Transcript Emergency Reversal of Coagulopathy and Platelet

Evaluating the Evidence Behind
the Surgical Treatment of Lumbar
Stenosis
Joseph Beshay, MD
Assistant Professor of Neurological Surgery
University of Texas Southwestern
Lumbar Stenosis





Narrowing of the Spinal canal with
encroachment on the neural structures by
surrounding bone or soft tissue.
Typical presentation: Radicular symptoms or
neurogenic claudication.
Back pain common but not a result of lumbar
stenosis per se.
May be asymptomatic
Most common reason for lumbar surgery in
adults over 65
Deyo et al. Spine 1993
Lumbar Stenosis
Lumbar Stenosis

Treatment


Surgical: Laminectomy / laminotomy / foraminotomy/
Open vs MIS technique, Interspinous process device,
Extension limiting device, Interspinous
decompression, “chimney” laminectomy……
Non-Surgical: Epidural steroids, Physical therapy,
NSAIDS, Narcotic analgesics, spinal manipulation,
back exercises…..
Lumbar Stenosis


In any given week I see at least one patient with
‘lumbar stenosis’ and at least one patient with
‘LOW BACK PAIN’
Preconceived ideas……….
Lumbar Stenosis
Thank you for
seeing me
today, Doctor
I won’t take up
much of your
time
If you can help
with my leg
pains, I would
be thankful
If you can’t
help me, I
thank you
anyways
I baked you this
pie…
Back Pain
Back Pain
Hey! Fix
my back
pain.
I want to be
able to go
bowling
again!
Don’t you
tell me to
lose weight!
Hey! Refill
my Lortab
Fill out this
stack of
disability
forms
Lumbar Stenosis
The Evidence

Treatment





Where is the data?
Paucity of head to head randomized trials
Many small non-randomized observational type
studies.
Most randomized studies are small
Two large studies


Maine Lumbar Spine Study
SPORT (Spine patients outcomes research trial) lumbar
stenosis arm.
Lumbar Stenosis
Main Lumbar Spine Study





Observational Study
Surgical vs Non-Surgical treatment determined
by Surgeon and Patient
148 patients enrolled between 1990-1992
Excluded patients with prior lumbar surgery,
cauda equina, fractures, infections,
malignancies…etc.
Reported results at 1, 4 and 10 years
Atlas et al. Spine 2005
Lumbar Stenosis
Main Lumbar Spine Study




97 patients available for follow-up
Surgical group had more severe baseline
symptoms and worse functional status
Leg pain relief and back related functional status
favored surgery
Low back pain relief similar between both
groups
Atlas et al. Spine 2005
Lumbar Stenosis
Main Lumbar Spine Study
Atlas et al. Spine 2005
Lumbar Stenosis
Main Lumbar Spine Study
Atlas et al. Spine 2005
SPORT - Background




The Spine Patient Outcomes Research Trial
(SPORT) was designed to assess the relative
efficacy and cost-effectiveness of surgical and
non-surgical approaches to the treatment of
common conditions associated with low back
and leg pain.
Three arms studying lumbar HNP, Stenosis and
degenerative spondylolisthesis
13 medical centers in 11 States
Randomized as well as observational cohorts
Weinstein et al. NEJM 2008
SPORT – Lumbar Stenosis

Inclusion Criteria




Neurogenic claudication or radicular symptoms for at
least 12 weeks
Cross sectional imaging confirming dx
Surgical Candidate
Exclusion Criteria
Spondylolisthesis (studied separately)
 Instability (4mm translation or 10 degree angulation)
Most patients had received some form of non-surgical tx
for their symptoms prior to enrollment (PT 68%, ESI
56%, Chiropractor 28%, NSAIDs 55%, Opioids 27%).


