- Email: [email protected]
Factors Associated with Nonunion, Delayed Union, and Malunion in Foot and Ankle Surgery in Diabetic Patients
The Journal of Foot & Ankle Surgery 52 (2013) 207–211
Contents lists available at ScienceDirect
The Journal of Foot & Ankle Surgery journal homepage: www.jfas.org
Factors Associated with Nonunion, Delayed Union, and Malunion in Foot and Ankle Surgery in Diabetic Patients Naohiro Shibuya, DPM, MS, FACFAS 1, Jon M. Humphers, DPM 2, Benjamin L. Fluhman, DPM 2, Daniel C. Jupiter, PhD 3 1 Associate Professor, Department of Surgery, Texas A&M Health and Science Center College of Medicine; Acting Chief, Section of Podiatry, Central Texas Veterans Affairs Health Care System; and Staff, Scott and White Memorial Hospital and Clinics, Temple, TX 2 Third Year Resident, Scott and White Memorial Hospital, Texas A&M Health and Science Center College of Medicine, Temple, TX 3 Assistant Professor, Department of Surgery, Texas A&M Health and Science Center College of Medicine; and Research Scientist, Scott and White Memorial Hospital and Clinics, Central Texas Veterans Affairs Health Care System, Temple, TX
a r t i c l e i n f o
a b s t r a c t
Level of Clinical Evidence: 3 Keywords: bone Charcot foot complication diabetes mellitus hyperglycemia peripheral neuropathy surgery
The incidence of bone healing complications in diabetic patients is believed to be high after foot and ankle surgery. Although the association of hyperglycemia with bone healing complications has been well documented, little clinical information is available to show which diabetes-related comorbidities directly affect bone healing. Our goal was to better understand the risk factors associated with poor bone healing in the diabetic population through an exploratory, observational, retrospective, cohort study. To this end, 165 diabetic patients who had undergone arthrodesis, osteotomy, or fracture reduction were enrolled in the study to assess the risk factors associated with nonunion, delayed union, and malunion after elective and nonelective foot and/or ankle surgery. Bivariate analyses showed that a history of foot ulcer, peripheral neuropathy, and surgery duration were statistically significantly associated with bone healing complications. After adjusting for other covariates, only peripheral neuropathy, surgery duration, and hemoglobin A1c levels >7% were significantly associated statistically with bone healing complications. Of the risk factors we considered, peripheral neuropathy had the strongest association with bone healing complications. Ó 2013 by the American College of Foot and Ankle Surgeons. All rights reserved.
The relationship between hyperglycemia and adverse outcomes after surgery has been widely documented (1–11). Long-term glucose control, as measured by glycosylated hemoglobin (hemoglobin A1c [HbA1c]), has been recognized as a major risk factor for negative outcomes after major surgery, including coronary artery procedures, and, in particular, for myocardial infarction and cardiac events after vascular surgery (1,2,5,10). In foot and ankle surgery, poorly controlled and complicated diabetes have also been shown to be significant risk factors for postoperative soft tissue and bone healing complications (12–16). Diabetic bone healing complications, in particular, have been extensively studied in animal models (17–41). In humans, the incidence of bone healing complication in diabetic patients is also believed to be high after foot and ankle surgery (13,42–48). However, few have
Financial Disclosure: None reported. Conflict of Interest: None reported. Address correspondence to: Naohiro Shibuya, DPM, MS, FACFAS, Associate Professor, Department of Surgery, Texas A&M Health and Science Center College of Medicine, Section of Podiatry, Central Texas Veterans Affairs Health Care System, Scott and White Memorial Hospital and Clinics, Temple, TX 76502. E-mail address: [email protected] (N. Shibuya).
conducted risk factor assessments within a diabetic population. Although the association of hyperglycemia with bone healing complications has been well documented (13,18,24,25,32,33,35,39,49,50), little clinical information is available to show which diabetes-related comorbidities directly affect bone healing. In diabetic animal models, many theories have been proposed to explain what causes the disturbance in normal bone metabolism. However, translational research to link each theory, often described in terms of molecular pathways to diabetic sequelae and complications in humans, is lacking. Our goal with the present investigation was to identify the risk factors associated with bone healing complications in a diabetic population using an exploratory, observational, retrospective, cohort study design. Understanding such risk factors could serve as a bridge between clinical and laboratory or animal research and promote a better understanding of the pathophysiology of diabetic bone healing. Patients and Methods A query of electronic medical records was performed to find diabetic patients who had undergone foot and ankle surgery during a 5-year period from January 2007 to December 2011 at the Scott and White Health Care System. The data from patients with an International Classification of Diseases, 9th Revision, code for all subcategories of
1067-2516/$ - see front matter Ó 2013 by the American College of Foot and Ankle Surgeons. All rights reserved. http://dx.doi.org/10.1053/j.jfas.2012.11.012
208
N. Shibuya et al. / The Journal of Foot & Ankle Surgery 52 (2013) 207–211
diabetes mellitus (code 250) and with Common Procedural Terminology codes of 27600 to 28899 were acquired by an information technology application analyst at our research department. Our institution’s internal review board approved our observational study protocol. The medical charts were then examined for the preoperative HbA1c values (recorded within 3 months before the date of surgery), other pre- and intraoperative risk factors, and outcome measures, including bone healing complications (Table 1). Patients with well-established diabetes, regardless of subtype, who had been treated pharmacologically, were included in the present study. The exclusion criteria included more than a 2-day course of antibiotic use within 1 week before surgery, the current use of immunosuppressant medications (including glucocorticoids, antibodies, cytostatics, drugs acting on immunophilins, and other drugs specifically used for immunosuppression), age older than 80 or younger than 12 years, postoperative follow-up duration less than 3 months, and a lack of radiographs to determine the progress of bone healing. The pooled diabetic patients were then screened for the type of foot and ankle procedures that had been performed. Nonosseous procedures were excluded. Those who had undergone both osseous and soft tissue surgery were included in the present study. Of those who had undergone osseous procedures, only arthrodesis, osteotomy, and/or reduction of fracture or dislocation were included in the study. Those who had undergone ostectomy (i.e., retrocalcaneal exostectomy, cheilectomy) or implant arthroplasty (i.e., total ankle replacement, first metatarsophalangeal implant arthroplasty), were excluded. For the remaining patients, the prevalence of bone healing complications was determined. For the purposes of our study, we defined bone healing complications as 1 or more of the following: nonunion, delayed union, or malunion. Delayed union was defined as no radiographic signs of bone healing within 3 months postoperatively. Nonunion was defined as no radiographic signs of bone healing for more than 6 months. Radiographic signs of bone healing were defined as the loss of visibility of osteotomy, arthrodesis, or fracture site in at least 2 of 3 views on the radiographs. To achieve consistency, the radiographs were reviewed by a single rater (J.H.). Malunion was defined as progression of deformity after original fixation, regardless of patient compliance. If an inadequate reduction or correction was seen in the initial radiographs and remained misaligned, the case was not considered malunion. Development of Charcot was recorded if a surgeon recognized the disease and documented it in their progress notes, regardless of method by which the condition was diagnosed. In this study, each patient was considered 1 case or event. When more than 2 procedures were performed on the same patient, only the earlier procedure was recorded. This measure was taken to eliminate consideration of revisional procedures and duplication of patient characteristics. However, if a patient underwent multiple arthrodeses, osteotomies, or fracture reductions on the same day, only 1 of the joints, osteotomies, or fracture sites had to be affected for the patient to be registered as having a bone healing complication. Similarly, if a patient showed nonunion in multiple joints, osteotomies, or fracture sites, it was regarded as 1 bone healing complication event. Forefoot procedures were defined as procedures at, or distal to, the tarsometatarsal joints. For example, open reduction of Lisfranc’s joint was recorded as a forefoot procedure. The rest of the procedures were considered rearfoot or ankle procedures. If a patient had undergone forefoot and rearfoot/ankle procedures, the case was recorded as “both.” When the medical records indicated current use of tobacco products, regardless of amount, the patient was considered positive for tobacco use. The presence of neuropathy was also determined from the medical records. Any sensory deficit noted in the medical records was considered neuropathy, regardless of the magnitude, type, or diagnostic test used to identify the neuropathy. Other comorbidities, including peripheral vascular disease and a history of ulceration were determined from the medical records. The body mass index was calculated from the weight and height
Table 1 Preoperative, intraoperative, and postoperative variables recorded Pre- and intraoperative variables Age Gender Body mass index Hemoglobin A1c within 90 days before surgery Preoperative random serum glucose drawn within 30 days of surgery Documented peripheral neuropathy Tobacco use (yes vs no) Operative time (10-min increments) Documented peripheral arterial disease (yes vs no) History of ulceration or infection in foot and ankle (yes vs no) Location of surgical procedure (forefoot vs rearfoot or ankle vs both) Postoperative outcome variables Malunion, delayed union, or nonunion (yes vs no) Acute, surgically induced Charcot neuroarthropathy (yes vs no)
measured by the anesthesia department during the preoperative visit. The surgery duration was obtained from the operating room records. Statistical Analysis The prevalence of the overall complications and each type of bone healing complication in our diabetic cohort was calculated. Bivariate analyses using t tests or analysis of variance or chi-square or Fisher’s exact tests were performed to compare those with and without a bone healing complication in terms of the various preoperative and intraoperative variables. A logistic regression model for the occurrence of bone healing complications was then built, including covariates that reached a significance level of 0.2 on bivariate analysis. Age, gender, and HbA1c were included in the model, regardless of the results from the bivariate analysis. For both bivariate and regression analyses, p .05 was considered statistically significant. The statistical analyses were performed using the R statistical package (R Developmental Core Team, R: A Language and Environment for Statistical Computing 2009, available from http:// www.R-project.org).
