Arthroscopy: The Journal of Arthroscopic and Related Surgery 9(2): 143-148 Published by Raven Press, Ltd. © 1993 Arthroscopy Association of North America

A Comparison of the Doubled Semitendinosus/Gracilis and Central Third of the Patellar Tendon Autografts in

Arthroscopic Anterior Cruciate Ligament Reconstruction

Angelo L. Otero, M.D., and Lonn Hutcheson, M.S.

Summary: The purpose of this study was to compare the postoperative success and stability of two different autografts used to reconstruct the anterior cru- ciate ligament (ACL): doubled semitendinosus/gracilis (DST&G) and bone- patellar tendon-bone (PAT). Ninety-one young (x = 25.4 years), active pa- tients were available for an average follow-up of 36.4 months and included 55 patients in the PAT group and 36 in the DST&G group. No patients had previously undergone ACL reconstruction. Knee stability data were obtained yearly and included scores from the Lysholm questionnaire, Lachman exam, and KT-1000 arthrometer at 30 pounds (KT30) and maximum pull (KTMAX). Both autografts were of comparable size, tension, and isometricity. Two sep- arate factorial multivariate analyses of variance (MANOVA) were used to compare the two series for 3 follow-up years in the categories acute and chronic (A versus C) and meniscectomy and no meniscectomy (M versus NM). Results indicated that in all categories and follow-up years, PAT patients had consistently greater knee stability compared with the DST&G group. Overall MANOVA results showed significantly lower (p < 0.01) Lachman scores in PAT versus DST&G in each of the 3 follow-up years. Significantly lower (p < 0.05) KT30 and KTMAX values were also observed for PAT compared with DST&G in year l. MANOVA results also showed lower (p < 0.01) Lachman scores in PAT-A versus DST&G-A for 3 follow-up years. Lachman scores in PAT-NM patients were lower (p < 0.01) for postop years 1 and 2 compared with DST&G-NM. PAT-C and PAT-M patients had greater but insignificant (p > 0.0l) knee stability compared with DST&G-C and DST&G-M patients, re- spectively. Strict criteria were established to rate success for 3 postoperative years. The DST&G series had 94%, 83%, and 81% success rates compared with 98%, 96%, and 96% in PAT over 3 years, respectively. These results indicate that both grafts are good choices in ACL reconstruction, but the PAT graft provides more overall knee stability than DST&G. Key Words: ACL autografts--ACL reconstruction techniques.

A current issue in o r thoped ic surgery is the treat- men t o f the an ter ior c ruc ia te l i gamen t - (ACL) defi- cient knee. N u m e r o u s surgical t echniques o f A C L recons t ruc t ion are avai lable to the o r thoped ic sur-

From the Texas Arthroscopic Surgery Knee and Sports Med- icine Clinic, Fort Worth, Texas, U.S.A.

Address correspondence and rewint requests to Angelo L. Otero, M.D., at Texas Arthroscopic Surgery Knee and Sports Medicine Clinic, 750 8th Ave., Suite 400, Fort Worth, TX 76104, U.S.A.

geon and are generally classified as extraarticular, intraarticular, or a combination. Excellent histori- cal reviews of ACL reconstruction have been pub- lished elsewhere and include information concern- ing the development of various techniques as well as several synthetic ligament substitutions (1-3).

Jensen et al. (2) report that 65 methods of ACL reconstruction have been documented from 1963 to 1983. The techniques include grafts from the iliotib- ial tract, gracilis tendon, semitendinosus tendon,

143

144 A . L. OTERO A N D L. H U T C H E S O N

meniscus, and the patellar tendon as substitutes for the ACL.

Some tendon substitutes are inherently weaker than others, and many do not approximate the strength of the normal ACL (4,5). The central bone- patellar tendon-bone specimens were found to withstand the greatest loadings (168% of a normal ACL). The semitendinosus and gracilis tendons performed at 70% and 44% of the ACL loads, re- spectively. Other structures used in ACL substitu- tions included the fascia lata and the distal iliotibial tract with various widths of accompanying fascia lata.

The purpose of this study was to compare the postoperative success and stability of ACL recon- structions using the central bone-patellar tendon- bone (PAT) graft and a doubled semitendinosus/ gracilis (DST&G) tendon graft.

METHODS AND MATERIALS

Between 1985 and 1989, 118 patients were diag- nosed with ACL insufficiency and underwent re- construction at the Texas Arthroscopic Surgery Knee and Sports Medicine Clinic in Fort Worth, Texas. Ninety-one patients were available for fol- low-up examinations for an average of 36.4 months after surgery. All data were obtained from patients' records. Fifty-five patients were surgically treated

with the PAT graft and 36 patients with the DST&G graft. No patient had previously undergone ACL reconstruction.

