Limb Salvage and Reconstruction

Scoring Systems

In 1987, Dr. Sigvard T. Hansen challenged the orthopaedic community “to define clear, concise, acceptable guidelines to help decide which severely damaged extremities are handled by immediate amputation.” In attempts to make the decision easier and more calculated, several scoring systems have been developed during the years. Unfortunately, all of these scoring systems were developed and applied retrospectively in a small number of patients. Further scrutiny in a large prospective study demonstrated a lack of utility in published scoring systems.

In 1987, Howe and colleagues published the Predicted Salvage Index (PSI) after combined orthopaedic and vascular injury. They found that limb survival was related to the interval from injury to arrival in the operating room (OR); the level of arterial injury; and the quantitative degree of muscle, bone, and skin injury. They combined these variables into the PSI and showed 78% sensitivity and 100% specificity. Johansen and colleagues published the Mangled Extremity Severity Score (MESS) in 1990. The MESS is a simple scoring system that stratifies four variables: skeletal or soft tissue damage, limb ischemia, shock, and patient age. They retrospectively reviewed 25 cases and showed that a MESS score of 7 or above predicted amputation with 100% accuracy. However, a recent prospective study involving 230 patients with lower extremity arterial injuries found that a MESS of 8 predicted in-hospital amputation in only 43.2% of patients. In 1991, Russell and colleagues published a seven-part predictive index, the Limb Salvage Index (LSI), to facilitate the decision of whether to amputate or to salvage a mangled extremity. They found that limb salvage was related to warm ischemia time and the quantitative degree of arterial, nerve, bone, muscle, skin, and venous injury. Limb salvage was not related to the presence or absence of shock and order of repair (orthopaedic or vascular). Reviewing 70 limbs in 67 patients, they showed that all of the patients with LSI scores of less than 6 had successful limb salvage, and all of the patients with LSI scores of 6 or more had amputations. McNamara and colleagues reviewed their results of treatment of open tibial shaft fracture classified as Gustilo type 3B or 3C using the MESS and their modified scoring system. They found that their modification of the MESS score, the Nerve Injury, Ischemia, Soft-Tissue Injury, Skeletal Injury, Shock, and Age of Patient Score (NISSSA), was found to be more sensitive (81.8% vs. 63.6%) and more specific (92.3 vs. 69.2%) than the MESS, although both scores were highly accurate in predicting amputations. Tscherne and Oestern published the Hanover Fracture Scale (HFS) in 1982 as the first score for severely injured limbs. It has undergone multiple modifications, the most recent of which was the HFS-98, which incorporates concerns for soft tissue damage and bone loss. Krettek and colleagues evaluated the modified score and compared it retrospectively with the MESS and NISSSA scores, finding it more specific and sensitive for predicting the need for amputation.

The presence of many scoring systems proves that no single scoring system is efficient enough to use as a good predictor of the results of either amputation or salvage of the mangled extremity. In a retrospective study in 1993, Bonanni and colleagues demonstrated that available scoring systems (MESS, PSI, LSI, and Mangled Extremity Syndrome Index) are not predictive of successful limb salvage. Despite this, these scorings systems continued to be used or in some cases modified in an attempt to improve their usefulness. The best available data regarding the usefulness, or lack thereof, of scoring systems were published in 2001 in a prospective evaluation of the utility of lower extremity injury severity scores. Bosse and colleagues studied 556 “high-energy” lower extremity injuries and found that the use of injury severity scores, including the PSI, the NISSSA, Hanover Fracture Scale-97, and the MESS, were not predictive of limb salvage potential. They concluded that the scoring systems have limited usefulness and should not be the sole criterion on which amputation decisions are based. However, the authors did not conclude that very low (good) scores were associated with limb retention. The same conclusion has also been reached on a review of the application of scoring systems to predict the need for amputation in combat-related injuries.

Mangled upper extremity injuries are another major issue. All of the proposed scoring systems are based on injury to the lower limb and, in most cases, the tibial shaft. The assessment of the mangled upper extremity can be very different and is often not frequently discussed in the literature. In fact, there is no separate classification or scoring system that has been developed to specifically evaluate the mangled upper extremity. In 1994, Slauterbeck and colleagues reviewed their results of managing mangled upper extremity injuries and concluded that the MESS can be applied to the mangled upper extremity and is a good early predictor of the need for amputation. Other studies have failed to corroborate these results. Evaluating 52 patients with upper extremity vascular injuries, Prichayudh and colleagues found that a MESS score of 7 or greater did not correlate with amputation. However, a MESS score of 7 and lower was found to be a better predictor of patients who do not need an amputation. Additional authors have concluded that use of the MESS is not appropriate for the mangled upper extremity.

The pediatric population is an underestimated yet important patient subgroup that is usually not addressed differently than adults when concerning mangled extremity injuries. That is despite their superior physical healing capabilities. Venkatadass and his colleagues evaluated the results of 52 pediatric patients with Gustilo type IIIb tibial fractures evaluated by the Ganga Hospital Open Injury Severity Scoring (GHOISS). They assessed the value of this score in the decision making of whether to salvage or to amputate a mangled lower limb. Their retrospective study compared the evaluation with the MESS score and the GHOISS. In total, 48 limbs were salvaged, and 4 were primary amputated. They found that the GHOISS was more valuable in determining amputation, with a specificity of 75% and a sensitivity of 93.75%. They concluded that the GHOISS is a reliable predictor of injury severity in type IIIB open fractures in children and can be used as a guide for decision making. The use of the MESS score in children has a lower predictive value compared with the GHOISS in deciding on amputation versus salvage.

Recently, Schirò et al. reviewed the published literature in pursuing an answer as to whether the current scoring systems can predict the results of primary amputation or limb reconstruction in a mangled extremity. They concluded that the current scoring systems may help the surgeons in making their decisions; however, they cannot be used as the sole criteria by which amputation decisions are made, and in cases of successful limb salvage, they are not predictive of the functional recovery.

Nerve Injury and Plantar Sensation

The absence of plantar sensation has been historically used as a predictor of the need for an amputation. In a survey of surgeons of the most important factor typically considered in the decision to amputate or reconstruct a limb published in 2002, absent plantar sensation was the number one reason for orthopaedic surgeons and the number three reason for general surgeons, behind a high Injury Severity Scale (ISS) score and limb ischemia. Recent data, however, suggest that it should not be an indication for amputation. The Lower Extremity Assessment Project (LEAP) study group identified 55 patients who presented with an insensate extremity and divided them into two groups: insensate and had an amputation ( n = 26) and insensate and had salvage ( n = 29). These patients were then compared with a sensate matched control group ( n = 29). There was no difference between the sensate control group and insensate salvage group at 2 years in terms of outcome or the need for late amputation. In fact, foot sensation was normal in more than half of the insensate control group (55.6%) at 2 years, which was similar to the sensate control group (55%). Only one patient in the insensate salvage group had absent plantar sensation at 2 years.

Momoh and his colleagues from the Ganga hospital in India published their experience of eight cases of lower-limb-threatening injuries in patients who suffered posterior tibial nerve injury with damaged plantar sensation but were treated with salvage procedures. Their patients suffered mild pain, decreased ankle range of motion, and decreased strength, but their scores were not different from those in the literature review they made for comparison. They concluded that there is a value for limb salvage even with posterior tibial nerve injuries ( Fig. 71.4 ).