Keywords anterior cruciate ligament injuries - anterior cruciate ligament reconstruction -
rehabilitation
Introduction
Knee injuries are common musculoskeletal problems worldwide, with a prevalence as
high as 35 cases for every 100 thousand patients.[1 ] In any case of trauma to a knee joint, the anterior cruciate ligament (ACL) is the
most common ligament to be injured.[2 ] Around 200 thousand ACL reconstruction (ACL-R) surgeries are performed annually
in the United States.[3 ]
[4 ] In India, ACL-R is usually performed using a hamstring tendon (semitendinosus and
gracilis) or a bone-patellar tendon-bone (BPTB) autograft. The former is more commonly
used than the BPTB autograft because of the ease of harvesting and lower donor-site
morbidity.[5 ] In order to achieve a successful outcome, postoperative rehabilitation following
ACL-R is an essential part of the management.[6 ] Different surgeons follow different protocols, and consensus is lacking.[7 ] But these protocols can broadly be divided into early accelerated rehabilitation
(EAR) and delayed conservative rehabilitation (DCR). Studies in the Western literature
suggest no functional differences between them, whereas other studies advocate for
the DCR protocol.[8 ]
[9 ]
The purpose of the present study was to compare the effectiveness of the EAR and DCR
protocols in the achievement of optimal outcomes by patients undergoing ACL-R surgeries
with an autologous hamstring graft in our tertiary health care hospital.
Materials and Methods
Study Design
The present prospective, randomized, single-blinded study with two parallel arms was
conducted at our tertiary care hospital from April 2019 to April 2020. The institutional
ethical committee (T/IM-NF/T&EM/18/45) approved the study, which was registered in
India's clinical trial registry (CTRI/2019/02/017726) before patient recruitment.
Study Population
We included adult patients aged between 18 and 60 years admitted for ACL-R. Those
with meniscus injuries, multiligament injuries, or associated damages to the back,
hip, or ankle joint, as well as bilateral limb injuries, were excluded. We also excluded
patients in whom the surgeons failed to harvest a right hamstring graft (minimum width
of 9 mm), or any other graft (peroneus, BPTB). Patients with radiological osteoarthritic
changes were excluded as well.
Randomization and Allocation Concealment
After fulfilling the eligibility criteria, the patients included were randomized into
two groups per computer-generated sequence using an online software randomizer. An
independent coordinator (NS) concealed the allocation numbers in sealed envelopes.
The baseline characteristics of the patients were taken after group allocation. The
independent statistician (CRM) performing the analysis was blinded in the present
study.
Sample Size
Based on a previous study by Christensen et al.[10 ] (2013), we calculated the sample size 39 patients in each group, assuming a significance
level of 5% (alpha error) with a 90% probability of achieving statistical significance
(power).
Surgical Technique
Two surgeons (SKP or BPP) operated on all included patients. After creating standard
arthroscopic portals, an arthroscopic examination was performed in all cases to confirm
the diagnosis. Double autologous semitendinosus and gracilis grafts were used in all
cases. Depending on the diameter of the harvested graft, bone tunnels were prepared
using standard jigs. The femoral side of the graft was fixed with the titanium Endobutton
adjustable fixation device (Smith & Nephew, London United Kingdom) through the inside-out
technique. A BioScrew (Linvatec Corp., Largo, FL, United States) measuring 1 mm more
than the tunnel diameter was used as a fixation device on the tibial side. Any associated
meniscal or chondral lesions were properly treated, but these cases were excluded
from the study. After thorough lavage of the knee joint, the wound was closed, and
an extended knee brace was applied. After the ACL-R, all patients were sent to rehabilitation
as per the allocated group.
EAR group – The patients performed closed kinetic chain (CKC) range of motion (ROM) exercises,
and were submitted to complete weight-bearing mobilization with the long-knee brace
from postoperative day one as per tolerance, followed by open kinetic chain (OKC)
ROM exercises and full weight-bearing walking without knee braces after two weeks.
In two to ten weeks, the patients usually followed a home-based protocol ([Fig. 1 ]). After ten weeks, both groups followed the same rehabilitation protocol.
Fig. 1 Flowchart of the protocols followed by each study group.
DCR group
— The patients kept the leg in an extended knee brace under non-weight-bearing mobilization
for the first two weeks. This was followed by CKC ROM exercises and full weight-bearing
with the long-knee brace for up to six weeks; OKC ROM exercises and full weight-bearing
mobilization without knee braces were only started after six weeks. As per the home-based
schedule ([Fig. 1 ]), after ten weeks, both groups followed the same rehabilitation protocol, described
as follows:
10 to 14 weeks: stage-III exercises
Forward and backward slow running;
Lunges and squats;
Slide board;
Ladder drills;
Aquatic program;
Progressive isokinetic quadriceps;
Progressive hamstring strengthening.
