Introduction

Autologous bone grafting is a long-established method, first performed on soldiers after war wounds in the seventeenth century [5, 6, 15, 26, 32]. This changed to bone defect and non-union treatment [35] and is now equivalent to allograft bone or bone substitutes [11]. Harvesting from anterior iliac crest has the advantage of access in supine position or beach chair position. A disadvantage is the higher complication rate compared to posterior iliac crest bone harvesting [1] but overall, the rates are moderate to low. The most common complications are infections, hematoma/bleedings, and sensory disturbances of the lateral femoral cutaneous nerve. Fractures or severe complications are rare [3, 18, 30]. Distal radius fracture is the most frequent fracture in humans [8, 21, 23]. The AO classification is mostly used to describe fracture pattern, and gives hints for treatment decision-making [27]. In fracture treatment, correct restoration of joint anatomy and bone length including bone losses should be achieved, regardless of the used implant or technique [10, 12, 14, 25, 36].

In this study, all fractures showed a metaphyseal defect after fracture. After proper reduction, the dorsal comminution zone does not provide a stable bony wall. Bone grafting or use of bone substitute was indicated. An autologous bone graft is osteogenetic, osteoinductive and osteoconductive and superior to allograft or bone substitutes. However, harvesting has complications, as mentioned before [17].

The aim of this study was to evaluate the donor site morbidity following anterior iliac crest harvesting for autograft in the treatment of distal radius fractures and to identify potential risk factors for complications. Unlike other localizations, only a small amount of bone graft is needed for the treatment of distal radius fractures. It could, therefore, be cheaper than the use of allograft bone or bone substitutes, and in view of this economic impact the results of this study might lead to a re-assessment of how distal radius fractures are treated.

Patients and methods

Patients with acute distal radius fracture who received an open reduction and fixation with locking plate and additional autograft bone transplantation from the anterior iliac crest between January 2008 and December 2012 were included in this study. Non-union therapy or late bone grafting were exclusion criteria. Patient’s demographic and clinical data, co-morbidities, fracture classification, duration of hospital admission, and complications as documented in patients’ records (revision surgery, nerve injuries, fractures, hematoma, bleeding, wound healing disturbances, infections, pain, etc.) were evaluated from the patient’s records.

Harvesting of the iliac crest bone was performed using an incision starting minimum 3 cm dorsal of the spina iliaca anterior superior along the iliac crest and exposure of the bone (Figs. 1, 2). Cortical bone, cancellous bone or chips were removed using an oscilliating saw, a curved osteotome or spoons (Fig. 3). The wound was irrigated and bleedings were stopped with or without the use of hemostatics. The fascia was closed and a subcutaneous drain was placed. The skin was closed with stitches and infiltration with 10 ml Bupivacaine 0.5% was applied. Every step described including the surgeon’s experience and duration of surgery was evaluated as a risk factor for complications.

Fig. 1
figure 1

Landmarks and skin incision on anterior iliac crest

Full size image
Fig. 2
figure 2

Incision to and through fascia with exposed iliac crest

Full size image
Fig. 3
figure 3

Harvesting of a bone block using a saw. Iliac crest after harvesting

Full size image

Complications were classified as intra- and post-operative major or minor complications. These included revision surgery, fractures, bleeding, nerve lesions, pain, hematoma, infection, wound healing disturbances, delayed discharge or death. At last follow-up, a re-evaluation of all parameters, including risk factors such as co-morbidities, drugs, alcohol and smoking, was performed.

Statistical analysis and descriptive statistics were performed using Microsoft® Excel® Vers. 14.1.0 and SPSS 22.0 (IBM, Armonk, USA). A power analysis showed 31 patients as a cut-off point for statistical significance regarding H0- (no increased risk of complications compared to literature) and H1-hypothesis. Chi-squared test, Student’s t test and Mann–Whitney U test were used. p < 0.05 was defined as statistically significant.

The local and institutional ethics committee approved this study.