Weinstein et al. NEJM 2008
SPORT – Lumbar Stenosis


After verifying eligibility, patients were
allowed to enroll in the randomized or
observational cohort. (Observational group
avoids ‘enrollment bias’)
To aid in decision making, patients shown
“evidence-based videotapes” with
“standardized information” regarding
surgical and non-surgical treatments.
Weinstein et al. NEJM 2008
SPORT – Interventions


Surgical: “Standard posterior
decompressive laminectomy.”
Non-Surgical:


“Usual care”
“….at least active physical therapy, education
or counseling with home exercise instruction,
and the administration of non-steroidal
antiinflammatory drugs, if tolerated.
Weinstein et al. NEJM 2008
SPORT – Outcome measures




Outcomes of bodily pain and disability
SF-36, Modified Oswestry Disability Index, other
patient indices including overall satisfaction,
pain, “Leg pain bothersomeness scale”
Follow-up at 6 weeks, 3 months, 6 months, 1
year and 2 years.
Effect of treatment was defined as the difference
in the mean change from baseline between
surgical and non-surgical groups.
Weinstein et al. NEJM 2008
SPORT – Patients
So far so good…..
SPORT – Patients
What really happened..

Randomized Cohort





Surgical arm: At two years 67% had undergone surgery
Non-Surgical arm: At two years 43% of patients had crossed
over and had surgery.
~55% of the randomized patients had surgery.
~40% of the randomized pt’s crossed over
Observational Cohort



Surgical arm: At two years 96% had undergone surgery
Non Surgical arm: 22% crossed over to surgery
~66% of observational patients had surgery
Weinstein et al. NEJM 2008
SPORT – Patients
What really happened..

Bottom line:



Lots of crossover – the Achilles heal of studies comparing
surgery to non-surgical tx.
Investigators maintained good follow-up
Interesting findings



Pt’s in observational cohort had more ‘signs of nerve root
tension’
Exhibited a stronger treatment preference
Group undergoing surgery tended to be younger, more likely
to be working, had more pain and lower level of function.
Also exhibited ‘more severe’ stenosis on imaging.
Weinstein et al. NEJM 2008
SPORT – Patients
What really happened..

Non surgical treatment
Similar between observational and randomized
cohorts
 More patients in the randomized group reported
visits to a surgeon (45% vs 32%). Also more use
of ‘injections’ (52% vs 39%) - ? An attempt to
keep them in their randomized group?
 More observational patients used ‘other’
medications such as gabapentin

Weinstein et al. NEJM 2008
SPORT – Patients
What really happened..

Surgical Treatment
Similar between both groups
 Instrumentation in 6% of patients
 9% durotomy rate
 8% re-operation rate at 2 years with <50% for
recurrent stenosis
 No mortality directly attributable to surgery

Weinstein et al. NEJM 2008
SPORT – Patients
What really happened..

Crossover largely ‘predictable’
Pts who crossed over to surgery had more severe
symptoms and self-rated disability
 Pts who crossed over to non-surgical tx had less
bothersome sx’s and favored non-surgery at
baseline.


Observational cohort very similar to
randomized group though.
Weinstein et al. NEJM 2008
SPORT
How They Analyzed the Data

Randomized group:
Surgery vs no surgery with ‘intent to treat’ analysis
 Surgery vs no surgery with ‘as treated’ analysis


Observation group:



Surgery vs no surgery
Looked at primary outcomes of bodily pain,
physical function and mean ODI.
Secondary outcome data
SPORT
What they found
Randomized Group
Weinstein et al. NEJM 2008
Weinstein et al. NEJM 2008
SPORT
What they found
Observational Group
Weinstein et al. NEJM 2008
Weinstein et al. NEJM 2008
Statistical Analysis

In RCT’s not all patients adhere to protocol they
were assigned to.




Non compliance
Crossover
Lost to follow-up / drop-out
This makes your data set imperfect and more
difficult to analyze.



