Leech Therapy for Patients With Surgically Unsalvageable Venous Obstruction After Revascularized Free Tissue Transfer

Objective  To assess the efficacy and associated complications of a leech therapy protocol used for patients with a head and neck free tissue transfer in whom flap viability is threatened because of surgically unsalvageable venous obstruction.

Design  Medical record review of a prospective protocol.

Setting  Tertiary care academic medical center.

Patients  Of the 450 free tissue transfers to the head and neck region performed by our microvascular program from January 1, 1995, to October 31, 2000, 8 patients (1.8%) developed venous obstruction not considered salvageable by conventional surgical or thrombolytic therapy.

Interventions  All 8 patients were placed on a protocol using leeches (Hirudo medicinalis), intensive care unit monitoring, antithrombotic pharmacotherapy, frequent hematologic evaluation, blood transfusions as needed, and antibiotic prophylaxis for Aeromonas hydrophila.

Main Outcome Measures  Flap salvage rate, number of leeches used per patient, time needed for inosculation, duration of intensive care unit admission, transfusion requirement per patient, and complications during leech therapy.

Results  All 8 flaps survived with the application of this protocol. An average of 215 leeches were used per patient, and the average time needed for inosculation was 6.6 days. The average duration in the intensive care unit was 9.6 days. The morbidity of our protocol was substantial, with intensive care unit psychosis, prerenal azotemia, and large transfusion requirements being the most frequent complications. An average of 13 U of packed red blood cells per patient was necessary.

Conclusions  Aggressive application of the presented leech therapy protocol can salvage free tissue transfers with venous obstruction that are otherwise unsalvageable. The associated morbidity can be marked. Thus, judicious application of this protocol for flap preservation is essential.

ALTHOUGH MANY microvascular programs have independently reported excellent success rates for free tissue transfer to the head and neck region, occasional flap failures do occur. The most common cause of these failures is venous obstruction.^1-3 Once venous obstruction is identified, these patients require emergency exploration to reestablish venous outflow. Reported salvage rates from emergent exploration vary from 19% to 100%.^1-7 A recent survey of 95 microsurgeons reported a 41% salvage rate in 192 threatened flaps.⁸ Occassionally, despite the surgeon's best efforts, venous obstruction may not be surgically salvageable because of microcirculatory problems within the flap or lack of alternate recipient venous access. For these cases, when flap failure is certain, leeches can be used as an alternative method for reestablishing venous outflow until inosculation occurs.

Some microvascular surgeons believe that leech therapy is a temporizing measure but, alone, will not save a free flap.^9,10 However, previous reports describe the successful use of leeches for treating flaps with venous congestion.^11-14 These reports do not concentrate solely on head and neck free tissue transfer, and there is limited information on the protocols used for these cases. In addition, other than Aeromonas hydrophila infections, the morbidity of leech therapy, specifically the transfusion requirement, has been infrequently reported.

This retrospective study describes the application of a leech therapy protocol for head and neck free tissue transfers with surgically unsalvageable venous obstruction. This study addresses the flap survival rate and associated morbidity with the application of this aggressive protocol.

The surgeons in the microvascular program in the Department of Otolaryngology at the University of Michigan, Ann Arbor, performed approximately 450 free tissue transfers to the head and neck region from January 1, 1995, to October 31, 2000. A medical record review identified 8 patients (1.8%) who developed venous outflow obstruction determined to be not salvageable by surgical or thrombolytic therapy. Arterial inflow was maintained for all of these patients.

There were 6 men and 2 women whose average age was 58 years, with a range of 49 to 73 years. The defects requiring revascularized free tissue reconstruction resulted from 4 oral cavity squamous cell carcinomas, a parapharyngeal space sarcoma, a midface basal cell carcinoma, an orocutaneous fistula closure, and a secondary reconstruction for a failed fibula free flap skin paddle. The revascularized free tissue transfers included 3 fasciocutaneous radial forearm flaps, 2 osteocutaneous parascapular flaps, a fasciocutaneous parascapular flap, an osteocutaneous fibula flap, and an osteomyocutaneous iliac crest flap.

