Gritti-stokes (through-knee) amputation: should it be reintroduced?

ABSTRACT

Background. When there is doubt about perfusion in the distal part of the leg and around the ankle, an above-knee amputation (AKA) site has been the most commonly selected to assure primary wound healing.

Methods. The Gritti-Stokes amputation (GSA), a modified through-knee amputation, seems to have value for selected patients, according to a literature review and our success in a small group of patients. We have investigated the proportion of patients who have had AKA and who might have been suitable candidates for GSA.

Results. In a retrospective study of 66 patients who had had AKA, at least one third could have been considered for a GSA and might have benefitted from this procedure. This suggests that some consideration should be given to the reintroduction of the Gritti-Stokes operation.

Conclusion. Our expectation is that patients so treated could achieve a high proportion of primary wound healing, an end-bearing stump, and more rapid rehabilitation than found in patients undergoing the classic AKA.

DESPITE ADVANCES in peripheral vascular surgery and limb salvage procedures, surgical amputation of parts of the lower limb remain a common operative procedure in the United States. It has been estimated that more than 50,000 lower limb amputations are done each year in the United States. (1) The functional results depend on maximizing the limb length, which can be preserved with sound wound healing. Thus, the challenge for the surgeon is to select the lowest possible level for an amputation that offers the greatest likelihood of primary wound healing. The multiple objective methods advocated to determine the appropriate level for leg amputation include bleeding from skin flap margins, segmental Doppler blood pressure measurements, photoplethysmography, skin perfusion pressure, skin temperature demarkation, fluorescein delivery, xenon clearance, transcutaneous oximetry, and laser Doppler velocimetry, (2,3) but none of these measures reliably predicts wound healing.

The preservation of the knee joint for successful patient rehabilitation cannot be overemphasized. Thus, the below-knee amputation (BKA) with long posterior flap, as popularized by Burgess et al (4) in the early 1970s has been shown to heal successfully in more than 85% of appropriately selected patients. However, for peripheral vascular disease (PVD), the level of vascular insufficiency must be distal because much of the posterior compartment of the lower leg is used in the formation of the stump.

When there is doubt concerning the viability of the long posterior flap BKA, most surgeons will elect to carry out a classic AKA, though alternative procedures are currently available. One alternative is a through-knee amputation, (5) with the patellar ligament sutured to the posterior capsule of the knee joint. (6) However, the bulbous ends of the femoral condyles rub against the prosthesis, frequently causing skin breakdown. In addition, the patella is unstable when sutured to the posterior knee joint capsule, and this configuration may cause pain and be associated with patellar dislocation even for those patients who are nonambulatory. For these two reasons, the reamputation rate after a through-knee disarticulation is high; a revision rate of nearly 10% (7 of 72 patients) was reported by Baumgartner (7) and a rate of 13% (6 of 46 patients) by Pinzur et al. (8)

The AKA (9) is the most widely used procedure when a BKA is thought, on clinical grounds, to be contraindicated. The AKA procedure requires much muscle transection and careful fascial plane reconstruction to obtain a good stump and wound drainage.

Despite attention to these details, stump nonhealing of rates of 2% to 10% have been reported. (3) In addition, rehabilitation is usually slow and requires upper body strength, a good sense of balance, and a complex and heavy pelvis-based prosthesis. Consequently, only 40% to 50% of all patients having an AKA are able to walk. (3) For bedridden patients, the short stump is of little benefit for moving around the bed or for transfer from bed to wheelchair, and a fixed flexion deformity becomes common.

In our experience, the Gritti-Stokes form of through-knee amputation has been a useful option for three patients who would otherwise have undergone AKA.

Operative Technique

A U-shaped incision is made at the adductor tubercle, across the tibial tuberosity to the lateral aspect of the lower end of the femur. This is connected posteriorly by a convex downward incision. The knee is then flexed and the patellar ligament released from the tibial tuberosity as near the bone as possible. The collateral and cruciate ligaments are divided. The hamstring attachments to the medial aspects of the tibia and laterally to the head of the fibula are divided, thus completing the disarticulation. With the tibia dislocated anteriorly, the posterior capsule of the knee joint is divided from the front. The vessels and nerves in the popliteal fossa are divided using standard techniques.

The crucial part of the operation follows. The periosteum covering the distal aspect of the femur is elevated, and the femur is cut at the level of the adductor tubercle to create an anterior to posterior angle of 10[degrees] (Fig 1, insert). This angulation allows the patella, after its articular surface has been resected, to fit snugly onto the end of the femur. In addition, this angulation reduces the possibility of anterior dislocation of the patella. The power of the quadriceps is greater than that of the hamstrings, and the 10[degrees] angulation helps to "lock" the patella in place.

