Treatment of Delayed and Nonunited Fractures and Osteotomies with Pulsed Electromagnetic Field in Children and Adolescents

n Feature Article

Treatment of Delayed and Nonunited Fractures and Osteotomies With Pulsed Electromagnetic Field in Children and Adolescents

Melissa Y. Boyette, MD; Jose A. Herrera-Soto, MD

abstract Full article available online at Healio.com/Orthopedics. Search: 20120621-20

Nonunion of fractures or osteotomies in the pediatric population is rare. The gold standard for the treatment of nonunions involves harvesting autologous iliac crest bone graft and sometimes internal fixation, which are invasive procedures. The purpose of this study was to evaluate the effectiveness of pulsed electromagnetic field on a nonunited fracture or osteotomy in the pediatric population. 1A 1B A retrospective study was performed on all patients at the authors’ institution who used Figure 1: Anteroposterior (A) and lateral (B) radio- pulsed electromagnetic field as part of their treatment for nonunion or delayed union. graphs of the distal radius of an 11-year-old patient 8 months after open reduction and internal fixation Success of the initial nonunion treatment was defined as complete union of the frac- showing fracture nonunion, for which pulsed elec- ture or osteotomy site. Two types of treatment were administered once delayed bone tromagnetic field was initiated. healing was identified: pulsed electromagnetic field alone or pulsed electromagnetic field plus an adjunct treatment. Twenty-one patients were included; 8 osteotomies and 14 fractures developed a nonunion. Average patient age was 11.7 years. Average age for patients who healed with the initial treatment was 10.7 years, whereas nonhealers had an average age of 14 years. Eighty-nine percent of osteotomy nonunions healed with their first management. Fifty-seven percent of fracture nonunions healed at the first attempt.

The use of pulsed electromagnetic field is a good option for the initial treatment of pediatric nonunions, especially for patients who develop nonunions secondary to osteotomies. Adding bone marrow aspiration improves the outcomes and is minimally 2A 2B invasive compared with autologous iliac crest bone graft, with no complications. Figure 2: Anteroposterior (A) and lateral (B) radiographs of the distal radius 3 months after initiat- ing pulsed electromagnetic field showing fracture healing. Dr Boyette is from the Orlando Health Orthopedic Residency Program, and Dr Herrera-Soto is from the Department of Pediatric Orthopedics, Arnold Palmer Hospital for Children, Orlando, Florida. Drs Boyette and Herrera-Soto have no relevant financial relationships to disclose. The authors thank Jonathan H. Phillips for his editorial contribution to this article. Correspondence should be addressed to: Jose A. Herrera-Soto, MD, Department of Pediatric Or- thopedics, Arnold Palmer Hospital for Children, 83 W Columbia St, Orlando, FL 32806 (jose.herrera@ orlandohealth.com). doi: 10.3928/01477447-20120621-20