14 to 18 weeks: stage-IV exercises
18 to 24 weeks: stage-V exercises
Agility training;
Figure-of-eight jogging;
Sport-specific drills, such as figure of eight and carioca, under supervision of a
physical therapist;
Continue total body fitness.
24 weeks onwards: return to sports.
Outcome Measures
All patients were followed up at two weeks, six weeks, six months, and one year. Stich
removal was performed during the follow up at two weeks. In order to perform the statistical
analysis, we measured the outcomes of the two groups during the follow-up visit after
one year. The functional outcome was assessed through the International Knee Documentation
Committee (IKDC) score and analyses of the ROM and laxity of the knee joint; the postoperative
pain was assessed through the visual analog scale (VAS).
The IKDC score is obtained by adding the scores on individual items and converting
the crude number to a scaled number ranging from 0 to 100. The final score is evaluated
as a measure of the functional status, with higher scores representing a higher functional
level. Scores < 2 on the VAS were graded as mild pain, from 2to 4, as moderate pain,
and > 4, as severe pain. Laxity was measured using an arthrometer (KT1000, Medmetric
Corp., San Diego, CA, United States); then, it was compared to that of the normal
opposite knee. Anterior tibial translation from 0 to 2 mm is considered no laxity;
from 3 mm to 5 mm, grade I; from 6 mm to 10 mm, grade II; and > 10 mm, grade-III laxity.
Knee ROM was classified as follows: < 90° – low; between 90° and 120° – moderate;
and > 120° –good.
Statistical Analysis
An independent statistician (CRM) performed the statistical analysis using the the
R software (R Foundation for Statistical Computing, Vienna, Austria), version 3.6.1.
The categorical variables were expressed as percentages, and the numerical variables
(non-parametric), as medians with interquartile ranges (IQRs). The Chi-squared test
was used for the bivariate analysis regarding the categorical variables, whereas the
Wilcoxon rank-sum test (non-parametric) was used for the categorical and numeric variables.
Results
A total of 87 subjects were selected, but 1 patient did not provide consent and was
excluded. The remaining 86 patients were divided into 2 groups (of 43 patients each).
One patient in the EAR group was lost to follow-up, and two patients had associated
meniscus tears; they were excluded. Three patients in the DCR group had associated
meniscus tears and were also excluded. Thus, 80 patients (40 patients in each group)
were analyzed, as shown in [Figure 2 ].
Fig. 2 Consolidated Standards of Reporting Trials (CONSORT) flow diagram showing randomization
and group allocation.
The baseline characteristics of the sample were comparable, as depicted in [Tables 1 ] and [2 ]. We observed a remarkable improvement in the IKDC score compared to the preoperative
values in both groups. During the follow-up after one year, the ROM and the VAS and
IKDC scores in both groups were comparable, but this was not statistically significant,
as seen in [Figure 3 ] and [Table 3 ] (p = 0.36, 0.51, and 0.91 respectively). The one-year postoperative laxity was higher
in the EAR group in comparison to the DCR group (p = 0.039; [Table 3 ] and [Fig. 4 ]).
Fig. 3 Bar diagram showing the postoperative range of motion of the study groups.
Fig. 4 Bar diagram showing the postoperative laxity of the study groups.
Table 1
Variables
Early rehabilitation
(n = 40)
Delayed rehabilitation
(n = 40)
p -value
Age in years:
median (IQR) [range]
34 (28–39) [18–60]
33 (26–38) [18–60]
0.66
Male gender: mean (%)
37 (92.5%)
36 (90%)
0.99
Height in cm:
median (IQR) [range]
165.3 (159.1–172) [152–179.2]
164.8 (158.4–171.8) [151.9–179]
0.57
Weight in kg:
median (IQR) [range]
65 (58–70) [54.4–82]
65.6 (60.1–72.2) [53.4–82.6]
0.76
BMI in kg/m2 :
median (IQR) [range]
24 (22–27) [16–38]
26 (22.6–27.4) [15.8–29]
0.43
Table 2
Variables
Early rehabilitation
(n = 40)
Delayed rehabilitation
(n = 40)
p -value
IKDC score: median (IQR)
49 (45–52.1)
48.2 (44.8–52.1)
0.75
Visual Analog Scale score: mean (%)
0.78
Mild
34 (85%)
32 (80%)
Moderate
6 (15%)
8 (20%)
Severe
0
0
Range of motion: mean (%)
0.216
< 90° (low)
0
0
90°–120° (medium)
8 (20%)
15 (37.5%)
> 120° (good)
32 (80%)
25 (62.5%)
Table 3
Variables
Early rehabilitation
(n = 40)
Delayed rehabilitation
(n = 40)
p -value
IKDC score: median (interquartile range)
89.5 (85.2–92.1)
89.6 (85.4–92.5)
0.91
Visual Analog Scale score: mean (%)
0.51
Mild
38 (95%)
38 (95%)
Moderate
1 (2.5%)
2 (5%)
Severe
1 (2.5%)
0
Range of motion: mean (%)
0.36
< 90° (low)
4 (10%)
5 (12.5%)
90°–120° (medium)
7 (17.5%)
13 (32.5%)
> 120° (good)
29 (72.5%)
22 (55%)
Laxity score: mean (%)
0.039*
No laxity
27 (67.5%)
37 (92.5%)
Grade 1
10 (25%)
3 (7.5%)
Grade 2
2 (5%)
0
Grade 3
1 (2.5%)
0
There were three cases of superficial wound infection in our series, one in the EAR
group and two in the DCR group. All responded to debridement and serial dressing.