Results A total of 322 diabetic patients had HbA1c values documented within 3 months of the date of their bone surgery. Of these, 269 patients had undergone at least 1 osseous procedure. After excluding those patients who had undergone osseous procedures that did not require bone healing, 177 patients who had undergone osteotomy, arthrodesis, and/or fracture reduction were included in the present study. Of these patients, 11 had undergone more than 1 foot and ankle procedure on different occasions; only the first was included in our analyses. One patient had undergone 3 separate foot and ankle procedures, and 2 of these procedures were excluded. This left a total of 165 cases in our final analysis. The mean age of the selected cohort was 62.1 9.71 (range 37 to 80) years. Of the 165 patients, 115 (69.7%) were women and 50 (30.3%) were men; 112 (67.9%) were white, 32 (19.4%) were other races, 18 (10.9%) were black, 2 (1.2%) were Hispanic, and 1 (0.6%) was unidentifiable (relative to race). One patient (0.61) was missing data necessary to calculate the body mass index. For the remainder, the mean body mass index was 34.2 7.40 (range 21.0 to 56.3) kg/m2. Of the patients, 59 (35.7%) had peripheral neuropathy, 21 (12.7%) had a history of foot ulcer, 20 (12.1%) had a history of tobacco use, 11 (6.7%) had peripheral vascular disease, and 77 (46.7%) had at least 1 of these comorbidities. The duration of surgery was not obtainable for 11 patients (6.67%); however, for the remainder, the mean duration was 96.1 44.54 (range 21 to 237) minutes. The mean HbA1c was 7.2% 1.53% (range 4.8% to 14.9%). The serum glucose values were missing for 12 patients (7.27%). For the remainder, the mean glucose value was 156.8 66.18 (range 41 to 572) mg/dL. Of the procedures, 84 (50.9%) were in the forefoot, 78 (47.3%) in the rearfoot or ankle, and 3 (1.8%) were in both. A total of 123 patients (74.6%) had no bone healing complication. Of the patients with complications, 13 (7.9%) had delayed union, 19 (11.5%) nonunion, and 10 (6.1%) malunion. The overall prevalence of bone healing complication in our diabetic cohort was 25.4% (42/165). Four patients (2.42%) developed Charcot arthropathy, which resulted in nonunion in 2 and malunion in 2. Of these 4 Charcot arthropathies, 2 developed after open reduction and internal fixation of an ankle fracture, 1 after Lisfranc’s fracture reduction, and 1 after tibiotalocalcaneal arthrodesis. The corresponding HbA1c levels were 9.1%, 6.1%, 9.6%, and 7.3%. The results of the bivariate analyses are presented in Table 2. Of the variables tested, the presence of 1 or more comorbidity, neuropathy, a history of ulcer, and the duration of surgery showed significant statistical differences between those with bone healing complications and those without bone healing complications. In a logistic regression model for the presence of a bone healing complication with the covariates of age, HbA1c level, duration of surgery, anatomic location of the procedure,
N. Shibuya et al. / The Journal of Foot & Ankle Surgery 52 (2013) 207–211
Variable
Age (y) Gender Female Male Race White Other Black Hispanic Unidentifiable Body mass index (kg/m2) Comorbidity Yes No Neuropathy Yes No Peripheral vascular disease Yes No Tobacco use Yes No History of ulcer Yes No Surgery time (min) Hemoglobin A1c (%) Perioperative serum glucose (mg/dL) Anatomic location of procedure Forefoot Rearfoot/ankle Both * y z
Bone Healing Complication No (n ¼ 123)
Yes (n ¼ 42)
63 10.20
60 7.99
88 35
p Value
.07* .38y
27 15 .37z
87 22 12 1 1 34.4 7.77
25 10 6 1 0 33.7 6.53
46 88
31 11
33 90
26 16
7 116
4 38
14 108
6 36
11 112 90 44.09 7.2 1.62 158.1 52.36
10 32 113 41.69 7.4 1.24 156.3 70.78
.61* .001y
.001y
.474z
.63y
.013y
67 53 3
.004 .42 .87 .14z
17 25 0
Probabilities of null hypothesis (p values) were computed using Student’s t test. Probabilities of null hypothesis (p values) were computed using the chi-square test. Probabilities of null hypothesis (p values) were computed using Fisher’s exact test.
neuropathy, and ulcer, the only significant factor was the duration of surgery (odds ratio [OR] 1.01, 95% confidence interval [CI] 1.00 to 1.02). After exploring the data further, we built a model that would be more clinically relevant by categorizing patient age by decade, dichotomizing the HbA1c into >7% or 7%, and coding the duration of surgery in increments of 10 minutes. Thus, our final model included age (by decade), the presence of neuropathy (yes or no), a history of ulcer (yes or no), HbA1c (>7% or 7%), and duration of surgery (by 10-minute increments). The factors that were statistically significantly associated statistically with the occurrence of a bone healing complication in this model included the duration of surgery (OR 1.15, 95% CI 1.04 to 1.26), peripheral neuropathy (OR 3.93, 95% CI 1.16 to 9.59), and HbA1c >7% (OR 2.88, 95% CI 1.23 to 6.72; Table 3). When we added the presence of a comorbidity and the location of surgery to this model, both the HbA1c level and the duration of surgery remained significant factors.
Discussion Our results have shown that approximately 1 of 4 diabetic patients (25.6%) will have 1 or more bone healing complications. Although the severity of the complication was not captured in the present study, the overall prevalence was striking and clinically significant. In our study, neither the perioperative serum glucose nor the HbA1c proportion differed statistically significantly between those with a bone healing complication and those without such a complication in our bivariate analysis. The mean HbA1c level for those with and without bone healing complications was 7.2% and 7.4%, respectively. After post hoc exploration of the data, we found that the group without bone healing complications had more apparent outliers in the upper HbA1c levels (Fig.). Those outliers were mostly nonelective cases (i.e., open reduction and internal fixation of ankle fractures). Because malleolar fractures have a relatively high union rate with fewer complications than elective arthrodesis or osteotomy procedures (12,13,43,51,52), this might have artificially elevated the mean HbA1c level in the no-complication group. Overall, only 4 bone healing complications (12.9%) occurred among 31 open reductions and internal fixations of malleolar fractures in our diabetic cohort. In the regression analysis, HbA1c showed a significant association with the presence of a postoperative bone healing complication (OR 2.88, 95% CI 1.23 to 6.72) when it was dichotomized as >7% or <7%. We suspect that this dichotomization minimized the effect of the outliers. Thus, we found that patients with HbA1c >7% were approximately 3 times as likely to have a bone healing complication than patients with HbA1c <7%. Having at least 1 comorbidity was also significantly associated with the development of bone healing complications. Specifically, on post hoc bivariate analysis, neuropathy (chi-square, p ¼ .001) and a history of ulcer (chi-square, p ¼ .013) were significantly associated statistically with the development of bone healing complications. In particular, 44% (26/59) of neuropathic patients had 1 or more bone healing complications (Table 2). More impressively, 50% (26/42) of those with bone healing complication had neuropathy. After adjusting for other covariates, a history of ulcer was not a significant factor for bone healing complications. Although a history of ulceration is often linked to amputation and mortality risk (53–58), we suspect that this factor might be confounded by neuropathy in the setting of bone
14
12 HA1c (%)
Table 2 Results of bivariate analysis of recorded variables vs occurrence of bone healing complications
209
10
8 Table 3 Results of multivariate logistic regression analyses (N ¼ 165) Variable
OR
95% CI
p value
Age (10-y increment) HbA1c (>7%) Surgery time (10-min increment) History of ulcer (positive) Neuropathy (positive)
0.86 2.88 1.15 1.24 3.93
0.58–1.28 1.23–6.72 1.04–1.26 0.38–4.03 1.16–9.59
.470 .014* .004* .720 .003*
Abbreviations: CI, confidence interval; HbA1c, hemoglobin A1c; OR, odds ratio. * Statistically significant.
6
No complication
Complication
Fig. Box plots revealing outliers mostly among those with no bone healing complication (N ¼ 165).