JULY 2012 | Volume 35 • Number 7 e1051 n Feature Article

onunion of fractures or osteotomies in the pediatric population Nis rare.1-4 The thicker periosteum found in younger patients makes them less prone to develop nonunions. Delayed union is defined in adults as a healing time of more than 12 weeks. However, no standard description exists of delayed or nonunion in the pediatric literature. Nonunion in children has been quoted as healing de- 1 2A 2B layed beyond 10 weeks.4 Figure 1: Anteroposterior radio- Figure 2: Anteroposterior (A) and lateral (B) radiographs of the distal The gold standard for the treatment of graph of an 11-year-old patient radius after open reduction and internal fixation. nonunions involves harvesting autologous showing a displaced distal ra- iliac crest bone graft.3-5 However, this dius fracture. method is invasive and associated with a high rate of complications, most com- fracture or osteotomy), patient age when without fixation. All patients were instruct- monly persistent donor-site pain.6 Some treatment began, and treatment type rec- ed to use the pulsed electromagnetic field fractures also require internal fixation to ommended to the patient once a delayed for 10 hours a day as part of their treatment stabilize the injury.3 This involves an ad- union or a nonunion was identified. regimen. Bone marrow aspiration or injec- ditional procedure to remove the implants Inclusion criteria were patients younger tion was considered a minimally invasive in children. than 16 years with documented delayed or procedure because all were performed per- At the authors’ institution, pulsed elec- nonunion and the use of pulsed electro- cutaneously. Three- to 4-mL aliquots were tromagnetic field has been used as the first magnetic field as part of the treatment plan. drawn from different bony sites through the line of treatment for nonunion for oste- Success of the initial nonunion treatment same skin entry point to obtain a total of 12 otomies and fractures. This noninvasive was defined as complete union of the frac- to 15 mL per procedure.7 method enhances bone healing. To the au- ture or osteotomy site. The latest follow-up thors’ knowledge, no studies have report- was at least to nonunion consolidation. The Results ed the use of pulsed electromagnetic field fracture or osteotomy site was considered All Patients in the treatment of pediatric nonunions. united when 4 cortical bridges were pres- Twenty-one patients were included; 1 The purpose of this study was to evaluate ent or the nonunited area was described patient had a bilateral osteotomy. Eight the effectiveness of pulsed electromagnet- as healed in the medical records. The time osteotomies and 14 fractures developed a ic field on a nonunited or delayed fracture of healing or failure was determined from nonunion (Figures 1, 2). Average patient or osteotomy in the pediatric population. the time the pulsed electromagnetic field age was 11.7 years (range, 3-15.7 years). was prescribed to cortical healing or to Average age was 13 years in the fracture Materials and Methods the need to use an additional procedure. group and 9.6 years in the osteotomy Institutional Review Board approval Failure was described as persistent non- group. Pulsed electromagnetic field alone was obtained to perform a retrospective union of the fracture or osteotomy site re- was used in 17 limbs, and 5 limbs had database search from July 2002 to June sulting in a secondary procedure. Patients some type of adjunct used (4 bone mar- 2009 at a single orthopedic practice on all lost to follow-up, older than 16 years, and row injections and 1 injection with nail patients who used pulsed electromagnetic treated with pulsed electromagnetic field dynamization). Average follow-up from field as part of their treatment. Nonunion for other reasons were excluded. Patients the moment pulsed electromagnetic field was defined as a lack of fracture healing with pathologic fractures, stress fractures, was administered was 12 months. progression on sequential radiographs or refractures were also excluded (n54). Nonunion or lack of fracture improve- or no evidence of healing more than 10 Two types of treatment were admin- ment was diagnosed at an average of 19.6 weeks following the injury.4 Delayed istered once a nonunion was identified: weeks (range, 9.1-42 weeks) (Figure 3). union was defined as no evidence of cal- pulsed electromagnetic field alone or pulsed One patient was less than 10 weeks from lus progression on sequential radiographs. electromagnetic field plus an adjunct treat- fracture or osteotomy at the time of de- Medical records for all patients were ment. Adjunct treatment was anterior iliac layed union diagnosis. He was a 4-year- reviewed. The collected data included crest bone marrow injection or autologous old with myelomeningocele who under- type of injury that led to nonunion (ie, iliac crest bone graft application with or went bilateral distal tibia osteotomies.

e1052 ORTHOPEDICS | Healio.com/Orthopedics Pulsed Electromagnetic Field | Boyette & Herrera-Soto

