Surgical treatment of specific Unified Classification System B fractures: potentially destabilising lesser trochanter periprosthetic fractures

To investigate the clinical effects of specific Unified Classification System B (UCS B)-lesser trochanter periprosthetic fractures and determine whether they occur only with non-cemented stems. A retrospective analysis of 28 patients with specific UCS B2 fractures who underwent two surgical treatments, longer stem revision and internal fixation (LSRIF) and open reduction and internal fixation (ORIF), was performed. The patients were assessed at 1, 3, 6, 12, and 24 months and annually thereafter. Fracture healing, complications, Harris Hip Score (HHS), and the Short Form Health Survey questionnaire (SF-36) quality of life score were assessed at each follow-up. At the time of the last follow-up, seven patients had been lost: three were lost to contact, two died, and two were hospitalised elsewhere and unavailable for follow-up. The remaining 21 patients were followed for an average of 49.3 ± 15.4 (range: 24–74.4) months. Their average fracture healing time was 13.5 ± 1.1 (12–15.4) weeks. Complications included three cases (10.71%) of thrombus, one (3.57%) of heart failure, and one (3.57%) of pulmonary infection. There were no revisions due to prosthesis loosening, subsidence, or infection. At the last follow-up, the HHS, SF-36 mental score, and SF-36 physical score were recorded, LSRIF vs. ORIF (82.9 ± 6.6 vs. 74.7 ± 3.9, p = 0.059; 50.9 ± 7.6 vs. 38 ± 1.4, p = 0.012, and 51.7 ± 8.4 vs. 39.7 ± 3.4, p = 0.032, respectively). Specific UCS B2 fractures mostly occur with non-cemented stems. LSRIF with cables is the main treatment, while ORIF is an option for those elderly in poor condition.

Periprosthetic femur fracture (PPFF) is an acute complication that often occurs after hip arthroplasty.Due to the increasing use of arthroplasty surgery and the aging population, the incidence of PPFF is increasing [1][2][3][4] .
The Vancouver Classification System (VCS) 5 and subsequent Unified Classification System (UCS) 6 were intended as simple systems for classifying fractures and guiding management.However, these two classifications do not include some special types of fracture.For example, Van Houwelingen and Duncan 7 reported a type of periprosthetic fracture of the lesser trochanter that involved a segment of the proximal medial femoral cortex.This fracture is easily misidentified as an A LT fracture, but is actually type B2.Consequently, it is called a "pseudo A LT " or "new B2" fracture.Pseudo A LT fractures often occur in non-cemented stems [7][8][9][10][11] ; the mechanism may involve an occult fracture resulting from surgical insertion, and the fractures occur during early weight-bearing rehabilitation.
To assess the importance to this type of fracture and improve the surgical outcome, we hypothesized pseudo A LT fractures occur only with non-cemented stems and examined how different surgical treatments of pseudo A LT fractures affect the clinical outcome.
Follow-up.All patients were followed as outpatients at 1, 3, 6, 12, and 24 months and yearly thereafter.All surviving patients with complete follow-up were analysed.Fracture healing and the position of the prosthesis loosening were evaluated by X-rays at each visit.The Harris Hip Score (HHS) 12 and Short Form Health Survey questionnaire (SF-36) quality of life score 13 were also recorded.Fracture healing was assessed in terms of local tenderness, longitudinal percussion pain on the injured leg, local abnormalities, and blurred fracture lines with continuous callus passing through the fracture line on X-ray at the last follow-up.The fracture healing time was defined as the time from the second postoperative day to the end of follow-up if there were no special circumstances.
Data are expressed as mean ± standard deviation (SD).The groups were compared using t-tests.Count data are expressed as numbers or percentages.P-values < 0.05 were considered significant.

Results
Postoperative pseudo A LT fractures were seen in 28 patients (7 males [25% and 21 females [75%]).The age of the patients at surgery was 73.7 (range: 52-92) years.Of the cases, 27 (96.43%)occurred with non-cemented stems, and 1 (3.57%) with cemented stems (Fig. 4).The basic patient data are summarised in Table 1.At the time of the last follow-up, three patients were lost to contact, two died, and two were hospitalised elsewhere.The remaining 21 patients were followed for 49.3 ± 15.4 (24-74.4)months.These patients had a fracture healing time of 13.5 ± 1.1 (12-15.4)weeks.
Postoperative complications included thrombus in three cases (10.71%) and heart failure and pulmonary infection in one case each (both 3.57%).The patients with postoperative thrombosis all recovered on treatment with low-molecular-weight heparin.There were no revisions for prosthesis loosening, subsidence, or infection.The HHS and SF-36 scores were obtained scores those with complete follow-up (Table 2).