210
N. Shibuya et al. / The Journal of Foot & Ankle Surgery 52 (2013) 207–211
healing. This result has confirmed the results from many animal and bench studies and reports that focused on diabetic bone healing complications, which have often described the complication as a malfunction of bone metabolism resulting from neuropathy (45,59– 62). This is an interesting finding, especially considering that osteolytic activities, including Charcot arthropathy in patients with diabetes, are believed to be strongly linked to neuropathy rather than to hyperglycemia itself. Many Charcot cases have presented in nondiabetic neuropathic patients (60,63–71). The lack of adequate neuropeptide release in the neuropathic patient might upregulate osteoclastogenesis while downregulating osteoblastic activities (59,72–74). Although the precise relationship between neuropathy and poor bone healing in patients with diabetes undergoing foot and ankle surgery is uncertain, our results have indicated that it is the single most important factor among all the variables we tested. In our study, we found that those with neuropathy had an OR of 4 in favor of the development of bone healing complications compared with those without neuropathy. The odds of experiencing a bone healing complication in the presence of neuropathy were greater than those associated with hyperglycemia, as measured by the serum glucose or HbA1c level. A longer duration of surgery was also associated with complicated bone healing after adjusting for other covariates. A longer duration of surgery might indicate, but is not restricted to, a more complex case, more dissection, an adverse event occurring during surgery, or a procedure conducted by a slow surgeon. Although we did not investigate the reasons for the longer operative time, all these reasons are, intuitively, reasonable causes of bone healing complications. In our analysis, we found that each additional 10 minutes of surgery increased the odds of a bone healing complication by a factor of 15%. We recognize several shortcomings of the present study. First, we relied on progress notes written by each surgeon to detect neuropathy, peripheral vascular disease, a history of foot ulcer, tobacco use, and the development of Charcot. We had no systematic method to measure these variables. We suspect that these conditions were therefore underreported. Also, the magnitude of these risk factors was not captured. All these variables were dichotomized as “yes” or “no,” and we did not make any distinction between severe and minor neuropathy. Second, we designated 1 rater to review all the radiographic outcomes. The rater was not unaware of the other variables. Although measurement error could have occurred at random, bias stemming from knowing the other findings could have affected our final result. Third, the surgeons were not unaware of each patient’s medical history. Naturally, high-risk patients are less likely to undergo elective surgery. This would not only lead to lower enrollment of patients with uncontrolled diabetes, but would also oversample emergency cases among the high-risk patients. Also, surgeons naturally change their fixation method, dissection technique, and followup protocol according to the findings from their risk assessment. Therefore, the effect of potential risk factors might have been underestimated in the present study, reflecting current ideas about the effect of the HbA1c level on outcomes. Finally, extracting patient data from the electronic medical record system using “International Classification of Disease, 9th revision,” and Common Procedural Terminology codes could have potentially underrepresented the patients with borderline or undiagnosed diabetes. However, we believe that it is more reliable to use the “International Classification of Disease, 9th revision,” data from our institution’s medical record system to detect diabetes than to rely on the surgeons’ progress notes or billing information. Because any other service can record the diagnosis of diabetes in our system, the risk of missing a diagnosis of diabetes would be less for patients undergoing foot and ankle surgery. In conclusion, within the limitations of the present study, that factors that affected bone healing in diabetic patients after foot and
ankle surgery were HbA1c >7%, duration of surgery, and peripheral neuropathy. Among all the variables we analyzed, neuropathy was the single most significant factor that affected bone healing in patients with diabetes. Although randomized clinical trials might not be practical or ethical in this setting, a longitudinal study with better, systematically defined risk factors is warranted to confirm our findings. Furthermore, translational research to confirm the laboratory and animal study findings are needed to better understand the pathophysiology of diabetic bone healing. Future studies should also focus on quantifying each risk factor to determine the clinically useful cutoff line at which the risk of surgery exceeds the benefit. References 1. Acott AA, Theus SA, Kim LT. Long-term glucose control and risk of perioperative complications. Am J Surg 198:596–599, 2009. 2. Dronge AS, Perkal MF, Kancir S, Concato J, Aslan M, Rosenthal RA. Long-term glycemic control and postoperative infectious complications. Arch Surg 141:375–380, 2006. 3. Glassman SD, Alegre G, Carreon L, Dimar JR, Johnson JR. Perioperative complications of lumbar instrumentation and fusion in patients with diabetes mellitus. Spine J 3:496–501, 2003. 4. Golden SH, Peart-Vigilance C, Kao WH, Brancati FL. Perioperative glycemic control and the risk of infectious complications in a cohort of adults with diabetes. Diabetes Care 22:1408–1414, 1999. 5. Halkos ME, Puskas JD, Lattouf OM, Kilgo P, Kerendi F, Song HK, Guyton RA, Thourani VH. Elevated preoperative hemoglobin A1c level is predictive of adverse events after coronary artery bypass surgery. J Thorac Cardiovasc Surg 136:631–640, 2008. 6. Lamloum SM, Mobasher LA, Karar AH, Basiony L, Abdallah TH, Al-Saleh AI, Al-Shamali NA. Relationship between postoperative infectious complications and glycemic control for diabetic patients in an orthopedic hospital in Kuwait. Med Princ Pract 18:447–452, 2009. 7. Marchant MH Jr, Viens NA, Cook C, Vail TP, Bolognesi MP. The impact of glycemic control and diabetes mellitus on perioperative outcomes after total joint arthroplasty. J Bone Joint Surg Am 91:1621–1629, 2009. 8. Mraovic B, Suh D, Jacovides C, Parvizi J. Perioperative hyperglycemia and postoperative infection after lower limb arthroplasty. J Diabetes Sci Technol 5:412–418, 2011. 9. Noordzij PG, Boersma E, Schreiner F, Kertai MD, Feringa HH, Dunkelgrun M, Bax JJ, Klein J, Poldermans D. Increased preoperative glucose levels are associated with perioperative mortality in patients undergoing noncardiac, nonvascular surgery. Eur J Endocrinol 156:137–142, 2007. 10. O’Sullivan CJ, Hynes N, Mahendran B, Andrews EJ, Avalos G, Tawfik S, Lowery A, Sultan S. Haemoglobin A1c (HbA1C) in non-diabetic and diabetic vascular patients: is HbA1C an independent risk factor and predictor of adverse outcome? Eur J Vasc Endovasc Surg 32:188–197, 2006. 11. Richards JE, Kauffmann RM, Zuckerman SL, Obremskey WT, May AK. Relationship of hyperglycemia and surgical-site infection in orthopaedic surgery. J Bone Joint Surg Am 94:1181–1186, 2012. 12. Myers TG, Lowery NJ, Frykberg RG, Wukich DK. Ankle and hindfoot fusions: comparison of outcomes in patients with and without diabetes. Foot Ankle Int 33:20–28, 2012. 13. Perlman MH, Thordarson DB. Ankle fusion in a high risk population: an assessment of nonunion risk factors. Foot Ankle Int 20:491–496, 1999. 14. Wukich DK, Belczyk RJ, Burns PR, Frykberg RG. Complications encountered with circular ring fixation in persons with diabetes mellitus. Foot Ankle Int 29:994–1000, 2008. 15. Wukich DK, Shen JY, Ramirez CP, Irrgang JJ. Retrograde ankle arthrodesis using an intramedullary nail: a comparison of patients with and without diabetes mellitus. J Foot Ankle Surg 50:299–306, 2011. 16. Younger AS, Awwad MA, Kalla TP, de Vries G. Risk factors for failure of transmetatarsal amputation in diabetic patients: a cohort study. Foot Ankle Int 30: 1177–1182, 2009. 17. Alblowi J, Kayal RA, Siqueira M, McKenzie E, Krothapalli N, McLean J, Conn J, Nikolajczyk B, Einhorn TA, Gerstenfeld L, Graves DT. High levels of tumor necrosis factor-alpha contribute to accelerated loss of cartilage in diabetic fracture healing. Am J Pathol 175:1574–1585, 2009. 18. Beam HA, Parsons JR, Lin SS. The effects of blood glucose control upon fracture healing in the BB Wistar rat with diabetes mellitus. J Orthop Res 20:1210–1216, 2002. 19. Black CT, Hennessey PJ, Ford EG, Andrassy RJ. Protein glycosylation and collagen metabolism in normal and diabetic rats. J Surg Res 47:200–202, 1989. 20. Coords M, Breitbart E, Paglia D, Kappy N, Gandhi A, Cottrell J, Cedeno N, Pounder N, O’Connor JP, Lin SS. The effects of low-intensity pulsed ultrasound upon diabetic fracture healing. J Orthop Res 29:181–188, 2011. 21. Follak N, Kloting I, Merk H. Influence of diabetic metabolic state on fracture healing in spontaneously diabetic rats. Diabetes Metab Res Rev 21:288–296, 2005.
