Keywords
hip arthroscopy - patient-reported outcomes - tetracycline - doxycycline
Femoroacetabular impingement syndrome (FAIS) describes a condition in which the anatomy
of the hip allows abnormal contact between the femoral head and the acetabulum. FAIS
may lead to pain, cartilage damage, and labral tears, resulting in hip dysfunction.[1] FAIS is associated with two main anatomical deformities: the cam deformity and the
pincer deformity. The cam deformity is associated with an increased α angle, and is
described as an unnaturally aspherical shape of the femoral head. The pincer deformity
is described as overcoverage of the femoral head by the acetabulum, and is associated
with lateral center-edge angle (LCEA).[1]
While open surgical treatment has been reported to be successful in treating FAIS,[1] hip arthroscopy has evolved into the primary surgical option for the treatment of
FAIS. Systematic reviews comparing the arthroscopic approach to open procedures have
demonstrated arthroscopic procedures to be a safe and effective method of treatment,
with a reported complication rate of 0 to 5%.[2] Specifically, persistent postoperative pain and restricted range of motion, which
may be caused by postoperative adhesion formation,[3] are of exceeding importance in the young, active FAIS population.[4] Adhesions form as fibrin deposits secondary to the coagulation cascade and lead
to fibroblast chemotaxis, which deposit the extracellular matrix (ECM).[5]
Resolution of postoperative pain and discomfort continues to be a limiting factor
in patient-reported outcome (PRO) scores. Numerous modalities have been evaluated
to combat postprocedural pain in this population, including perioperative nerve blocks
and local anesthetic injections as well as preoperative anti-inflammatory medications.[6] While not yet labeled for such a use, doxycycline has several potential mechanisms
to reduce postoperative pain after hip arthroscopy, including inhibiting painful adhesions.
Many adhesions are asymptomatic, but some can be a significant source of postoperative
pain.[3] The pathogenesis of arthrofibrosis is largely due to the activation and differentiation
of fibroblasts.[7] Doxycycline reduces fibroblast activity and the inflammatory dysregulation of matrix
metalloproteinases (MMPs) that can lead to ECM contraction in a variety of pathologies.[8]
[9]
[10] In addition to inhibiting painful adhesions, the anti-inflammatory effects of tetracyclines
have been shown to suppress secondary pain pathways, resulting in reduced central
nervous system inflammation.[11] However, the literature is sparse in documenting the modulation of the neuroinflammatory
response in the surgical setting.
The modulation of secondary pain pathways is of particular interest in the hip arthroscopy
patient population that has higher prevalence of chronic opioid use, substance abuse,
and concomitant psychiatric conditions such as anxiety and depression.[12] Furthermore, anxiety and depression are independently predictive of poor PROs and
postoperative pain scores in patient undergoing hip arthroscopy.[13] The neuromodulatory and secondary pain pathway effects of doxycycline may have an
added benefit in this patient population. This may be specifically measured in common
PROs in questions that specifically relate to pain and in functional measures that
may improve due to decreased pain.
The purpose of this retrospective case–control study was to determine if those who
underwent a hip arthroscopy procedure for femoroacetabular instability (FAIS) and
took a 7-day postoperative course of doxycycline had improved PRO scores compared
with those who did not take doxycycline postoperatively. Our hypothesis was that postoperative
doxycycline 100 mg twice daily for 7 days would improve early PROs after hip arthroscopy.
We secondarily hypothesized that this effect would be intensified in those patients
with a mood disorder.
Methods and Materials
The University of Kentucky institutional review board–approved prospective hip arthroscopy
outcomes registry was queried for patients who had undergone arthroscopic femoral
osteochondroplasty and/or labral repair or reconstruction secondary to FAIS performed
by a single, fellowship-trained surgeon between December 30, 2013, and May 26, 2020.
Patients were excluded if they underwent open hip procedures, revision hip arthroscopy,
or the indication for surgery was anything other than FAIS. Starting in November 2018,
patients with FAIS treated with an arthroscopic procedure were given a 7-day course
of doxycycline (100 mg twice daily).
