Abstract: The objective of this study was to determine if plantar forefoot ulcers would heal and not recur if treated with tendon lengthenings. Patients with neuropathy (usually from diabetes mellitus) and plantar forefoot ulcers were treated with tendon lengthening and followed. Of 34 forefoot ulcers treated with tendon lengthenings, 1 did not heal, and 4 recurred at the same location by 36 months average follow-up. Three patients developed transfer ulcers at other locations on the foot. Two patients subsequently required leg amputation for gangrene (1 transfemoral and 1 transtibial). No amputations were completed for progressive infection from ulcers. Tendon lengthening appears to be an effective treatment for neuropathic plantar forefoot ulcers.

A

pproximately 3% of the United States population have diagnosed diabetes mellitus.¹ Diabetic foot problems, however, are the leading cause of amputation.^2,3 The risk of amputation is 15 times greater in patients with diabetes than other people.² Up to 15% of patients with diabetes will require amputation.^1,3 Over 50,000 amputations in patients with diabetes occur annually in the US.⁴ In 1 study of patients with diabetes mellitus, 84% of lower-extremity amputations (67/80) were preceded by foot ulcers.²
       Among people with diabetes, about 15% will eventually have foot ulcers.^1,5 In the absence of large-vessel disease, diabetic forefoot ulcers result from the combination of neuropathy and abnormal mechanical stress.^1,6,7 Deformity that increases stress on a portion of the foot can instigate ulceration in a patient with diabetic neuropathy.^1,6,7 Peripheral neuropathy results in the loss of protective sensation and a lack of recognition of repetitive mechanical stress, which often cause forefoot ulcers in patients with diabetes.^1,7
       Diabetic motor neuropathy^7,8 and glycosylation of collagen⁷ may contribute to calf tightness and decreased dorsiflexion of the ankle, which increase pressure on the forefoot. The high forefoot pressure is consistent with the most common locations of foot ulcers, which are the plantar surface of the metatarsal heads and the hallux.^7,8
       Foot ulcers can cause deep spreading infection, which can result in leg amputation.^5,9 Prevention and cure of foot ulcers should prevent most leg amputations in patients with diabetes.⁸
       Patients who have experienced previous foot ulceration have abnormally high pressures at healed ulcer sites.¹⁰ Plantar foot ulcers occur at sites of high pressure.^7,11 The combination of neuropathy and decreased ankle dorsiflexion has been implicated in the cause of forefoot ulcers.^7,12,13 Reducing the risk of neuropathic ulceration of the forefoot should be accomplished by decreasing pressure on the forefoot.¹⁰ Pressure relief has been accomplished by shoe modification, total contact casts, and Achilles tendon lengthening (TL).^9,14
       Total contact casting (TCC) is an effective technique for healing diabetic neuropathic foot ulcerations.^5,15,16 However, 20% (20/102) to 81% (21/26) of the ulcers healed by TCC have recurred in 2 years.^15,16
       Achilles TL promotes healing of chronic foot ulcers in patients with neuropathy (Figures 1 and 2).^12,17,18 Healing occurred in patients who did not heal by TCC and with fewer ulcer recurrences than ulcers healed by TCC.¹⁷ Dorsiflexion metatarsal osteotomy also is effective in healing chronic neuropathic forefoot ulcers but has a much higher complication rate.¹⁹

Figure 2

Figure 1

       Healing of forefoot ulcers and prevention of ulcer recurrence and amputation caused by ulcers are important objectives for those who manage patients with neuropathic forefoot ulcers. This report describes the results of a group of patients who had forefoot ulcers treated with TLs.

