Key words
glucocorticoid-induced osteoporosis - umbrella review - efficacy - safety - pharmacotherapy
Introduction
Glucocorticoids (GCs) are widely used in clinical work due to their effectiveness in
achieving immunosuppression, an anti-inflammatory response, and other
pharmacological effects [1]
[2]; in fact, GCs are used to treat many types
of diseases. However, long-term use of GCs can induce osteoblast regulation,
increase osteoclast activation, and reduce calcium absorption by the digestive
system [1]
[2]
[3]. Glucocorticoid-induced
osteoporosis (GIOP), a metabolic bone disease caused by endogenous or exogenous GCs,
is the most common type of secondary osteoporosis [4]. Studies have shown that approximately 1% of the population in
the United States requires the long-term use of GCs [5]. The incidence of osteoporotic fractures in patients with long-term
use of GCs at doses beyond the physiological levels will reach
30–50%, and the risk of refracture after the initial fracture will
increase significantly [6]. Existing
epidemiological data show that continuous oral administration of GCs for 3–6
months (or longer), high-dose inhaled GCs or intermittent use of oral GCs can lead
to decreased bone density and an increased fracture risk [7]
[8].
The incidence rate of GIOP is high, making it the third most common form of
osteoporosis, and this incidence is second only to that of postmenopausal
osteoporosis and senile osteoporosis [9].
Therefore, treatment for GIOP requires the attention of patients and medical
professionals.
At present, the most commonly used therapeutic drugs for GIOP are calcium (Ca) and
vitamin D (Vit D); bisphosphonates (BPs), teriparatide, and other drugs are also
used to treat GIOP [10]
[11]. However, there is a lack of advanced
evidence-based studies on GIOP drugs, and this deficiency is not conducive to the
application of clinical drugs. According to the American College of Rheumatology
(ACR), the prevention and treatment guidelines for GIOP show that evidence on
existing drugs used to treat GIOP is limited; therefore, the application of
anti-GIOP drugs has specific usage conditions [12]. In recent years, clinical randomized controlled trials (RCTs) and
systematic reviews and/or meta-analyses (SRs) of drug treatment for GIOP
have been studied and disclosed, thus confirming that there are high-level
evidence-based studies on drug treatment for GIOP. Umbrella reviews, also known as
systematic reviews of systematic reviews, systematic reviews of meta-analyses, and
overviews of reviews [13], provide healthcare
decision-makers with current comprehensive evidence on specific issues by
systematically retrieving SRs and extracting, analyzing, and summarizing the results
of the existing evidence [14]. In this
context, we reviewed published SRs of RCTs for inclusion in this umbrella review to
further evaluate the efficacy of pharmacological interventions for GIOP. Another
objective of this study is to provide guidance for improving the clinical study
design, a reference for the clinical application of drug therapy for GIOP and a plan
for clinical guidelines.
Materials and Methods
Inclusion and exclusion criteria
The following are the inclusion criteria of this umbrella review: 1) the included
studies were SRs of RCTs; 2) the cases included in the SR were osteoporosis
secondary to taking GCs, and there was no restriction on the duration of the
primary disease or the dose of GC; 3) the experimental group (EG) was treated
with any drug, combined with other drugs on the basis of the control group, or
evaluated for a certain class of drugs (such as BPs); 4) the control group was a
placebo, blank group, positive drug or basic drug treatment (such as Ca and a
vitamin); and 5) the main outcome measures were the bone mineral density (BMD)
change rate. Secondary outcome measures were risk of infection, adverse events
(AEs), risk of a new nontraumatic fracture (NTF), incidence of vertebral (VF) or
nonvertebral fractures (NVF), N-terminal propeptide of type I collagen (PINP),
and C-telopeptide of type I collagen (CTX).
The exclusion criteria were as follows: 1) narrative reviews, 2) network
meta-analyses, 3) animal experiments, 4) repeated published literature, and 5)
literature published in a language other than English.
Retrieval strategy
We searched PubMed, Embase, and the Cochrane Library for SRs of drug therapy for
GIOP. We searched for literature published from database inception to November
2022. In addition, we manually searched the references of the included studies
to supplement SRs that might meet the inclusion criteria. The literature was
searched by using a combination of subject words and free words, and the
retrieval strategy was adjusted according to the retrieval characteristics of
each database. The key words included glucocorticoids, osteoporosis,
glucocorticosteroids, glucocorticoid-induced osteoporosis, meta-analysis and
systematic review. The retrieval formulas of the above three databases are shown
in Supplementary Material 1.