Preoperative Veterans RAND 12-Item Health Survey (VR-12) mental component scores (MCSs),
6-week postoperative Hip Disability and Osteoarthritis Outcome Score (HOOS) global
(HOOSglobal), and International Hip Outcome Tool (iHOT) scores were then compared between the
consecutive series treated with versus without doxycycline. Additional subgroup analysis
was performed on patients with low preoperative VR-12 MCSs to determine possible additional
benefit of postoperative doxycycline in this group. Consistent with previous studies,
the threshold to categorize patients as having low preoperative VR-12 MCSs was a score
of less than 42.9.[14]
[15]
Demographic measures collected included age, sex, and body mass index (BMI). Radiographic
measures collected included pre- and postoperative α angle and LCEA. Alpha angles
were measured by drawing a best-fit circle around the femoral head in either the Dunn
view or the frog-leg lateral view of the operative leg. One component of the angle
was made by a line in the trajectory of the femoral neck running through the center
of the best-fit circle. The second component was comprised of the line from the center
of the best-fit circle to the point at which the femoral head or neck exceeds the
circumference of the best-fit circle.[16] The LCEA was measured in the standing anteroposterior pelvis view by drawing a best-fit
circle around the femoral head. The angle was formed by the line perpendicular to
the horizontal axis through the center of the best-fit circle and a line intersecting
with the lateral edge of the acetabulum.[17] Operative data collected included number of anchors used, Outerbridge classification
of femoral, acetabular chondromalacia observed during arthroscopy, and size of labral
tear visualized during arthroscopy as measured on a clock face. Further postoperative
variables collected included number of reoperations.
Proportions of patients who achieved a patient-acceptable symptom state (PASS) were
calculated for each group and compared in both the overall and subgroup analysis.
PASS is defined by a HOOSglobal score of at least 62.5, which was calculated to represent the threshold at which
the majority of patients were satisfied with their procedure.[18] With the primary end point of proportion achieving the HOOSglobal threshold, a sensitivity analysis was performed to determine the minimum percentage
difference detectable with the existent data. With α set at 0.05 and β at 0.80, the
cohort of 132 patients who completed a postoperative HOOSglobal, the minimum PASS threshold proportion difference detectable was about 23.5%. In
addition, to isolate a postoperative measure of pain, a component of the iHOT was
analyzed separately, which consisted of a 100-point visual analogue scale (VAS) for
pain, labeled in the following as iHOT VAS.
Two-sample t-tests were used to compare the postoperative outcome means. The z-test for proportions
was used to analyze the difference between the proportion of patients who achieved
the PASS threshold and those who underwent reoperation. A chi-square test was used
to determine differences between the doxycycline and no-doxycycline groups in terms
of Tonnis and Outerbridge classifications. R software version 4.0.2 (R Foundation
for Statistical Computing, Vienna, Austria)[19] was used for data analysis. Statistical significance was set at p < 0.05.
Results
Demographics
In total, 134 patients were identified who underwent hip arthroscopy for FAIS and
completed either postoperative iHOT or postoperative HOOSglobal surveys. The average age of all patients was 36.2 ± 13.4 years. There were 98 female
patients (73.1%), and the average BMI was 27.2 ± 5 kg/m2. The group that received doxycycline did not significantly differ from the group
that did not receive doxycycline in age, sex, or BMI ([Table 1]). The groups also did not differ in distribution of acetabular versus femoral osteochondroplasty,
labral repair versus reconstruction, frequency of capsular closure, or use of knotless
versus knotted anchors ([Table 2]).
