Intrapelvic migration of the hip screw of a proximal femoral nail: report of two cases and review of the literature

INTRODUCTION

Proximal femoral nailing for the treatment of inter-subtrochanteric fractures was developed based on the principles of Dr Kutscher, it entered widespread clinical use and a number of devices with similar features were developed to cover the market needs. Currently, these devices are considered the main workhorse for the management of inter-subtrochanteric fractures in many parts of the world. Despite the clinical success of the proximal femoral nailing, a number of complications were associated with these devices. One of the most uncommon is intrapelvic migration of the hip screw.^1-18 Apart from the two presented cases in this article, twenty cases have been documented in the international literature in English language. Weil, Gardner et al attempted in 2008 to give a hypothesis in order to understand the biomechanical contributors to this phenomenon.¹⁹ In the laboratory, they simulated a basicervical fracture and tested a wide variety of commercial implants. They concluded that none of the tested implants was immune from medial migration and that a number of reasons can contribute to the phenomenon, such as the quality of the bone, the morphology of the fracture, intraoperative complications such as drill penetration of the articular surface, as well as the correct placement of the device. In the current article we would like to present two more case reports of medial migration of the femoral nail screw, review, and expand the existing literature in order to identify the trends in dealing with this rare complication.

CASE REPORT

An 88-year-old gentleman was seen in our hospital following a fall while at home. Despite his age, his overall health was good and claimed that he was able to mobilize unassisted before his fall. On radiographs he appeared to have a reverse oblique comminuted fracture proximal femur 31-A3.³ According to the AO classification with a long medial butterfly segment extending to the subtrochanteric region (Figure 1a and 1b). Due to the morphology of the fracture, it was decided by the operating surgeon to treat it with a long gamma nail (Stryker Howmedica, Kalamazoo, MI, USA) that was performed the following day (Figure 2). The operation was documented as uneventful, the fracture was reduced by traction and a 370 mm long gamma nail was inserted that was locked statically distally. The immediate postoperative period was uncomplicated. Due to his age, the patient was allowed to fully weight-bear and was discharged home three days following his operation. The stiches were removed in the community and the patient was seen upon his request 3 weeks following his discharge due to gradual increasing pain at his operated hip while ambulating.

On radiographs, medial intraabdominal migration of the femoral screw was noted (Figure 3a and 3b). He was readmitted in the orthopedic ward. On readmission, the patient was well, claimed he did not have any change of his bowel habits, he was afebrile, his abdomen was soft without evidence of any pathology and his lower limbs pulses were palpable and symmetrical arterial perfusion. Regarding his re-admission hematological profile, his hemoglobin was 800 g/dl and White Blood Cells 8,20×103 /μl of which 71.4% neutrophils. His C-Reactive Protein (CRP) was 157. The patient was taken to the operating theaters two days following his readmission once he had been hematologic ally optimized. Through an extensile lateral approach, the femoral nail and the femoral screw were removed with surprising relative ease. It was noted that the locking screw of the gamma nail was jammed and had never engaged the femoral-neck screw. There was no local evidence of any infection. The removed metalwork as well as tissue samples from the region were sent to microbiology. It was decided by the operating surgeon to proceed with a definite procedure and following reconstruction of the femoral shaft with cerclage wires, a cemented hemiarthroplasty was inserted (Figure 4a and 4b). His immediate postoperative period was uneventful. The microbiology reports of the intraoperative samples were received after six days that identified rare colonies of Enterobacter cloacae with multiple sensitivities. He was administered appropriate antibiotic therapy, however soon after there was wound dehiscence with increase of the inflammatory markers. During the following seven weeks the patient was taken to the operating theaters 3 times for wound debridement and removal of any loose metalwork. Microbiology samples taken during the reoperations revealed coagulase negative staphylococcus and Acinetobacter baumannii, both with multiple resistance in antibiotics. His general health progressively deteriorated due to continuing sepsis that resulted in respiratory and kidney failure. The patient deceased while in the hospital 62 days following his revision surgery.