Literature screening and data extraction
Two researchers (HL and JZ) independently read the titles and abstracts as well
as the full text of the literature to determine whether the publications met the
inclusion criteria. If there was any disagreement, it was resolved through
consultation with the third researcher (TT). The data that were collected
included the author, the year of publication, the number of included studies,
the number of samples, the intervention measures, the quality evaluation methods
of the included studies, and the outcome indicators. If there were multiple SRs
focused on the same subject or drug therapy, one systematic review was reserved
for subsequent analysis according to the principle of the highest quality of SR
methodology and the largest number of RCTs included.
Methodology and evidence quality evaluation
We used A Measurement Tool to Assess Systematic Reviews (AMSTAR-2) to evaluate
the methodological quality of the included SRs [15]. The AMSTAR-2 includes 16 items (Supplementary Material
2), of which items 2, 4, 7, 9, 11, 13, and 15 are key items and the
remaining items are non-key items [15].
According to the AMSTAR-2 evaluation standard, the methodological quality of
each SR can be evaluated as high, moderate, low and critically low quality.
The GRADE (Grades of Recommendations Assessment, Development and Evaluation)
grading system was used to evaluate the quality of evidence for the outcome
indicators in the SR [16]. The factors
that reduce the level of evidence are divided into five dimensions: limitation,
inconsistency, indirection, accuracy and publication bias. According to the
degree of compliance with the degradation factors, the evidence level of the
outcome indicators can be rated as high, moderate, low and very low. To help
readers understand our research conclusions, we generated an evidence map
according to the comparative results of the combined effect values and GRADE
score.
Statistical method
We conducted a descriptive analysis to summarize the evidence results of the
included SRs. Based on the primary and secondary outcome measures, the efficacy
and safety outcomes of pharmacological interventions for GIOP were
re-evaluated.
Results
Retrieval results of literature
After double checking and reading the title and abstract of the results, we
included 38 SRs for full-text reading. After excluding a narrative review, a
network meta-analysis, and animal experiments, 6 SRs [17]
[18]
[19]
[20]
[21]
[22] of pharmacological
interventions for GIOP were ultimately included. The list of excluded documents
and reasons are shown in Supplementary Material 3. The literature
screening process and results are shown in [Fig. 1].
Fig. 1 Flow diagram of the umbrella review.
Basic characteristics of the included SRs
Six SRs [17]
[18]
[19]
[20]
[21]
[22] involving 59 RCTs with 7225 patients exhibiting GIOP were
included in this umbrella review. All the included patients were diagnosed with
GIOP. SRs published between 2010 and 2022 were included. The drug therapies
covered in this umbrella review included BPs, Ca+Vit D, alendronate,
denosumab and teriparatide. The specific characteristics of the included SRs are
shown in [Table 1].
Table 1 Characteristics of the systematic reviews and
meta-analyses included in the umbrella review.
|
Study
|
No. of RCTs (Sample size)
|
Participants
|
Descriptions of Interventions
|
Methodological quality evaluation tool
|
GRADE evaluation
|
Outcomes assessed
|
|
EG
|
CG
|
|
CS Allen 2016 [17]
|
27 (3075)
|
Adults taking a mean steroid dose of
5.0 mg/day or more
|
Standard-dose BPs
|
Low-dose BPs
|
ROB
|
Yes
|
Percent change in BMD
|
|
J Homik 2010 [18]
|
5 (274)
|
Patients (older than age of 18) taking systemic
corticosteroids
|
Ca and Vit D
|
Ca alone or placebo
|
Jadad scores
|
No
|
Percent change in BMD, fracture incidence
|
|
ZM Liu 2022 [19]
|
5 (1460)
|
Patients were at least 21 years old
|
Alendronate
|
Teriparatide
|
ROB
|
No
|
Percent change in BMD, fracture incidence, AE, changes in
turnover markers
|
|
YK Wang 2018 [20]
|
10 (1002)
|
Adult patients with GIOP taking alendronate for at least 6
months.