One osteotomy was diagnosed with an infection of the pin tracts with lack of union progression. This patient developed a wound dehiscence and an infected nonunited osteotomy by 4 weeks. The other leg developed a methicillin-resistant Staphylococcus aureus infection from oti- tis media at 6 weeks. He underwent pin re- moval, intravenous antibiotics, whirlpool therapy, and then casting once the wound 3A 3B 4A 4B had healed, with pulsed electromagnetic Figure 3: Anteroposterior (A) and lateral (B) radio- Figure 4: Anteroposterior (A) and lateral (B) radio- field for the nonunion. graphs of the distal radius 8 months postopera- graphs of the distal radius 3 months after initiat- Twelve (71%) of 17 nonunions treated tively showing fracture nonunion, for which pulsed ing pulsed electromagnetic field showing fracture with pulsed electromagnetic field only electromagnetic field was initiated. healing. healed at an average of 15.8 weeks (100% [6/6] in the osteotomy group and 55% [6/11] in the fracture group) (Figure 4). and bone grafting to successfully heal the Complications Three of 4 (75%) patients treated with osteotomy. Therefore, 88% of nonunited No patients who underwent bone mar- pulsed electromagnetic field and adjunct osteotomies healed with minimally to row aspiration or injection presented any treatment healed at 19.8 weeks (50% [1/2] noninvasive procedures. Four (50%) oste- type of morbidity from the harvested or in the osteotomy group and 100% [2/2] otomies were tibial derotation osteotomies. grafted sites. No complications occurred in the fracture group). Therefore, 68% A proximal femoral osteotomy required a with the use of pulsed electromagnetic (15/22) healed after initial treatment with second procedure. Autologous iliac crest field. No patient reported persistent pain minimally to noninvasive techniques at an bone graft was used to augment after a when open bone graft procedures were average of 15.8 weeks. The patient treated failed attempt with bone marrow injection performed. No compliance issues were with pulsed electromagnetic field, injec- and pulsed electromagnetic field. detected in the clinical notes because the tion, and dynamization did not heal due to use of the pulsed electromagnetic field poor fixation that eventually required an Fracture Group occurred during sleeping hours, which exchange for a more rigid nail. Fifty-seven percent (8/14) of fracture anecdotally was well-tolerated by the pa- Average age for patients who healed nonunions healed after the first treatment tients. with initial treatment was 10.7 years, (6/11 with pulsed electromagnetic field whereas the nonhealers had an average only, 2/2 with pulsed electromagnetic field Discussion age of 14 years (range, 12.8-15.5 years). plus bone marrow injection, and 0/1 with Two studies have evaluated delayed All patients younger than 12 years healed pulsed electromagnetic field with nail dy- unions or nonunions following fractures after the first attempt (n56). All patients namization). Average time to healing was in children.1,2 This complication is rare.1- who failed to unite with pulsed electro- 13.2 weeks (range, 6-22 weeks). Four 3,5 The current authors found no report magnetic field alone or with an adjunct (29%) fractures were open. Three of the 4 addressing the treatment of nonunion fol- healed with a secondary procedure. Those open fractures had been treated originally lowing osteotomy in children. The most treated with a secondary procedure healed with fixation and healed with pulsed elec- common causes of nonunion or incom- 12 weeks (range, 9-16 weeks) after the tromagnetic field alone or with bone mar- plete bone healing in children are com- secondary procedure. row injection. The open fracture that did not minuted or open fractures, severe soft unite despite pulsed electromagnetic field tissue loss, infection, fractures undergo- Osteotomy Group had been casted and healed with fixation, ing open reduction and internal fixation, Eighty-eight percent (7/8) of osteotomy pulsed electromagnetic field, and injection. and inadequately fixed fractures.3,5 In the nonunions healed after the first treatment All persistent nonunions healed after a sec- adult literature, delayed union has been (6/6 with pulsed electromagnetic field only ond surgical procedure, including internal defined as a healing time of 12 weeks or and 1/2 with bone plus added bone mar- fixation (n52), fibular osteotomy (n52), more. However, no standard description row injection). Average time to healing and nail exchange (n52). Therefore, 57% exists of delayed or nonunion in the pedi- was 18.8 weeks (range, 7-39 weeks). One of nonunited fractures healed with mini- atric literature. Greenbaum et al4 reported patient required a second open procedure mally to noninvasive procedures. that fractures in children younger than 9