Case report 2

A 66-year-old lady attended the emergency department of our hospital after injuring her left hip following a simple fall while at home. She had a history of Diabetes mellitus (DM), depression and hypothyroidism, all of which were controlled by oral medication. Clinically and radiologically she had sustained a multi-fragmentary pertrochanteric fracture, which according to the AO classification classified as 31-A.2.2. (Figure 5a and 5b). The patient was operated the following day with a proximal femoral nail (spectrum femoral nail, SanaMetal, Hungary) (Figure 6). The intraoperative and immediate postoperative period was uncomplicated, and the patient was allowed to fully weight bear the first postoperative day and discharged home three days following her operation. The patient was seen at the outpatient department at two weeks following the operation for stitch removal. No radiograph was obtained at that visit. Six weeks after her operation, the patient attended the clinic complaining for increasing pain at her operated hip and inability to walk. Radiographs demonstrated intraabdominal migration of the hip screw (Figure 7a and 7b). Computer tomography (CT) of the region showed that the femoral screw had been placed fairly superior and posterior in relation to the center of the hip, and intact bone was noted inferior and central-anterior regions of the femoral head. The patient was taken to the operation theater three days following the admission, and on the traction table, the old incision was opened, and the nail was removed. It was noted that the locking screw was missing, suggesting that the surgeon forgot to put it. The femoral screw was removed with relative ease, after managing to forward the guide wire in the screw stem (Figure 8a and 8b). The femoral screw was sent for microbiology. A gamma third generation nail (Stryker Howmedica, Kalamazoo, MI, USA) was inserted inferiorly and central-anteriorly in comparison to the previous track (Figure 9). A solid purchase was achieved intraoperatively. After the locking screw insertion, a small amount of cement was inserted at the supero-lateral aspect of the femoral screw at its junction with the femoral nail, in order to decrease any movement between the femoral screw and the nail. The immediate postoperative period was uncomplicated, and the patient was discharged home with instruction to mobilize only in bed for six weeks. Following this period, the patient was allowed to weight bear. At its latest follow up at two years post-revision surgery, the patient is happy, mobilizing well, without pain and without limitation.

DISCUSSION

DISCUSSION

To the best of our knowledge, our first presented case is the second described in the existing literature describing medial migration of a long reconstruction nail (long gamma nail, stryker, Minnesota, USA).¹⁰ In our case, the fracture involved the pertrochanteric area of the femur with multi-comminution of the medial buttress, extending distally. The surgery was documented as uneventful and the hip screw was inserted with a Tip-to-Apex Distance (TAD) marginally over 25. The operating surgeon did not insert the set screw properly and this probably contributed to the dislodgement of the hip screw. A one-stage revision was decided by the operating surgeon despite the presence of a high CRP on the grounds of the complete clinical absence of gastrointestinal pathology, his general good health and the absence of any intraoperative findings suggesting of an infection. However, the femoral nail was found to be infected. It is difficult to associate the failure of the implant with the infection and whether this contributed to the subsequent infection of the hemiarthroplasty as a variety of different microbes were cultured during the debridement of the infected wound. In respect of the second case there was significant comminution involving the medial buttress of the femur. The set screw was not inserted, allowing again mobility at the hip screw- femoral nail interface. During revision of the devise, a small quantity of cement was introduced at the superomedial junction of the hip screw with the femoral nail in order to reinforce the region and obstruct any mobility of the hip screw.

*DePuy ACE - Johnson&Johnson , Warsaw, IN, **Gamma nail system, Stryker, Mahwah, New Jersey, ***Intramedullary Hip Screw, Smith&Nephew, Memphis, Tennessee, ****PFNA Synthes, Oberdorf, Switzerland, *****Dyna locking trochanteric (DLT) nail, U&I, Uijeongbu, Korea, ****** Spectrum Femoral Nail, SanaMetal, Hungary

The complementation of the construct with cement has not been previously described in the literature to the best of our knowledge. One could argue that the cement could prevent the bony consolidation of the fracture at the area, but we felt that its limited use increased the stability of the construct especially in view of the tract of the previous implant and the excessive weight of the specific patient.

