Intraosseous microdialysis for bone free flap monitoring in head and neck reconstructive surgery: A prospective pilot study

Abstract Background Although some researchers have positioned microdialysis catheters in the soft tissue surrounding bone, the results did not accurately reflect bone metabolism. The present study's objective was to establish the feasibility of microdialysis with a catheter positioned directly in bone. Methods Thirty‐four patients (19 males, 15 females; median age: 59) were included in a prospective, nonrandomized clinical trial in the Department of Maxillofacial Surgery at Amiens‐Picardie University Hospital (Amiens, France). Fibula or iliac crest free flaps were used in reconstructive head and neck surgery (for cancer, osteoradionecrosis, trauma, or ameloblastoma) and monitored with microdialysis catheters positioned in a hole drilled into the bone. Glucose, lactate, pyruvate, and glycerol concentrations were analyzed for 5 days. Results All catheters were positioned successfully, and thrombosis did not occur during the monitoring. In two patients, an increase in the lactate concentration and a glucose level close to 0 were associated with signs of flap necrosis, with removal on Days 9 and 50. In viable flaps, the mean glucose level was 2.02 mmol/L, the mean lactate level was 8.36 mmol/L, and the mean lactate/pyruvate ratio was 53. Forty percent of the glucose values were below 1 mmol/L, and 50% of the lactate/pyruvate ratio values were above 50—suggesting a specific metabolic pattern because these values would be considered as alert values in soft tissue. Conclusion Monitoring bone free flaps with intraosseous microdialysis is feasible. This technique specifically assesses bone viability, and further studies are now necessary to define the alert values in bone.

Microdialysis provides an accurate, objective measurement of the flap's metabolism during ischemia, and thus produces an early, sensitive indication of flap failure (Chae et al., 2015).
The concept of microdialysis is based on the sampling of metabolites present in the interstitial liquid (glucose, lactate, pyruvate, and glycerol) and whose concentrations vary in the event of tissue hypoxia (Delgado et al., 1972). When perfusion is impaired (e.g., by ischemia), the glucose and pyruvate levels decrease and the lactate level increases as a result of anaerobic metabolism. These changes are ultimately followed by an increase in glycerol levels related to cell lysis. Ungerstedt and Rostami (2004) established the usual normal values for soft tissue based on the studies of the brain: 2 mmol/L for glucose, 120 μmol/L for pyruvate, 2 mmol/L for lactate, and 15 to 20 for the lactate/pyruvate ratio. The researchers further stated that the lactate/pyruvate ratio appears to be a reliable marker of tissue ischemia since there is a statistically significant correlation between the lactate/pyruvate ratio and clinical outcomes. Hence, Ungerstedt and Rostami defined a lactate/pyruvate ratio >25 as a warning sign for metabolic crisis and energy deficiency. In the same year, Setala et al.'s study of a microvascular flap model showed that a decrease in the glucose concentration and an increase in the lactate concentration were associated with arterial and venous occlusion (Setälä et al., 2004). Some years later, Birke-Sørensen, Toft, and Bengaard (2010) specified the alert values in the case of soft tissues transfer. Compromise of a muscle-free flap can be suspected when the glucose level falls below 1 mmol/L, when the lactate level exceeds 15 mmol/L, and when the lactate/pyruvate ratio increases above 25. Although a few researchers have used microdialysis to monitor bone free flaps (Laure, Sury, Bayol, & Goga, 2009;Mourouzis, Anand, Bowden, & Brennan, 2007), they all placed the catheter in the surrounding muscle or in a skin paddle; these sites do not, however, reflect bone vitality. Only Bøgehøj, Emmeluth and Overgaard (2007) used intraosseous microdialysis to evidence ischemia in human femoral heads removed during total hip replacement. Nevertheless, an accurate assessment of bone vascularization is essential for improving bone free flap survival rates.
We hypothesized that a catheter placed directly into a bone free flap can assess local perfusion by monitoring glucose, lactate, and pyruvate levels and lactate/pyruvate ratio. To test this hypothesis, we conducted a pilot clinical study of the feasibility and reliability of intraosseous microdialysis for bone free flap monitoring in head and neck reconstructive surgery.

| PATIENTS AND METHODS
This prospective, single-center, nonrandomized, pilot clinical study (ClinicalTrials.gov identifier: NCT01879384) was conducted in the Department of Maxillofacial Surgery at Amiens-Picardie University Medical Center (Amiens, France). The study was approved by the local institutional review board (CPP Nord-Ouest II: reference 2010/42, ID-RCB 2010-A01176-33). All participants provided their written, informed consent prior to inclusion. The study was conducted in accordance with the tenets of the 1975 Declaration of Helsinki and its subsequent amendments.
Each patient due to undergo reconstructive facial surgery involving a bone free flap was invited to participate in the study. The main inclusion criteria were aged over 18, an indication for head and neck reconstructive surgery with an iliac crest or fibula free flap, and the provision of informed consent. The main exclusion criteria were aged under 18, and reconstructive surgery with another type of flap.
The venous connection was the thyrolinguofacial trunk in 44.12% of the T A B L E 1 Characteristics of the study population Variable Population (n = 34) cases (n = 15), the external jugular vein and the facial vein in 20.59% each (n = 7 each), and the internal jugular vein in 14.71% (n = 5). The mean ± SD duration of ischemia was 90.48 ± 35.15 min (Table 2).