Materials and Methods

       Between May of 1995 and October of 2002, all patients seen with plantar forefoot ulcers without active infection or gangrene were offered TL. Twenty-four patients with 1 or more plantar forefoot ulcers agreed to surgical TL.
       Ulcers of metatarsal heads were treated with lengthening of the gastrocnemius-soleus mechanism. The Vulpius technique²⁰ was used, transecting the proximal tendon of the gastrocnemius muscle and underlying aponeurosis of the soleus muscle in the mid-calf. Varus-valgus alignment was improved by peroneus longus (z-type) lengthening for first metatarsal ulcers and posterior tibial (intramuscular) lengthening for fifth metatarsal and cuboid ulcers. The patient’s ankle was then placed in neutral in a short leg walking cast for 6 weeks bearing full weight. The Vulpius procedure lengthens the gastrocnemius tendon and soleus aponeurosis proximal to the actual Achilles tendon. Both the Vulpius procedure and Achilles TL increase ankle dorsiflexion in a similar manner.
       Toe ulcers were treated with toe flexor tendon tenotomy percutaneously at the proximal phalanx. Toe tenotomy was usually performed in the office, whereas calf surgery was usually completed on an outpatient basis. Due to a higher recurrence rate of first toe ulcers, both calf and toe procedures were recommended to patients with first toe ulcers. Patients were allowed to bear full weight on the operated side and were released the day of surgery, unless they were in the hospital for another reason.
       The amount of active ankle dorsiflexion with the knee in full extension was measured pre-operatively with a goniometer. Pulses were measured by palpation of the dorsalis pedis and posterior tibial arteries. All patients without pulses were referred for evaluation and treatment by a vascular surgeon, but none were thought to be candidates for vascular surgery.
       Ulcers were graded using Wagner’s²¹ classification: Grade 1 is superficial; Grade 2 is deep, extending to ligament, tendon, joint capsule, fascia, or bone; Grade 3 is Grade 2 with infection (abscess, osteitis, or osteomyelitis); Grade 4 is gangrene of the toe or forefoot; and Grade 5 is gangrene of the entire foot. Grades 3 to 5 were not included unless Grade 3 could be transformed to Grade 1 or 2 by antibiotics. Some patients with prior ulcers and infections had been treated in the past by other physicians with amputation of toes and/or metatarsal heads but were not excluded from this study.
       Durations of the ulcers were obtained from the patients. If patients were unable to feel the 5.07 nylon monofilament in multiple areas without callus or ulcer on the plantar foot, they were considered neuropathic.²² The Touro Infirmary Investigational Review Board approved this study.

Results

       Each patient was assigned a number in the order in which he or she had the first TL surgery (Tables 1 and 2). All 24 patients had neuropathy by the monofilament test. Twenty had diabetes mellitus, 2 had lumbar radiculopathy, 1 had hemiplegia, and 1 had alcoholism. Thirty-four ulcers were treated. There were 17 metatarsal ulcers (14 metatarsal head ulcers), 11 first toe ulcers, 5 lesser toe ulcers, and 1 cuboid ulcer.

Table 1 (continued)

Table 1

       Ages of the 24 patients ranged from 33 to 81 at the time of surgery, with an average age of 60. There were 11 males and 13 females. All patients with ulcers volar to metatarsals had Vulpius procedures. Toe ulcers were treated with percutaneous toe flexor tenotomies. All patients had calf tightness with inability to dorsiflex the ankle beyond 10^o with the knee extended, and the average active dorsiflexion was 6^o less than neutral (-6^o) for the 34 ulcers.
       Twenty ulcers were Grade 1, and 14 were Grade 2. Thirteen ulcers had associated cellulitis treated with antibiotics prior to surgery. Once infection appeared to be under control, TLs were performed. Five patients had had 1 or more toes previously removed, and 3 had had 1 or more metatarsal heads removed.
       Ulcer duration prior to surgery ranged from 1 month to 7 years (average 10 months). Most patients had had unsuccessful attempts to heal the ulcers with decreased weight bearing, debridement, and shoe modification.

Table 2 (continued)

Table 2

       Of ulcers that healed, all healed in less than 2 months except the ulcer of Patient 24 (cuboid ulcer), which healed in 5 months. Pulses were not palpable in 7 patients. Patient 11 had no pulse and was on dialysis, and the ulcer did not heal. He later developed gangrene of the foot and required transfemoral amputation.
       Of 34 ulcers, 33 eventually healed (97%). All incisions healed primarily without any infections. Of the 33 ulcers that healed, 4 recurred (12%). One patient with a recurrent toe ulcer (Patient 22) agreed to repeat toe TL, and the ulcer healed. There were no transfer metatarsal head ulcers. The transfer ulcer on the heel (Patient 13) occurred when the patient got the cast wet and broke the heel of the cast but failed to return to the office immediately. This ulcer also healed. There were 2 transfer ulcers in toes (Patient 1, first toe; Patient 16, third toe); both later healed after toe tenotomy. No new deformities developed, and no patient had development or progression of Charcot arthritis after the TLs. Last follow-up for 24 patients was in person for 12 patients and by telephone for 12. The follow-up ranged from 12 months to 81 months; the average follow-up was 36 months for the 34 ulcers.
       There were 3 complications in addition to those previously described. Patient 1 had a pulmonary embolus more than 3 months after the first surgery but recovered completely. Patient 7 developed gangrene without a foot ulcer 12 months after surgery and had transtibial amputation. Patient 15 had a traumatic open dislocation of her fifth toe treated with amputation.
       Results of smaller, more homogeneous groups are presented as follows (Figure 3). There were 11 ulcers plantar to metatarsal heads in 9 patients with diabetes. All had gastrocnemius-soleus lengthening. Ten ulcers healed after TL, and 1 of 10 recurred after an average follow-up of 41 months. There was 1 transfer ulcer to the heel and 1 to a toe, which later healed.
       There were 14 ulcers plantar to the tips of the toes in 10 patients with diabetes. All healed after TL. All had toe flexor tenotomy. Two of 7 first toe ulcers recurred, but none of the 7 second to fifth toe ulcers recurred. Of 6 patients without pulses with healed ulcers, 1 recurred.