JULY 2012 | Volume 35 • Number 7 e1053 n Feature Article

years unite sooner. As a result, they de- electromagnetic field for the treatment Several studies report on the technique fined a delayed union in children as im- of pediatric fracture or osteotomy non- and effectiveness of bone marrow injec- paired healing at 10 weeks. In the current unions. Schnoke and Midura14 examined tion.7,19,20 Connolly et al7 treated 20 non- study, the time to diagnose a nonunion or the phosphorylation kinetics and signal- united tibial fractures with bone marrow delayed union was more than 10 weeks in ing pathways induced by pulsed electro- injection and adequate fixation. Half of all but 1 patient. This patient was treated magnetic field and compared the intensity the patients were treated with casts and before due to infection of his osteotomy with that induced by other known anabol- the other half with intramedullary nails. that showed no progression of bone heal- ic agents parathyroid hormone and insu- Eighteen of the 20 tibial fractures healed, ing. Therefore, the diagnosis of nonunion lin. The authors found “that exposure of with the 2 nonunited fractures occurring in children should be made at 10 weeks bone cells to PEMF [pulsed electromag- in the cast immobilization group. The au- following the inciting event. netic field] induce immediate intracellular thors recommended aspirating between Schrader et al2 reviewed 43 pediatric signal transduction events consistent with 1- to 5-mL aliquots at a time because con- nonunions. When comparing diaphyseal those associated with anabolic bone cell tinued aspiration draws off mainly venous vs elbow nonunions, diaphyseal non- responses.”14 blood.7 Aspirating twice from the same unions had a shorter mean time to bony The effective dosage treatment of location decreases cell counts of mono- union (9 vs 11 months, respectively). pulsed electromagnetic field for fracture nuclear cells and mesenchymal stem cells Schmittenbecher et al1 reported a delayed nonunions has been studied. The reported in the second aspirate; however, it does union rate of 1.9% in pediatric forearm success rate was 35.7% for patients who not affect the biological characteristics. fractures treated with elastic intramedul- used the device less than an average of 3 Although an increased aspiration volume lary nails. All of these ultimately healed hours a day and 80% for patients who used higher than 8 mL reduced the cellular within 13 months. Although the current the device for more than 3 hours a day.15 concentration, the admixture of peripheral authors agree that most of these delayed The patients in the current study used the blood did not affect the biological charac- unions and nonunions may eventually device approximately 8 to 10 hours per teristics.7 heal, it seems unreasonable to prolong day (during sleep) as recommended by the The current authors considered the use closed treatment in a cast for such a long manufacturer. of a noninvasive treatment method, such time in lieu of minimally or noninvasive Nolte et al16 and Rutten et al17 reported as pulsed electromagnetic field, to avoid treatment. Avoidable and unnecessary the treatment results of using low-intensity the morbidity associated with harvesting joint stiffness from prolonged casting is pulsed ultrasound on nonunions. Rutten et autogenous iliac crest bone graft, which is another factor that influences the decision al17 reported healing rates over 70%. The the gold standard. Conway21 compared the to act earlier. The authors’ institution of- current study’s success rate with this de- use of iliac crest bone graft (anterior and fers both types of treatment. vice was 71% when used initially without posterior) with intramedullary canal graft- Pulsed electromagnetic field has been the need to perform open bone graft. ing. He found that iliac crest grafting had used successfully in the adult population.8,9 Connolly18 reported his experience donor-site complications, such as nerve One report exists in the literature of an iso- with various forms of electrical stimulation injury and persistent pain. Intramedullary lated case of a nonunited lateral condyle and evaluated how to select which type of canal grafting resulted in the fewest com- fracture treated by this method.10 Other re- nonunited fractures would be appropriate plications. No patient in the current study ports describe using pulsed electromagnet- for surgical intervention, electrical stimu- who underwent bone marrow aspiration ic field in animal models.11,12 These studies lation, or no treatment. Although he was or injection or open grafting presented demonstrate less bone volume loss, faster unable to justify a firm conclusion due to any type of morbidity from the harvested recovery of load bearing, increased new the small size, he concluded that electrical or grafted sites. bone formation, and higher mechanical stimulation does not compensate for in- In a series of 208 patients undergoing strength compared with the control side. adequate fracture fixation. This may have iliac crest bone graft harvest for anterior Abdelrahim et al13 reported similar find- been the case in the current study’s fracture lumbar interbody fusion, 41% reported ings when comparing mandible fractures patients because 43% of them required pain at 6 months, 33% at 1 year, and 31% treated with and without the use of pulsed fixation or revision of the fixation, whereas at 24 months.6 No significant differences electromagnetic field; those treated with 88% osteotomies healed after pulsed elec- existed between the anterior and posterior pulsed electromagnetic field showed better tromagnetic field alone or with bone mar- harvest sites or bicortical grafts. However, bone mineral density and stronger healing. row injection. The osteotomy patient that when Ahlmann et al22 compared anterior To the authors’ knowledge, the cur- required a second procedure healed with and posterior iliac crest bone graft har- rent study is the first to examine pulsed adding bone graft to the osteotomy site. vest, they found that postoperative site