Surgical technique All patients were initially examined under anes-

thesia before harvesting the graft. In the DST&G series, both tendons were dissected as a free graft and sutured together. The graft was then folded once and tubularized. An auxiliary no. 5 Tycron suture was interwoven to each end of the graft, as described by Krackow et al. (6). Firm tension was applied to the DST&G graft via the auxiliary su- tures during fixation to the tibia and femur. Proxi- mally, the no. 5 Tycron Krackow stitch attached to the graft was tied to the post of a screw located in the femur. The distal aspect of the graft was secured to the tibia with an acrylic spiked washer and cor- tical screw (Fig. 1).

In the PAT series; the central third ( - 1 cm) of the patellar tendon was identified and resected, includ- ing a 3-cm bone plug at the proximal and distal as- pects of the graft (Fig. 2). The graft was then placed in a 10-mm tendon passer. The PAT graft was fixed using interference Kurosaka screws (DePuy Inc., Warsaw, IN, U.S.A.) . A Kurosaka screw was placed on the cortical side of the bone block at the proximal end of the graft. The distal end of the graft was then rotated 180 ° , and the distal bone block was

FIG. 1. Fixation of double semi- tendinosus/gracilis graft.

Arthroscopy, VoL 9, No. 2, 1993

COMPARISON OF TWO ACL RECONSTRUCTION GRAFTS 145

FIG. 2. Fixation of patellar ten- don graft.

also fixed with an interference Kurosaka screw while tension was applied.

Isometricity was verified in both graft techniques by the Acufex Isometer (Acufex Microsurgical, Inc., Norwood, MA, U.S.A.) using the technique of Noyes et al. (4).

Postoperative and rehabilitative care The affected knees in t he DST&G group re-

mained immobilized in cast braces for 4 weeks, at which time passive range of motion (ROM) exer- cises from 0 ° to 90 ° were started. At 6 weeks, weight-bearing began progressively, as tolerated.

Early in the PAT series 22 patients were immo- bilized in cast braces for 3 weeks at 45 ° of knee flexion, and passive ROM exercises were pre- scribed at 4 weeks postsurgery. Weight-bearing be- gan progressively 6 weeks after surgery. Acceler- ated rehabilitation was implemented as confidence in the fixation of the PAT graft improved. Thirty- three patients in the PAT series underwent imme- diate postoperative knee ROM with a constant pas- sive motion machine as well as immediate weight- bearing of 25% of body weight with 25% weekly increases. By the end of the 4th or 5th postoperative week, patients were bearing full body weight on the affected lower limb.

Method of evaluation The assessment of knee stability included subjec-

tive and objective measurements. Knee stability

was evaluated using scores from the Lysholm ques- tionnaire (7), KT Arthrometer (MED Metric Corp, San Diego, CA, U,S.A.) readings at 30 pounds (KT30) and maximum pull (KTMAX), and Lach- man exam (8) scores obtained during follow-up clin- ical examinations. All clinical examinations were pe r fo rmed exc lus ive ly by the senior au thor (A.L.O.).

For purposes of a more detailed comparison, pa- tients in each series were categorized as acutely injured (DST&G-A and PAT-A) if reconstruction was performed within 6 weeks after ACL injury, while all other patients were categorized as chronic (DST&G-C and PAT-C). The meniscectomy groups (DST&G-M and PAT-M) included patients who had a history of meniscectomy or underwent meniscec- tomy at the time of ACL reconstruction. Patients who underwent meniscal repair at the time of ACL reconstruction or who had no meniscal problems were categorized in the no meniscectomy groups (DST&G-NM and PAT-NM). Table 1 shows the number of patients in each category of the DST&G and PAT series.

Statistical analyses Multivariate analysis of variance (MANOVA)

was used in the statistical analysis to calculate sep- arate 2 x 2 between-groups factorial MANOVAS using the four postoperative knee stability measure- ments as dependent variables. M A N O V A was used as the statistical technique because there were

Arthroscopy, Vol. 9, No. 2, 1993

146 A. L. OTERO A N D L. H U T C H E S O N

T A B L E 1. Sample sizes for patient categories

D S T & G PAT (n = 36) (n = 55)

No. of No. of patients % patients %

Acute 10 28 27 49 Chronic 26 72 28 51 Meniscec tomy 28 78 32 58 No meni scec tomy 8 22 23 42

multiple postoperative variables in affecting the as- sessment of knee stability. Separate MANOVAs were calculated to compare surgical results of the two grafts in A/C patients and in M/NM patients.

