Cementless Primary or Revision Stem in the First-time Revision Hip Arthroplasty for Aseptic Stem Loosening? A Retrospective Comparative Study.

Materials and Methods


Introduction
Total hip arthroplasty (THA) is one of the most successful orthopedic procedure. With an aging population, the demand for THA is estimated to increase by 174% between 2007 and 2030 in the United States 1 . The increased number of primary THA procedures have led to an increase of revision THA procedures 1,2 . Aseptic loosening remains as one of the most common indications for revision THA 2 . The choice between cemented or cementless stems during the rst-time revision THA procedure remains controversial, since both types of stems have led to satisfactory long-term implant survival 3 . When using a cementless femoral stem, primary stability is of paramount importance, which can be challenging owing to varying degrees of bone loss 4 . Using a primary stem during the revision procedure is associated with advantages including easier implantation, sparing of diaphysis invasion and while preserving native bone stock 5,6 . On the contrary, bene ts of using a longer, revision stem includes increased contact area for osteointegration, which allows the bypass of metaphysis and achieve a more reliable xation in the diaphysis 4,6 .

Materials And Methods
This was a retrospective, cohort study conducted in a single tertiary referral hospital of xxxxxx, xxxxxxx.
The ethical approval was granted by the institutional review board of our hospital. Informed consent for participation was obtained from all patients and their legal guardians. All procedures were conducted in accordance to the Declaration of Helsinki and performed according to relevant guidelines and regulations. Our study period was from January 2010 to May 2019. We obtained medical record and images from xxxxxx xxxxxx xxxxxxx Hospital Orthopedic database. First, we collected the number of revision hip arthroplasty procedures (including total and partial revision) during this period according to xxxxxxx's National Health Insurance procedure codes: "PCS-64258B, PCS-64201B". Second, we reviewed patients that underwent surgery for aseptic stem loosening according to the ICD-10-CM code: "T84.03, T84.030, T84.031, T84.038 or T84.039". We reviewed medical records and images, and included patients ful lled the following criteria: (1) undergone the rst-time revision procedure for aseptic stem loosening of primary cementless THA, (2) above 18 years of age, (3) follow-up duration more than 24 months, (4) bone defect classi ed as Paprosky type I or II, and (5) revision with either cementless primary or revision stems. We have excluded patients who had undergone revision procedure for (1) revision cup (N=456) or liner only (N=71), (2) periprosthetic joint infection (N=100), (3) periprosthetic fracture (N=89), (4) recurrent dislocation (N=34), (5) broken stem (N=7), (6) revision with cemented stem (N=13), (7) with Paprosky type III or IV bone defect (N=6) (Fig. 1) Of the 78 patients included this study, 28 patients had undergone the revision procedure using a cementless primary stem (Fig. 2), while the other 50 patients had a cementless revision stem (Fig. 3). All of the procedures were performed by fellowship trained, orthopaedic surgeons. Decision of using a primary or revision stem was made based on surgeon's preference. The