Transcript Rational Use of DEXA - Alberta Medical Association

RATIONAL USE OF
DXA-BMD
A CASE BASED APPROACH - GREGORY KLINE
MD FRCPC
Disclosures (since 2009)
no relationships or conflicts to declare
Objectives
discussion of what BMD does and does not add to
“screening/diagnosis” in osteoporosis
discuss indications for and limitations of serial BMD
tests
review the related 2016 TOP Osteoporosis
guidelines in clinical context
most important! Learn to think about what BMD
might add BEFORE you order it.
Cases for discussion
Case 1: when to do the “first” bone density
Case 2: serial “screening” bone density
Bonus material
CASE 1:
PRINCIPLES,
NOT RULES
THE “FIRST” BMD
Mrs. B
Mrs. B is a 50 year old healthy woman, no
medications, no family history of hip fracture.
Height 152.6 cm, weight 61 kg
She is concerned about developing osteoporosis
and requests a “baseline” BMD
How do you respond and what reasons would you
give either for or against?
Thinking it through
What are we looking for with a “baseline” BMD?
remember: bone loss is universal with age - don’t
need a test to prove ageing.
What will you do if it shows T score > -2.0?
What will you do if it shows T score -2.3?
Thinking it through: if the
baseline BMD is “good”





A “baseline” measure implies that there will be a future measure which
may be different than baseline AND
The rate of change between baseline and repeat measure is of
relevance to management.
REALITY: the baseline BMD will simply reflect the time that has passed
since menopause
REALITY: 100% chance that the next BMD will be lower
REALITY: management decisions will depend on the next BMD....no
relevance to “baseline”
Thinking it through:
what if the baseline BMD is
“bad”?