N. Shibuya et al. / The Journal of Foot & Ankle Surgery 52 (2013) 207–211
22. Follak N, Kloting I, Wolf E, Merk H. Improving metabolic control reverses the histomorphometric and biomechanical abnormalities of an experimentally induced bone defect in spontaneously diabetic rats. Calcif Tissue Int 74:551–560, 2004. 23. Follak N, Kloting I, Wolf E, Merk H. Histomorphometric evaluation of the influence of the diabetic metabolic state on bone defect healing depending on the defect size in spontaneously diabetic BB/OK rats. Bone 35:144–152, 2004. 24. Follak N, Kloting L, Wolf E, Merk H. Delayed remodeling in the early period of fracture healing in spontaneously diabetic BB/OK rats depending on the diabetic metabolic state. Histol Histopathol 19:473–486, 2004. 25. Gandhi A, Beam HA, O’Connor JP, Parsons JR, Lin SS. The effects of local insulin delivery on diabetic fracture healing. Bone 37:482–490, 2005. 26. Gandhi A, Doumas C, O’Connor JP, Parsons JR, Lin SS. The effects of local platelet rich plasma delivery on diabetic fracture healing. Bone 38:540–546, 2006. 27. Gandhi A, Liporace F, Azad V, Mattie J, Lin SS. Diabetic fracture healing. Foot Ankle Clin 11:805–824, 2006. 28. Gebauer GP, Lin SS, Beam HA, Vieira P, Parsons JR. Low-intensity pulsed ultrasound increases the fracture callus strength in diabetic BB Wistar rats but does not affect cellular proliferation. J Orthop Res 20:587–592, 2002. 29. Goto S, Fujii H, Kono K, Nakai K, Hamada Y, Yamato H, Shinohara M, Kitazawa R, Kitazawa S, Nishi S, Fukagawa M. Carvedilol ameliorates low-turnover bone disease in non-obese type 2 diabetes. Am J Nephrol 34:281–290, 2011. 30. Hamada Y, Kitazawa S, Kitazawa R, Fujii H, Kasuga M, Fukagawa M. Histomorphometric analysis of diabetic osteopenia in streptozotocin-induced diabetic mice: a possible role of oxidative stress. Bone 40:1408–1414, 2007. 31. Hamada Y, Kitazawa S, Kitazawa R, Kono K, Goto S, Komaba H, Fujii H, Yamamoto Y, Yamamoto H, Usami M, Fukagawa M. The effects of the receptor for advanced glycation end products (RAGE) on bone metabolism under physiological and diabetic conditions. Endocrine 38:369–376, 2010. 32. Kayal RA, Alblowi J, McKenzie E, Krothapalli N, Silkman L, Gerstenfeld L, Einhorn TA, Graves DT. Diabetes causes the accelerated loss of cartilage during fracture repair which is reversed by insulin treatment. Bone 44:357–363, 2009. 33. Kayal RA, Tsatsas D, Bauer MA, Allen B, Al-Sebaei MO, Kakar S, Leone CW, Morgan EF, Gerstenfeld LC, Einhorn TA, Graves DT. Diminished bone formation during diabetic fracture healing is related to the premature resorption of cartilage associated with increased osteoclast activity. J Bone Miner Res 22:560–568, 2007. 34. Lange J, Barz T, Ekkernkamp A, Kloting I, Follak N. Gene expression profile in bone of diabetes-prone BB/OK rats fed a high-fat diet. Genes Nutr Epub 2012 May 26. 35. Lu H, Kraut D, Gerstenfeld LC, Graves DT. Diabetes interferes with the bone formation by affecting the expression of transcription factors that regulate osteoblast differentiation. Endocrinology 144:346–352, 2003. 36. Macey LR, Kana SM, Jingushi S, Terek RM, Borretos J, Bolander ME. Defects of early fracture-healing in experimental diabetes. J Bone Joint Surg Am 71:722–733, 1989. 37. Mathey J, Horcajada-Molteni MN, Chanteranne B, Picherit C, Puel C, Lebecque P, Cubizoles C, Davicco MJ, Coxam V, Barlet JP. Bone mass in obese diabetic Zucker rats: influence of treadmill running. Calcif Tissue Int 70:305–311, 2002. 38. Ogasawara A, Nakajima A, Nakajima F, Goto K, Yamazaki M. Molecular basis for affected cartilage formation and bone union in fracture healing of the streptozotocin-induced diabetic rat. Bone 43:832–839, 2008. 39. Santana RB, Xu L, Chase HB, Amar S, Graves DT, Trackman PC. A role for advanced glycation end products in diminished bone healing in type 1 diabetes. Diabetes 52:1502–1510, 2003. 40. Topping RE, Bolander ME, Balian G. Type X collagen in fracture callus and the effects of experimental diabetes. Clin Orthop Relat Res 308:220–228, 1994. 41. Tyndall WA, Beam HA, Zarro C, O’Connor JP, Lin SS. Decreased platelet derived growth factor expression during fracture healing in diabetic animals. Clin Orthop Relat Res 408:319–330, 2003. 42. Bibbo C, Lin SS, Beam HA, Behrens FF. Complications of ankle fractures in diabetic patients. Orthop Clin North Am 32:113–133, 2001. 43. Blotter RH, Connolly E, Wasan A, Chapman MW. Acute complications in the operative treatment of isolated ankle fractures in patients with diabetes mellitus. Foot Ankle Int 20:687–694, 1999. 44. Connolly JF, Csencsitz TA. Limb threatening neuropathic complications from ankle fractures in patients with diabetes. Clin Orthop Relat Res 348:212–219, 1998. 45. Costigan W, Thordarson DB, Debnath UK. Operative management of ankle fractures in patients with diabetes mellitus. Foot Ankle Int 28:32–37, 2007. 46. Jones KB, Maiers-Yelden KA, Marsh JL, Zimmerman MB, Estin M, Saltzman CL. Ankle fractures in patients with diabetes mellitus. J Bone Joint Surg Br 87:489–495, 2005.
211
47. Kristiansen B. Ankle and foot fractures in diabetics provoking neuropathic joint changes. Acta Orthop Scand 51:975–979, 1980. 48. Prisk VR, Wukich DK. Ankle fractures in diabetics. Foot Ankle Clin 11:849–863, 2006. 49. Tang SY, Vashishth D. Non-enzymatic glycation alters microdamage formation in human cancellous bone. Bone 46:148–154, 2010. 50. Mehta SK, Breitbart EA, Berberian WS, Liporace FA, Lin SS. Bone and wound healing in the diabetic patient. Foot Ankle Clin 15:411–437, 2010. 51. Kristiansen B. Results of surgical treatment of malleolar fractures in patients with diabetes mellitus. Dan Med Bull 30:272–274, 1983. 52. Low CK, Tan SK. Infection in diabetic patients with ankle fractures. Ann Acad Med Singapore 24:353–355, 1995. 53. Adler AI, Boyko EJ, Ahroni JH, Smith DG. Lower-extremity amputation in diabetes: the independent effects of peripheral vascular disease, sensory neuropathy, and foot ulcers. Diabetes Care 22:1029–1035, 1999. 54. Boyko EJ, Ahroni JH, Smith DG, Davignon D. Increased mortality associated with diabetic foot ulcer. Diabet Med 13:967–972, 1996. 55. Junrungsee S, Kosachunhanun N, Wongthanee A, Rerkasem K. History of foot ulcers increases mortality among patients with diabetes in Northern Thailand. Diabet Med 28:608–611, 2011. 56. Lavery LA, Peters EJ, Williams JR, Murdoch DP, Hudson A, Lavery DC. Reevaluating the way we classify the diabetic foot: restructuring the diabetic foot risk classification system of the International Working Group on the Diabetic Foot. Diabetes Care 31:154–156, 2008. 57. Moss SE, Klein R, Klein BE. The prevalence and incidence of lower extremity amputation in a diabetic population. Arch Intern Med 152:610–616, 1992. 58. Ndip A, Lavery LA, Lafontaine J, Rutter MK, Vardhan A, Vileikyte L, Boulton AJ. High levels of foot ulceration and amputation risk in a multiracial cohort of diabetic patients on dialysis therapy. Diabetes Care 33:878–880, 2010. 59. La Fontaine J, Harkless LB, Sylvia VL, Carnes D, Heim-Hall J, Jude E. Levels of endothelial nitric oxide synthase and calcitonin gene-related peptide in the Charcot foot: a pilot study. J Foot Ankle Surg 47:424–429, 2008. 60. Shibuya N, La Fontaine J, Frania SJ. Alcohol-induced neuroarthropathy in the foot: a case series and review of literature. J Foot Ankle Surg 47:118–124, 2008. 61. Jeffcoate WJ. Theories concerning the pathogenesis of the acute Charcot foot suggest future therapy. Curr Diab Rep 5:430–435, 2005. 62. Sinacore DR, Hastings MK, Bohnert KL, Fielder FA, Villareal DT, Blair VP III, Johnson JE. Inflammatory osteolysis in diabetic neuropathic (Charcot) arthropathies of the foot. Phys Ther 88:1399–1407, 2008. 63. Radhakrishnan K, Vijayan VP, Ashok PP, Sridharan R, Mousa ME. Syphilitic spinal neuroarthropathy with paraplegia. Clin Neurol Neurosurg 87:61–64, 1985. 64. Crim JR, Bassett LW, Gold RH, Mirra JM, Mikulics M, Dawson EG, Eckhardt JJ. Spinal neuroarthropathy after traumatic paraplegia. AJNR Am J Neuroradiol 9:359–362, 1988. 65. Heylen Y. Neuropathic arthropathy of the shoulder secondary to syringomyelia. J Belge Radiol 76:232–233, 1993. 66. Thornhill HL, Richter RW, Shelton ML, Johnson CA. Neuropathic arthropathy (Charcot forefeet) in alcoholics. Orthop Clin North Am 4:7–20, 1973. 67. Sudanese A, Paderni S, Guerra E, Bertoni F. Neurogenic arthropathy of the knee due to chronic alcoholism: two case reports. Chir Organi Mov 88:427–434, 2003. 68. Barber DB, Janus RB, Wade WH. Neuroarthropathy: an overuse injury of the shoulder in quadriplegia. J Spinal Cord Med 19:9–11, 1996. 69. Horibe S, Tada K, Nagano J. Neuroarthropathy of the foot in leprosy. J Bone Joint Surg Br 70:481–485, 1988. 70. Bjorkengren AG, Weisman M, Pathria MN, Zlatkin MB, Pate D, Resnick D. Neuroarthropathy associated with chronic alcoholism. AJR Am J Roentgenol 151:743– 745, 1988. 71. Nagano J, Tada K, Masatomi T, Horibe S. Arthropathy of the wrist in leprosydwhat changes are caused by long-standing peripheral nerve palsy? Arch Orthop Trauma Surg 108:210–217, 1989. 72. Zaidi M. Neural surveillance of skeletal homeostasis. Cell Metab 1:219–221, 2005. 73. Irie K, Hara-Irie F, Ozawa H, Yajima T. Calcitonin gene-related peptide (CGRP)containing nerve fibers in bone tissue and their involvement in bone remodeling. Microsc Res Tech 58:85–90, 2002. 74. Young MJ, Marshall A, Adams JE, Selby PL, Boulton AJ. Osteopenia, neurological dysfunction, and the development of Charcot neuroarthropathy. Diabetes Care 18:34–38, 1995.