Results

136 patients were identified, and 42 (22 females, 20 males) of them fitted the inclusion criteria. Mean follow-up was 75 months (6.3 years; range 4.7–8.6 years). Mean age was 56.3 years (range 19–81). 8 patients were between 19 and 40 years old, 17 patients were between 41 and 60 years old. 17 patients (40.5%) had normal BMI; 13 (32%) were above normal. There were 22 patients (52.4%) with left and 20 patients (47.6%) with right radius fractures. Bone graft was harvested from the same side in 41 (97.6%) of cases. 31 patients (73.8%) had type C3 fractures of the radius, 7 (16.7%) showed type C2, and 4 (9.5%) an A3-type fracture. Mean hospital admission was 12.3 days (range 4–27) due to multiple trauma and revision surgery in one case. Four patients (9.5%) suffered from pain at the iliac crest. 26 patients (61.9%) showed hematoma in the first post-operative day, but at discharge, the hematoma was almost resolved in all cases. No nerve injuries, wound healing disturbances, fractures, deaths or infections were seen. One patient (2.4%) had relevant bleeding and underwent revision surgery twice. There was no relationship between the surgeon’s experience and complications. Mean duration of procedure was 71 min (range 34–270) including multiple injury treatment. In 32 patients (76.2%), cancellous bone (15 patients), bone chips or a small block was harvested. In ten patients (23.8%), a combination of the techniques was performed. Documented hemostasis was performed in 13 patients (31%), in 38 patients (90.5%) a hemostyptic agent and in 39 patients (93%) a drain was placed. In 24 patients, (57.1%) irrigation and application of local anesthesia were performed. Six patients (14.3) % were treated with NOAKs or aspirin pre-operatively. 35 of them (83.3%) developed hematomas, whereas 21 (58.3%) of the 36 patients without the use of NOAKS or aspirin showed hematoma that was not statistically significant (p = 0.24%). One patient took immunosuppressive medication.

The cohort was relatively healthy: ten patients (23.8%) had arterial hypertony, only one had peripheral arterial disease, another one cardiac insufficiency. Five patients (11.9%) had diabetes, four (9.52%) had osteoporosis. Seven patients (17%) had allergies (two of them drug-associated allergies). Three (7.1%) of the patients abused alcohol, and two (4.8%) were smokers.

A hemostyptic agent was used in 38 (90.5%) procedures. 22 (50.0%) patients developed documented post-operative hematoma including all four patients without use of the hemostyptic agent, which was not statistically significant (p = 0.055).

In 13 patients (31.0%), intra-operative hemostasis was documented. Only five of these patients (38.5% of study population) showed hematoma. In 29 patients (69.0%), no hemostasis was documented intra-operatively. Only eight patients (27.6%) did not develop hematoma. Documented hemostasis showed significantly lower risk for hematoma (p < 0.05). In 38 patients (92.9%), a drain was inserted. 25 of these patients (64.1%) developed a hematoma. The use of a drain could not avert the development of hematoma which was not tested because of the inhomogeneity of the groups. Irrigation before wound closure showed a non-significant reduction of hematoma (p = 0.582).

Wound closure layer by layer was documented in 41 patients (97.6%) and showed hematoma in 15 of them. No correlation could be found regarding hematoma and surgeon’s experience (p = 0.38). Patients with or without bed rest for 1–3 days post-operatively showed no difference in the development of hematoma (p = 0.47). The rate of development of hematoma or pain did not differ between the pattern of the harvested iliac crest bone (p = 0.10, Fig. 4), age (p = 0.47) or BMI (p = 0.31). Patients with or without pain or hematoma did not differ in the length of hospital admission. There was no significant difference in surgery duration between patients with or without post-operative pain (p = 0.08). The use of local anesthetics did not influence post-operative pain (p = 0.45). One patient had a major complication with bleeding and indication for revision surgery. No sensory deficit was found at last follow-up.

Fig. 4
figure 4

Rates of hematoma dependent on the pattern of harvested bone graft

Full size image

Discussion

This study could show for the first time that donor site morbidity after harvesting a small quantity of iliac crest bone graft for distal radius fractures is lower than the morbidity published for harvesting larger grafts for other indications.