All patients who received leech therapy had surgically unsalvageable venous obstruction. The sites of the venous obstruction were grouped into 3 locations: the flap microcirculation (2 patients), the cutaneous perforators in an osteocutaneous flap (3 patients), and diffuse venous thrombosis (3 patients). Both patients with microcirculatory obstruction underwent immediate reexploration. In these cases, there appeared to be adequate arterial inflow and an unobstructed venous anastomosis but inadequate venous outflow. In the second group, only 1 of the 3 patients (patient 5) with obstructed cutaneous perforators underwent immediate reexploration. Patent anastomoses were found, with signs of appropriate inflow and outflow to the muscle and periosteal portions of the flap. The other 2 patients in this group (patients 2 and 3) did not undergo reexploration because of the clinical impression that the skin paddle was the only portion of the flap threatened with venous congestion. Finally, in the patients with diffuse venous thrombosis, 2 of the 3 patients (patients 4 and 7) underwent immediate reexploration. Both patients were found to have diffuse thrombosis of all potential recipient neck veins in both the superficial and deep venous systems and of the flap donor veins such that adequate outflow could not be restored. One patient in this group (patient 8) did not undergo reexploration because of bilateral internal jugular vein thrombosis and severe facial edema. In this case, the likelihood of a successful surgical salvage was low and leech therapy was initiated.

In all patients who underwent reexploration for venous congestion, a series of steps was generally followed during the attempted salvage. First, arterial patency was confirmed by both Doppler and direct inspection of the anastomosis. Various techniques were then used to relieve venous obstruction, depending on the cause. When the recipient vein was thrombotic, the anastomosis was revised with the use of alternative recipient veins. Vein grafts were used when necessary. When the venous obstruction was in the flap microcirculation, thrombolytic therapy with 50 000 U of streptokinase in divided doses was injected into the flap artery. Leeches were used to temporize the venous obstruction and then continued postoperatively when anastomosis revision or thrombolytics could not relieve the venous obstruction.

All patients identified above were placed on a protocol of leech therapy. Five patients received extraoral and 3 patients received intraoral leech therapy. Hirudo medicinalis was the leech species used. This protocol is given in Table 1 and the contraindications are listed in Table 2. No patient identified for this leech therapy protocol had any of the contraindications listed in Table 2. No patients had leech therapy prematurely stopped because of inability to tolerate the treatment.

Outcome variables assessed included successful salvage of the free tissue transfer and the associated morbidity of implementing this leech therapy protocol. The degree of flap survival and need for further wound care were documented by the attending physician in the medical record. Morbidity was measured by the length of time needed for intensive care unit monitoring, blood transfusion requirement, and complications arising during leech therapy.

Table 3 summarizes the patient data. All 8 flaps survived with application of the leech therapy protocol. Of the 8 patients, the entire skin paddle survived in 5 patients, greater than 90% of the skin paddle survived in 2 patients, and 1 patient had partial loss of the fibula skin paddle but complete viability of the underlying subcutaneous fat and fascia. This wound healed by secondary intention without bone exposure, fistula formation, or subsequent functional consequences.

There were 4 bone-containing flaps. Although no bone scans were performed to assess viability of the bone segment of the osseous free tissue transfers, there has been no clinical evidence of wound breakdown, extrusion, resorption, or alteration of the bony contour that would be anticipated with nonviable bone.

An average of 215 leeches were required for each patient, with a range of 80 to 350. Since leeching began on a continuous schedule and each leech remained attached to the flap for approximately 20 minutes, this equated to approximately 3 leeches per hour. Typically, this continuous leeching schedule lasted for a minimum of 24 hours. Subsequently, the leech schedule was based on a clinical assessment of the degree of venous congestion, rather than the volume or surface area of the flap. The minimum number of leeches was used to relieve clinical evidence of venous congestion. The average length of time required for leeching was 6.6 days, with a range of 3 to 10 days. The average duration of intensive care unit monitoring was 9.6 days, with a range of 5 to 14 days.

Patients placed on this protocol required frequent, skilled hemodynamic assessment because of the large amount of blood loss. Clinical and hematologic evaluation were performed every 4 hours. Even with attentive care, the morbidity related to this protocol was substantial. Table 3 displays the blood transfusion requirement for all patients. An average of 13 U of packed red blood cells was necessary, with a range of 5 to 28 U. No patient experienced a significant transfusion-related complication. Other complications arising during treatment were intensive care unit psychosis in 5 patients, prerenal azotemia in 4, and congestive heart failure, pneumonia, decubitus ulcer, and non-Aeromonas wound infection in 1 each. No patient developed complications from leech migration into the aerodigestive tract. No patient developed a local infection related to the leech therapy. No long-term scarring of the skin paddles from leech bites was evident.

All patients received systemic pharmacotherapy to alter thrombogenic potential during the time required for leech therapy. Aspirin was used for all patients, with the addition of intravenous heparin in 4 patients, subcutaneous heparin in 1 patient, dextran 40 with subcutaneous heparin in 1 patient, or dextran 40 followed by heparin in 1 patient. The choice of anticoagulants depended on the preference of the attending physician and the clinical situation. The 2 patients who developed a neck hematoma as the cause of the venous obstruction (patients 1 and 4) did not receive intravenous heparin in the immediate postoperative period. No patient experienced a hematoma related to these anticoagulative measures.

