A Regional Pedal Ischemia Scoring Systemfor Decision Analysis in Patients with

Heel UlcerationAndrew T. Gentile, MD, Scott S. Berman, MD, Kurt R. Reinke, MD, Christopher P. Demas, MD,

Daniel H. Ihnat, MD, John D. Hughes, MD, Joseph L. Mills, MD, Tucson, Arizona

PURPOSE: The objective of this study was toevaluate patients undergoing operative debride-ment for heel ulceration and to categorize pedalperfusion and its influence on therapeutic alter-natives.

METHODS: Patients with heel ulceration werestratified by arteriography and graded I (patentposterior tibial, PT), II (occluded PT/reconstitutedfrom peroneal), III (PT reconstituted from dorsalpedal), IV (no PT reconstitution but visible heeltributaries), and V (avascular heel).

RESULTS: From May 1992 through January 1997,23 patients underwent operative treatment for 25heel ulcers. The heel ischemia score stratifiedpatients into two groups: 1, revascularization/debridement (71% grades I to III, 29% grade IV,0% grade V); and 2, free tissue transfer with orwithout revascularization (100% grades IV, V).Cumulative functional limb salvage was 91%(BP), 60% (BP1TT), and 81% (TT) at 24 months(P 5 0.15 log rank).

CONCLUSION: The heel ischemia score may directtreatment of heel ulceration by identifying pa-tients who will need vascularized tissue transferearly in their treatment regimen. Am J Surg.1998;176:109–114. © 1998 by Excerpta Medica,Inc.

Ischemic tissue loss of the heel is commonly encoun-tered in chronically ill, debilitated, and hospitalizedpatients. We have noted a disturbingly high incidence

of heel necrosis in diabetic patients on hemodialysis forrenal replacement as they spend considerable time withpressure on their unprotected heels. These patients suffer adisproportionately high rate of major limb loss. They areoften marginally ambulatory or are thought to be too highrisk for extensive foot debridement, revascularization, andwound coverage. Patients with heel necrosis may suffer

limb loss despite a palpable pedal pulse. Adequate forefootperfusion does not ensure healing of the traumatized heelpad, especially in patients with peripheral vascular disease.Andros et al1 reported a case series of diabetic patients withpedal gangrene despite palpable foot pulses, and recom-mended that arteriography be performed in these patientswhen the foot is salvageable irrespective of their pulsestatus. These observations suggest that clinically importantregional circulatory differences exist in the foot. The pres-ence of a patent pedal arterial arch and collateral pathwaysdetermine regional differences in foot circulation. Otherfactors such as diabetic neuropathy2,3 or heel-pad mechan-ical load4 are known to contribute to tissue breakdown.Limb salvage in these patients depends on adequate hind-foot perfusion, debridement of all devitalized tissue, oftenwith partial calcanectomy, and durable wound coveragethat must be at least partially weight bearing.

With the advent of microsurgery and the successful use offree tissue transfer techniques for wound coverage, we andother investigators have reported functional limb salvagedespite extensive tissue loss when combining free tissuetransfer with distal revascularization in appropriately se-lected patients.5,6 Flap viability and limb salvage ratesbetween 70% and 90% are reported, even in elderly anddiabetic patients often with recalcitrant wounds of theweight bearing surfaces of the foot.7,8 The convergence ofdistal arterial reconstruction and free tissue transfer clearlyextends the limits of limb salvage in patients with exten-sive ischemic tissue loss, albeit at a considerable effort.Karp et al8 reported in 1991 an average in-hospital stay of47 days at an average cost of $58,000 for microvascular freetissue transfer for foot salvage in diabetic patients. A clin-ical algorithm for delineating which patients require iso-lated distal revascularization, or microvascular free tissuetransfer, alone or in combination for coverage of heelulceration remains to be defined. Our objective was toreview a contemporary series of patients undergoing oper-ative treatment for heel gangrene and devise a scoringsystem based on objective evaluation of the regional footperfusion. This information could be used to aid in thedecision analysis for foot salvage in patients with advancedischemic heel necrosis.