What are the chances that a BMD, done now, in the absence of other risk
factors, will be so low that it actually warrants medical therapy to be
started right now?
Assume that medical therapy is warranted for a 10 year fracture risk of
20% or higher....
She would need to have a T-score of -4.5
The odds of that, in a healthy 50 year old < 1/10,000
And by the way, if she did have a T-score of -4.5 at age 50...she likely
has some kind of metabolic bone disease…..
GUIDANCE
know the paradigm
When we screen for osteoporosis:
we should NOT be screening to “find” low BMD
because many will not warrant treatment anyways
what is the point of screening for something that is a
natural process of ageing and affects 100% of people
rather, appropriate osteoporosis screening should:
be designed to find people who are at such high
fracture risk that drug intervention is warranted.
PRACTICAL MEANING
how to think in practice
the idea is to “find” people who have at least 20%
chance of fracture in the next 10 years
BMD is a tool that may help find some of those
people
EXAMPLES
exhibits
A) a 60 year old woman with polymyalgia rheumatica on
prednisone 10 mg daily x 9 months. Smoker, height 168
cm weight 62 kg.
B) a 70 year old woman with mild kyphosis and maternal
history of hip fracture. Height 165 cm, weight 59 kg.
C) a 59 year old female smoker with family history of hip
fracture. Height 165 cm, weight 73 kg.
D) 62 year old healthy woman with no risk factors, height
166 cm, weight 75 kg.
exhibits
A) a 60 year old woman with polymyalgia rheumatica on prednisone 10
mg daily x 9 months, previous wrist fracture. Smoker, height 168 cm
weight 62 kg. 10 year major OP # risk: 23% (FRAX score, no BMD)
B) a 70 year old woman with mild kyphosis and maternal history of hip
fracture. Height 165 cm, weight 59 kg. 10 year major OP # risk: 31%
(FRAX score, no BMD)
C) a 59 year old female smoker with family history of hip fracture. Height
165 cm, weight 73 kg. 10 year major OP # risk: 11.5% (FRAX score,
no BMD)
D) 62 year old healthy woman with no risk factors, height 166 cm, weight
75 kg. 10 year major OP # risk: 7% (FRAX score, no BMD)
exhibits
A) a 60 year old woman with polymyalgia rheumatica on prednisone 10 mg daily x 9 months,
previous wrist fracture. Smoker, height 168 cm weight 62 kg. 10 year major OP # risk: 23% (no
BMD)
• OP # risk with T-score -2.6: 26%
• OP # risk with T-score - 1.5: 18%
B) a 70 year old woman with mild kyphosis and maternal history of hip fracture. Height 165 cm,
weight 59 kg. 10 year major OP # risk: 31% (no BMD)
OP risk with T-score -2.6: 31%
OP # risk with T-score -1.5: 20%
C) a 59 year old female smoker with family history of hip fracture. Height 165 cm, weight 73 kg. 10
year major OP # risk: 11.5% (no BMD)
OP risk with T-score -2.6: 19%
OP risk with T-score -1.5: 12.5%
D) 62 year old healthy woman with no risk factors, height 166 cm, weight 75 kg. 10 year major OP #
risk: 7% (no BMD)
OP # risk with T-score - 2.6: 11%
OP # risk with T-score -1.5: 7%
exhibits
A) a 60 year old woman with polymyalgia rheumatica on prednisone 10 mg daily x 9 months, previous
wrist fracture. Smoker, height 168 cm weight 62 kg. 10 year major OP # risk: 23% (no BMD)
• OP # risk with T-score -2.6: 26%
• OP # risk with T-score - 1.5: 18%
BMD changes nothing!
B) a 70 year old woman with mild kyphosis and maternal history of hip fracture. Height 165 cm, weight
59 kg. 10 year major OP # risk: 31% (no BMD)
OP risk with T-score -2.6: 31%
OP # risk with T-score -1.5: 20%
BMD changes nothing!
C) a 59 year old female smoker with family history of hip fracture. Height 165 cm, weight 73 kg. 10 year
major OP # risk: 11.5% (no BMD)
OP risk with T-score -2.6: 19%
*BMD will be useful*
OP risk with T-score -1.5: 12.5%
D) 62 year old healthy woman with no risk factors, height 166 cm, weight 75 kg. 10 year major OP #
risk: 7% (no BMD)
OP # risk with T-score - 2.6: 11%
OP # risk with T-score -1.5: 7%
BMD changes nothing!
GUIDELINES IN
CONTEXT
TOP guidelines Feb 2016
suggest screening BMD for women age > 65
because advanced age is a major risk factor
suggest screening BMD for men or women >50
with risk factors (long list)
memorize the list OR
use FRAX score without BMD
www.topalbertadoctors.org/download/1907/Osteoporosis%20CPG.pdf
Case-based messages
Case 1: the first bone density
“baseline” bone density concept is of no medical use
and should be discouraged
bone density is an occasionally useful tool for fracture
risk estimation in patients who are not obviously at very
high (>20%) or very low (<10%) risk.
goal is to find people at high fracture risk
goal is not to find people with low BMD
Case 2 - serial “screening”
Mrs. C is 66 years old. She has no fracture risk
factors. She had a screening BMD done last year,
showing femoral neck T-score -1.7. No specific
therapy was started. She asks if she should have a
repeat BMD this year to see if it has changed.
Will it have changed?
What is the chance she will have crossed a treatment
threshold on the basis of bone loss in the past year?
GUIDANCE
interpretation
a BMD is done (“rightly” or “wrongly”) and the
patient is far away from an interventional threshold
either because
They lack other risk factors OR
Their current risk factors + BMD still puts them at low
risk
It will take somewhere between 2 and 15 years for even
10% of such persons to see enough BMD loss to actually
cross an interventional threshold (all other risks being
equal)

