Monitoring free flap venous congestion using continuous tissue glucose monitoring: A case report.

Blood glucose levels (BGLs) are a good indicator of postoperative venous congestion caused by a thrombus at the anastomotic site of a free flap. Tissue glucose levels (TGLs) are believed to be superior to BGLs for two reasons: TGLs are thought to represent a tissue's congestive status more directly than BGLs and are able to be measured by a continuous tissue glucose monitoring device (CTGMD), whereas BGLs must be measured manually by sampling the flap, hindering the patient's sleep and increasing the nurse's workload. A case is described in which a postoperative thrombus developed in a free flap vein three times. TGL in the flap was monitored by a CTGMD (Free Style Libre®, Abbott, U.S.A.), and BGL was monitored in parallel by conventional sampling of the flap. When venous congestion developed at the anastomotic site, TGLs decreased faster than BGLs; after the congestion was ameliorated by exsanguination, BGLs increased faster than TGLs, indicating that TGLs are a better indicator of venous thrombosis at the anastomotic site than BGLs.

Introduction

Blood glucose monitoring after microsurgical anastomosis of a flap is a simple and reliable method to detect thrombosis at the anastomotic site1, 2, 3. We revised this method using a continuous tissue glucose monitoring device (CTGMD; Libre®, Abbott, U.S.A.), which only requires tissue fluids for measurements of tissue glucose levels (TGLs). During such measurements, the flap would not need to be sampled for conventional measurements of blood glucose level (BGL). To evaluate this new monitoring technique, the device was used to measure TGLs of free flaps postoperatively. Both continuous tissue glucose monitoring and conventional blood glucose monitoring were performed, and the data were compared.

Case Report

A 29-year-old female with a chronic ulcer on her right lateral malleolus, which was caused by postoperative infection after orthopaedic surgery, was referred to our hospital. Debridement of the ulcer and coverage with a superficial inferior epigastric artery perforator flap was performed with an end-to-end arterial anastomosis to the dorsalis pedis artery and two end-to-end anastomoses to the subcutaneous and dorsalis pedis veins. A CTGMD was applied to the flap after surgery, and TGLs were continuously monitored for six days (Figure 1). The device was designed to automatically obtain TGL every 15 min and work for two weeks without replacement. Conventional blood glucose monitoring of the flap was also performed by measuring BGL of the flap via pin pricks every three hours. When the flap became pale or BGL dropped to <40 mg/dL, the cut-off value for flap congestion, a doctor would be called for an evaluation.

Figure 1

The CTGMD was applied to the centre of the flap.

Venous thrombosis occurred at the anastomotic site three times postoperatively. By 14.5 h after surgery, TGL began to decrease; two hours later, the CTGMD indicated a “low” level, meaning below 40 mg/dL, whereas BGL was 75 mg/dL. Because the flap had apparently developed congestion, the patient was taken to the examination room; the comitant vein of the flap was found to be thrombosed and was subsequently exsanguinated. After approximately 45 minutes, BGL increased to 124 mg/dL, whereas TGL was 41 mg/dL (Figure 2). A second surgery was performed, and the comitant vein was re-anastomosed to the greater saphenous vein. Subsequently, BGLs stayed above 80 mg/dL, and TGLs improved but stayed relatively low. More than 9 h later, TGL dropped, and BGL decreased to 57 mg/dL. The flap again developed congestion; exsanguination was again performed 3 min later. Within 20 min, TGL increased to 52 mg/dL; BGL increased to 106 mg/dL in less than 2 h (Figure 2). After the second surgery, TGL stayed above 50 mg/dL for an hour but eventually decreased to 49 mg/dL 2 h later, whereas BGL stayed high. In >2 h, BGL dropped to 59 mg/dL, and the flap again developed congestion. The anastomotic site was found to be thrombosed and was exsanguinated within the hour. Within minutes, TGL was 43 mg/dL and BGL increased to 134 mg/dL (Figure 2). After the third surgery, both BGL and TGL increased to normal levels. The postoperative course was uneventful, and both TGL and BGL stabilized above 80 mg/dL; however, TGL began to decrease on postoperative day 5, while BGL stayed above 100 mg/dL (Figure 3). The flap colour was normal and there were no signs of congestion. The flap was found to be compressed by a cushion during sleep. The cushion was removed on postoperative day 6, and TGL increased to 71 mg/dL within approximately 20 min (Figure 3).