|
Alendronate plus EG
|
Ca and Vit D
|
Jadad scores
|
No
|
Percent change in BMD, fracture incidence, AE
|
|
J Wang 2019 [21]
|
9 (545)
|
Eastern Asians
|
BPs Alone
|
Vit D Alone or a Combination
|
ROB
|
No
|
Percent change in BMD and turnover markers
|
|
ZA Yanbeiy 2019 [22]
|
3 (869)
|
Subjects taking systemic glucocorticoid therapy
|
Denosumab
|
BPs
|
No
|
No
|
Percent change in BMD, fracture incidence, infection
|
RCTs: Randomized controlled trials; EG: Experimental Group; CG: Control
Group; BPs: Bisphosphonates; Ca: Calcium; Vit D: Vitamin D; ROB:
Cochrane Risk of Bias Tool; BMD: Bone Mineral Density; AE: Adverse
Events.
Methodological quality evaluation
According to the evaluation results of the AMSTAR-2 tool, the 6 SRs included in
this review included 2 high-quality [17]
[18], 2 low-quality [20]
[21], and 2 critically low-quality SRs [19]
[22]. The specific details of the methodological quality evaluation
are shown in [Table 2].
Table 2 AMSTAR scoring results of the included systematic
reviews and meta-analysis.
|
Study
|
Q1
|
Q2*
|
Q3
|
Q4*
|
Q5
|
Q6
|
Q7*
|
Q8
|
Q9*
|
Q10
|
Q11*
|
Q12
|
Q13*
|
Q14
|
Q15*
|
Q16
|
Ranking of quality
|
|
CS Allen 2016 [17]
|
Y
|
Y
|
Y
|
Y
|
Y
|
Y
|
Y
|
Y
|
Y
|
Y
|
Y
|
Y
|
Y
|
Y
|
Y
|
Y
|
High
|
|
J Homik 2010 [18]
|
Y
|
Y
|
Y
|
Y
|
Y
|
Y
|
Y
|
Y
|
Y
|
N
|
Y
|
Y
|
Y
|
Y
|
Y
|
Y
|
High
|
|
ZM Liu 2022 [19]
|
Y
|
N
|
Y
|
Y
|
Y
|
Y
|
PY
|
PY
|
Y
|
N
|
Y
|
PY
|
N
|
N
|
N
|
Y
|
Critically Low
|
|
YK Wang 2018 [20]
|
Y
|
N
|
Y
|
Y
|
Y
|
Y
|
PY
|
Y
|
PY
|
N
|
Y
|
Y
|
PY
|
PY
|
PY
|
Y
|
Low
|
|
J Wang 2019 [21]
|
Y
|
N
|
Y
|
Y
|
Y
|
Y
|
PY
|
PY
|
Y
|
N
|
Y
|
Y
|
Y
|
Y
|
PY
|
Y
|
Low
|
|
ZA Yanbeiy 2019 [22]
|
Y
|
Y
|
Y
|
Y
|
Y
|
Y
|
PY
|
Y
|
Y
|
N
|
Y
|
Y
|
N
|
N
|
N
|
Y
|
Critically Low
|
* Key entry; PY: Partial yes; Y: Yes; N: No. The
specific contents of 16 items are shown in Supplementary Material
2.
Results of evidence quality evaluation of outcome indicators
In this umbrella review, we evaluated 46 quality studies of 11 outcome indicators
([Table 3] and [Table 4]), among which the outcome
indicators mainly included the BMD, fracture incidence, bone turnover markers
and AEs. According to the GRADE evaluation criteria, this review included 3
high-level studies, 20 moderate-level studies, 15 low-level studies, and 8 very
low-level studies. The evidence map of pharmacological interventions for GIOP is
shown in [Fig. 2].
Fig. 2 Heat map of pharmacological Interventions on GIOP. BPs:
Bisphosphonates; Ca: Calcium; Vit D: Vitamin D; BMD: Bone Mineral
Density; AE: Adverse Events; LSBMD: BMD of Lumbar Spine; FNBMD: BMD of
femoral neck; THBMD: BMD of total hip; DRBMD: BMD of distal radius; NTF:
Nontraumatic fracture; VF: Vertebral fractures; NVF: Nonvertebral
fractures; PINP: N-terminal propeptide of type I collagen; CTX:
C-telopeptide of type I collagen.