e1054 ORTHOPEDICS | Healio.com/Orthopedics Pulsed Electromagnetic Field | Boyette & Herrera-Soto

pain was significantly more severe and of electromagnetic field will not be neces- 13. Abdelrahim A, Hassanein HR, Dahaba M. Effect of pulsed electromagnetic field on longer duration after anterior harvest. The sary, and a revision of the fixation should healing of mandibular fracture: a preliminary increased complication rate with anterior be performed to enhance stability. clinical study. J Oral Maxillofac Surg. 2011; harvest was likely due to the increased risk 69(6):1708-1717. of nerve injury and subsequent numbness eferences 14. Schnoke M, Midura RJ. Pulsed electromag- R netic fields rapidly modulate intracellular (lateral femoral cutaneous nerve). Other 1. Schmittenbecher PP, Fitze G, Gödeke J, signaling events in osteoblastic cells: com- studies have shown similar results follow- Kraus R, Schneidmüller D. Delayed healing parison to parathyroid hormone and insulin. 21-25 of forearm shaft fractures in children after in- J Orthop Res. 2007; 25(7):933-940. ing iliac crest bone graft harvest. tramedullary nailing. J Pediatr Orthop. 2008; 15. Garland DE, Moses B, Salyer W. Long-term 28(3):303-306. The current study had some limitations. follow-up of fracture nonunions treated with The number of patients was low; therefore, 2. Shrader MW, Stans AA, Shaughnessy WJ, PEMFs. Contemp Orthop. 1991; 22(3):295-302. Haidukewych GJ. Nonunion of fractures in no statistical significance could be estab- 16. Nolte PA, van der Krans A, Patka P, Jans- pediatric patients: 15-year experience at a level sen IM, Ryaby JP, Albers GH. Low-intensity lished. However, the authors reported their I trauma center. Orthopedics. 2009; 32(6):410. pulsed ultrasound in the treatment of non- doi: 10.3928/01477447-20090511-11 observations with 2 treatments. Also, the unions. J Trauma. 2001; 51(4):693-702. 3. Arslan H, Subasý M, Kesemenli C, Ersuz definition of nonunion in children is not 17. Rutten S, Nolte PA, Guit GL, Bouman DE, H. Occurrence and treatment of nonunion in Albers GH. Use of low-intensity pulsed ul- clearly established in the literature. No pa- long bone fractures in children. Arch Orthop trasound for posttraumatic nonunions of the tient was metabolically screened to assess Trauma Surg. 2002; 122(9-10):494-498. tibia: a review of patients treated in the Neth- nutrition or calcium or vitamin D content. 4. Greenbaum B, Zionts LE, Ebramzadeh E. erlands. J Trauma. 2007; 62(4):902-908. Open fractures of the forearm in children. J There was no way to know whether the pa- 18. Connolly JF. Selection, evaluation and indica- Orthop Trauma. 2001; 15(2):111-118. tients were fully compliant with the pulsed tions for electrical stimulation of ununited frac- 5. Lewallen RP, Peterson HA. Nonunion of long tures. Clin Orthop Relat Res. 1981; (161):39-53. electromagnetic field. Finally, no control bone fractures in children: a review of 30 cas- 19. McLain RF, Fleming JE, Boehm CA, Mus- group existed with which to compare the es. J Pediatr Orthop. 