PATIENTS AND METHODSAs part of our ongoing vascular surgical database registry

at the University of Arizona Health Sciences Center andaffiliated hospitals, demographics and wound characteris-tics of all patients requiring lower extremity revasculariza-tion for ischemic tissue loss were reviewed. Patients under-

From the Sections of Vascular (ATG, SSB, DHI, JDH, JLM) andPlastic Surgery (KRR, CPP), the University of Arizona HealthSciences Center, the Southern Arizona Vascular Institute (ATG,SSB), and Carondelet St. Mary’s Hospital (SSB), Tucson, Arizona.

Requests for reprints should be addressed to Scott S. Berman,MD, Southern Arizona Vascular Institute, PO Box 85727, Tucson,Arizona 85754-5727.

Presented at the 26th Annual Meeting of The Society for ClinicalVascular Surgery, Coronado, California, March 25–29, 1998.

SCIENTIFIC PAPERS

© 1998 by Excerpta Medica, Inc. 0002-9610/98/$19.00 109All rights reserved. PII S0002-9610(98)00168-8

going primary amputation or simple wound debridementwithout revascularization were excluded from further anal-ysis, and have been described in a previous publication.5

Patients presenting with pedal ulceration underwent stan-dard peripheral vascular pulse examination, noninvasivevascular laboratory testing with segmental pulse volumerecordings and Doppler-derived blood pressure measure-ments, ankle-brachial indices, and toe plethysmography.Patients with limited ischemic tissue loss underwent stan-dard extremity revascularization with conventional meth-ods of wound coverage.

Free flap candidates were evaluated and treated through acollaborative effort of the vascular and plastic surgicalservices. Evaluation of patients for free tissue transfer withor without revascularization included complete history andphysical examination with attention to the extent of tissueloss, overall functional status, and current circulatory sta-tus. Often prompt surgical drainage of deep space infec-tion or local debridement of devitalized tissue was neededbefore the full extent of tissue loss could be determined.Selection criteria for patients undergoing free tissue trans-fer included large, limb-threatening soft tissue defectsnot amenable to revascularization alone with conventionalmethods of wound management due to their size, exposureof deep structures, underlying osteomyelitis, or locationover weight-bearing surfaces of the foot. Patients with

neurologic impairment preventing functional ambulation,those nonambulatory, and those with limited life expect-ancy were not considered for free tissue transfer and wereusually treated conservatively or with primary amputation.

Records and preoperative arteriograms of all patients withextensive heel necrosis requiring surgical debridement be-tween May 1992 and January 1997 were reviewed. Demo-graphics and comorbid medical risk factors were comparedbetween patients requiring extremity revascularizationalone or in combination with free tissue transfer for heelcoverage. An arteriographic scoring system was devisedbased on the status and contribution of the pedal vessels tothe subcutaneous perfusion of the heel. Patients were strat-ified by regional pedal perfusion and graded I (patentposterior tibial, PT), II (occluded proximal PT/distal PTreconstituted from peroneal artery), III (occluded proximalPT/distal PT reconstituted from dorsal pedal artery), IV(no PT reconstitution but visible heel tributaries), and V(avascular heel) (Figure 1). Comparison of demographicsand treatment groups were performed with the Student’s ttest and chi-square analysis. Outcomes were assessed forfree flap viability and time for wound healing, and ex-pressed as life-table functional limb salvage. All statisticswere analyzed using standard PC compatible software(Statview, Abacus Concepts, Inc., Berkeley, California)

Figure 1. Representative pedal arteriograms for each category used to describe the regional pedal ischemia score for patients withextensive heel ulceration: I (patent posterior tibial, PT), II (occluded proximal PT/distal PT reconstituted from peroneal artery), III(occluded proximal PT/distal PT reconstituted from dorsal pedal artery), IV (no PT reconstitution but visible heel tributaries), and V(avascular heel).

PEDAL ISCHEMIA SCORING SYSTEM/GENTILE ET AL

110 THE AMERICAN JOURNAL OF SURGERY® VOLUME 176 AUGUST 1998

and were considered significant when the P value was,0.05.

RESULTSDuring the study period, 23 patients (13 male, 10 female,

mean age of 60 years) underwent operative debridement of25 heel ulcers. This represents approximately 4% of pa-tients undergoing operative treatment for lower extremitytissue loss during the same 4.5-year interval at our institu-tions.5 Comorbid medical risk factors in these 23 patientsincluded hypertension 56%, smoking 48%, diabetes 74%,coronary artery disease 30%, congestive heart failure 22%,and end stage renal disease (ESRD) requiring hemodialysis35%. Eighty-seven percent of patients were fully ambula-tory preoperatively, the remainder were able to stand ortransfer. There were 3 patients with previous contralateralmajor amputation. All patients included in the analysishad extensive heel necrosis requiring heel debridement(range 1 to 5 procedures).