Table 3 GRADE quality of evidence score for outcomes
reported in the systematic reviews included in the umbrella review
of pharmacological interventions for GIOP (primary
outcomes).
|
Outcome
|
Intervention and comparator
|
Follow-up
|
Effect Size (95% CI)
|
I2 (%)
|
p
|
Risk of bias
|
Inconsistency
|
Indirectness
|
Imprecision
|
Publication bias
|
GRADE quality
|
|
LSBMD
|
Standard-dose vs. Low-does BPs
|
12 months
|
MD: 0.95% (0.37% to 1.53%)
|
0
|
0.0014
|
No
|
No
|
No
|
No
|
Serious
|
Moderate
|
|
Ca + Vit D vs. Ca (or Placebo)
|
12 months
|
MD: 2.63% (0.74% to 4.53%)
|
0
|
0.0065
|
No
|
No
|
No
|
No
|
Serious
|
Moderate
|
|
Teriparatide vs. Alendronate
|
6 months
|
SMD: 0.30% (0.19% to 0.42%)
|
0
|
<0.001
|
No
|
No
|
No
|
No
|
No
|
High
|
|
12 months
|
SMD: 0.48% (0.36% to 0.60%)
|
45
|
<0.001
|
No
|
No
|
No
|
No
|
No
|
High
|
|
18 months
|
SMD: 0.53% (0.42% to 0.64%)
|
48
|
<0.001
|
No
|
No
|
No
|
No
|
No
|
High
|
|
Alendronate + Ca + Vit D vs. Ca + Vit
D
|
6 months
|
SMD: 0.67% (–0.02% to
1.36%)
|
81
|
0.06
|
Serious
|
Serious
|
No
|
No
|
Serious
|
Very Low
|
|
12 months
|
SMD: 0.83% (0.58% to 1.08%)
|
54
|
<0.001
|
Serious
|
Serious
|
No
|
No
|
Serious
|
Very Low
|
|
24 months
|
SMD: 0.80% (0.49% to 1.10%)
|
38
|
<0.001
|
Serious
|
No
|
No
|
No
|
Serious
|
Low
|
|
BPs vs. Vit D
|
Unspecified
|
MD: 4.11% (3.11% to 5.11%)
|
34
|
<0.001
|
No
|
No
|
No
|
No
|
Serious
|
Moderate
|
|
BPs vs. Vit D + BPs
|
Unspecified
|
MD: –2.09% (–3.72% to
–0.46%)
|
54
|
0.01
|
No
|
Serious
|
No
|
No
|
Serious
|
Low
|
|
Vit D vs. Vit D + BPs
|
Unspecified
|
MD: –6.83% (–8.63% to
–5.03%)
|
53
|
<0.001
|
No
|
Serious
|
No
|
Serious
|
Serious
|
Very Low
|
|
Risedronate vs. Vit D
|
Unspecified
|
MD: 4.00% (2.79% to 5.22%)
|
0
|
<0.001
|
No
|
No
|
No
|
No
|
Serious
|
Moderate
|
|
Alendronate vs. Vit D
|
Unspecified
|
MD: 4.49% (2.91% to 6.06%)
|
0
|
<0.001
|
No
|
No
|
No
|
No
|
Serious
|
Moderate
|
|
Ibandronate vs. Vit D
|
Unspecified
|
MD: 3.77% (0.05% to 7.49%)
|
88
|
0.05
|
No
|
Serious
|
No
|
Serious
|
Serious
|
Very Low
|
|
Denosumab vs. BPs
|
Unspecified
|
MD: 2.32% (1.72% to 2.91%)
|
0
|
<0.001
|
No
|
No
|
No
|
No
|
Serious
|
Moderate
|
|
FNBMD
|
Standard-dose vs. Low-does BPs
|
12 months
|
MD: 0.74% (–-0.42% to
1.90%)
|
54
|
0.21
|
No
|
Serious
|
No
|
No
|
Serious
|
Low
|
|
Ca + Vit D vs. Ca (or Placebo)
|
12 months
|
MD: 0.37% (–1.09% to
1.83%)
|
0
|
0.62
|
No
|
No
|
No
|
No
|
Serious
|
Moderate
|
|
Teriparatide vs. Alendronate
|
18 months
|
SMD: 0.17% (0.05% to 0.29%)