Small holes are drilled in the edge of the patella and the femur to connect these two bony structures with wire sutures, which then fuse with a synchondrosis rather than bony fusion. The hamstring tendons are then sutured with 0 Vicryl (polyglactin 910) to the patellar ligament, and any additional soft tissues can then be brought around in front of this anastomosis to further secure the patella and provide additional stability. Because of the minimal division of muscle, it is usually possible to obtain sufficient hemostasis so that a drain is not necessary. Once the deep fascia has been approximated with carefully placed interrupted sutures, staples are applied to the skin, followed by a gentle pressure dressing to the stump.

MATERIALS AND METHODS

We retrospectively reviewed the charts of all patients who had had AKA over an 18-month period. Of the 66 patients identified, 16 had had complex multiple procedures and were excluded, leaving 50 consecutive patients who had AKA for PVD and 6 who were operated on after trauma. These 56 formed the study group.

The data points collected were patient's age, ambulatory status, physical examination, and perioperative studies. We found that in many patients the decision for amputation was made on clinical grounds. Thus, few had had arteriography, and the details of the physical findings were frequently brief.

RESULTS

The average age of the patients with PVD was 73 years (range, 42 to 99 years); 22 were men and 28 were women. The average age of the 6 patients (3 men and 3 women) having surgery for trauma was 51 years (range, 18 to 101 years).

On the basis of the clinical record, the patients were classified as follows: (1) GSA positive (when we were confident that a Gritti-Stokes procedure could have been used); (2) GSA negative (when we were confident that a Gritti-Stokes could not have been used); and (3) GSA possible (when we were uncertain whether the patient might have been a candidate for the Gritti-Stokes procedure).

In the cases of PVD, one third of patients fit each of the three categories (Table 1). In the trauma group, 2 of the 6 patients were considered to have been suitable for GSA, while the remaining 4 patients were considered unsuitable for GSA (Table 1).

The features by which we judged that GSA might have been possible are summarized in Table 2. In the GSA positive category, the majority (13/16, 81%) had lesions that were confined to the foot and ankle only. We noted that of these 16 patients, 10 (63%) were nonambulatory; only 2 (12.5%) had had a previous BKA, and none had had ipsilateral vascular bypass surgery. Thus, our approach to placing a patient in the GSA positive category was conservative.

In the GSA possible group, 5 of the 17 patients (29%) had lesions confined to the foot and ankle, 8 (47%) were nonambulatory, 5 (29%) had a nonhealing BKA stump, and 4 (24%) had had previous ipsilateral arterial bypass surgery.

In the 17 patients in whom GSA was deemed not suitable, the AKA appeared to be the best procedure because of more extensive disease. In this group, 7 patients were ambulatory pre-operatively, and the ambulatory status of the remaining 10 patients was not documented.

DISCUSSION

Rocco Gritti, (10) a surgeon at the Ospedale Maggiore in Milan, Italy, described a through-knee procedure in 1857 in which the patella and surrounding tissues were used as an osteo-plastic flap to provide an end-bearing stump. Thirteen years later, the procedure was modified by Sir William Stokes, (11) and the operation is now known as the Gritti-Stokes amputation (Fig 1.) The objectives of this operation are to preserve as much of the lower limb function and length of femur as possible, to achieve a stable end-bearing stump, and to achieve a high percentage of primary wound healing. The reported postoperative complications are non-union of the patella to the femoral shaft, resulting in patellar pain (12); some long-term stump problems (13); and some difficulty in fitting a cosmetically attractive prosthesis. (14-16) These difficulties have been overcome by angulation of the femoral condyle transection and the great advances in modern prosthetics.

However, the eventual ambulatory rate after GSA has ranged from 52% in one reported series (14) to 60% in another, (13) compared with a substantially lower rate found after AKA (40% to 50%). (3) This greater proportion of patients achieving ambulatory rehabilitation after GSA may be as a consequence of the mechanically simpler prosthesis required for GSA as compared with that for AKA. The GSA provides the patient with an end-bearing conical stump that fits firmly and snugly into the lace-up prosthesis and requires only a simple waist strap to secure its position. Thus, the proprioceptive components surrounding the hip joint are more readily maintained by this end- and side-bearing stump configuration. By contrast, for AKA much of the patient's weight is borne through the ischiotuberosity into the prosthesis, though some weight-bearing occurs laterally through the upper thigh into the prosthesis. Consequently, a patient's sense of balance and the proprioceptive issues of coordinated movements around the hip ca n be impaired.

