutralise this change, and so material is lost from the metal components. This process is known as mechanically assisted crevice corrosion (MACC).

1.2. Trunnionosis and its risk factors

Corrosion at the neck-head interface and subsequent local tissue reaction has been identified as the cause of revision in 1.8%–3.3% of all revision THA's.¹ It is felt to be underdiagnosed, often mistaken for prosthetic joint infection (presenting with joint pain, reduced range of motion, raised inflammatory markers) and managed with long-term antibiotics, where a revision of modular components may have returned the patient to baseline function more efficiently.²

1.3. Trunnionosis is theorised to be due to a number of factors

Implant-based factors. Larger diameter heads in theory reduce risk of dislocation and allow a greater range of impingement-free movement.¹ However, they have been found to show increased wear at the head-neck junction in a series of hybrid metal-on-metal THA's.³ A further series of 74 retrieved implants found that 36 mm heads showed a significantly greater corrosion score at the head-neck taper compared to 28 mm heads, hypothesised to have occurred due to greater torque acting on the interface with the larger diameter head.⁴ However, a retrieval study of 154 metal-on-polyethylene THA's observing for fretting/corrosion at the taper/trunnion found no correlation of severity with femoral head size.⁵

Head length and offset of the femoral stem may also affect rates of trunnionosis. An in vivo study found from a study of 56 femoral heads and 17 femoral stems, that increased fretting occurs in longer head lengths and high-offset stems.⁶ Martin et al. found elevated serum metal ion levels to correlate with high offset stems, however this wasn't directly correlated with adverse tissue reactions.⁷

Mixed metal pairings can give rise to galvanic corrosion, and therefore greater rates of trunnionosis. A retrieval study of 231 modular hip implants found neck and head corrosion scores to be significantly higher in mixed alloy couples (42%) compared with similar alloy couples (28%).⁸ This same study also noted a correlation between higher flexural rigidity and reduced corrosion.

Trunnion geometry has been investigated in relation to risk of trunnion corrosion. Trunnions have been made shorter to reduce the risk of impingement, however this increases risk of edge loading.^1,9 One study of 44 cobalt-chrome implants found that there were significantly higher fretting and corrosion scores in 11/13 tapers.¹⁰ This was hypothesised to be due to the thinner taper allowing more fluid ingress and higher torsional forces and micromotion. A further retrieval study found a trend of trunnions being made shorter and less rigid with time, hypothesising these features to be a factor explaining the increasing rates of trunnionosis being seen.¹¹

Surgical factors. Cleanliness of the modular component, or a lack thereof, is a logical risk factor for corrosion and trunnionosis due to fretting or pitting corrosion. A study of head/taper assemblies has shown dry assembled components have greater resistance to fretting when compared to wet components.¹² Force of impaction in the assembly of modular components has been found to have an effect on rates of fretting corrosion, with a reduced MACC following impaction of 8 kN compared to forces of 2 or 4 kN.¹³

Patient factors. A laboratory study found that increasing weight directly correlates to increased micromotion between modular femoral components, and therefore fretting wear.¹⁴ In vivo, this can result in an adverse tissue reaction following release of local immune mediators, with tissue necrosis, osteolysis, and thereby loosening occurring secondarily.

1.4. Case report

We encountered a case of a patient with a metal-on-metal (MOM) THA using a Cobalt Chrome alloy. An Accolade (Stryker, Kalamazoo, Michigan) stem was used, in combination with a V40 (Stryker) femoral head. The patient had previously been annually reviewed, with no clinical concerns. Whilst walking, the patient felt a sudden ‘giving way’, and subsequent investigation confirmed dissociation at the taper-head modular interface (Fig. 1). This occurred at 12 years after primary surgery. Subsequently the prosthesis was found to have catastrophic trunnion wear at time of revision surgery (Fig. 2).

Fig. 1.

Anteroposterior radiograph showing dissociation at the left taper-head modular interface.

Fig. 2.

Significant trunnion wear of the removed femoral component at time of revision surgery.

We set out to quantify the relative risk of this complication, and identify possible correlations or epidemiological risk factors from the few published case reports on the subject.

2. Review of reported cases

Following the Cochrane collaboration, an extensive literature search of PubMed, Medline, and Google scholar was conducted. Randomised controlled trials, comparative studies, and cohort studies and case reports were included. This systematic review adhered to the preferred reporting items for systematic reviews and meta-analyses (PRISMA) guidelines.

The search terms included key words and Medical Subject Headings (MeSH) terms related to arthroplasty, replacement, hip, total hip arthroplasty, total hip replacement, THA, THR, dissociation, trunnion, and wear. This analysis did not involve direct contact with individual patients; therefore, no ethics approval was needed.

Six published case reports were identified describing modular dissociation of the femoral component. Two of these reports were found to have originated from the same patient case, but had been re-published, and one paper was unavailable for review. A total of five patients were included in the data analysis, including our own case report.15, 16, 17, 18, 19

The age range for patients was 42–76, all patients were male. The date range from primary surgery to dissociation of the component was 1 month–10 years. The mean age was 5 years since primary surgery. The mechanism of action was recorded for all 5 patients and did not include any significant traumatic event. Instead all cases were described as ‘rising from sitting’ or a limb twist (presumed as either extreme internal or external rotation of the leg while standing).