Contents lists available at ScienceDirect
The Journal of Foot & Ankle Surgery journal homepage: www.jfas.org
Factors Associated with Nonunion, Delayed Union, and Malunion in Foot and Ankle Surgery in Diabetic Patients Naohiro Shibuya, DPM, MS, FACFAS 1, Jon M. Humphers, DPM 2, Benjamin L. Fluhman, DPM 2, Daniel C. Jupiter, PhD 3 1 Associate Professor, Department of Surgery, Texas A&M Health and Science Center College of Medicine; Acting Chief, Section of Podiatry, Central Texas Veterans Affairs Health Care System; and Staff, Scott and White Memorial Hospital and Clinics, Temple, TX 2 Third Year Resident, Scott and White Memorial Hospital, Texas A&M Health and Science Center College of Medicine, Temple, TX 3 Assistant Professor, Department of Surgery, Texas A&M Health and Science Center College of Medicine; and Research Scientist, Scott and White Memorial Hospital and Clinics, Central Texas Veterans Affairs Health Care System, Temple, TX
a r t i c l e i n f o
a b s t r a c t
Level of Clinical Evidence: 3 Keywords: bone Charcot foot complication diabetes mellitus hyperglycemia peripheral neuropathy surgery
The incidence of bone healing complications in diabetic patients is believed to be high after foot and ankle surgery. Although the association of hyperglycemia with bone healing complications has been well documented, little clinical information is available to show which diabetes-related comorbidities directly affect bone healing. Our goal was to better understand the risk factors associated with poor bone healing in the diabetic population through an exploratory, observational, retrospective, cohort study. To this end, 165 diabetic patients who had undergone arthrodesis, osteotomy, or fracture reduction were enrolled in the study to assess the risk factors associated with nonunion, delayed union, and malunion after elective and nonelective foot and/or ankle surgery. Bivariate analyses showed that a history of foot ulcer, peripheral neuropathy, and surgery duration were statistically significantly associated with bone healing complications. After adjusting for other covariates, only peripheral neuropathy, surgery duration, and hemoglobin A1c levels >7% were significantly associated statistically with bone healing complications. Of the risk factors we considered, peripheral neuropathy had the strongest association with bone healing complications. Ó 2013 by the American College of Foot and Ankle Surgeons. All rights reserved.
The relationship between hyperglycemia and adverse outcomes after surgery has been widely documented (1–11). Long-term glucose control, as measured by glycosylated hemoglobin (hemoglobin A1c [HbA1c]), has been recognized as a major risk factor for negative outcomes after major surgery, including coronary artery procedures, and, in particular, for myocardial infarction and cardiac events after vascular surgery (1,2,5,10). In foot and ankle surgery, poorly controlled and complicated diabetes have also been shown to be significant risk factors for postoperative soft tissue and bone healing complications (12–16). Diabetic bone healing complications, in particular, have been extensively studied in animal models (17–41). In humans, the incidence of bone healing complication in diabetic patients is also believed to be high after foot and ankle surgery (13,42–48). However, few have
Financial Disclosure: None reported. Conflict of Interest: None reported. Address correspondence to: Naohiro Shibuya, DPM, MS, FACFAS, Associate Professor, Department of Surgery, Texas A&M Health and Science Center College of Medicine, Section of Podiatry, Central Texas Veterans Affairs Health Care System, Scott and White Memorial Hospital and Clinics, Temple, TX 76502. E-mail address: [email protected] (N. Shibuya).
conducted risk factor assessments within a diabetic population. Although the association of hyperglycemia with bone healing complications has been well documented (13,18,24,25,32,33,35,39,49,50), little clinical information is available to show which diabetes-related comorbidities directly affect bone healing. In diabetic animal models, many theories have been proposed to explain what causes the disturbance in normal bone metabolism. However, translational research to link each theory, often described in terms of molecular pathways to diabetic sequelae and complications in humans, is lacking. Our goal with the present investigation was to identify the risk factors associated with bone healing complications in a diabetic population using an exploratory, observational, retrospective, cohort study design. Understanding such risk factors could serve as a bridge between clinical and laboratory or animal research and promote a better understanding of the pathophysiology of diabetic bone healing. Patients and Methods A query of electronic medical records was performed to find diabetic patients who had undergone foot and ankle surgery during a 5-year period from January 2007 to December 2011 at the Scott and White Health Care System. The data from patients with an International Classification of Diseases, 9th Revision, code for all subcategories of
1067-2516/$ - see front matter Ó 2013 by the American College of Foot and Ankle Surgeons. All rights reserved. http://dx.doi.org/10.1053/j.jfas.2012.11.012
208
N. Shibuya et al. / The Journal of Foot & Ankle Surgery 52 (2013) 207–211
diabetes mellitus (code 250) and with Common Procedural Terminology codes of 27600 to 28899 were acquired by an information technology application analyst at our research department. Our institution’s internal review board approved our observational study protocol. The medical charts were then examined for the preoperative HbA1c values (recorded within 3 months before the date of surgery), other pre- and intraoperative risk factors, and outcome measures, including bone healing complications (Table 1). Patients with well-established diabetes, regardless of subtype, who had been treated pharmacologically, were included in the present study. The exclusion criteria included more than a 2-day course of antibiotic use within 1 week before surgery, the current use of immunosuppressant medications (including glucocorticoids, antibodies, cytostatics, drugs acting on immunophilins, and other drugs specifically used for immunosuppression), age older than 80 or younger than 12 years, postoperative follow-up duration less than 3 months, and a lack of radiographs to determine the progress of bone healing. The pooled diabetic patients were then screened for the type of foot and ankle procedures that had been performed. Nonosseous procedures were excluded. Those who had undergone both osseous and soft tissue surgery were included in the present study. Of those who had undergone osseous procedures, only arthrodesis, osteotomy, and/or reduction of fracture or dislocation were included in the study. Those who had undergone ostectomy (i.e., retrocalcaneal exostectomy, cheilectomy) or implant arthroplasty (i.e., total ankle replacement, first metatarsophalangeal implant arthroplasty), were excluded. For the remaining patients, the prevalence of bone healing complications was determined. For the purposes of our study, we defined bone healing complications as 1 or more of the following: nonunion, delayed union, or malunion. Delayed union was defined as no radiographic signs of bone healing within 3 months postoperatively. Nonunion was defined as no radiographic signs of bone healing for more than 6 months. Radiographic signs of bone healing were defined as the loss of visibility of osteotomy, arthrodesis, or fracture site in at least 2 of 3 views on the radiographs. To achieve consistency, the radiographs were reviewed by a single rater (J.H.). Malunion was defined as progression of deformity after original fixation, regardless of patient compliance. If an inadequate reduction or correction was seen in the initial radiographs and remained misaligned, the case was not considered malunion. Development of Charcot was recorded if a surgeon recognized the disease and documented it in their progress notes, regardless of method by which the condition was diagnosed. In this study, each patient was considered 1 case or event. When more than 2 procedures were performed on the same patient, only the earlier procedure was recorded. This measure was taken to eliminate consideration of revisional procedures and duplication of patient characteristics. However, if a patient underwent multiple arthrodeses, osteotomies, or fracture reductions on the same day, only 1 of the joints, osteotomies, or fracture sites had to be affected for the patient to be registered as having a bone healing complication. Similarly, if a patient showed nonunion in multiple joints, osteotomies, or fracture sites, it was regarded as 1 bone healing complication event. Forefoot procedures were defined as procedures at, or distal to, the tarsometatarsal joints. For example, open reduction of Lisfranc’s joint was recorded as a forefoot procedure. The rest of the procedures were considered rearfoot or ankle procedures. If a patient had undergone forefoot and rearfoot/ankle procedures, the case was recorded as “both.” When the medical records indicated current use of tobacco products, regardless of amount, the patient was considered positive for tobacco use. The presence of neuropathy was also determined from the medical records. Any sensory deficit noted in the medical records was considered neuropathy, regardless of the magnitude, type, or diagnostic test used to identify the neuropathy. Other comorbidities, including peripheral vascular disease and a history of ulceration were determined from the medical records. The body mass index was calculated from the weight and height
Table 1 Preoperative, intraoperative, and postoperative variables recorded Pre- and intraoperative variables Age Gender Body mass index Hemoglobin A1c within 90 days before surgery Preoperative random serum glucose drawn within 30 days of surgery Documented peripheral neuropathy Tobacco use (yes vs no) Operative time (10-min increments) Documented peripheral arterial disease (yes vs no) History of ulceration or infection in foot and ankle (yes vs no) Location of surgical procedure (forefoot vs rearfoot or ankle vs both) Postoperative outcome variables Malunion, delayed union, or nonunion (yes vs no) Acute, surgically induced Charcot neuroarthropathy (yes vs no)
measured by the anesthesia department during the preoperative visit. The surgery duration was obtained from the operating room records. Statistical Analysis The prevalence of the overall complications and each type of bone healing complication in our diabetic cohort was calculated. Bivariate analyses using t tests or analysis of variance or chi-square or Fisher’s exact tests were performed to compare those with and without a bone healing complication in terms of the various preoperative and intraoperative variables. A logistic regression model for the occurrence of bone healing complications was then built, including covariates that reached a significance level of 0.2 on bivariate analysis. Age, gender, and HbA1c were included in the model, regardless of the results from the bivariate analysis. For both bivariate and regression analyses, p .05 was considered statistically significant. The statistical analyses were performed using the R statistical package (R Developmental Core Team, R: A Language and Environment for Statistical Computing 2009, available from http:// www.R-project.org).