Patients were treated after evaluation of the circulatorystatus and potential for wound coverage by either isolatedextremity bypass (BP, 14), combined bypass and free tissuetransfer (BP 1 TT, 5), or free flap alone (TT, 6) dependingon surgeon preference. Heel ulcers were present for a meanof 3.4 months (BP), 5.6 months (BP 1 TT), and 5.2months (TT); P 5 nonsignificant). Eighteen of 23 patientsunderwent 19 lower extremity revascularization for clinicalgrade III, category 5 (SVS/ISCVS) limb ischemia.9 Preop-erative ankle-brachial indices were frequently inaccurateand unhelpful in determining the degree of extremity isch-emia because of arterial incompressibility due to medialcalcinosis. Many examinations demonstrated diminishedpulse volume recording and flat toe plethysmographic trac-ings. Extremity bypass configurations included femoral-above-knee popliteal (1), femoral-below-knee popliteal(3), femoral-tibial (9), femoral-pedal (2), and popliteal-tibial (4). All bypass graft conduits were autogenous andconsisted of 15 reversed greater saphenous vein (GSV), 3in situ GSV, and 1 spliced vein graft. Five patients under-went isolated free flap for wound coverage of 6 heel ulcers,each of these patients had palpable foot pulses preopera-tively.

Regional Ischemia Scoring SystemFourteen patients (BP) underwent revascularization alone

and heel debridement (50% diabetic, 25% ESRD); 5 pa-tients (BP 1 TT) underwent revascularization and freetissue transfer (100% diabetic, 50% ESRD); and 6 patients(TT) underwent free tissue transfer alone (100% diabetic,60% ESRD). The heel ischemia score stratified patientsinto two groups (Table I) (1) revascularization/debride-ment (71% grades I to III, 29% grade IV, 0% grade V); and(2) free tissue transfer with or without revascularization(100% grades IV, V). As is apparent from Table I, patientswith poor regional circulation to the heel as judged bypreoperative arteriography (ischemia scores IV, V) bene-fited by tissue transfer for wound coverage of their heeldefects.

Outcomes of Treatment for Heel UlcerationAll patients had evaluation of their wounds/flaps by both

the vascular and plastic surgical teams. Bypass grafts un-

derwent standard duplex ultrasound graft surveillancewithin 6 weeks of implantation (usually before hospitaldischarge) and at 3-month intervals. Free flap viability wasassessed clinically. There were 7 postoperative wound in-fections (2 of saphenous vein harvest incisions). All healedwith antibiotics and local wound care. Two patients (1 BP,1 TT) had persistent or recurrent extensive deep spaceinfection of the foot with progressive foot sepsis after thereference operation culminating in below-knee amputationfor uncontrollable infection. There were two free flap sep-arations and partial flap edge necrosis requiring reoperationfor debridement.

Characteristics of heel ulcers are presented in Table II. Ingeneral, patients undergoing free tissue transfer for heelcoverage had larger wounds that were present for longerperiods of time and had slightly shorter times to achieve ahealed wound and regain ambulation. Patients undergoingfree tissue transfer with or without extremity revasculariza-tion were more likely to be diabetic (P 5 0.05) thanpatients judged to need bypass alone for wound coveragefor their heel ulceration. Many of these patients also hadESRD. The mean patient follow-up was 14 months (range2 to 54) for the entire series. The mean time to healingwith functional ambulation was 5 months (BP), 4.3months (BP 1 TT), and 4 months (TT). Life-table cumu-lative functional limb salvage was 91% (BP), 60% (BP 1TT), and 81% (TT) at 24 months (Figure 2, P 5 0.15 logrank).

COMMENTSReconstruction of soft tissue defects of the ankle and heel

presents a significant ongoing challenge. The difficulty is

TABLE IResults of Regional Pedal Ischemia Scoring System for

Patients with Heel Ulceration

IschemiaScore I II III IV V

Bypass alone(n 5 14) 3 4 3 4 (29%) 0

Flap 1 bypass(n 5 5) 0 0 0 2 (40%) 3 (60%)

Free flap(n 5 6) 0 0 0 1 (17%) 5 (83%)

* All free flap patients ischemia scores IV, V.