Table 1
Group demographics
|
Doxycycline (n = 50)
|
No doxycycline (n = 84)
|
p-Value
|
|
Age (y)
|
35.9 ± 11.2
|
35.5 ± 11.4
|
0.840
|
|
Sex (M/F)
|
16/34
|
20/64
|
0.410
|
|
Body mass index
|
26.9 ± 5.1
|
27.4 ± 4.9
|
0.580
|
Table 2
Procedure breakdown and analysis; Fisher's exact test for osteochondroplasty breakdown,
repair versus reconstruction, knotless versus knotted anchors, and specific anchor
types; z-test for proportions for percentage of capsular closure
|
Doxycycline (N = 50)
|
No doxycycline (N = 84)
|
p-Value
|
|
Osteochondroplasty location
|
|
Acetabular osteochondroplasty
|
0
|
13
|
> 0.999
|
|
Femoral osteochondroplasty
|
48
|
81
|
|
Labral treatment
|
|
Labral repair
|
50
|
71
|
> 0.999
|
|
Labral reconstruction
|
8
|
8
|
|
Capsular closure (%)
|
49 (98.0)
|
82 (97.6)
|
0.889
|
|
Knotted versus knotless
|
|
Knotted anchors
|
31
|
73
|
0.538
|
|
Knotless anchors
|
1
|
1
|
|
Specific anchor types
|
|
Pivot NanoTack
|
26
|
67
|
0.514
|
|
Smith and Nephew Bioraptor
|
0
|
3
|
|
Smith and Nephew Healicoil and Footprint
|
0
|
1
|
|
Smith and Nephew Suture Fix
|
0
|
2
|
|
Smith and Nephew Q Fix
|
13
|
0
|
|
Smith and Nephew Bioraptor
|
1
|
0
|
|
Smith and Nephew Healicoil
|
1
|
0
|
Overall Outcomes
Preoperative MCSs were significantly lower in the group that received doxycycline
(38.2 vs 48.8, p < 0.001). However, there was no significant difference in preoperative α angle, LCEA,
or Tonnis grade ([Table 3]). In addition, there was no difference in average number of anchors used intraoperatively,
nor was there a difference in the Outerbridge grade of acetabular or femoral chondromalacia
described intraoperatively. Ultimately, the groups did not differ significantly with
respect to postoperative iHOT, HOOSglobal, α angle, LCEA, iHOT VAS, the proportion above HOOSglobal PASS threshold, or reoperation rate ([Table 4]).
Table 3
Tonnis osteoarthritis classification and Outerbridge acetabular and femoral chondromalacia
classifications
|
Group
|
0
|
1
|
2
|
3
|
4
|
p-Value
|
|
Hip osteoarthritis
Tonnis grade
|
Doxycycline
|
37
|
13
|
0
|
0
|
0
|
0.544
|
|
No doxycycline
|
66
|
18
|
0
|
0
|
0
|
|
Acetabular chondromalacia
Outerbridge grade
|
Doxycycline
|
5
|
5
|
32
|
3
|
5
|
0.152
|
|
No doxycycline
|
15
|
1
|
52
|
5
|
9
|
|
Femoral chondromalacia
Outerbridge grade
|
Doxycycline
|
41
|
5
|
2
|
2
|
0
|
0.716
|
|
No doxycycline
|
64
|
5
|
5
|
6
|
1
|
Table 4
Overall group analyses
|
Doxycycline
|
No doxycycline
|
p-Value
|
|
Metric
|
Mean ± SD
|
N
|
Mean ± SD
|
N
|
|
Reoperation
|
–
|
8
|
–
|
8
|
0.399
|
|
Pre-op α angle
|
55.4 ± 10.5
|
48
|
60.0 ± 13.9
|
76
|
0.052
|
|
Post-op α angle
|
44.3 ± 7.4
|
48
|
47.5 ± 9.9
|
73
|
0.065
|
|
Pre-op LCEA
|
33.0 ± 7.0
|
49
|
31.9 ± 6.8
|
79
|
0.384
|
|
Post-op LCEA
|
31.8 ± 6.5
|
49
|
31.2 ± 6.8
|
79
|
0.619
|
|
Number of anchors
|
3.7 ± 1.8
|
42
|
3.2 ± 1.2
|
71
|
0.056
|
|
Size of labral defect
|
9:52 to 2:20
|
50
|
10:45 to 1:23
|
78
|
0.130
|
|
Pre-op MCS
|
38.2 ± 1 7.0
|
30
|
48.8 ± 11.8
|
58
|
< 0.001
|
|
Post-op iHOT
|
52.3 ± 24.5
|
50
|
54.6 ± 21.4
|
81
|
0.590
|
|
Post-op iHOT VAS
|
31.2 ± 26.9
|
45
|
34.6 ± 23.5
|
80
|
0.463
|
|
Post-op HOOSglobal
|
58.7 ± 14.6
|
48
|
59.0 ± 12.5
|
84
|
0.910
|
|
> HOOSglobal PASS threshold
|
–
|
13
|
–
|
28
|
0.580
|
Abbreviations: HOOSglobal, Hip Disability and Osteoarthritis Outcome Score (HOOS) global; iHOT, International
Hip Outcome Tool; LCEA, lateral center-edge angle; MCS, mental component score; PASS,
patient-acceptable symptom state; VAS, visual analogue scale. Bold values highlight
statistical significance.