Despite the high success rate of free tissue transfer, approximately 8% to 11% of cases require reexploration for threatened flap viability.^2,4,5,7 For unsalvageable flaps, a second free flap often needs to be performed if the surgeon does not want to abandon the original goals of the reconstruction.^2,15-17 This study focused on patients with venous obstruction whose flaps could not be salvaged by surgical means. These patients had extensive thrombosis of all the alternative recipient veins, secondary microcirculatory damage, or occluded venous cutaneous perforators.

This study shows that head and neck free tissue transfers with surgically unsalvageable venous obstruction can be reliably salvaged by means of the described leech therapy protocol, thus sparing the patient a second major reconstructive procedure with its inherent morbidities. This finding is contrary to the belief that leeches alone cannot salvage a free flap, but rather are only useful for providing a temporary alternative method for venous outflow. Leech protocols are helpful because revision of the venous anastomosis does not guarantee correction of the outflow problem because of resultant microcirculatory abnormalities. Studies in rat and pig models show that secondary ischemia related to venous obstruction is more detrimental to flap survival than is arterial ischemia or complete pedicle obstruction.^18,19 This is true because, as proved histologically, venous obstruction causes microcirculatory thrombosis, platelet trapping, and stasis.^19,20 Thus, even after successful reanastomosis, secondary changes in the microcirculation can persist and prevent adequate outflow from being reestablished.

The effectiveness of medicinal leech therapy in relieving venous congestion is due to both mechanical and biological effects. Each leech directly extracts 5 to 15 mL of blood from the flap, but an extra 20 to 50 mL of blood is lost from oozing from the bite site after the leech detaches.²¹ The saliva of H medicinalis contains coagulation inhibitors (hirudin, factor Xa inhibitor, and bdellin), a platelet aggregation inhibitor (apyrase), and a vasodilator (histamine) that get directly injected into the flap through the bite site. The leech saliva also contains collagenase and hyaluronidase, which facilitate local infiltration of the antithrombotic mediators into the congested tissue.^22,23 Thus, leech therapy addresses both the venous outflow and microcirculation.

At the time of initial tissue transfer, the use of leeches when no venous anastomosis is possible has rarely been reported for free flaps, but it is well established for digit,²⁴ auricle,^25,26 nose,²⁷ and lip²⁸ replantation. Leech therapy used for salvaging head and neck free flaps that would otherwise require another major reconstruction is not as well described. De Chalain¹² described 4 head and neck free flaps that required leeching, but 2 of these treated flaps likely did not have complete venous obstruction because of the short duration of leeching required (2 hours). We learned from our experience that when leeches provide the only source of venous outflow, the treatment must be continued until inosculation occurs. On the basis of experimental animal models, this process needs 3 to 5 days.^29,30 Clinical data suggest that 6 to 10 days is necessary.^11,14,31 Our patients required leech therapy for an average of 6.6 days.

Guidelines for the safe use of medicinal leeches are well described,^22,31 but there are few published protocols.^21,32 The optimal leech application schedule is unknown because there is little consistency in the reported frequency or length of time needed for leech application.¹² Our experience for these salvage cases dictates that significantly more leeches are required than is generally reported in the literature. In this study, an average of 215 leeches per patient were used. Previous studies report that 2 to 50 leeches are required for flap salvage.^13,33 Some reconstructive surgeons believe that the frequency of applying leeches to the flap should be minimized as long as oozing from previous leech bites is providing continued venous drainage.¹² This discrepancy can also be explained if leeches were being used as an adjunct to treat lesser degrees of venous congestion rather than as a true salvage therapy.

Although all patients' flaps were successfully salvaged by means of our protocol, we found that marked associated morbidities occurred during leech therapy. Large transfusion requirements were common, with patients requiring an average of 13 U of packed red blood cells. The need for blood transfusions during leech therapy is not consistently reported in the literature. Of 108 cases reviewed from the published literature, de Chalain¹² found that only 6 cases reported the transfusions requirements during therapy. In these 6 cases, an average of 6.8 U of packed red blood cells was needed. Of his own 18 cases, de Chalian found an average transfusion requirement of 4.4 U of packed red blood cells. As many as 14 U for free flaps¹³ and 17 U for replants²⁸ have been reported. Thus, patients undergoing leech therapy need to be properly counseled about the likely need for blood transfusions and the associated risks.