1985; 5(2):135-142. chler GF. Aspiration of osteoprogenitor cells study group. Some may argue that a pa- 6. Sasso RC, LeHuec JC, Shaffrey C. Iliac crest for augmenting spinal fusion: comparison of tient can be kept in a cast for as long as bone graft donor site pain after anterior lum- progenitor cell concentrations from the ver- bar interbody fusion: a prospective patient tebral body and iliac crest. J Bone Joint Surg 6 months prior to calling such a fracture satisfaction outcome assessment. J Spinal Am. 2005; 87(12):2655-2661. Disord Tech. 2005; 18(suppl):S77-S81. or osteotomy nonunited. In the authors’ in- 20. Muschler GF, Boehm C, Easley K. Aspira- stitution, this seems unreasonable because 7. Connolly JF, Guse R, Tiedeman J, Dehne R. tion to obtain osteoblast progenitor cells from other minimally invasive modalities exist Autologous marrow injection as a substitute human bone marrow: the influence of aspi- for operative grafting of tibial nonunions. ration volume. J Bone Joint Surg Am. 1997; to help patients. Clin Orthop Relat Res. 1991; (266):259-270. 79(11):1699-1709. 8. Bassett CA. The development and application 21. Conway JD. Autograft and nonunions: mor- Conclusion of pulsed electromagnetic fields (PEMFs) bidity with intramedullary bone graft versus The use of pulsed electromagnetic for ununited fractures and arthrodeses. Clin iliac crest bone graft. Orthop Clin North Am. Plast Surg. 1985; 12(2):259-277. 41(1):75-84. field is a good option for the initial treat- 9. Gossling HR, Bernstein RA, Abbott J. Treat- 22. Ahlmann E, Patzakis M, Roidis N, Shepherd ment of pediatric nonunions, especially ment of ununited tibial fractures: a com- L, Holtom P. Comparison of anterior and pos- for patients who develop nonunions sec- parison of surgery and pulsed electromag- terior iliac crest bone grafts in terms of har- netic fields (PEMF). Orthopedics. 1992; vest-site morbidity and functional outcomes. ondary to osteotomies. It is noninvasive 15(6):711-719. J Bone Joint Surg Am. 2002; 84(5):716-720. and demonstrates no long- or short-term 10. Das Sarkar S, Bassett CA. Healing of non- 23. Schwartz CE, Martha JF, Kowalski P, et al. Pro- morbidity. Adding bone marrow aspira- union of a fractured lateral condyle of the hu- spective evaluation of chronic pain associated tion improves outcomes and is minimally merus by pulsing electromagnetic induction. with posterior autologous iliac crest bone graft Contemp Orthop. 1991; 22(1):47-51. harvest and its effect on postoperative outcome. invasive when compared with autologous 11. Ibiwoye MO, Powell KA, Grabiner MD, et Health Qual Life Outcomes. 2009; 7:49. iliac crest bone graft, with no complica- al. Bone mass is preserved in a critical-sized 24. Silber JS, Anderson DG, Daffner SD, et al. tions. However, the source of the non- osteotomy by low energy pulsed electromag- Donor site morbidity after anterior iliac crest union and the fixation stability should be netic fields as quantitated by in vivo micro- bone harvest for single-level anterior cervi- computed tomography. J Orthop Res. 2004; cal discectomy and fusion. Spine (Phila Pa carefully scrutinized, especially in frac- 22(5):1086-1093. 1976). 2003; 28(2):134-139. ture cases, because some of these may 12. Midura RJ, Ibiwoye MO, Powell KA, et al. 25. Pieske O, Wittmann A, Zaspel J, et al. Autolo- benefit from a more rigid immobilization. Pulsed electromagnetic field treatments en- gous bone graft versus demineralized bone ma- hance the healing of fibular osteotomies. J trix in internal fixation of ununited long bones. If evidence exists of poor fixation, pulsed Orthop Res. 2005; 23(5):1035-1046. J Trauma Manag Outcomes. 2009; 3:11.

JULY 2012 | Volume 35 • Number 7 e1055