Figure 2

TGLs and BGLs during three venous thromboses after flap transplantation were plotted.

Figure 3

TGLs and BGLs during flap compression were plotted.

Discussion

Free flaps have a risk of thrombosis at the anastomotic site postoperatively. Therefore, postoperative monitoring of the flap is vital. Hara et al. reported that BGL was an indicator of flap venous congestion. We propose that TGL is superior to BGL as an indicator of flap hypoxia caused by thrombosis at the anastomotic site. As each event of venous congestion occurred in our patient, TGL decreased earlier and faster than BGL. However, after exsanguination, BGL increased earlier and faster than TGL. This phenomenon indicates that TGL represents tissue status more directly than BGL. Our hypothesis is as follows: When venous congestion occurs, the tissue becomes hypoxic, and anaerobic glycolysis is accelerated, resulting in tissue glucose depletion. As the venous network functions as a glucose reservoir, blood glucose depletion is delayed, resulting in a slower decrease in BGL than in TGL. However, after exsanguination, hypoxia causes reversible tissue damage that is restored by the consumption of tissue glucose, resulting in slower TGL recovery. Additionally, hypoxic and hypoglycaemic blood is flushed from the flap by the recovered blood circulation, resulting in rapid BGL recovery.

The sudden decrease in TGL on postoperative day 5 was caused by compression: TGL was low and BGL was normal, because the CTGMD was applied to the flap during its compression, and TGL directly represented the condition of the hypoxia caused by the compression. However, BGL was measured from a flap sample after the flap was released from the compression.

To summarise, TGL could be superior to BGL as an indicator of flap venous congestion. However, BGL could be superior to TGL as an indicator of flap recovery from congestion after exsanguination or venous re-anastomosis.

To our knowledge, this is the first report describing the postoperative monitoring of free flap TGLs. We believe that the CTGMD would detect an anastomotic failure much earlier than would an experienced clinician.

The CTGMD was originally developed and used for monitoring TGLs in diabetic patients. The device consists of a coin-form sensor and a reader that graphically displays the data. The sensor is designed to be placed on the skin for 2 weeks without the need for replacement. It has a metallic filament that selectively binds tissue glucose molecules, thus working as a glucose detector. The cost of the sensor is 6800 JPY (61.8 USD), and the cost of the reader is 7089 JPY (64.4 USD).

This report describes a case of venous occlusion; arterial occlusion is not discussed. We experienced arterial occlusion of the free flap caused by arterial spasm while TGLs were being monitored. TGLs began to decrease to “low” levels after approximately 2 h; eventually, the device indicated a “sensor error,” making further analysis difficult.

TGLs decreased with compression, even when the circulation of the flap was normal. This finding indicates that TGLs would be useful for the detection of subclinical flap compromise, including a kink in the vessel or a hematoma that compromises blood flow but has not yet resulted in complete occlusion. On the contrary, a low TGL would be detected in the absence of congestion at the anastomotic site, triggering a false alarm by the device.

The cut-off BGL was set to 40 mg/dL based on our experience. Considering that BGLs below 50 mg/dL are considered indicative of hypoglycaemia, using 50 mg/dL instead of 40 mg/dL as a cut-off value would increase sensitivity without also triggering a false alarm.

Theoretically, the device could be used on a non-skin flap (e.g., muscle flap), as the device does not require blood for the measurement of tissue glucose levels. However, the sensor could not be easily affixed to a wet flap and would require an alternative form of attachment.

Conclusions

TGL could be superior to BGL as an indicator of flap venous congestion at the anastomotic site.