There were no cases of deep infection.
Discussion
In the follow-up visit after one year, the knee laxity was significantly higher in
the EAR group compared to the DCR group, even as the pain, ROM and functional outcomes
remained the same.
Christensen et al.[10 ] did not find any differences in the subjective IKDC score, knee laxity and ROM between
their two study groups. The essential goal of the rehabilitation program following
ACL-R should be to restore the full knee ROM.[11 ] Although there are enough studies in the literature to enable the conclusion that
early recovery of the ROM is required to achieve better outcomes following ACL-R,
it is still inconclusive whether an accelerated rehabilitation protocol helps achieve
this more rapidly.[12 ] However, the present study demonstrates a higher potential risk of residual laxity
with the EAR protocol. Some authors[13 ]
[14 ] have mentioned an increase in the diameter of the bone tunnel after the EAR protocol
with semitendinous grafts, but without any conclusive evidence on the anteroposterior
and subjective outcomes.
Similarly, Beynnon et al.[15 ] concluded that there were no significant differences regarding both protocols in
terms of muscle strength and knee laxity. In contrast, we recorded substantial anteroposterior
laxity during the follow-up after one year in the EAR group. Osteointegration initially
occurs with fibrovascular interface tissue between bone and tendon and, subsequently,
bony ingrowth takes around three to six weeks.[16 ] This could be affected by OKC exercises if they start early. Moreover, activities
like squatting and cross-leg sitting may strain the newly-reconstructed ACL. Escamilla
et al.[17 ] found that, in early squatting between 0° and 60°, the shear forces were low and
primarily restricted by the ACL. With near-maximum knee flexion, the shear forces
also peak, putting a lot of stress on the new ACL.[17 ]
Andersson et al.[18 ] performed a systematic review of four randomized studies. They found that in ACL-R
with BPTB graft, early CKC exercises generate lower levels of pain, lower risk of
increased laxity, and better self-reported knee function than OKC quadriceps exercises.
In contrast, a recent study[19 ] found no difference between the groups. Glass et al.[20 ] also advised against using OKC exercises within the first six weeks. All these studies
involved BPTB grafts. Heijne et al.[21 ] compared BPTB and hamstring grafts in early (4 weeks) and late (12 weeks) OKC exercises
and found that laxity was higher in the hamstring group. In the present study, patients
in the DCR group performed delayed (after six weeks) OKC exercises, and fared better
in terms of laxity. Van Grinsven et al.[22 ] performed a systemic review and found that an accelerated protocol without postoperative
bracing does not lead to stability problems, and offers the advantage of reduction
in pain, swelling, and inflammation, as well as ROM recovery. Morrissey et al.[23 ] did not observe differences in the VAS pain scores of the OKC and CKC groups. In
the present study, the VAS scores were lower in the EAR group, which performed CKC
exercises, both three and six months postoperatively. Kruse et al.[24 ] have stated that further investigations are warranted in order to draw conclusions
regarding the rehabilitation protocol.
Tyler et al.[25 ] compared immediate weight-bearing as tolerated versus a delay of two weeks; they
found a statistically significant difference in anterior knee pain in the delayed
group, and concluded that early weight-bearing did not cause harmful effects on stability
or function. Schenck et al.[26 ] compared clinic-based and home-based rehabilitation protocols and found that minimal
supervision during rehabilitation could result in equivalent outcomes following ACL-R.
We distributed printouts explaining the exercises to our patients, who were free to
inquire about any difficulties on any day outside the follow-up visits.
In the present study, the median IKDC score (of 89.5) was similar in both groups one
year postoperatively, which is in line with the study by Grindem et al.,[27 ] who found a score of 89/100 score two years postoperatively. However, Hopper et
al.[28 ] found that knee scores continue to improve up to six years after surgery and only
attain approximately 86% of their value after one year.
The strength of the present study is that it is the first of its kind conducted in
the Indian population. The limitation was the short follow-up, of only one year. Hence,
a larger, multicentric, and longer study in the same population may help to validate
our findings.
Conclusion
The EAR protocol yields similar ROM, IKDC scores, and postoperative pain relief compared
to the DCR protocol. However, at the one-year follow-up visit, we observed significantly
higher knee laxity in the EAR group.