Despite the unequal knee centers associated with a GSA, the development of lightweight polycentric hydraulic knee joints and endoskeletal systems has alleviated many of these problems. Indeed, all prosthetic devices in recent years have benefited by weight reduction, which has helped to diminish energy expenditure for walking and balance.

There is little mention of the GSA in North America, though it has been discussed in a small number of papers in the English-language literature over the past 40 years. (12-21) In these reports, GSA has been used most frequently in patients with PVD. Reported advantages include (1) better primary and overall healing rates, (12,18) (2) low mortality in the immediate postoperative period, (12) (3) low reamputation rate, (18) (4) improved stump length in which the stump provides better leverage for both bedridden and mobile patients, (18,20) (5) excellent collateral blood supply around the knee, which helps to preserve the skin flaps, (20) (6) minimal muscle tissue division, (20) (7) an end-bearing stump without the risk of retraction, (20) (8) preservation of function of both posterior and anterior muscle groups; (9) avoidance of flexion contracture at the hip seen in many patients with the classic AKA, and (10) rapid postoperative mobility and rehabilitation. (15) Thus, for patients who require a lower limb am putation but who are not suitable for a BKA, we suggest that the GSA be considered and that this "old" operation might be reintroduced to our surgical armamentarium.

Proposals for Prospective Study

An optimal amputation level is important for good clinical results in wound healing and patient rehabilitation, but the determination of that level is controversial. Blood supply and healthy tissue are critical to the success of the procedure, and in patients with PVD, these factors may already be compromised. Our algorithm for the level of leg amputation in end-stage PVD (Fig 2) is based on location of the lesion, presence or absence of pulses, and lowest level of healthy tissue.

* Having established that amputation is warranted, the limb should be examined for location of the lesion.

* If the lesion is confined to the foot or ankle and the patient is ambulatory, a BKA is the optimal treatment. If BKA is thought to have a low chance of success, then consider GSA. If the most distal part of the lower leg is involved, it is less likely that BKA will be suitable and GSA should be considered.

* If the ischemic changes involve the leg distal to a handbreadth below the tibial tuberosity, it is unlikely that BKA will be suitable. If the popliteal pulse is present, consider GSA; if the popliteal pulse is absent, AKA is probably indicated.

* If the ischemic signs are present at the knee, AKA will be needed.

We have designed a phase 2 clinical trial based on this algorithm to assess the hypothesis that selected patients with PVD can benefit from GSA. The patients to be enrolled in the trial will have end-stage PVD of sufficient severity to require a lower extremity amputation. The data collection will include patient's age, physical examination, preoperative segmental Doppler blood pressure measurements, ambulatory status, discharge status, and rehabilitation results.

CONCLUSION AND SUMMARY

Our data suggest that at least one third of the patients who currently have an AKA, and perhaps as many as two thirds of this group, might benefit from a GSA. In selected cases, the GSA procedure has been reported to have high healing rates; to result in an end-bearing stump; to allow greater maneuverability of patients in bed, in a chair, or in a prosthesis; and to allow a much simpler prosthesis to be constructed. All these features seem to be associated with a higher rate of rehabilitation to independent walking than found in patients having an AKA. Our results indicate that using a posterior 10 [degrees] tilt for the resection of the femur prevents an unstable patella and anterior patellar dislocation.

We propose to follow this retrospective review with a prospective phase 2 trial, in which the clinical algorithm (Fig 2) will be used to identify patients who should be considered for the GSA. In addition, we will record details of physical examination and segmental Doppler blood flow pressure measurements, healing, and reoperative rates, and rehabilitation complications.

We conclude that if this phase 2 study provides a substantial proportion of good clinical results, indicating that a large number of patients are suitable for GSA, a randomized trial would appear unnecessary. By contrast, if the results were more marginal, a randomized trial of Gritti-Stokes through-knee amputation versus AKA might be indicated. We encourage other interested surgeons to participate with us in this prospective investigation.