TABLE IICharacteristics of Heel Ulcer Patients

Patient Group Diabetes ESRD Ulcer SizeTime to

Heal

Bypass alone(n 5 14) 50% 25% 8 cm2 5 mo

Bypass 1 freeflap (n 5 5) 100% 50% 54 cm2 4.3 mo

Free flap(n 5 6) 100% 60% 49 cm2 4 mo

P value 0.05 NS NS NS

ESRD 5 end stage renal disease; NS 5 not significant.

PEDAL ISCHEMIA SCORING SYSTEM/GENTILE ET AL

THE AMERICAN JOURNAL OF SURGERY® VOLUME 176 AUGUST 1998 111

compounded by the lack of local tissue flaps large enoughto cover these defects and the exposed deep structures aswell as the necessity for shock-absorbing capacity of tissuesof the weight-bearing areas of the sole of the foot. Histor-ically, diabetic patients with ischemic tissue loss face highrates of amputation. More than one half of patients whodevelop a foot lesion will subsequently develop a contralat-eral foot lesion within 2 years, and about half of thesepatients will require contralateral major amputation.10

Management of elderly diabetic patients with ulcerativetissue loss of these weight-bearing surfaces of the foot hasimproved through better understanding of the interactionsof patient comorbid risk factors,11 peripheral neuropathy,and mechanical properties of the heel pad,4 and throughrecognition and treatment of peripheral vascular insuffi-ciency.2 Prompt surgical drainage of deep space foot infec-tions provides shorter healing time than medical manage-ment of diabetic foot ulcers.12 Unfortunately, heelulceration often extends into the deep subcutaneous tissueand large tissue defects result with problems of woundcoverage and long-term immobility. By supplying well-vascularized tissue from a remote, relatively nonatheroscle-rotic site, free tissue transfer provides prompt wound cov-erage for variety of tissue defects and has greatly expandedthe limits of limb salvage now obtainable in clinical prac-tice.5–8

Although pulsatile blood flow is optimal for healing isch-emic foot ulcerations, regional circulatory differences existbetween the forefoot and hindfoot, and the patency of thepedal arch is critical in providing collateral blood flowbetween different zones of the foot when occlusive vasculardisease limits pedal inflow (Figure 3). Andros et al1 de-scribed a cohort of patients with ischemic foot ulcers withinfrapopliteal arterial occlusive disease and the presence of

foot pulses who failed to heal without revascularization.The analysis of patients with heel ulceration similarlyrequires investigation of the regional perfusion to the heel.Limb salvage in patients with heel ulceration depends onadequate hindfoot perfusion, debridement of all devitalizedand infective tissue, and durable wound coverage. We areunaware of any scientific investigation documenting re-gional differences in pedal circulation. No clear-cut man-agement protocols exist in the medical literature defining asingle, concerted algorithm of the potential treatment op-tions for patients with extensive heel ulceration and itssecondary complications of infection and patient immobil-ity. The purpose of this report was to determine regionalcirculatory patterns in the foot as well as present severaltreatment options currently available to obtain a closed,functional wound in patients with heel necrosis.

We devised a scoring system linking regional perfusion ofthe heel with methods of surgical wound coverage. Allpatients require detailed circulatory evaluation with non-invasive pulse volume recordings and segmental bloodpressures. The presence of a dorsal foot pulse does notguarantee adequate hindfoot circulation in patients withrecalcitrant heel ulceration. In patients who require oper-ative debridement for heel necrosis, the regional circula-tion to the heel should be precisely defined by high-qualitystandard extremity arteriography. We are fortunate to haveexpert diagnostic angiographers who are dedicated to limitsystemic contrast amounts and will often provide single-legantegrade arteriograms or use digital subtraction arterio-grams and magnified views of the heel (Figure 1). Thearterial inflow to the areas at risk, the status of the pedalarch, and the contribution of the subcutaneous arterialtributaries to the heel are important in determiningwhether the circulation to the tissue defect will be ade-

Figure 2. Life-table results of cumulative limb salvage in patients undergoing operative debridement for extensive heel necrosis. BP 5isolated extremity bypass, n 5 14; FF 5 isolated free tissue transfer, n 5 6; BP/FF 5 combined bypass and free tissue transfer, n 5 5.