If
PRACTICAL MEANING
practical meaning

The value to annual or biannual BMD testing (all
other risks being constant) is virtually zero since
overall fracture risks (and therefore, distance to
treatment threshold) changes minimally in less
than 3-5 years.
EXAMPLES
long term monitoring?
66 year old woman, height 170 cm, wt 73 kg, no
risks, screening BMD T-score -1.7 OP # risk 8.6%
67 years old, BMD T-score -1.9 OP # risk 9.5%
68 years old, BMD T-score -1.9 OP # risk 9.8%
69 years old, BMD T-score -2.2 OP # risk 11.4%
70 years old, BMD T-score -2.3 OP # risk 12.1%
71 years old, BMD T-score -2.6 OP # risk 14.3%
how much did it add?
66 year old woman, height 170 cm, wt 73 kg, no
risks, screening BMD T-score -1.7 OP # risk 8.6%
67 years old, BMD T-score -1.9 OP # riskNOT
9.5%
VALUABLE
68 years old, BMD T-score -1.9 OP # riskNOT
9.8%
VALUABLE
69 years old, BMD T-score -2.2 OP # riskNOT
11.4%
VALUABLE
70 years old, BMD T-score -2.3 OP # riskNOT
12.1%
VALUABLE
71 years old, BMD T-score -2.6 OP # riskNOT
14.3%
VALUABLE?
Thinking about what you might find first…
65 year old woman, height 170 cm, wt 73 kg, no risks, screening BMD
T-score -1.7 OP # risk 8.6%
Age
other risks
66
67
maternal hip #
reported
risk (no
BMD)
risk (T-2.0)
risk (T-3.3)
8.6%
9.6%
17.5%*
15.1%
highly unlikely!
17.2%*
29%*
most likely but
adds little
useful if the first
BMD
68
15.3%
17.2%
29.8%
69
15.5%
17.1%
30.8%
how much does BMD
add…?
if a prior BMD exists from the past 3-5 years, a repeat measure
rarely adds to your updated FRAX assessment (use the old BMD
result!)
if over age 50 and a new major risk factor acquired, consider new
BMD if:
no prior BMD within 5 years AND
intervention not already obviously indicated by risk factors
alone
General principle: before ordering BMD, consider whether it is
likely to really change anything; FRAX calculation may help.
GUIDELINES IN
CONTEXT
TOP guidelines 2016
repeated BMD testing alone provides little value to
decision-making in osteoporosis
www.topalbertadoctors.org/download/1907/Osteoporosis%20CPG.pdf
Case-based messages (2)
Case 2: serial “screening” bone density
intervals for repeat screening depend upon fracture risk at
last screening visit, not necessarily time-standardized.
fracture risk and BMD change very slowly
annual BMD almost never useful unless patient has also
acquired new major risk factors as well
since goal is to find people who are at intervention
threshold, serial screening need not necessarily be BMDdriven
BONUS MATERIAL
BMD measures “on
therapy”
BMD “change” in either direction not really
predictive of anything while on therapy
BMD / scoring systems on therapy not useful for
fracture risk estimation
BMD - driven changes to therapy (medication
switching) not supported by evidence
Undetermined role of BMD monitoring around
bisphosphonate “holiday”
Monitoring Osteoporosis Therapy With Bone Densitometry
Misleading Changes and Regression to the Mean
JAMA. 2000;283(10):1318-1321. doi:10.1001/jama.283.10.1318.
examples
patient on alendronate, follow up BMD shows 2% decrease. Action:
nothing.
patient on alendronate, follow up BMD shows 3% decrease. Action:
probably obligated to repeat next year…..BMD no change….action:
nothing.
patient on alendronate, follow up BMD shows no change. Action: nothing.
patient on alendronate, follow up BMD shows 10% decrease.
Action:investigate……but very rare situation!
Controversy: is serial BMD broadly warranted to find very rare cases of
occult metabolic bone disease / cancer?
exhibits (2)
patient on bisphosphonate has repeat BMD (no
change), report says “high risk of fracture”
is this true? is this a surprise? does this change
anything?
when does repeat BMD on
therapy help?
little evidence to answer this question
possibly if suspected malabsorption of oral
bisphosphonate
possibly during bisphosphonate holiday
and therefore, possibly at the “end” of
bisphosphonate treatment, prior to bisphosphonate
holiday
SUMMARY
BMD uses
baseline BMD at menopause - discourage
choice to do BMD should FOLLOW, not precede, a review of risk
factors
BMD is most useful in cases of indeterminate risk
…which usually means people with fracture risk factors or
advanced age without risk factors
serial BMD in untreated patients should likely be 3 - 5 years apart
in order to be useful
serial BMD in treated patients is of very low value in most cases