Subgroup Analysis
Among those with low preoperative MCSs (< 42.9, N = 83), the groups did not differ with respect to any postoperative outcome, including
proportion above HOOSglobal PASS threshold and reoperation rate. However, while pre- and postoperative LCEAs
were not significantly different between the two groups, α angles were lower on average
in both timeframes in the group that received doxycycline (55.3 vs 63.5 degrees, p = 0.009; 44.3 vs 48.7 degrees, p = 0.042). Additionally, more anchors were used on average in the doxycycline group
(four vs three, p = 0.018) ([Table 5]).
Table 5
Subgroup analyses of patients with preoperative MCS less than 42.9
|
Doxycycline
|
No doxycycline
|
p-Value
|
|
PRO
|
Mean ± SD
|
N
|
Mean ± SD
|
N
|
|
Reoperation
|
8
|
41
|
7
|
42
|
0.337
|
|
Pre-op α angle
|
55.3 ± 11.0
|
40
|
63.5 ± 16.1
|
41
|
0.009
|
|
Post-op α angle
|
44.3 ± 7.5
|
40
|
48.7 ± 12.3
|
40
|
0.042
|
|
Pre-op LCEA
|
33.4 ± 7.4
|
41
|
31.7 ± 8.0
|
42
|
0.318
|
|
Post-op LCEA
|
32.0 ± 6.8
|
41
|
31.6 ± 7.5
|
42
|
0.800
|
|
Number of anchors
|
3.8 ± 1.8
|
36
|
2.9 ± 1.3
|
36
|
0.018
|
|
Size of labral defect
|
10:00 to 2:20
|
50
|
10:45 to 1:13
|
40
|
0.401
|
|
Post-op iHOT
|
53.6 ± 23.1
|
22
|
45.5 ± 20.3
|
19
|
0.250
|
|
Post-op iHOT VAS
|
30.7 ± 27.2
|
38
|
36.8 ± 26.5
|
41
|
0.316
|
|
Post-op HOOSglobal
|
57.3 ± 7.4
|
21
|
51.6 ± 13.3
|
19
|
0.120
|
|
> HOOSglobal PASS threshold
|
–
|
4
|
–
|
4
|
> 0.999
|
Abbreviations: HOOSglobal, Hip Disability and Osteoarthritis Outcome Score (HOOS)
global; iHOT, International Hip Outcome Tool; LCEA, lateral center-edge angle; MCS,
mental component score; PASS, patient-acceptable symptom state; PRO, patient-reported
outcome; VAS, visual analogue scale. Bold values highlight statistical significance.
Discussion
The purpose of this study was to assess the potential benefits of postoperative doxycycline
use in the FAIS population treated with hip arthroscopy. The primary finding of this
study was that a 7-day course of postoperative doxycycline did not improve early clinical
outcome scores after hip arthroscopy, regardless of preoperative mental health status.
This study is the first to examine the effects of doxycycline use in modifying postoperative
pain. Tetracyclines, while initially used for their bacteriostatic effects, have recently
been studied for their antihypernociceptive and anti-inflammatory effects.[20] The anti-inflammatory effects of another tetracycline, minocycline, has already
been well established in its effectiveness for rheumatoid arthritis.[21]
[22]
[23]
[24]
[25] In addition, tetracyclines' ability to inhibit microglial cell and astrocyte activation
has shown the potential to decrease chronic pain in an animal model of osteoarthritis.[11]
Our analyses suggested doxycycline was not effective in combating postoperative pain.
Overall, the groups did not differ significantly in any postoperative PROs, including
the proportion of those who reported outcomes above the PASS threshold (HOOSglobal > 62.5),[18] the iHOT VAS, and reoperation rate. The treatment groups did differ in preoperative
MCSs, which could have served as a confounder causing the outcomes of the doxycycline
group to be suppressed postoperatively. Chronic hip pain and postoperative pain scores
have been shown to be significantly correlated with depression.[26]
[27] As such, the experimental group would be expected to experience more postoperative
pain and therefore report lower functional scores.