Per protocol analysis – Deviate from protocol and you’re out
Intent to treat analysis – As randomized, so analyzed
As treated analysis – Analyze based on tx not randomization
Statistical Analysis
Origin of Randomization





R.A. Fisher in 1923 studied the effects of
different fertilizers on potato yields.
Some fields (or even parts of a field) are more
fertile than others though.
Decided to apply the fertilizer to small plots.
Randomly assign fertilizers to plots/rows.
Randomization destroys any connection between
soil fertility and treatment.
Statistical Analysis
Randomization


Any difference between groups that arises
after randomization should be due to
consequences of the randomized
treatment assignment
Adjusting the analysis of treatment effect
by post-randomization group differences
could introduce bias
Statistical Analysis
Intention to Treat Analysis

Includes all randomized patients in the groups to
which they were randomly assigned, regardless
of their adherence with the entry criteria,
regardless of the treatment they actually
received, and regardless of subsequent
withdrawal from treatment or deviation from the
protocol
(Lloyd) Fisher et al., 1990
Statistical Analysis
Intention to Treat Analysis
What happens when the crossover is great and there is, in
fact, a difference between the two treatments?
Statistical Analysis
Intention to Treat Analysis

LARD
Laparoscopic Antroplasty vs Restricted Diet
1000 patients

Inclusion criteria



Exclusion criteria



Obese - BMI >40
Surgical Candidate
Prior surgical intervention
Psychiatric issues
Randomized to Bariatric
surgery or diet only
Beshay et al. Bogus Journal 2010
Statistical Analysis
Intention to Treat Analysis
LARD
Laparoscopic Antroplasty vs Restricted Diet
• 1000 patients
•500 randomized to surgery
•500 randomized to diet
• At the end of the study
•350 surgical patients underwent surgery
•250 diet patients underwent surgery
Beshay et al. Bogus Journal 2010
Statistical Analysis
Intent to Treat Analysis
LARD
Laparoscopic Antroplasty vs Restricted Diet

In this study 60% of surgical patients experienced a significant drop in BMI
while 30% of diet patients experienced a similar drop
AT
550 pt’s
500 randomized
to surgery
200 patients 250 patients
1000 patients
500 randomized
to diet
330 lost
weight
60%
IT
500 patients
300 x 0.6 +
200 x 0.3= 240
240/500=
48%
450 pt’s
500 patients
135 lost
weight
250 x 0.6 +
30%
250 x 0.3 =
Intent to treat analysis with large crossover favors the null
hypothesis.
45%
SPORT – Lumbar Stenosis
Conclusions






Lumbar stenosis is common among elderly patients and
is the number one cause of lumbar surgery in those
older than 65.
When treating neurogenic claudication or radicular
symptoms, patients will improve with operative or nonoperative strategies
There is significantly more improvement with surgical
intervention
Surgical intervention in this group of patients was safe
Beware of Intent to Treat Analysis in studies in which
there was a large amount of crossover.
Don’t underestimate the value of observational groups.
Surgical Treatment of Lumbar
Stenosis

‘Standard’ open laminectomy/medial
facetectomy/foraminotomy – unilateral or
bilateral
Surgical Treatment of Lumbar
Stenosis

Minimally invasive approach




Goal is to achieve same results with little
muscle disruption
Utilize a tubular retractor
Incision off the midline
Muscle splitting technique
Surgical Treatment of Lumbar
Stenosis

Minimally invasive approach
Surgical Treatment of Lumbar
Stenosis

Minimally invasive approach
Surgical Treatment of Lumbar
Stenosis

Minimally invasive approach
Surgical Treatment of Lumbar
Stenosis
Surgical Treatment of Lumbar
Stenosis
Surgical Treatment of Lumbar
Stenosis
Papavero et al. 2009
Surgical Treatment of Lumbar
Stenosis
Papavero et al. 2009
Surgical Treatment of Lumbar
Stenosis

Are minimally invasive techniques better?





In experienced hands same results as open
surgery
Less hospital stay – modest
Less post op narcotics
Less muscle disruption as assessed by CK and
follow-up MRI
? Less post op instability ? – Maybe..
Surgical Treatment of Lumbar
Stenosis

What about interspinous process spacers?
Surgical Treatment of Lumbar
Stenosis

What about interspinous process spacers?
Surgical Treatment of Lumbar
Stenosis

What about interspinous process spacers?

Do they work? Yes, in selected patients.
Kuchta et al. Eur Spine J 2009
Surgical Treatment of Lumbar
Stenosis

Extension Limiting Devices
Surgical Treatment of Lumbar
Stenosis

“Other”
Swei-Ming et al. JNS 2006
Questions
Thank You