Discussion
The VCS divides fractures into three subtypes.Type A consists of fractures in the greater (A GT ) and lesser (A LT ) trochanteric areas.Type B fractures involve the metaphyseal or diaphyseal femur around the prosthetic stem: type B1 fractures occur around a stable stem; B2 fractures involve a loose stem with good bone quality; and B3 fractures involve bone loss.Type C fractures occur distal to the stem.
The UCS expanded and updated the VCS by adding three types.Type D is a femoral fracture occurring after hip or knee arthroplasty.Type E is a fracture of both the acetabulum and femur after hip arthroplasty.Type F is a fracture of the acetabulum after hemi-arthroplasty of the hip 4,6 (Table 4).
However, some special fractures are not included.For example, a pseudo A LT /new B2 fracture occurs at the lesser trochanter extending into the medial cortex of the proximal femur 7,8, 14-18 (Table 5).Consequently, modified classifications based on the VCS and UCS were proposed.Capello     9 .These are all actually B2/pseudo A LT fractures.Pseudo A LT fractures often occur with non-cemented stems, especially tapered stems 7,10,11 .This may be due to an unrecognised intraoperative fracture that is subsequently displaced under load or during rehabilitation [7][8][9]15,16 . Capell et al. reported nine newly described "clamshell" fractures that might be related to unrecognised fractures during operation and believed that they were directly related to the geometry of the tapered stem used 8 .Huang et al. mentioned that this type of fracture occurred at the lesser trochanter, including a segment of the proximal medial femoral cortex, and was associated with destabilisation of the stem 15 .Karam et al. showed that such fractures were significantly associated with non-cemented supports with an anatomical wedge-shaped design 16 .Egrise et al. thought this type of fracture can occur intraoperatively or in the early postoperative period with non-cemented implants, and is an occult intraoperative fracture 9 .
In our series, 27 patients with non-cemented stems (96.43%) sustained a PPFF.The mechanism was as described above.One fracture involved a cemented stem (3.57%); this patient had osteoporosis and it might have been an occult fracture that was not found intraoperatively.
There are three common treatments for new B2 fractures in PPFF: LSRIF, ORIF, and conservative treatment.When Van Houwelingen and Duncan recognised an undisplaced cortical crack, they performed cerclage cable fixation and revision with a longer stem, with protected weight-bearing for 6 weeks 7 .Capello et al. reported nine cases of pseudo A LT fracture caused by non-cemented femoral stems.They treated three cases conservatively without surgery because the stem subsidence stabilised during rehabilitation; the remaining six were treated by removing the loose stems and fixing the fracture and all healed well 8 ; the remaining six were treated by removing the loose stems and fixing the fracture and all healed well 8 .Karam et al. adopted wedge-shaped (10 cases) and straight (20 cases) fixation treatments 16 .Egrise et al. treated nine VB-LT1 and six VB-LT2 in patients in poor health nonoperatively, 13 VB-T2 stems with exchange and cerclage, three with isolated stem exchange, and two with internal fixation by cerclage 9 .Huang et al. 15 and González-Martín et al. 17 treated this type with LSRIF.
In this study, of the 21 patients with complete follow-up, 18 underwent LSRIF and ORIF was used in 3 cases.The patient outcomes were determined by postoperative fracture-healing and functional assessment scores (HHS and SF-36).LSRIF had the better curative effect, not only in terms of the above but also from the average healing time of 13.4 ± 1.1 weeks.The LSRIF patients were aged 71.8 ± 10.9 (52-86) years and were younger than the ORIF patients (88 [range: 85-91] years).The mental and physical SF-36 scores were better with LSRIF (50.9 ± 7.6 and   2), indicating good functional recovery after surgery.The HHS of case 1 was 68, which we attributed to the patient's poor basic condition and severe osteoporosis.Moreover, revision surgery with a cemented stem was more difficult than conventional LSRIF.Compared with LSRIF, ORIF is less invasive and is a suitable choice for the elderly with poor physical condition and low living needs 3 .Smitham et al. confirmed that the anatomical reduction and ORIF of type B2 fractures should be considered an appropriate treatment for frail elderly patients with a PPFF around cemented polished double-tapered stems 14 .This study has some limitations.The patients were from a single hospital, limiting the generalisability of the results.Given the retrospective nature of the study, multicentre prospective research is needed.

Conclusion
Pseudo A LT fractures result mainly from unrecognised occult fractures intraoperatively and improper exercise postoperatively.They involve a posteromedial cortical fragment around the lesser trochanter and mainly occur with non-cemented stems.While the main treatment is LSRIF with cables/plates, simple ORIF can achieve good healing, especially for elderly patients in poor general condition.

Figure 1 .
Figure 1.Case 2, a 60-year-old woman with a left femoral neck fracture due to a fall.(A) Preoperative lateral radiograph.(B) Postoperative anterior-posterior radiograph after longer stem revision and internal fixation with cables.
Ethical approval.Approved by the Ethics Committee of the First Affiliated Hospital of Zhejiang Chinese Medicine University, Ethics No. 2016-K-143-01.Informed consent.Informed consent was obtained from all participants.

Figure 2 .
Figure 2. Case 8, an 81-year-old woman.(A) Anterior-posterior radiograph after hemi-arthroplasty.(B) After a fall in the ward, the continuity of the left medial femoral cortex was disrupted on anterior-posterior and lateral radiographs.(C) Postoperative anterior-posterior radiograph shows the patient with a longer stem revision and internal fixation with cables.

Figure 3 .
Figure 3. Case 19, an 88-year-old woman.(A) Anterior-posterior radiograph after hemi-arthroplasty.(B-C) Two coronal CT images show the periprosthetic fracture.(D) Anterior-posterior radiograph shows the surgical treatment of open reduction and internal fixation with cables.

Figure 4 .
Figure 4. Case 1, an 82-year-old man underwent left hemi-arthroplasty with a cemented stem because of a left femoral neck fracture.(A-B) Two intraoperative radiographs taken with a C-arm machine.(C) After a fall, the continuity of the left medial femoral cortex was interrupted on an anterior-posterior radiograph.(D) Postoperative anterior-posterior radiograph after longer stem revision and internal fixation with cables.

Table 1 .
Basic patient data.LSRIF, longer stem revision and internal fixation with cables; ORIF, open reduction and internal fixation with cables.

Table 2 .
Patient demographic and medical data.PPFF, periprosthetic femur fracture; M, male; F, female; HHS, Harris Hip Score; ORIF, open reduction and internal fixation with cables; LSRIF, longer stem revision and internal fixation with cables.

Table 4 .
Summary of the Unified Classification System.THA, total hip arthroplasty.