RESULTS

Description of patients The DST&G group included 29 men and seven

women (n = 36) with a mean age of 25.4 years (SD = 5.5) and an age range from 15 to 39 years. There were 42 men and 13 women (n = 55) in the PAT group, with an average age of 25.8 years (SD = 6.1). Ages also range from 15 to 39 years. The average time from injury to ACL reconstruction surgery in the PAT group was 12.4 months (SD = 20.8), ranging from 0.1 to 104 months, compared with an average time of 13.6 months (SD = 17.3) and a range of 0.1-59 months in the DST&G group.

MANOVA results In each of" the 3 follow-up years, patients in the

PAT series had overall consistently greater knee stability for all variables compared with the DST&G group. There were no significant differ- ences (p > 0.05) in mean Lysholm scores between DST&G and PAT patients for any of the three post- operative years (Fig. 3). Lachman scores, however, were significantly (p < 0.01) higher in the DST&G 90l

80 t ~

70

60 Year 1 Year 2

i

[] PAT

i

Yeay 3

FIG. 3. Lysho lm m e a n s in D S T & G and P A T for 3 pos top years.

8

1.0

0.8

0.6

0,4

0.2

0.0 Year 1" Year 2* Year 3*

* sig, diff. (p<0.0t)

FIG. 4. L a c h m a n means in DST&G and PAT for 3 pos top years .

group compared with the PAT group for all 3 postop years (Fig. 4). Significantly higher (p < 0.05) KT-30 and KTMAX values at postop year 1 were also noted in the DST&G series compared with PAT (see Figs. 5 and 6, respectively).

A versus C patients Knee stability was greater for all variables in PAT

patients compared with DST&G patients in both injury categories. Significantly higher (p < 0.01) Lachman scores were observed in DST&G-A ver- sus PAT-A patients in all 3 postop years. Lower but insignificant (p > 0.01) knee stability was noted throughout the 3-year postop follow-up in PAT-C patients versus DST&G-C.

M versus NM patients Lachman scores were significantly lower (p <

0.01) in PAT-NM patients compared with DST&G- NM patients for postop years 1 and 2, with no sig- nificant differences at postop year 3. Comparisons of M categories in the two series showed that greater knee stability existed in PAT-M patients versus DST&G-M patients, but the difference was

2.0- [] DST&G I

I [ ] P A T / A

1.5

1.0 /

0.5'

0.0 Year 1" Year 2 Year 3

* s ig , d i f f . ( I ) < 0 . 0 5 )

FIG. 5. KT30 m e a n s in DST&G and PAT for 3 pos top years .

Arthroscopy, Vol. 9, No. 2, 1993

COMPARISON OF TWO ACL RECONSTRUCTION GRAFTS 147

1.5"

1.0"

0.5"

0.0 , Year 1"

* slg. diff. (p<0,05)

Year 2 Year

FIG. 6. K T M A X me a n s in D S T & G and PAT for 3 postop years.

not significant (p > 0.01) during the 3-year fol- low-up.

Postoperative success Four criteria were established to assess postop-

erative success in the two series: Lachman values negative or + 1, negative pivot shift, KT-1000 scores <4 mm, and extension lag of <10 °. Table 2 shows the yearly success rate based on these crite- ria. Higher success rates were observed in the PAT series compared with the DST&G series.

A m o r e strict assessment of surgical success in the two series was devised, which included the above mentioned criteria as well as Lysholm scores of I>77%. Table 3 shows yearly success rates in the two series with the addition of Lysholm scores.

Complications In our series there were no infections, deep ve-

nous thromboses, or nerve or artery involvement. There was no extension lag in DST&G patients of > 10 °, but 2% of the PAT group had an extension lag of 1>10 °. Anterior knee pain was reported in 14% of the patients in the DST&G series and 5% of PAT patients. Patellofemoral crepitus was seen in 19% of DST&G and in 29% of overall PAT patients.

As mentioned earlier, 22 patients in the PAT se- ries (40%) were immobilized for 3 weeks (PAT- LROM), whereas later in the series 33 patients (60%) began motion early in the recovery room (PAT-EROM). Table 4 shows the higher incidence of manipulations in PAT-LROM patients compared

T A B L E 2. Success rate without Lysholrn

Year 1 Year 2 Year 3

DST&G 94% 83% 81% PAT 98% 96% 96%

T A B L E 3. Success rate with Lysholm

Year 1 Year 2 Year 3

DST&G 86% 73% 69% PAT 96% 87% 87%

with early range of motion PAT-EROM and DST&G patients. The incidence of patellofemoral crepitus in DST&G, PAT-EROM, and PAT-LROM patients is shown in Table 5.

DISCUSSION

The purpose of our study was to determine which graft provided more stability--DST&G or PAT. We were fortunate that we were able to study two grafts using the same technique in a very homogeneous population, in whom isometricity was tested by the same surgeon.