Following review of the existing literature, a common feature in most reported cases involves the comminution or insufficiency of the medial buttress of the femur. There are only two exceptions. In the first case that was published by Thein et al, the fracture was classified as 31-A3 according to the AO classification system with intact medial cortex, however the fracture was mal-reduced with a slight rotational component and one can speculate that this mal-reduction resulted in insufficiency in the medial buttress.¹¹ In the second case however, that was reported by Gomes PL et al, there was no mal-reduction and the medial cortex was intact.¹³ We can speculate therefore that insufficiency of the medial buttress of the femur is only one of the contributing factors that results in medial migration. Werner-Tutschku et al were the first to suggest a possible mechanism for medial migration in the PFN (Synthes Produktions, GmbH, Bettlach, Switzerland).²⁰ The specific nail has two proximal screws, a superior anti-rotational screw and an inferior larger lag screw. They noticed that in 7.1 % of their reported cases there was a proximal migration of the superior anti-rotational screw and lateral migration of the inferior lag screw. They called this phenomenon as the “Z-effect”. They believed that the most important single factor for the creation of the “Z-effect” was varus mal-reduction before nail insertion. Weil, Gardner et al produced a biomechanical study that attempted to reproduce the medial movement of the femoral screw in vitro.¹⁹ In their study, they tested five commercially used implant, the PFN-a (Synthes, Switzerland), the gamma-3 (Stryker, NJ, USA), the IMHS (Smith & Nephew, Memphis, TN, USA), the TFN (Synthes, Paoli, PN, USA) and the PFN (Synthes, Switzerland) in order to identify implant related factors that could contribute in the migration of the hip screw. They presumed a medial calcar insufficiency and mechanically loaded a hip simulator. According to their results, medial migration of the femoral neck component could be reliably reproduced in all tested devices. They felt that this phenomenon is an inherent property of the tested devices regardless of the number of femoral neck screw, different sliding interfaces, or the use of either lag screws or helical blades. On the clinical setting, according to the published cases, there are only five commercially used devises that have been associated with medial migration of the hip screw. One more commercial devise is reported in the current article. The devises are summarized in table 1. This discrepancy may be due to non-reporting of similar cases and not due to an inherent property of described cephalon-medullary devices. In our opinion a detrimental cause for the medial migration of the cervical screw, is the incorrect engagement of the set screw that fixes the cervical to the femoral component, present in most but not all devices. In our cases the set screw was either absent or jammed. In the existing literature there is mention of the set screw only in 11 reports, in one case the device was designed not to have a set screw and in two of them the authors mentioned that the set screw was not used in the primary procedure.¹ Overall, few of the authors focused on its existence or its insufficiency, something that is difficult to judge if not carefully inspected as the cervical component had already dislodged from the medullary implant. We would suggest a ring-construct at the lateral aspect of the cervical component, of a diameter that would be bigger than the hole of the medullary component that normally engages the cervical sliding screw. This mechanical feature would not forbid the medial displacement but would prevent the complete dislodgement of the cervical screw and would not allow excessive migration beyond the cussing layer of the iliacus muscle. The time period between the primary operation and the development of clinical symptoms varies widely and rages between 19 days 3 and 11 months 7 as seen in table 1. Many health authorities around the world do not review the elderly that have sustained an intertrochanteric fracture beyond the early postoperative period. We feel that the inconsistency of the presenting clinical symptom and the rarity of this complication does not justify regular radiographic surveillance as most of them have significant comorbidities and difficulties in travelling to their local hospital. Treatment options vary widely and tend to be individualized per case as seen in table 1.

In most cases the intramedullary device was revised in a hemiarthroplasty or a total hip arthroplasty whether in one or two stages.^1-7,10-12 Other options included revision to a Dynamic Hip Screw (DHS), exchange of the lag screw into a shorted one or removal of the metalwork.^{4,7,10,13,14,16,18} The quoted reasons for this variation included the morphology of the fracture, the progression of the healing process, the destruction of the femoral head and the acetabulum by the cervical component, the fear of laceration of intraabdominal content and the general health of the patient. Other factors that could have influenced the decision of the treatment could include the surgeons experience and the availability of appropriate alternative devises. Complications that have been reported include intra-operative death, infection intraoperative bleeding during the removal of the femoral screw and laceration of the intestines and early dislocation of a total hip replacement.^2,5,16 Follow up of these cases has not been systemically documented and it is very difficult to reach any safe conclusions. Out of the eight cases that the treating surgeon attempted to preserve the femoral head, four were documented to have varus malunion in the long term.^9,10,13,18 A summary in respect of follow up period and documented outcomes are presented in Table 2.

CONCLUSION

To conclude, we need to emphasize that medial migration of the hip screw in cases of cephalon-medullary nailing has not as yet been fully understood either on clinical grounds or biomechanically. Further reporting is of paramount importance and should be actively encouraged in order to reach safe conclusion as to the pathology of this rare complication. The medical industry should take in account the current and future observations in order to modify the commercial implants so as to avoid this potentially lethal complication.