PEDAL ISCHEMIA SCORING SYSTEM/GENTILE ET AL

112 THE AMERICAN JOURNAL OF SURGERY® VOLUME 176 AUGUST 1998

quate to support granulation and obtain ulcer coveragewith conventional methods (ischemia scores I, II, III), or iffree tissue transfer may be required to cover large avascularheel defects (ischemia scores IV, V).

Investigators agree on the importance of adequate pedalperfusion to heal ischemic tissue loss, especially in thepresence of foot infection.10,11 In a series of 79 patientsundergoing operative drainage of diabetic foot infectionswithout extremity revascularization, Criado et al13 re-ported major amputations in 21 of 79 patients duringfollow-up. Although limited areas of tissue loss of the heelcan be treated with minor debridement, occasionally localrotational tissue flaps may be used to cover chronic neuro-trophic pedal ulcerations.14 Heel ulcerations may extendmuch deeper than originally thought based on inspection.These deeper ulcerations of the heel often require aggres-sive debridement and can be treated with partial calca-nectomy and split thickness skin grafting.15 Calcanectomywith microvascular free tissue transfers for foot reconstruc-tion of large areas of tissue loss have been well described.16

Lastly, the combination of distal revascularization and freetissue transfer is a well-accepted treatment for extensiveischemic tissue loss.5–8

The assessment of heel perfusion and documentation ofthe circulatory status in patients who may have regionalpedal ischemia has not been previously emphasized. Withcareful selection of patients requiring revascularizationand/or free tissue transfer for heel ulceration, we were ableto achieve an overall 90% initial limb salvage rate inpatients with extensive heel ulceration. Life-table, func-

tional limb salvage was 91% in patients who had limitedareas of heel ulceration treated with extremity revascular-ization and local wound care alone. Patients with accept-able functional medical status with larger areas of heelulceration (approximately 50 cm2) benefited from free tis-sue transfer and achieved prompt wound healing and ear-lier return to an ambulatory status than patients undergo-ing isolated leg bypass (Table II). Patients undergoing freetissue transfer for heel coverage were uniformly diabeticwith a high prevalence of dialysis-dependent ESRD. Thesubgroup of patients at greatest risk for major limb loss fromcomplications of heel necrosis were those with combineddistal ischemia requiring extremity bypass and large areas ofischemic heel tissue loss undergoing free flap heel coverage(BP 1 TT, limb salvage 5 60% at 25 months).

The regional pedal ischemia scoring system applied ret-rospectively stratified patients into two groups: (1) Patientswith limited heel ulceration and intact vascular heel trib-utaries through tibial artery collaterals through the foot(ischemia scores I to III) were most often treated withextremity revascularization and wound debridement alone;and (2) patients with extensive heel tissue loss and no

Figure 3. Paramalleolar circulation, dorsal. Inset: Primary pedalarch: AT 5 anterior tibial-dorsalis pedis; DP 5 deep plantar; PA 5plantar arch; LP 5 lateral plantar; PT 5 posterior tibial. Reprintedwith permission from Andros G, Bypasses to the ankle and foot,in The Surgical Clinics of North America, Philadelphia, PA: WBSaunders, Aug 1995;75(4):717.

Figure 4. Intraoperative femoral-pedal arteriogram in a patientwith a 3 3 4-cm ischemic heel ulcer. Patient’s preoperativearteriogram appears in Figure 1, ischemia score III. Heel perfusionwas restored through pedal collaterals after bypass to the dorsalpedal artery.

PEDAL ISCHEMIA SCORING SYSTEM/GENTILE ET AL

THE AMERICAN JOURNAL OF SURGERY® VOLUME 176 AUGUST 1998 113

reconstitution of the posterior tibial artery at the heelregion or with an entirely avascular heel area (ischemiascores IV, V) underwent vascularized tissue transfer toachieve functional wound coverage. We believe evaluationof patients with extensive heel ulceration requiring oper-ative debridement should include complete assessment ofthe regional pedal perfusion, regardless of pulse status, todetermine if extremity revascularization will be required forwound healing. Large tissue defects may benefit from freetissue transfer as prompt wound coverage may provideearlier ambulation and functional limb salvage in thesehigh-risk patients.