Results A total of 322 diabetic patients had HbA1c values documented within 3 months of the date of their bone surgery. Of these, 269 patients had undergone at least 1 osseous procedure. After excluding those patients who had undergone osseous procedures that did not require bone healing, 177 patients who had undergone osteotomy, arthrodesis, and/or fracture reduction were included in the present study. Of these patients, 11 had undergone more than 1 foot and ankle procedure on different occasions; only the first was included in our analyses. One patient had undergone 3 separate foot and ankle procedures, and 2 of these procedures were excluded. This left a total of 165 cases in our final analysis. The mean age of the selected cohort was 62.1 9.71 (range 37 to 80) years. Of the 165 patients, 115 (69.7%) were women and 50 (30.3%) were men; 112 (67.9%) were white, 32 (19.4%) were other races, 18 (10.9%) were black, 2 (1.2%) were Hispanic, and 1 (0.6%) was unidentifiable (relative to race). One patient (0.61) was missing data necessary to calculate the body mass index. For the remainder, the mean body mass index was 34.2 7.40 (range 21.0 to 56.3) kg/m2. Of the patients, 59 (35.7%) had peripheral neuropathy, 21 (12.7%) had a history of foot ulcer, 20 (12.1%) had a history of tobacco use, 11 (6.7%) had peripheral vascular disease, and 77 (46.7%) had at least 1 of these comorbidities. The duration of surgery was not obtainable for 11 patients (6.67%); however, for the remainder, the mean duration was 96.1 44.54 (range 21 to 237) minutes. The mean HbA1c was 7.2% 1.53% (range 4.8% to 14.9%). The serum glucose values were missing for 12 patients (7.27%). For the remainder, the mean glucose value was 156.8 66.18 (range 41 to 572) mg/dL. Of the procedures, 84 (50.9%) were in the forefoot, 78 (47.3%) in the rearfoot or ankle, and 3 (1.8%) were in both. A total of 123 patients (74.6%) had no bone healing complication. Of the patients with complications, 13 (7.9%) had delayed union, 19 (11.5%) nonunion, and 10 (6.1%) malunion. The overall prevalence of bone healing complication in our diabetic cohort was 25.4% (42/165). Four patients (2.42%) developed Charcot arthropathy, which resulted in nonunion in 2 and malunion in 2. Of these 4 Charcot arthropathies, 2 developed after open reduction and internal fixation of an ankle fracture, 1 after Lisfranc’s fracture reduction, and 1 after tibiotalocalcaneal arthrodesis. The corresponding HbA1c levels were 9.1%, 6.1%, 9.6%, and 7.3%. The results of the bivariate analyses are presented in Table 2. Of the variables tested, the presence of 1 or more comorbidity, neuropathy, a history of ulcer, and the duration of surgery showed significant statistical differences between those with bone healing complications and those without bone healing complications. In a logistic regression model for the presence of a bone healing complication with the covariates of age, HbA1c level, duration of surgery, anatomic location of the procedure,
N. Shibuya et al. / The Journal of Foot & Ankle Surgery 52 (2013) 207–211
Variable
Age (y) Gender Female Male Race White Other Black Hispanic Unidentifiable Body mass index (kg/m2) Comorbidity Yes No Neuropathy Yes No Peripheral vascular disease Yes No Tobacco use Yes No History of ulcer Yes No Surgery time (min) Hemoglobin A1c (%) Perioperative serum glucose (mg/dL) Anatomic location of procedure Forefoot Rearfoot/ankle Both * y z
Bone Healing Complication No (n ¼ 123)
Yes (n ¼ 42)
63 10.20
60 7.99
88 35
p Value
.07* .38y
27 15 .37z
87 22 12 1 1 34.4 7.77
25 10 6 1 0 33.7 6.53
46 88
31 11
33 90
26 16
7 116
4 38
14 108
6 36
11 112 90 44.09 7.2 1.62 158.1 52.36
10 32 113 41.69 7.4 1.24 156.3 70.78
.61* .001y
.001y
.474z
.63y
.013y
67 53 3
.004 .42 .87 .14z
17 25 0
Probabilities of null hypothesis (p values) were computed using Student’s t test. Probabilities of null hypothesis (p values) were computed using the chi-square test. Probabilities of null hypothesis (p values) were computed using Fisher’s exact test.
neuropathy, and ulcer, the only significant factor was the duration of surgery (odds ratio [OR] 1.01, 95% confidence interval [CI] 1.00 to 1.02). After exploring the data further, we built a model that would be more clinically relevant by categorizing patient age by decade, dichotomizing the HbA1c into >7% or 7%, and coding the duration of surgery in increments of 10 minutes. Thus, our final model included age (by decade), the presence of neuropathy (yes or no), a history of ulcer (yes or no), HbA1c (>7% or 7%), and duration of surgery (by 10-minute increments). The factors that were statistically significantly associated statistically with the occurrence of a bone healing complication in this model included the duration of surgery (OR 1.15, 95% CI 1.04 to 1.26), peripheral neuropathy (OR 3.93, 95% CI 1.16 to 9.59), and HbA1c >7% (OR 2.88, 95% CI 1.23 to 6.72; Table 3). When we added the presence of a comorbidity and the location of surgery to this model, both the HbA1c level and the duration of surgery remained significant factors.
Discussion Our results have shown that approximately 1 of 4 diabetic patients (25.6%) will have 1 or more bone healing complications. Although the severity of the complication was not captured in the present study, the overall prevalence was striking and clinically significant. In our study, neither the perioperative serum glucose nor the HbA1c proportion differed statistically significantly between those with a bone healing complication and those without such a complication in our bivariate analysis. The mean HbA1c level for those with and without bone healing complications was 7.2% and 7.4%, respectively. After post hoc exploration of the data, we found that the group without bone healing complications had more apparent outliers in the upper HbA1c levels (Fig.). Those outliers were mostly nonelective cases (i.e., open reduction and internal fixation of ankle fractures). Because malleolar fractures have a relatively high union rate with fewer complications than elective arthrodesis or osteotomy procedures (12,13,43,51,52), this might have artificially elevated the mean HbA1c level in the no-complication group. Overall, only 4 bone healing complications (12.9%) occurred among 31 open reductions and internal fixations of malleolar fractures in our diabetic cohort. In the regression analysis, HbA1c showed a significant association with the presence of a postoperative bone healing complication (OR 2.88, 95% CI 1.23 to 6.72) when it was dichotomized as >7% or <7%. We suspect that this dichotomization minimized the effect of the outliers. Thus, we found that patients with HbA1c >7% were approximately 3 times as likely to have a bone healing complication than patients with HbA1c <7%. Having at least 1 comorbidity was also significantly associated with the development of bone healing complications. Specifically, on post hoc bivariate analysis, neuropathy (chi-square, p ¼ .001) and a history of ulcer (chi-square, p ¼ .013) were significantly associated statistically with the development of bone healing complications. In particular, 44% (26/59) of neuropathic patients had 1 or more bone healing complications (Table 2). More impressively, 50% (26/42) of those with bone healing complication had neuropathy. After adjusting for other covariates, a history of ulcer was not a significant factor for bone healing complications. Although a history of ulceration is often linked to amputation and mortality risk (53–58), we suspect that this factor might be confounded by neuropathy in the setting of bone
14
12 HA1c (%)
Table 2 Results of bivariate analysis of recorded variables vs occurrence of bone healing complications
209
10
8 Table 3 Results of multivariate logistic regression analyses (N ¼ 165) Variable
OR
95% CI
p value
Age (10-y increment) HbA1c (>7%) Surgery time (10-min increment) History of ulcer (positive) Neuropathy (positive)
0.86 2.88 1.15 1.24 3.93
0.58–1.28 1.23–6.72 1.04–1.26 0.38–4.03 1.16–9.59
.470 .014* .004* .720 .003*
Abbreviations: CI, confidence interval; HbA1c, hemoglobin A1c; OR, odds ratio. * Statistically significant.
6
No complication
Complication
Fig. Box plots revealing outliers mostly among those with no bone healing complication (N ¼ 165).