The use of multiple, concomitant antithrombotic medications with leech therapy likely increases the risk of significant blood loss. Replanted digits postoperatively treated with aspirin, intravenous heparin, and leeches had an average decrease in hemoglobin of 5.3 g/dL.³⁴ An additional risk related to postoperative anticoagulant use is development of hematomas. The risk of major postoperative bleeding is difficult to determine, but an estimate of 3% for each 2 days of intravenous heparin therapy has been suggested.³⁵ Kroll et al³⁶ found a 20% incidence of hematoma in 30 patients treated with conventional doses of heparin (500 to 1200 U/h) after free flap surgery. We had no hematomas in our cohort of patients, who were on a systemic anticoagulation regimen despite extensive ablative and reconstructive procedures.

Because of large transfusion requirements, the question arises whether systemic anticoagulation is necessary during leech therapy. Microvascular surgeons generally reject routine use of full anticoagulation after uncomplicated flap surgery. In salvage situations after vascular thrombosis, however, intravenous heparin use may be beneficial. In a rat model, Hirigoyen et al³⁷ found a significantly better flap survival rate when intravenous heparin was used for 7 days after thrombolytic salvage. That study also demonstrated notable differences in the venous microcirculation by electron microscopy. Although microvascular surgeons commonly use a postoperative heparin drip for "high-risk" flaps, it is difficult to demonstrate an improved flap survival rate.⁴

Other morbidities associated with this protocol were encountered. The average duration of intensive care unit monitoring was 9.6 days. As a result, intensive care unit psychosis was a frequent occurrence in our patients (62%). Volume management was especially challenging in these patients. Generally, the blood loss was underestimated related to the oozing from the leech bite sites. Even with attentive physician care and frequent hematologic assessments, 50% of patients developed prerenal azotemia.

A complication of leech therapy that we did not encounter, but that is frequently described in the literature, is Aeromonas infection. The bacterium A hydrophila resides in the leech gut, and infections complicate 7% to 20% of cases when H medicinalis is used.³⁸ These infections can occur acutely (within 24 hours) or in a delayed fashion (up to 26 days) after starting leech therapy. Lineaweaver et al³⁸ reviewed data from 8 patients with flaps in whom leech-related Aeromonas infections developed and found that 50% of infections occurred more than 10 days after leech application began. Clinical presentations of infections can range from a minor wound infection to extensive tissue loss to sepsis. Infected tissues commonly are not salvaged. De Chalain¹² performed a meta-analysis of 19 Aeromonas infections occurring in 9 replantations, 3 microsurgical flaps, and 7 pedicled flaps. There was a 31% tissue survival rate, as compared with an 83% tissue survival rate when no infections were associated with leeching 43 replants, 7 microsurgical flaps, and 39 pedicled flaps.

Prophylactic antimicrobial therapy is an important measure for preventing Aeromonas infections from leech therapy. Lineaweaver et al³⁹ found that Aeromonas could be cultured from the leech gut in only 12% of cases after application to patients receiving appropriate prophylactic antibiotics. This compares with the 100% positive culture rate that was found when leeches were not exposed to antibiotics. Lineaweaver et al³⁸ and de Chalain¹² also found that most leech-related Aeromonas infections occurred while patients were receiving inadequate or no prophylactic antibiotics. Aeromonas hydrophila is sensitive to second- and third-generation cephalosporins, ciprofloxacin, sulfamethoxazole-trimethoprim, tetracycline, and aminoglycosides. Notably, Aeromonas is resistant to penicillin, ampicillin, first-generation cephalosporins, and erythromycin. Five of our patients were treated with single-agent prophylaxis, but 3 patients were treated with double antibiotic coverage. This treatment decision was based on the recommendations of our infectious disease colleagues. We felt that because of this aggressive prophylaxis regimen, none of our patients experienced a leech-related wound infection.

This study shows that application of a leech therapy protocol can reliably salvage head and neck free tissue transfers with venous obstruction not otherwise salvageable by surgical means. Flaps undergoing leech therapy must have continued arterial inflow. Eliminating the need for a second reconstruction, likely with another microsurgical flap, is the benefit of successful salvage with leech therapy. The risks from transfusions, Aeromonas infections, intravascular volume shifts, and prolonged admission to the intensive care unit must be seriously considered. Thus, judicious application of this protocol for flap preservation is essential.

Accepted for publication January 17, 2002.

Corresponding author and reprints: Douglas B. Chepeha, MD, MSPH, Department of Otolaryngology, University of Michigan Health System, 1904 Taubman, 1500 E Medical Center Dr, Ann Arbor, MI 48109-0312 (e-mail: dchepeha@umich.edu).