| Intervention and surgical procedure
All patients underwent facial reconstructive surgery using either an iliac crest free flap or a fibula free flap. In addition to routine clinical monitoring, bone free flaps were monitored using the microdialysis catheter positioned directly in the bone tissue. Flap harvesting and shaping (including osteosynthesis) and surgery were performed according to standard techniques. In the cases of cancer resection, the neck was dissected. Microsurgical anastomoses were end-to-end or side-to-end. Before closure, a hole (diameter: 2.3 mm) was drilled into the bone part of the flap. A sterile, double lumen microdialysis catheter was placed through the skin into the bone using a splittable introducer and was anchored to the skin with a suture stitch. The microdialysis pump and perfusion solution were connected immediately after implantation of the catheter. The dialysate was monitored for 5 days; the samples were analyzed every hour on Day 0, every 2 hr on Day 1, and every 3 hr on the following days. The dialysis fluid collected in the microvials was analyzed using an ISCUSflex ® analyzer (CMA Microdialysis AB), with an automatic, real-time graphical display for the level of each metabolite (including glucose, lactate, pyruvate, and glycerol).
The data were processed using LABpilot ® software (CMA Microdialysis AB). At the end of the monitoring period, the catheter was easily removed. The mean levels and kinetic parameters were calculated for the glucose and lactate concentrations and the lactate/pyruvate ratio.
Finally, the equilibration period (defined as the time required for the achievement of stable metabolite concentrations, as monitored by microdialysis) was estimated for each patient by the retrospective observation of the time curves (Figure 1).

| RESULTS
The catheter was successfully positioned in all cases, and 88.2%   Table 3 and Figure 3A-C, respectively. The mean    Figure 5A-C, respectively. Forty percent of the glucose values measured during the monitoring period were below 1 mmol/L, and 50% of the lactate/pyruvate ratio values were above 50 ( Figure 5).

| DISCUSSION
To the best of our knowledge, this is the first prospective study to have used intraosseous microdialysis for the intraoperative and postoperative monitoring of bone free flaps. For buried bone flaps, monitoring the viability of the transferred skeletal tissue is challenging.
Moreover, all of the latter studies featured catheters implanted in the surrounding muscle tissue because a reliable, well-perfused cuff of tissue around the catheter is required. For bone flaps or composite flap, the assessment of this adjacent tissue may not be sufficient and might not accurately reflect the bone metabolism. In a study of a fibular free flap, Mourouzis et al. (2007) performed microdialysis on residual fragments of the flexor hallucis longus and concluded that microdialysis catheters cannot be inserted directly into bone. Our present results show that direct monitoring of bone tissue perfusion with microdialysis is possible. Bøgehøj et al. (2007) used microdialysis to evidence the development of ischemia in femoral heads removed from patients undergoing total hip replacement. Next, the researchers explored the phenomenon in the dead space around the catheter in the drill canal and concluded that (a) an equilibration period of 2 hr is necessary so that the measured values are representative of the bone metabolism, and (b) a reference measurement in healthy bone must be made (Bøgehøj, Emmeluth, & Overgaard, 2011). In the present study, the mean equilibration period was 5 hr and 5 min; however, the specific metabolic pattern of bone tissue must be taken into account. Indeed, flaps mainly composed of muscle and those mainly composed of fat have different metabolic patterns. Although it is known that variables involved in glycolysis change faster in muscle than in other tissues in the flap, a specific analysis in bone has yet to be performed (Röjdmark, Ungerstedt, Blomqvist, Ungerstedt, & Hedén, 2002). Our study revealed the metabolic changes in bone tissue after a period of ischemia (intraoperatively and during the first 5 days postsurgery). In a recent review of objective methods primarily the lack of a more secure way to anchor the catheter. Our loss rate of 11.8% might be acceptable for a feasibility study but not for routine clinical practice. Moreover, the clinical use of microdialysis will certainly generate additional costs. However, Setälä, Koskenvuori, Gudaviciene, Berg, and Mustonen (2009) considered that the extra costs of using microdialysis would be covered if one or two flaps per year were saved by more effective monitoring. Furthermore, an equilibrium period of about 5 hr seems to be required before relevant data can be collected for bone free flaps. Although this may be a disadvantage for clinical use, it corroborates the fact that bone free flaps have a particular metabolic pattern. Finally, the use of microdialysis certainly requires additional work but the staff compliance rate for sample collection of 80% in our study was more than acceptable-especially since microdialysis provides objective, quantitative values that are likely to reassure medical teams with little experience of monitoring buried flaps. In view of (a) the abovementioned factors, (b) the challenging nature of monitoring of bone free flaps, and (c) the financial and human costs of surgical revision, it seems reasonable to take advantage of the additional assistance provided by microdialysis. This is why further studies of a larger numbers of patients will now be required to define

| CONCLUSION
Our present results demonstrate the feasibility of intraosseous microdialysis for safely exploring bone perfusion in free flaps. This technique allows the specific assessment of bone tissue viability. Given that our prospective measurements were correlated with the clinical outcomes, microdialysis data may be of value to clinicians in the early diagnosis of ischemic events. With a view to broadening the clinical use of this technique, it is now necessary to define the specific alert value for each metabolite in bone free flaps and to specify the latter's metabolic pattern, specifically after reperfusion.