Despite the overall difference in preoperative MCSs between groups, subgroup analysis
of only those with low preoperative MCSs also showed no difference in postoperative
outcomes between groups. This finding suggests that the doxycycline group's insignificantly
different results in the overall analysis were not due to preoperative MCS differences.
Furthermore, the nondoxycycline group exhibited higher preoperative α angles on average
and had fewer anchors placed on average. While a higher α angle suggests the potential
for more severe FAIS, the fewer anchors placed may suggest less difficult labral tear
fixation intraoperatively. While it is difficult to fully isolate doxycycline as an
independent variable, our findings suggest that, regardless of preoperative symptoms,
“acceptable” symptomatic improvement was achieved in roughly the same proportions
in each group.
In addition, the subgroup analysis further subverts the potential for doxycycline
to be of added benefit to those patients with low MCS who would theoretically be most
at risk for pain-related complications. Its added benefit was thought to be through
a dual mechanism of inhibiting MMPs and potentially acting as an antidepressant via
a reduction in neuroinflammation.[28]
[29]
[30] Such a hypothesis has been supported by a placebo-controlled randomized trial with
minocycline as the treatment drug.[31] In addition, the chronic central nervous system inflammation now implicated in certain
depression etiologies is also causative of chronic osteoarthritis pain in animal models,[11] which minocycline has similarly been found to reduce.[32] Our findings suggest that while tetracyclines may have future clinical utility as
antineuroinflammatory drugs, doxycycline in the post–hip arthroscopy setting did not
provide a secondary benefit to patients with low MCSs in the early postoperative setting.
Our study contains common limitations of retrospective designs. The PRO instruments
used in our outcomes registry changed over time and, as such, many patients did not
have preoperative scores available. To maintain adequate power, only postoperative
outcome comparisons were included in the analysis, but we are then limited in our
ability to assess pre- to postoperative changes. This was compensated for by using
the HOOSglobal PASS criteria to assess for the proportion of patients who achieved acceptable postoperative
outcomes in each group. Even with this adjustment, the sample size collected allowed
for determination of a minimum of over 20% difference in those achieving PASS threshold.
There may exist a true difference between the groups of less than 20%.
In addition, the retrospective nature of the study would potentially introduce significant
selection bias, which allowed the potential for the doxycycline and no-doxycycline
groups to differ. The risk of this bias was minimized by comparing the group demographics
and using subgroup analysis to account for the measure in which the two groups did
significantly differ. Further, the groups were not shown to differ significantly in
capsular management, anchor types, or distribution of osteochondroplasty. Additionally,
no patients before November 2018 received doxycycline, while all patients after November
2018 received doxycycline. Although this was a retrospective study, the doxycycline
administration was in practice similar to a trial using blocked randomization. This
design does lead to potential surgical skill differences, however, as hip arthroscopy
complications have been shown to decrease with increased surgeon experience.[33] The learning curve as described by Mehta et al suggests that the surgeon included
in the present study may have advanced into a higher stratum during the course of
the 134 hip arthroscopies performed,[33] to improved outcomes in the more recent doxycycline group.
While these limitations may not allow for the potential benefits of postoperative
doxycycline to be fully captured, our results did not demonstrate an early postoperative
benefit with the use of doxycycline. Further analysis and long-term follow-up studies
are indicated to evaluate for long-term benefits of doxycycline in potentiating the
neuroinflammatory response associated with hip arthroscopy. Furthermore, prospective
studies should analyze opioid use postoperatively with doxycycline treatment. The
present study was not able to analyze opioid use due to the retrospective nature of
the study, but a design prospectively including pill counts along with PROs for pain
may illuminate benefit of tetracyclines, decreasing need for opioids.
Conclusion
Mitigating postoperative pain is of importance after arthroscopic FAIS procedures.
Our study aimed to see if postoperatively administered doxycycline attenuated pain,
resulting in better postoperative outcomes. The PRO scores did not significantly differ
between those who did and did not receive a 7-day postoperative course of doxycycline.
There was also no difference in the proportion of patients who achieved the postoperative
HOOSglobal PASS threshold. Our study suggests that doxycycline does not decrease postoperative
pain in hip arthroscopy patients.