Funding: No funding sources

Conflict of interest: None declared

Ethical approval: Not required

REFERENCES

1. 1. Ramkumar U, Saffar N, Thinakarajan TR, Parmar HV. Pelvic migration of lag screw from a nailing device. Injury Extra. 2006;37:53-55.
2. 2. Mark Tauber, Herbert Resch. Sigmoid perforation after medial migration of lag screw in gamma nailing. Arch Orthop Trauma Surg. 2006;26:118-122.
3. 3. Lucke M, Burghardt RD, Siebenlist S, Ganslmeier A, Stöckle U. Medial migration of lag screw with intrapelvic dislocation in gamma nailing-a unique problem? A report of 2 cases. J Orthop Trauma. 2010;24(2):e6-e1.
4. 4. Xinning Li, Heffernan MJ, Kane C and Leclair W. Medial pelvic migration of the lag screw in a short gamma nail after hip fracture fixation: a case report and review of the literature. J Orthop Surg Res. 2010;5:62.
5. 5. Heineman DJ, van Buijtenen JM, Heuff G, Derksen EJ, Pöll RG. Intra-abdominal migration of a lag screw in gamma nailing: report of a case. J Orthop Trauma. 2010;24(12):e119-e122.
6. 6. Flint JH, Sanchez-Navarro CF, Buckwalter JA, Marsh JL. Intrapelvic migration of a gamma nail lag screw: review of the possible mechanisms. Orthopedics. 2010;33(4).
7. 7. Robinson SJ, Fountain JR, Torella F, Pennie BH. Intrapelvic migration of a lag screw from a cephalomedullary femoral nail: a case report. Injury. 2011; 42(11):1384-6.
8. 8. Takigami I, Ohnishi K, Ito Y, Nagano A, Sumida H, Tanaka K, Shimizu Acetabular perforation after medial migration of the helical blade through the femoral head after treatment of an unstable trochanteric fracture with proximal femoral nail antirotation (PFNA): a case report. J Orthop Trauma. 2011;25(9):e86-9.
9. 9. Lal H, Sharma DK, and Mittal D. Intrapelvic migration of hip lag screw of proximal femoral nail-sequele to a paradoxical reverse Z effect and their critical analysis. J Clin Orthop Trauma. 2012;3(1):4853.
10. 10. Lozano-Alvarez C, Alier A, Pelfort X, Martínez-Díaz S, Puig L. Cervicocephalic medial screw migration after intertrochanteric fracture fixation, OTA/AO 31-A2, using intramedullary nail Gamma3: report of 2 cases and literature review. J Orthop Trauma. 2013;27(11):e264-7.
11. 11. Thein E, De Cannière A, Burn A, Borens O. Medial migration of lag screw after gamma nailing. Injury. 2014;45(8):1275-9.
12. 12. Takasago T, Goto T, Toki S, Hamada D, Yoshioka S, Tonogai I et al. Migration of the Lag Screw in Intramedullary Nailing. Case Rep Orthop. 2014;2014:519045.
13. 13. Gomes PL, Castelo LS, Lopes AL, Maio M, Miranda A, Dias AM. Pelvic migration of the helical blade after treatment of transtrochanteric fracture using a proximal femoral nail. Rev Bras Ortop. 2016;51(4):482-5.
14. 14. Pinheiro AC, Alpoim B, Félix A, Alves C, Sousa C, Rodrigues A. Medial migration of the intramedullary Gamma 3nail- a case report. Rev. bras. ortop. 2016;51(6).
15. 15. van Hoef S, Fuchs MC1, Ten Broeke RH. Late Occurring Medial Migration of a Lag Screw in Gamma Nailing. Case Rep Orthop. 2016;2016:5201674.
16. 16. Lee JW, Cho HM, Seo JW. Intrapelvic Penetration of Lag Screw in Proximal Femoral Nailing - A Case Report. Journal of the Korean Fracture Society. 2017;30(4).
17. 17. Kim YW, Kim PKJ, Lee SW. Intrapelvic Migration of the Lag Screw With Wedge Wing From Dyna Locking Trochanteric Nail: A Case Report and Literature Review. Hip Pelvis. 2019;31(2):110-119.
18. 18. Nayak M,Yadav R, Ganesh V and Digge V. An unusual case of femoral head perforation following fixation with proximal femoral nail antirotation (PFNA-II) for an unstable intertrochanteric fracture: Case report and literature review. Trauma Case Rep. 2019;20:100178.
19. 19. Weil YA1, Gardner MJ, Mikhail G, Pierson G, Helfet DL, Lorich DG. Medial migration of intramedullary hip fixation devices: a biomechanical analysis. Arch Orthop Trauma Surg. 2008;128(2):227-34.
20. 20. Werner-Tutschku W, Lajtai G, Schmiedhuber G, Lang T, Pirkl C, Orthner E. Intra- and perioperative complications in the stabilization of per- and subtrochanteric femoral fractures by means of PFN Unfallchirurg. 2002;105(10):881-5.