The THA bearing type was either metal on polyethylene (2 of the 5 cases) or MoM (3 of the 5 cases). One patient had a fully cemented implant and the remaining patients had fully uncemented HA coated implants. All patients were managed with open surgery. Trunnion damage was recorded in three of the five patients at time of revision surgery. Heterotopic ossification was recorded in one patient.

Revision surgery performed was either head exchange only (2 of the 5 of cases), open reduction only (1 of the 5 cases), or long stem modular restoration (2 of the 5 cases). At the time of revision surgery, metalosis with pseudotrumor was recorded in one case, ectopic bone in one case, no abnormal findings in one case and no information given in two cases. Implant size, bearing size, and whole blood metal ion levels were only recorded in one case and so not included in these results.

Follow up was recorded for two cases with no report of further complications.

3. Discussion

Dislocation of total hip arthroplasty remains the most common postoperative complication. Factors affecting this have been well described and reported including surgical approach, and prosthetic design and orientation. Modular design of hip prostheses was, in part, introduced to mitigate some of the risks for THA dislocation, allowing the surgeon choice in size and position of components where monoblock components had been previously unable to. There are multiple other benefits to a modular system, from manufacturing to stock and inventory management. Revision surgery of a modular design can be technically less challenging and bone preserving when appropriate.

Complications related to modularity of the femoral component are rare, and there are very few reported. In particular, dissociation of the modular head from the femoral stem will in almost all cases result in a catastrophic failure of the construct. Unlike a dislocation, management of a dissociated component is likely to result in open surgical management and as such is a significantly more serious complication.

We set out to quantify any commonalities between the few reported case reports and literature describing femoral component dissociation.

Trunnion damage was reported in 3 of the 5 cases and accounted for all the dissociations occurring over 3 years since primary procedure. In the case reported, there was significant flattening of the trunnion with almost complete obliteration of the 12/14 taper (Fig. 2). We have no doubt that this was the main causative factor in the final dissociation, however we must surmise that this damage occurred over months and years. Factors driving MACC include fretting, and the prevention of this is multifactorial, including material type and seating methods.

We do not feel the method of bone to implant fixation is relevant and merely represents the ratio of un-cemented to cemented in this epidemiological group. All reported cases were male and 3 of the 5 were large bearing surface MoM. This poses the question whether these factors alone are risks for dissociation or if they are a proxy measure for MACC of the taper. Notably one of the large bearing MoM dissociations did not show any signs of taper wear at revision; the single case was 5 years from the primary surgery date, and was treated with simple open reduction and retention of the original implants.¹⁸

The other reported case where no trunnion damage was recorded was that of dissociation at one month following primary procedure, the only early post-operative case report in the literature.¹⁶ The authors of this study proposed a mechanism coined the ‘pumping phenomenon’. This is described as an incorrectly seated head-taper interface which ‘bounces’ when impacted and therefore never forms a true interference fit between components.

Soft tissue factors including heterotopic bone formation and pseudo tumour were recorded in two separate case reports. The case of heterotopic ossification (HO) – Brooker class II, was recorded 3 years since primary surgery, and had no trunnion damage at time of revision.¹⁵ This led the authors to propose that in hip abduction, the leverage provided by the HO was enough to result in dissociation between the femoral neck taper and femoral head.

The case of pseudotumour had significant trunnion damage and so the soft tissue envelope was not associated in the mechanics of the dissociation by the authors, but rather a consequence of MACC.¹⁷

The management of dissociation was always open surgery. 4 of the 5 cases required component exchange and of these, two required femoral revision.15, 16, 17, 18 Open surgery is likely to be the only management in these scenarios. Relocating a trunnion within a femoral head while the head remains in a working bearing would prove technically challenging with image intensifier guidance only. Secondly analysis of the trunnion and modular components must be a priority in deciding management. We hypothesise that relocating a damaged or ‘sharpened’ trunnion such as our case report may actually be easier than that of an undamaged trunnion, and in performing closed reduction further damage may inadvertently be caused.

The particular stem and head in our reported case have been identified as having a propensity towards trunnionosis and failure. Wylde et al. found high rates of trunnion failure in retrieval study examining Accolade I stems in use with LFIT V40 or MITCH cobalt chromium heads (all Stryker). 6 of their 10 cases were revised for trunnion failure, and 4 of the remaining for trunnion fracture. They concluded that failure was primarily due to material composition mismatch and larger head sizes (greater than 36 mm) and recommended a recall on patients with these prostheses implanted for clinical and radiological review.²⁰

Our case report and study illustrate that dissociation of modular femoral components remains an extremely rare complication, but one which requires more investigation. Changes to taper design must include analysis into the effects of MACC and the resultant damage to the interference fit. We cannot formulate any correlations due to the limited nature of published reports, however it would appear that male patients with larger bearing surfaces may be more at risk. We encourage any surgeons with case experience to add to the literature for future analysis.

Author statement

Agneish Dutta – writing of manuscript, conducting review.

James Nutt - writing of manuscript, conducting review.

Guy Slater – editing of manuscript, supervision.

Syed Ahmed - editing of manuscript, supervision.