Figure 3

Discussion

       The strength of this study is that all foot ulcer patients who agreed to treatment with TL during the given time period are included. This results in more ulcers and/or longer follow-up than previously published studies. The weakness of this type of study is the heterogeneous location of ulcers and resulting heterogeneous type and number of TLs. Results of smaller more homogeneous groups were also presented (Figure 3). Recurrence rates of ulcers for all groups in this study were much lower than with no surgery²³ and TCCs with and without Achilles lengthening.¹⁶
       No patients were lost to follow-up. The length of follow-up was shortest for the 2 patients who had leg amputations for gangrene. The minimum follow-up for the other patients was 18 months.
       The association of gastrocnemius-soleus contracture, neuropathy, and chronic ulceration of the forefoot in this group of patients was previously reported by Yosipovitch and Sheskin¹² and by Lin et al.¹⁷ The high rate of successful healing of forefoot ulcers after Achilles lengthening in these studies,12,17 7 of 8 (88%) and 14 of 15 (93%), was similar to this report—33 of 34 (97%).
       Of ulcers that recurred in this study, 3 were on the first toes. The first toe ulcers may have recurred because the short flexors are not cut in a percutaneous tenotomy at the proximal phalanx in the first toe, whereas both short and long toe flexors are cut in the other toes. Both short and long first toe flexors are now cut percutaneously with no recurrences thus far. The recurrence of Patient 23 at the third metatarsal head may have been partially related to the prior first ray amputation.
       The reported recurrence rate of foot ulcers in patients with diabetes treated without TL after 3 years was 61% (280/458).²³ After Achilles TL, Yosipovitch and Sheskin¹² had 3 ulcer recurrences in 8 patients (38%) in 4- to 5-year follow-up, and Mueller et al.¹⁶ had recurrence in 38% (10/26) at 2-year follow-up. The low rate of 12% (4/33) for recurrent ulcers in this series compares favorably with treatment without TL,23 with TL,12,16 and with 20% (20/102)15 to 81% (21/26)16 recurrence after healing in TCCs at 2-year average follow-up. Lin et al.¹⁷ had no recurrences but had shorter follow-up (17 months vs. 36 months) and fewer ulcers (15 vs. 34).
       The controlled, randomized study of Mueller et al.¹⁶ showed a high rate of ulcer healing: 89% (29/33) with TCC and 100% (30/30) when TCC was combined with Achilles TL. Their results are similar to this study in which 97% (33/34) healed with TL alone; this study indicates that TCC is not necessary for forefoot ulcer healing. The ulcer recurrence rate with TL in their study of 38% (10/26) and this study of 12% (4/33) of healed ulcers or 10% (1/10) of healed metatarsal head ulcers in patients with diabetes is much less than with TCC alone of 81% (21/26). All their patients had diabetic forefoot ulcers and pulses. This study also shows TL can work on toe ulcers (15 of 33 healed ulcers), in patients without pulses (7 of 24 patients), and with nondiabetic neuropathy (4 of 24 patients).
       Dorsiflexion metatarsal osteotomy also had a high rate of successful healing of neuropathic forefoot ulcers of 95% (21/22).¹⁹ There was, however, a 68% complication rate with 7 patients developing acute Charcot disease, 3 developing midfoot ulcers, 3 deep wound infections, 2 transfer ulcers under adjacent metatarsal heads, and 1 below-knee amputation. This study revealed no new or worsening Charcot disease, new mid-foot ulcers, transfer metatarsal ulcers, or wound infections. There were 3 transfer lesions: a first toe ulcer in Patient 1, third toe ulcer in Patient 16, and a heel ulcer in Patient 12, all of which later healed. Two patients subsequently required amputation for gangrene (Patients 7 and 11) but none for recurrent ulcer or progressive infection.
       The value of noninvasive laboratory testing for peripheral vascular disease is controversial.¹ Vascular surgery evaluation was obtained in patients without pulses, but no patients in this study had vascular surgery either recently before or during follow-up after TL. Gangrene was thought to be a contraindication to TL. Patients without pulses, however, were considered potentially salvageable, since there were no incision problems, and ulcers healed in all but 1 of these patients.