Stam et al. [34] showed a complication rate of 7.3% after iliac crest bone graft harvesting. In a systematic review, Dimitriou et al. demonstrated up to 18.96% complications after anterior iliac crest bone graft harvesting including 8.13% hematoma (10% of these with revision surgery), 33.83% chronic pain, 27.36% sensory disturbances, and 9.45% infections. These complications were classified as minor complications. They found 5.8% nerve injuries and 1.49% fractures classified as major complications [9]. Kitzinger et al. described a low complication rate after iliac crest bone graft harvesting using a new specialized reamer device [20]. We found hematoma during hospital admission in 61.9% of patients’ records that were daily documented by nurses and/or doctors. No cases of hematoma were mentioned in the patients’ reports at discharge. One patient (2.38%) received revision surgery twice for hematoma. Several authors describe hematoma from 0.85 to 9% with no discrimination between hematoma with and without indication for revision surgery [4, 9, 19, 34]. Our cohort shows similar results.

Infections after iliac crest bone graft harvesting of between 0 and 7% are described in the literature [4, 18, 34]. Barone et al. [9] reported 235 patients in a retrospective study with complications after anterior iliac crest graft harvesting, such as pain, hematoma, and one pelvic fracture but no infection. Dimitriou et al. found 1.79% infections (57 patients) in 3180 patients. Armaghani et al. described 50 patients in their study after anterior iliac crest bone graft harvesting with a 2% infection rate and conservative treatment [2]. Loeffler et al. reported 3% deep infections treated with irrigation, debridement, and systemic antibiotic therapy in their study of 92 patients [24]. All patients had co-morbidities and high-risk factors for infection. Heneghan et al. analyzed morbidity and quality of life after anterior iliac crest bone graft harvesting and found 7% of patients with wound infection [18].

Persistent pain varies in the literature between 0 and 31% [4, 16, 24, 29, 33]. Armaghani et al. and Stam et al. [2, 4] postulate that pain should have disappeared after 6 weeks. Shin reported no pain at follow-up after 4.5 years, but almost 10% of the 37 patients in their study would not give consent for iliac crest bone graft harvesting again due to pain experienced [33]. In our study, pain was documented in the first few days post-operatively in 9.5% of patients. At discharge, pain was not mentioned. At follow-up, no patient showed pain.

Revision surgery was performed in one patient (2.4%), which is in the range of already published data: Schaaf et al., Calori et al. and Fasolis et al. reported pelvic fractures with following osteosynthesis [7, 13, 30]. Heneghan described bowel perforation with a bone fragment of the iliac crest as a major complication and following laparotomy [18]. Nerve injuries with irritation of the lateral cutaneous femoral nerve are reported between 0 and 10.3% at 1 year post-operatively [4, 9, 24, 30, 33]. In our study, we did not find any nerve injuries at last follow-up which is similar to the results reported by Barone and Shin. Although pain can be reduced using local anesthesia after wound closure, we did not see differences in patients with or without the use of local anesthesia regarding pain [38]. Similar to the results of Raposo-Amaral, we did not see differences in pain levels in patients with a different pattern of the harvested bone graft [28]. Vura et al. reported post-operative pain with a mean visual analog scale (VAS) of 7.68 for a mean period of 6.3 days. In that study, children were included, and so the results are not representative for adolescents. No nerve injuries or infections were evaluated and likewise no hematoma [37].

This study has several limitations: first, it has a retrospective design without randomization and some data were found in patients’ records. Second, the study population of 42 patients is relatively small. Although no other paper regarding complications after iliac crest bone graft harvesting for distal radius fractures has been published, different results with more complications might be obtained. Third, the use of allograft bone or bone substitutes reduces donor site morbidity to zero. However, the advantage of most substitutes is not evidence based, although their use reduces pain and complications of the iliac crest to zero [17, 22, 31]. Fourth, this study shows a very low complication rate compared to published literature. This could be due to the small amount of bone graft harvested with less trauma compared to other indications for bone graft.