210
N. Shibuya et al. / The Journal of Foot & Ankle Surgery 52 (2013) 207–211
healing. This result has confirmed the results from many animal and bench studies and reports that focused on diabetic bone healing complications, which have often described the complication as a malfunction of bone metabolism resulting from neuropathy (45,59– 62). This is an interesting finding, especially considering that osteolytic activities, including Charcot arthropathy in patients with diabetes, are believed to be strongly linked to neuropathy rather than to hyperglycemia itself. Many Charcot cases have presented in nondiabetic neuropathic patients (60,63–71). The lack of adequate neuropeptide release in the neuropathic patient might upregulate osteoclastogenesis while downregulating osteoblastic activities (59,72–74). Although the precise relationship between neuropathy and poor bone healing in patients with diabetes undergoing foot and ankle surgery is uncertain, our results have indicated that it is the single most important factor among all the variables we tested. In our study, we found that those with neuropathy had an OR of 4 in favor of the development of bone healing complications compared with those without neuropathy. The odds of experiencing a bone healing complication in the presence of neuropathy were greater than those associated with hyperglycemia, as measured by the serum glucose or HbA1c level. A longer duration of surgery was also associated with complicated bone healing after adjusting for other covariates. A longer duration of surgery might indicate, but is not restricted to, a more complex case, more dissection, an adverse event occurring during surgery, or a procedure conducted by a slow surgeon. Although we did not investigate the reasons for the longer operative time, all these reasons are, intuitively, reasonable causes of bone healing complications. In our analysis, we found that each additional 10 minutes of surgery increased the odds of a bone healing complication by a factor of 15%. We recognize several shortcomings of the present study. First, we relied on progress notes written by each surgeon to detect neuropathy, peripheral vascular disease, a history of foot ulcer, tobacco use, and the development of Charcot. We had no systematic method to measure these variables. We suspect that these conditions were therefore underreported. Also, the magnitude of these risk factors was not captured. All these variables were dichotomized as “yes” or “no,” and we did not make any distinction between severe and minor neuropathy. Second, we designated 1 rater to review all the radiographic outcomes. The rater was not unaware of the other variables. Although measurement error could have occurred at random, bias stemming from knowing the other findings could have affected our final result. Third, the surgeons were not unaware of each patient’s medical history. Naturally, high-risk patients are less likely to undergo elective surgery. This would not only lead to lower enrollment of patients with uncontrolled diabetes, but would also oversample emergency cases among the high-risk patients. Also, surgeons naturally change their fixation method, dissection technique, and followup protocol according to the findings from their risk assessment. Therefore, the effect of potential risk factors might have been underestimated in the present study, reflecting current ideas about the effect of the HbA1c level on outcomes. Finally, extracting patient data from the electronic medical record system using “International Classification of Disease, 9th revision,” and Common Procedural Terminology codes could have potentially underrepresented the patients with borderline or undiagnosed diabetes. However, we believe that it is more reliable to use the “International Classification of Disease, 9th revision,” data from our institution’s medical record system to detect diabetes than to rely on the surgeons’ progress notes or billing information. Because any other service can record the diagnosis of diabetes in our system, the risk of missing a diagnosis of diabetes would be less for patients undergoing foot and ankle surgery. In conclusion, within the limitations of the present study, that factors that affected bone healing in diabetic patients after foot and
ankle surgery were HbA1c >7%, duration of surgery, and peripheral neuropathy. Among all the variables we analyzed, neuropathy was the single most significant factor that affected bone healing in patients with diabetes. Although randomized clinical trials might not be practical or ethical in this setting, a longitudinal study with better, systematically defined risk factors is warranted to confirm our findings. Furthermore, translational research to confirm the laboratory and animal study findings are needed to better understand the pathophysiology of diabetic bone healing. Future studies should also focus on quantifying each risk factor to determine the clinically useful cutoff line at which the risk of surgery exceeds the benefit. References 1. Acott AA, Theus SA, Kim LT. Long-term glucose control and risk of perioperative complications. Am J Surg 198:596–599, 2009. 2. Dronge AS, Perkal MF, Kancir S, Concato J, Aslan M, Rosenthal RA. Long-term glycemic control and postoperative infectious complications. Arch Surg 141:375–380, 2006. 3. Glassman SD, Alegre G, Carreon L, Dimar JR, Johnson JR. Perioperative complications of lumbar instrumentation and fusion in patients with diabetes mellitus. Spine J 3:496–501, 2003. 4. Golden SH, Peart-Vigilance C, Kao WH, Brancati FL. Perioperative glycemic control and the risk of infectious complications in a cohort of adults with diabetes. Diabetes Care 22:1408–1414, 1999. 5. Halkos ME, Puskas JD, Lattouf OM, Kilgo P, Kerendi F, Song HK, Guyton RA, Thourani VH. Elevated preoperative hemoglobin A1c level is predictive of adverse events after coronary artery bypass surgery. J Thorac Cardiovasc Surg 136:631–640, 2008. 6. Lamloum SM, Mobasher LA, Karar AH, Basiony L, Abdallah TH, Al-Saleh AI, Al-Shamali NA. Relationship between postoperative infectious complications and glycemic control for diabetic patients in an orthopedic hospital in Kuwait. Med Princ Pract 18:447–452, 2009. 7. Marchant MH Jr, Viens NA, Cook C, Vail TP, Bolognesi MP. The impact of glycemic control and diabetes mellitus on perioperative outcomes after total joint arthroplasty. J Bone Joint Surg Am 91:1621–1629, 2009. 8. Mraovic B, Suh D, Jacovides C, Parvizi J. Perioperative hyperglycemia and postoperative infection after lower limb arthroplasty. J Diabetes Sci Technol 5:412–418, 2011. 9. Noordzij PG, Boersma E, Schreiner F, Kertai MD, Feringa HH, Dunkelgrun M, Bax JJ, Klein J, Poldermans D. Increased preoperative glucose levels are associated with perioperative mortality in patients undergoing noncardiac, nonvascular surgery. Eur J Endocrinol 156:137–142, 2007. 10. O’Sullivan CJ, Hynes N, Mahendran B, Andrews EJ, Avalos G, Tawfik S, Lowery A, Sultan S. Haemoglobin A1c (HbA1C) in non-diabetic and diabetic vascular patients: is HbA1C an independent risk factor and predictor of adverse outcome? Eur J Vasc Endovasc Surg 32:188–197, 2006. 11. Richards JE, Kauffmann RM, Zuckerman SL, Obremskey WT, May AK. Relationship of hyperglycemia and surgical-site infection in orthopaedic surgery. J Bone Joint Surg Am 94:1181–1186, 2012. 12. Myers TG, Lowery NJ, Frykberg RG, Wukich DK. Ankle and hindfoot fusions: comparison of outcomes in patients with and without diabetes. Foot Ankle Int 33:20–28, 2012. 13. Perlman MH, Thordarson DB. Ankle fusion in a high risk population: an assessment of nonunion risk factors. Foot Ankle Int 20:491–496, 1999. 14. Wukich DK, Belczyk RJ, Burns PR, Frykberg RG. Complications encountered with circular ring fixation in persons with diabetes mellitus. Foot Ankle Int 29:994–1000, 2008. 15. Wukich DK, Shen JY, Ramirez CP, Irrgang JJ. Retrograde ankle arthrodesis using an intramedullary nail: a comparison of patients with and without diabetes mellitus. J Foot Ankle Surg 50:299–306, 2011. 16. Younger AS, Awwad MA, Kalla TP, de Vries G. Risk factors for failure of transmetatarsal amputation in diabetic patients: a cohort study. Foot Ankle Int 30: 1177–1182, 2009. 17. Alblowi J, Kayal RA, Siqueira M, McKenzie E, Krothapalli N, McLean J, Conn J, Nikolajczyk B, Einhorn TA, Gerstenfeld L, Graves DT. High levels of tumor necrosis factor-alpha contribute to accelerated loss of cartilage in diabetic fracture healing. Am J Pathol 175:1574–1585, 2009. 18. Beam HA, Parsons JR, Lin SS. The effects of blood glucose control upon fracture healing in the BB Wistar rat with diabetes mellitus. J Orthop Res 20:1210–1216, 2002. 19. Black CT, Hennessey PJ, Ford EG, Andrassy RJ. Protein glycosylation and collagen metabolism in normal and diabetic rats. J Surg Res 47:200–202, 1989. 20. Coords M, Breitbart E, Paglia D, Kappy N, Gandhi A, Cottrell J, Cedeno N, Pounder N, O’Connor JP, Lin SS. The effects of low-intensity pulsed ultrasound upon diabetic fracture healing. J Orthop Res 29:181–188, 2011. 21. Follak N, Kloting I, Merk H. Influence of diabetic metabolic state on fracture healing in spontaneously diabetic rats. Diabetes Metab Res Rev 21:288–296, 2005.