(1.) Cutson TM, Bongiorni DR: Rehabilitation of the older lower limb amputee: a brief review. J Am Geriatr Soc 1996; 44:1388-1393

References

(2.) Kram FIB, Appel PL, Shoemaker WC: Prediction of below knee amputation wound healing using noninvasive laser Doppler velocimetry. Am J Surg 1989; 158:29-31

(3.) Malone JM: Lower extremity amputation. Vascular Surgery: A Comprehensive Review. Moore WS (ed). Philadelphia, WB Saunders Go, 4th Ed, 1993, pp 809-854

(4.) Burgess EM, Romano RI, Zettl JH, et al: Amputations of the leg for peripheral vascular insufficiency. J Bone Joint Surg 1971; 53:874-890

(5.) Neweombe JF, Marcuson RW: Through-knee amputation. Br J Surg 1972; 59:260-266

(6.) Ayoub MM, Solis MM, Rogers JJ, et al: Thru-knee amputation: the operation of choice for non-ambulatory patients. Am Surg 1993; 59:619-623

(7.) Baumgartner RF: Knee disarticulation versus above-knee amputation. Prosthet Orthol Int 1979; 3:15-19

(8.) Pinzur MS, Smith DC, Daluga DJ, et al: Selection of patients for through-the-knee amputation. J Bone Joint Surg Am 1988; 70:746-750

(9.) Tooms RE: General principles of amputations. Campbell's Operative Orthopaedics. Canale ST (ed). St. Louis, CV Moshy Co, 9th Ed, 1998, pp 521

(10.) Gritti R: Dell' amputazione del femore al terzo inferiore e della disartictilazione del ginochhio. Ann Unive Med Milano 1857; 161:5-32

(11.) Stokes W: On supra-condyloid amputation of the thigh. Med Chir Trans 1870; 53:175-186

(12.) Beacock CJW, Doran J, Hopkinson BR, et al: A modified Gritti-Stokes amputation: its place in the management of peripheral vascular disease. Ann R Coil Surg Engl 1983; 65:90-92

(13.) Doran J, Hopkinson BR, Makin GS: The Gritti-Stokes amputation in ischaemia: a review of 134 cases. Br J Surg 1978; 65:135-137

(14.) Houghton A, Allen A, Luff R, et al: Rehabilitation after lower limb amputation: a comparative study of above-knee and Gritti-Stokes amputations. Br J Surg 1989; 76:622-624

(15.) Middleton MD, Webster CU: Clinical review of the Gritti-Stokes amputation. BMJ 1962; 2:574-576

(16.) Siriwardena GJA, Bertrand PV: Factors influencing rehabilitation of arteriosclerotic lower limb amputees. J Rehabil Res Dev 1991; 28:35-44

(17.) Houghton A, Taylor PR, Thurlow S, et al: Success rates for rehabilitation of vascular amputees: implications for preoperative assessment and amputation level. Br J Surg 1992; 79:753-755

(18.) Martin P, Renwick S, Thomas EM: Gritti-Stokes amputation in atherosclerosis: a review of 237 cases. BMJ 1967; 3:837-838

(19.) Moran BJ, Buttenshaw P, Mulcahy M, et al: Through-knee amputation in high-risk patients with vascular disease: indications, complications and rehabilitation. Br J Surg 1990; 77:1118-1120

(20.) Shackleton ME: The Gritti-Stokes amputation: a reappraisal. N Z Med J 1966; 65:227-229

(21.) Yusuf SW, Baker DM, Wenham PW, et al: Role of Gritti-Stokes amputation in peripheral vascular disease. Ann R Coil Surg Engl 1997; 79:102-104

TABLE 1

Categorization of Patients Having Had Above-Knee Amputation

                    GSA           GSA            GSA
                Positive (*)  Possible (+)  Negative (**)

PVD (n = 50)      16 (32%)      17 (34%)      17 (34%)
Trauma (n = 6)     2 (33%)       0 (0%)        4 (67%)

(*)A Gritti-Stokes amputation (GSA) could have been used.

(+)Uncertain whether patient might have been a candidate for GSA.

(**)A GSA could not have been used.

PVD = Peripheral vascular disease
TABLE 2

Clinical Features in Each GSA Group

                          GSA           GSA         GSA
                       Positive (*)   Possible    Negative

Only ankle/foot       13/16 (81%)    5/17 (29%)  1/17 (5%)
 lesions
Nonambulatory         10/16 (63%)    8/17 (47%)   0/7 (0%) (*)
Nonhealing BKA         2/16 (13%)    5/17 (29%)  1/17 (6%)
 stump
Previous ipsilateral   0/16 (0%)     4/17 (24%)  9/17 (53%)
 bypass surgery

(*)Seven patients were ambulatory; ambulatory status of remaining 10
patients is unknown.

BKA = Below-knee amputation

KEY POINTS

* The Gritti-Stokes (GS) amputation is an underused alternative to some patients currently having an above-knee amputation (AKA).

* The primary healing rate after a GS amputation is greater than after AKA in appropriately selected patients.

* The GS procedure provides a better stump for a simpler prosthesis than that required after an AKA.

* Our data suggest that an excessive number of above-knee amputations are currently being done.