In spite of our attempt to have many similar fac- tors represented in both series, major differences existed in the fixation of the graft and in rehabilita- tion. The main differences between the two grafts was fixation. Although the diameters of the grafts were similar (x = 9.6 mm in DST&G and x = 9.9 mm in PAT), the strength of the reconstruction was controlled by the fixation. The DST&G graft fixa- tion site heals by tendon-bone healing, which re- sults in the development of fibrous scar tissue. This scar tissue has unreliable tensile strength and lim- ited elasticity. The bone blocks at the ends of the PAT graft are better suited for healing by primary cells, which results in a stronger union compared with DST&G (9).

Another difference between the two series was in rehabilitation. Most series of ACL reconstructions undergo a certain amount of evolution and change in the rehabilitation as more experience and knowl- edge is gained in the treatment of patients. Postop- erative care and the approach to rehabilitation was conservative at the time all DST&G patients and 22 of the PAT patients underwent ACL reconstruc- tion. Despite the conservative postop care in all DST&G patients, knee stability was less overall compared with PAT patients.

T A B L E 4. Incidence o f manipulations

No. of pat ients %

DST&G (n = 36) 4 11 P A T - E R O M (n = 33) 1 3 P A T - L R O M (n = 22) 16 73

Arthroscopy, Vol. 9, No. 2, 1993

148 A. L. OTERO A N D L. H U T C H E S O N

TABLE 5. Incidence o f patel lofemoral crepitus

No. of patients %

DST&G (n = 36) 7 19 PAT-EROM (n = 33) 5 15 PAT-LROM (n = 22) I 1 50

The results of these series suggested that EROM in PAT reduces the incidence of postop manipula- tion, patellofemoral crepitus, as well as anterior knee pain. Many investigators have noted a high incidence of patellofemoral crepitus and anterior knee pain with the PAT graft. We believe that EROM and a properly tensioned graft minimizes this complication.

In these series, there was no significant differ- ence in Lysholm scores for any of the 3 postoper- ative years. We logically expected the chronic pa- tients, especially those with a higher severity of chondromalacia, to have lower Lysholm scores compared with those acutely injured. Also, patients with a history of meniscectomy did not have lower Lysholm scores compared with those with intact menisci. We wonder if a more sensitive scoring re- gime should be developed to assess patients post- operatively.

We did not find negative KT-1000 values in any patient in these series, and all KT values were av- eraged for purposes of analysis. During fixation of both grafts, great care was taken to tension grafts at - 5 - 7 pounds of pull. Tension was tested intraartic- ularly with a probe. This qualitative method gener- ally worked for us, but establishing more exact ten- sioning measurements of ACL grafts remains an

important issue in future ACL reconstruction re- search.

In conclusion, a comprehensive 3-year statistical analysis showed that PAT provided more stability than DST&G. Furthermore, acute patients and those with intact menisci in the PAT series had sig- nificantly greater stability compared with DST&G patients. These series showed that PAT had excel- lent success, but rather than saying PAT is the gold standard, we should say that it is the best autolo- gous substitute currently available.

REFERENCES

1. Rovere GD, Adair DM. Anterior cruciate deficient knees: a review of the literature. Am J Sports Med 1983;11:412-9.

2. Jensen JE, Slocum DB, Larson RL, James SL, Singer KM. Reconstruction procedures for anterior cruciate ligament in- sufficiency: a computer analysis of clinical results. Am J Sports Med 1983 ;11:240-8.

3. Burnett QM, Fowler PJ. Reconstruction of the anterior cru- ciate ligament: historical review. Orthop Clin North Am t985;16:143-57.

4. Noyes FR, Butler DL, Paulos LE, Grood ES. Intraarticular cruciate reconstruction. Clin Orthop Ret Res 1983;172:71-7.

5. Noyes FR, Butler DL, Grood ES, Zernicke RF, Hefzy MS. Biomechanical analysis of human ligament grafts used in knee-ligament repairs and reconstructions. J Bone Joint Surg [Am] 1984;66A:344-52.

6. Krackow KA, Thomas SC, Jones LC. A new stitch for lig- ament-tendon fixation. J Bone Joint Surg [Am] 1986;68A: 764-6.

7. Lysholm J, Gillquist J. Evaluation of knee ligament surgery results with special emphasis on use of a scoring scale. Am J Sports Med 1982;10:150-4.

8. Torg JS, Conrad W, Kalen V. Clinical diagnosis of anterior cruciate ligament instability in the athlete. Am J Sports Med 1976;4:84-91.

9. Lambert KL. Vascularized patellar tendon graft with rigid internal fixation for anterior cruciate ligament insufficiency. Clin Orthop Rel Res 1983;172:85-9.

Arthroscopy, Vol. 9, No. 2, 1993