We reviewed the medical records of each patient and recorded age, sex, height, weight, body mass index, American Society of Anesthesiologists (ASA) grade, Charlson comorbidity index (CCI), type of revision procedure, surgical approach, Paprosky classi cation and follow-up duration. The mean age was 62.6±14.1 years. Twenty-three patients were female (29.5%) and 55 were male (70.5%). The mean body mass index was 26.9±4.4 kg/m 2 . Most of the patients were classi ed as ASA II (N=47, 60.3%) and III (N=23, 29.5%). The distribution of CCI were 0 (N=13, 16.7%), 1 (N=15, 19.2%), 2 (N=15, 19.2%), 3 (N=9, 11.5%), 4 (N=11, 14.1%) and more than 5 (N=15, 19.2%). Thirty-six patients (46.2%) underwent revision procedure for stem only, while the other 42 patients underwent revision THA (53.8%). All the patients were classi ed as Paprosky type I (N=67, 85.9%) or II (N=11, 14.1%) bone defect. The mean follow up duration after the revision procedure were 72.3±34.7 months (range, 24 to 132). (Table 1) All the images were examined by two senior authors (xxx, xxx). The diagnosis of aseptic loosening was made based on clinical symptoms, presence of radiolucent lines in three or more Gruen zones and/or stem subsidence more than 5mm on plain radiographs 25 , intra-operative ndings and multiple sets of intra-operative cultures. Stem subsidence of more than 5mm on serial plain radiographs was considered clinically relevant 26 . Paprosky classi cation 27 was used to evaluate proximal femoral bone defects.
During the perioperative period, we recorded operation time, intraoperative blood loss, preoperative and postoperative hemoglobin level, estimated blood loss, transfusion rate and amount, length of stay, 30days and 90-day readmission. After surgery, clinical condition and plain lms were evaluated monthly during the rst 3 months, then in 3-month intervals for the rst year, and annually thereafter. We evaluated the functional outcome of all the patients using Harris hip score (HHS) 28 at the last follow-up visit. We recorded both surgical and medication complications, number of patient having complications, reoperation rate and implant failure rate. Common surgical complications included surgical site infection (SSI), periprosthetic joint infection (PJI), intraoperative femur fracture, greater trochanter fracture, periprosthetic femur fracture, stem subsidence, aseptic stem loosening, dislocation, and nerve injury. The distance of stem subsidence of more than 5mm was considered with clinical relevance 29 . Common medical complications included acute coronary syndrome, congestive heart failure, acute kidney injury (AKI), deep vein thrombosis (DVT), pulmonary embolism, cerebrovascular disease, urinary tract infection, pneumonia and gastrointestinal bleeding.
Statistical analyses were performed using SPSS version 22 (IBM Corp., Armonk, NY). Descriptive statistics were performed for all available data. The Student's t test was used for comparing continuous variable. The Chi-square test was used for comparing discrete variable. When one or more of the cells in the contingency table had an expected frequency of less than 5, we performed the Fisher's exact test. Time-dependent analyses for implant failure was performed using Kaplan-Meier and differences between group curves were analyzed using the Log-rank test. Statistical signi cance was de ned as p-value < 0.05.