CONCLUSIONPatients with heel ulceration most often suffer from

chronic debilitating diseases, such as ESRD, and face highrates of major amputation of both the affected side and ofthe contralateral side in the near future. This subgroup ofpatients requires careful evaluation of their functional sta-tus before complex reconstructive surgery with revascular-ization and/or free tissue transfer for wound coverage isundertaken. In a select group of patients with heel ulcer-ation, we have demonstrated regional pedal perfusion dif-ferences based on high-quality contrast arteriography. Thisinitial report categorizing pedal perfusion in patients withheel ulcers demonstrates that vascularized tissue transfermay be a necessary adjunct to revascularization to accom-plish functional limb salvage in patients with regional heelischemia. The heel ischemia score should be applied pro-spectively to confirm its utility in directing the decisionanalysis in treating patients with heel ulceration.

REFERENCES1. Andros G, Harris RW, Dulawa LB, et al. The need for arteriog-raphy in diabetic patients with gangrene and palpable foot pulses.Arch Surg. 1984;119:1260–1263.2. Rosenblum BI, Pomposelli FB Jr., Biurine JM, et al. Maximizingfoot salvage by a combined approach to foot ischemia and neuro-pathic ulceration in patients with diabetes. A five-year experience.Diabet Care. 1994;17:983–987.

3. Murray HJ, Boulton AJ. The pathophysiology of diabetic footulceration. Clin Pod Med Surg. 1995;12:1–17.4. Kinoshita H, Francis PR, Murase T, et al. The mechanicalproperties of the heel pad in the elderly. Eur J Appl Phys Occ Phys.1996;73:404–409.5. Gooden MA, Gentile AT, Mills JL, et al. Free tissue transfer toextend the limits of limb salvage for extremity tissue loss. Am JSurg. 1997;174:644–648.6. Cronenwett JL, McDaniel MD, Zwolak RM, et al. Limb salvagedespite extensive tissue loss. Arch Surg. 1989;124:609–615.7. Oishi SN, Leven LS, Pederson WC. Microsurgical managementof extremity wounds in diabetics with peripheral vascular disease.Plast Recon Surg. 1993;92:485–492.8. Karp NS, Kasabian AK, Siebert JW, et al. Microsurgical free flapsalvage of the diabetic foot. A 5-year experience. Plast Recon Surg.1994;94:834–840.9. Rutherford RB, Baker JD, Ernst C, et al. Recommended stan-dards for reports dealing with lower extremity ischemia: revisedversion. J Vasc Surg. 1997;26:517–538.10. Klamer TW, Towne JB, Bandyk DF, Bonner M. The influenceof sepsis and ischemia on the natural history of the diabetic foot.Ann Surg. 1987;53:490–496.11. Apelqvist J, Larsson J, Agardh CD. Medical risk factors indiabetic patients with foot ulcers and severe peripheral vasculardisease and their influence on outcome. J Diabet Complicat. 1992;6:167–174.12. Frykberg RG, Piaggesi A, Donaghue VM, et al. Difference intreatment of foot ulcerations in Boston, USA and Pisa, Italy. DiabetRes Clin Pract. 1997;35:21–26.13. Criado E, De Stefan AA, Keagy BA, et al. The course of severefoot infection in patients with diabetes. Surg Gynecol Obstet. 1992;175:135–140.14. Graziano TA, Baratta JB, Menditto L, Ramirez L. A surgicalalternative in the management of chronic neurotrophic ulcerationsof the foot. J Foot Ankle Surg. 1993;32:295–298.15. Isenberg JS, Costigan WM, Thordarson DB. Subtotal calcane-ctomy for osteomyelitis of the os calcis: a reasonable alternative tofree tissue transfer. Ann Plastic Surg. 1995;35:660–663.16. Ferreira MC, Besteiro JM, Monteiro AA Jr., Zumiotti A. Re-construction of the foot with microvascular free flaps. Microsurgery.1994;15:33–34.

PEDAL ISCHEMIA SCORING SYSTEM/GENTILE ET AL

114 THE AMERICAN JOURNAL OF SURGERY® VOLUME 176 AUGUST 1998