       In a study of diabetic amputations, 84% (67/80) were attributed to ulcers.²³ Two of 24 patients (8%, Patients 7 and 11) in this study required amputation of the leg for gangrene, but none required amputation for progressive infection from an ulcer at average follow-up of 36 months. This is less than the reported total amputation rate of 22% (101/458) at 3 years of observation of patients with diabetes with prior foot ulcers.²³ Their rate of amputation for complications of recurrent ulcers, 21% (98/458), was much higher than this study, 0% (0/24). The number of patients in this study is small by comparison, and the decrease in amputation rate by TL needs to be confirmed by a study of more patients.
       Lin et al. and Mueller et al. lengthened the Achilles tendon by Hoke’s method²⁴ of hemisection at 3 levels of the tendon. Yospovitch and Sheskin used the subcutaneous tenotomy method of Strohmeyer.²⁵ Nishimoto et al.²⁶ preferred gastrocnemius recession for diabetic forefoot ulceration because of the lower risk of over lengthening, calcaneal gait, and plantar heel ulceration with a 16% recurrence rate. The author chose the Vulpius technique²⁰ because of prior favorable experience using this technique in children with cerebral palsy, which allows immediate full weightbearing in a cast with less over-correction and incision problems. The Vulpius procedure was used successfully by Takahashi and Shrestha²⁷ to correct equinus deformity of the ankle in 230 adults after cerebrovascular accident. The average age was 68, and 98 patients had diabetes. They had no wound or tendon problems and allowed standing in a brace (ankle foot orthosis) the next day. The technique of TL for forefoot ulcers should probably be left to the discretion of the surgeon. The addition of peroneus longus or posterior tibial and/or toe flexor lengthening may have been the reason for the lower ulcer recurrence rate in this study, 12% (4/34), compared to Mueller et al., 38% (10/26).
       Most of the factors recorded (age, sex, cause of neuropathy, ankle dorsiflexion, ulcer grade, ulcer size, ulcer location, ulcer duration, prior toe or metatarsal head resection) did not seem to obviously affect results, such as ulcer healing, recurrence, or other complications. Larger studies with longer follow-up could change this finding. The 2 patients who had amputation for gangrene, however, had no pedal pulses.
       The purpose of TL is to decrease stress on the area of ulceration. Tenotomy of the toe flexor tendons is done to decrease stress on the plantar surface of the toe. Peroneus longus lengthening should decrease pressure on the first metatarsal head, as the posterior tibial lengthening should decrease pressure on the fifth metatarsal. Lengthening the gastrocnemius-soleus mechanism should decrease stress on the entire plantar forefoot. Armstrong et al.¹⁴ confirmed that Achilles lengthening does in fact decrease pressure on the forefoot, recommending the procedure as an adjunctive therapeutic and prophylactic measure to reduce the risk of neuropathic ulceration. Tendon lengthening has also been performed prophylactically on these and other patients with prior ulcers, impending ulcers, or progressive callus with no recurrences thus far.
       Another study should be performed to determine if daily calf stretching can prevent calf tightness and progression to forefoot callus, forefoot ulceration, and Charcot arthritis in patients with diabetes. Since calf stretching might help and probably would not harm patients with diabetes, it now seems reasonable to recommend prophylactic calf stretching to these patients.

Conclusion

       The results of this series of patients suggest that lengthening of the tendon-muscle units is effective treatment for neuropathic forefoot ulcerations. Tendon lengthenings would not be expected to prevent amputation in patients with severe peripheral vascular disease and gangrene. By healing most forefoot ulcerations and lowering their recurrence rate, this procedure appears to lower the incidence of progression of forefoot ulceration to infection and subsequent amputation. A follow-up study to better document these findings with more patients and larger groups of different ulcer locations is under way.