N. Shibuya et al. / The Journal of Foot & Ankle Surgery 52 (2013) 207–211
22. Follak N, Kloting I, Wolf E, Merk H. Improving metabolic control reverses the histomorphometric and biomechanical abnormalities of an experimentally induced bone defect in spontaneously diabetic rats. Calcif Tissue Int 74:551–560, 2004. 23. Follak N, Kloting I, Wolf E, Merk H. Histomorphometric evaluation of the influence of the diabetic metabolic state on bone defect healing depending on the defect size in spontaneously diabetic BB/OK rats. Bone 35:144–152, 2004. 24. Follak N, Kloting L, Wolf E, Merk H. Delayed remodeling in the early period of fracture healing in spontaneously diabetic BB/OK rats depending on the diabetic metabolic state. Histol Histopathol 19:473–486, 2004. 25. Gandhi A, Beam HA, O’Connor JP, Parsons JR, Lin SS. The effects of local insulin delivery on diabetic fracture healing. Bone 37:482–490, 2005. 26. Gandhi A, Doumas C, O’Connor JP, Parsons JR, Lin SS. The effects of local platelet rich plasma delivery on diabetic fracture healing. Bone 38:540–546, 2006. 27. Gandhi A, Liporace F, Azad V, Mattie J, Lin SS. Diabetic fracture healing. Foot Ankle Clin 11:805–824, 2006. 28. Gebauer GP, Lin SS, Beam HA, Vieira P, Parsons JR. Low-intensity pulsed ultrasound increases the fracture callus strength in diabetic BB Wistar rats but does not affect cellular proliferation. J Orthop Res 20:587–592, 2002. 29. Goto S, Fujii H, Kono K, Nakai K, Hamada Y, Yamato H, Shinohara M, Kitazawa R, Kitazawa S, Nishi S, Fukagawa M. Carvedilol ameliorates low-turnover bone disease in non-obese type 2 diabetes. Am J Nephrol 34:281–290, 2011. 30. Hamada Y, Kitazawa S, Kitazawa R, Fujii H, Kasuga M, Fukagawa M. Histomorphometric analysis of diabetic osteopenia in streptozotocin-induced diabetic mice: a possible role of oxidative stress. Bone 40:1408–1414, 2007. 31. Hamada Y, Kitazawa S, Kitazawa R, Kono K, Goto S, Komaba H, Fujii H, Yamamoto Y, Yamamoto H, Usami M, Fukagawa M. The effects of the receptor for advanced glycation end products (RAGE) on bone metabolism under physiological and diabetic conditions. Endocrine 38:369–376, 2010. 32. Kayal RA, Alblowi J, McKenzie E, Krothapalli N, Silkman L, Gerstenfeld L, Einhorn TA, Graves DT. Diabetes causes the accelerated loss of cartilage during fracture repair which is reversed by insulin treatment. Bone 44:357–363, 2009. 33. Kayal RA, Tsatsas D, Bauer MA, Allen B, Al-Sebaei MO, Kakar S, Leone CW, Morgan EF, Gerstenfeld LC, Einhorn TA, Graves DT. Diminished bone formation during diabetic fracture healing is related to the premature resorption of cartilage associated with increased osteoclast activity. J Bone Miner Res 22:560–568, 2007. 34. Lange J, Barz T, Ekkernkamp A, Kloting I, Follak N. Gene expression profile in bone of diabetes-prone BB/OK rats fed a high-fat diet. Genes Nutr Epub 2012 May 26. 35. Lu H, Kraut D, Gerstenfeld LC, Graves DT. Diabetes interferes with the bone formation by affecting the expression of transcription factors that regulate osteoblast differentiation. Endocrinology 144:346–352, 2003. 36. Macey LR, Kana SM, Jingushi S, Terek RM, Borretos J, Bolander ME. Defects of early fracture-healing in experimental diabetes. J Bone Joint Surg Am 71:722–733, 1989. 37. Mathey J, Horcajada-Molteni MN, Chanteranne B, Picherit C, Puel C, Lebecque P, Cubizoles C, Davicco MJ, Coxam V, Barlet JP. Bone mass in obese diabetic Zucker rats: influence of treadmill running. Calcif Tissue Int 70:305–311, 2002. 38. Ogasawara A, Nakajima A, Nakajima F, Goto K, Yamazaki M. Molecular basis for affected cartilage formation and bone union in fracture healing of the streptozotocin-induced diabetic rat. Bone 43:832–839, 2008. 39. Santana RB, Xu L, Chase HB, Amar S, Graves DT, Trackman PC. A role for advanced glycation end products in diminished bone healing in type 1 diabetes. Diabetes 52:1502–1510, 2003. 40. Topping RE, Bolander ME, Balian G. Type X collagen in fracture callus and the effects of experimental diabetes. Clin Orthop Relat Res 308:220–228, 1994. 41. Tyndall WA, Beam HA, Zarro C, O’Connor JP, Lin SS. Decreased platelet derived growth factor expression during fracture healing in diabetic animals. Clin Orthop Relat Res 408:319–330, 2003. 42. Bibbo C, Lin SS, Beam HA, Behrens FF. Complications of ankle fractures in diabetic patients. Orthop Clin North Am 32:113–133, 2001. 43. Blotter RH, Connolly E, Wasan A, Chapman MW. Acute complications in the operative treatment of isolated ankle fractures in patients with diabetes mellitus. Foot Ankle Int 20:687–694, 1999. 44. Connolly JF, Csencsitz TA. Limb threatening neuropathic complications from ankle fractures in patients with diabetes. Clin Orthop Relat Res 348:212–219, 1998. 45. Costigan W, Thordarson DB, Debnath UK. Operative management of ankle fractures in patients with diabetes mellitus. Foot Ankle Int 28:32–37, 2007. 46. Jones KB, Maiers-Yelden KA, Marsh JL, Zimmerman MB, Estin M, Saltzman CL. Ankle fractures in patients with diabetes mellitus. J Bone Joint Surg Br 87:489–495, 2005.
211
47. Kristiansen B. Ankle and foot fractures in diabetics provoking neuropathic joint changes. Acta Orthop Scand 51:975–979, 1980. 48. Prisk VR, Wukich DK. Ankle fractures in diabetics. Foot Ankle Clin 11:849–863, 2006. 49. Tang SY, Vashishth D. Non-enzymatic glycation alters microdamage formation in human cancellous bone. Bone 46:148–154, 2010. 50. Mehta SK, Breitbart EA, Berberian WS, Liporace FA, Lin SS. Bone and wound healing in the diabetic patient. Foot Ankle Clin 15:411–437, 2010. 51. Kristiansen B. Results of surgical treatment of malleolar fractures in patients with diabetes mellitus. Dan Med Bull 30:272–274, 1983. 52. Low CK, Tan SK. Infection in diabetic patients with ankle fractures. Ann Acad Med Singapore 24:353–355, 1995. 53. Adler AI, Boyko EJ, Ahroni JH, Smith DG. Lower-extremity amputation in diabetes: the independent effects of peripheral vascular disease, sensory neuropathy, and foot ulcers. Diabetes Care 22:1029–1035, 1999. 54. Boyko EJ, Ahroni JH, Smith DG, Davignon D. Increased mortality associated with diabetic foot ulcer. Diabet Med 13:967–972, 1996. 55. Junrungsee S, Kosachunhanun N, Wongthanee A, Rerkasem K. History of foot ulcers increases mortality among patients with diabetes in Northern Thailand. Diabet Med 28:608–611, 2011. 56. Lavery LA, Peters EJ, Williams JR, Murdoch DP, Hudson A, Lavery DC. Reevaluating the way we classify the diabetic foot: restructuring the diabetic foot risk classification system of the International Working Group on the Diabetic Foot. Diabetes Care 31:154–156, 2008. 57. Moss SE, Klein R, Klein BE. The prevalence and incidence of lower extremity amputation in a diabetic population. Arch Intern Med 152:610–616, 1992. 58. Ndip A, Lavery LA, Lafontaine J, Rutter MK, Vardhan A, Vileikyte L, Boulton AJ. High levels of foot ulceration and amputation risk in a multiracial cohort of diabetic patients on dialysis therapy. Diabetes Care 33:878–880, 2010. 59. La Fontaine J, Harkless LB, Sylvia VL, Carnes D, Heim-Hall J, Jude E. Levels of endothelial nitric oxide synthase and calcitonin gene-related peptide in the Charcot foot: a pilot study. J Foot Ankle Surg 47:424–429, 2008. 60. Shibuya N, La Fontaine J, Frania SJ. Alcohol-induced neuroarthropathy in the foot: a case series and review of literature. J Foot Ankle Surg 47:118–124, 2008. 61. Jeffcoate WJ. Theories concerning the pathogenesis of the acute Charcot foot suggest future therapy. Curr Diab Rep 5:430–435, 2005. 62. Sinacore DR, Hastings MK, Bohnert KL, Fielder FA, Villareal DT, Blair VP III, Johnson JE. Inflammatory osteolysis in diabetic neuropathic (Charcot) arthropathies of the foot. Phys Ther 88:1399–1407, 2008. 63. Radhakrishnan K, Vijayan VP, Ashok PP, Sridharan R, Mousa ME. Syphilitic spinal neuroarthropathy with paraplegia. Clin Neurol Neurosurg 87:61–64, 1985. 64. Crim JR, Bassett LW, Gold RH, Mirra JM, Mikulics M, Dawson EG, Eckhardt JJ. Spinal neuroarthropathy after traumatic paraplegia. AJNR Am J Neuroradiol 9:359–362, 1988. 65. Heylen Y. Neuropathic arthropathy of the shoulder secondary to syringomyelia. J Belge Radiol 76:232–233, 1993. 66. Thornhill HL, Richter RW, Shelton ML, Johnson CA. Neuropathic arthropathy (Charcot forefeet) in alcoholics. Orthop Clin North Am 4:7–20, 1973. 67. Sudanese A, Paderni S, Guerra E, Bertoni F. Neurogenic arthropathy of the knee due to chronic alcoholism: two case reports. Chir Organi Mov 88:427–434, 2003. 68. Barber DB, Janus RB, Wade WH. Neuroarthropathy: an overuse injury of the shoulder in quadriplegia. J Spinal Cord Med 19:9–11, 1996. 69. Horibe S, Tada K, Nagano J. Neuroarthropathy of the foot in leprosy. J Bone Joint Surg Br 70:481–485, 1988. 70. Bjorkengren AG, Weisman M, Pathria MN, Zlatkin MB, Pate D, Resnick D. Neuroarthropathy associated with chronic alcoholism. AJR Am J Roentgenol 151:743– 745, 1988. 71. Nagano J, Tada K, Masatomi T, Horibe S. Arthropathy of the wrist in leprosydwhat changes are caused by long-standing peripheral nerve palsy? Arch Orthop Trauma Surg 108:210–217, 1989. 72. Zaidi M. Neural surveillance of skeletal homeostasis. Cell Metab 1:219–221, 2005. 73. Irie K, Hara-Irie F, Ozawa H, Yajima T. Calcitonin gene-related peptide (CGRP)containing nerve fibers in bone tissue and their involvement in bone remodeling. Microsc Res Tech 58:85–90, 2002. 74. Young MJ, Marshall A, Adams JE, Selby PL, Boulton AJ. Osteopenia, neurological dysfunction, and the development of Charcot neuroarthropathy. Diabetes Care 18:34–38, 1995.