Discussion
In this study, we have validated that during the rst-time revision hip arthroplasty for aseptic stem loosening, the use of primary stem can lead to similar, satisfactory mid-term implant survival, and a lower incidence of overall surgical complication, intraoperative femur fracture and greater trochanter fracture, compared with the use of revision stem. Since the bone defects during the rst-time revision procedure was usually small (mostly Paprosky type I or II), cementless primary stem might be an effective alternative to revision stem in this clinical scenario.
In our study, the overall incidence of patient having complication was higher in the revision stem group (60.0% vs. 32.1%, p=0.018). Notably, the incidence of intraoperative femur fracture (28.0% vs. 7.1%, p=0.029) and greater trochanter fracture (16.0% vs. 0%, p=0.045) was higher. The use of a cementless long stem with a large diameter during a revision hip procedure has been validated with higher risk intraoperative fracture 32,33 . Moreover, increased magnitude of sagittal and coronal femoral bowing in the Asian population might further increase the risk of breach over the distal cortex 34,35 . The reported incidence of intraoperative fracture or greater trochanter fracture varied 7,8,10,12,13,15,23 . But there was a trend toward a higher rate of intraoperative fracture (0-64% vs 0.7-20.2%) 7,10,12,23 or greater trochanter fracture (6-19.5% vs 3.3%) 8,13,15 of using revision stem than the primary stem. Additional xation procedures for the greater trochanter or femur fracture might lead to longer operation time, which might be associated with an increased risk of short-term complications, including surgical site infection 36,37 . In our study, there was a trend toward a higher rate of surgical site infection in the revision group (18.0% vs. 3.6%, p=0.086). Notably, none of these patients had developed PJI or undergone reoperation. In our study, PJI accounted for the most common reason for reoperation (N=5 of 9, 55.6%) and implant failure (N=3 of 4, 75.0%). The PJI rate was similar in the primary (7.1%) and revision stem (6.0%) groups, which was comparable with the reported PJI rate associated with the use of primary (0.7-10%) 7,9,10,24 or revision stem (0-14%) 12,15,16,19 .
In our study, the mid-term implant survival of both primary and revision stem groups was satisfactory. The 5-year implant survival rate of the primary and revision stem group were 95.2% and 93.3%, respectively, which was similar to other studies 7,9,10,12,14,19 . The mid-term survival rate of using primary and revision stem in revision THA procedure ranged from 95-100% 7,9,10 and 86-94% 12,14,19 , respectively. The leading causes for failure following a revision THA procedure included aseptic loosening, instability and PJI 38,39 . In our study, we observed reliable osteointegration with low rates of aseptic stem loosening in both primary (0%) and revision (N=1, 2.0%) stem groups, which was similar to other studies that reported mid-term results following revision THA with primary stem (0-5.2%) 7,9 and revision stem (0-5%) 13,14,17 . In addition, the incidence of stem subsidence in the primary (7.1%) and revision stem group (8.0%) was similar to the incidence reported in other studies (0-19.5%) 7,13,15,21 . Of the 84 patients who had the rst-time revision for aseptic stem loosening, 78 (92.9%) patients had small bone defects (Paprosky type I or II) (Fig. 1). As a result, we hypothesized and validated that there might be a role for a cementless primary stem during the revision THA procedure. The potential advantages of using cementless primary stem included preservation of diaphyseal bone stock and less stress-shielding effect compared with the longer, extensively coated revision stem, in which the long-term incidence of stressshielding around a revision stem can be up to 30% 18,23,40 . A cohort study with long-term follow-up is necessary to further validate these bene ts.
There are some limitations of this study. First, the retrospective design of this study could have led to potential biases, including: 1) surgeries performed by multiple surgeons; 2) decision of using primary or revision stem based on surgeon's preference; 3) mixed primary and revision implant brands and 4) not a prospective, randomized design. Second, we did not routinely check bone mineral density on every patient since osteoporosis might have an impact on complications such as intra-operative femur fractures 33 . Despite that we did not study on senile patients with osteoporotic fracture (e.g. fracture neck fracture), the age in our cohort was ranged from 40 to 93. Osteoporosis in some of our patients should be considered an important confounding factor. Third, based on the limited number of this study, it would be underpowered to detect the difference of events with a lower incidence, such as medical complications.

Conclusion
In conclusion, for patients that underwent rst-time revision procedure for aseptic stem loosening with Paprosky type I or II defect, cementless primary stem might be a better alternative to revision stem, with lower risk of overall surgical complication, intra-operative fracture and a similar, satisfactory mid-term implant survival.

Not applicable
Authors' contributions Fu-Yuan Pai and Shang-Wen Tsai were responsible for conception and design, publication screening, acquisition of data, analysis and interpretation, and drafting and revising the manuscript. Te-Feng Arthur Chou, Hsuan-Hsiao Ma and Wei-Lin Chang were initial analysis and prepared tables. Po-Kuei Wu prepared gures. Cheng-Fong Chen, Po-Kuei Wu and Wei-Ming Chen were responsible for reviewing and revising the manuscript. All authors were involved with interpretation of the data. All authors discussed the results and commented on the manuscript. The author(s) read and approved the nal manuscript.

Competing interests
The authors have no con icts of interest to declare that are relevant to the content of this article.

Availability of data and material
All data generated or analysed during this study are included in this published article and available upon request.

Ethical approval
The study was approved by the ethical committee of Taipei Veterans General Hospital.

Informed consent
Informed consent was obtained from all patients and their families for publication of this study and any accompanying images.