|
0
|
0.006
|
No
|
No
|
No
|
No
|
Serious
|
Moderate
|
|
Alendronate + Ca + Vit D vs. Ca + Vit
D
|
6 months
|
SMD: 0.94% (0.64% to 1.24%)
|
0
|
<0.001
|
Serious
|
No
|
No
|
No
|
Serious
|
Low
|
|
12 months
|
SMD: 0.29% (–0.28% to
0.87%)
|
92
|
0.32
|
Serious
|
Serious
|
No
|
No
|
Serious
|
Very Low
|
|
24 months
|
SMD: 0.60% (0.06% to 1.13%)
|
80
|
0.03
|
Serious
|
Serious
|
No
|
No
|
Serious
|
Very Low
|
|
BPs vs. Vit D
|
Unspecified
|
MD: –28.53% (–34.56% to
–22.50%)
|
0
|
<0.001
|
No
|
No
|
No
|
Serious
|
Serious
|
Low
|
|
BPs vs. Vit D + BPs
|
Unspecified
|
MD: 1.96% (–6.26% to
10.18%)
|
0
|
0.64
|
No
|
No
|
No
|
Serious
|
Serious
|
Low
|
|
Vit D vs. Vit D + BPs
|
Unspecified
|
MD: 36.20% (26.87% to 45.52%)
|
0
|
<0.001
|
No
|
No
|
No
|
Serious
|
Serious
|
Low
|
|
Risedronate vs. Vit D
|
Unspecified
|
MD: 2.20% (0.56% to 3.84%)
|
2
|
0.008
|
No
|
No
|
No
|
No
|
Serious
|
Moderate
|
|
Alendronate vs. Vit D
|
Unspecified
|
MD: 1.19% (–0.56% to
2.95%)
|
0
|
0.18
|
No
|
No
|
No
|
No
|
Serious
|
Moderate
|
|
Denosumab vs. BPs
|
Unspecified
|
MD: 1.35% (–1.59% to
4.30%)
|
46
|
0.37
|
No
|
No
|
No
|
No
|
Serious
|
Moderate
|
|
THBMD
|
Teriparatide vs. Alendronate
|
18 months
|
SMD: 0.17% (0.05% to 0.28%)
|
0
|
0.004
|
No
|
No
|
No
|
No
|
Serious
|
Moderate
|
|
Denosumab vs. BPs
|
Unspecified
|
MD: 1.52% (1.10% to 1.94%)
|
0
|
<0.001
|
No
|
No
|
No
|
No
|
Serious
|
Moderate
|
|
DRBMD
|
Ca + Vit D vs. Ca (or Placebo)
|
12 months
|
MD: 2.49% (0.62% to 4.36%)
|
54
|
0.0092
|
No
|
Serious
|
No
|
No
|
Serious
|
Low
|
BPs: Bisphosphonates; Ca: Calcium; Vit D: Vitamin D; BMD: Bone mineral
density; LSBMD: BMD of lumbar spine; FNBMD: BMD of femoral neck; THBMD:
BMD of total Hip; DRBMD: BMD of distal radius; MD: Weighted mean
difference; SMD: Standard mean difference; CI: Confidence intervals.
Table 4 GRADE quality of evidence score for outcomes
reported in the systematic reviews included in the umbrella review
of pharmacological interventions for GIOP (secondary
outcomes).
|
Outcome
|
Intervention and comparator
|
Follow-up
|
Effect Size (95% CI)
|
I2 (%)
|
p
|
Risk of bias
|
Inconsistency
|
Indirectness
|
Imprecision
|
Publication bias
|
GRADE quality
|
|
Risk of infection
|
Denosumab vs. BPs
|
Unspecified
|
RR: 2.16 (0.38 to 12.34)
|
66
|
0.39
|
No
|
Serious
|
No
|
Serious
|
Serious
|
Very Low
|
|
AE
|
Alendronate+Ca+Vit D vs.
Ca+Vit D
|
Unspecified
|
OR: 1.04 (0.72 to 1.51)
|
0
|
0.84
|
Serious
|
No
|
No
|
No
|
Serious
|
Low
|
|
Teriparatide vs. Alendronate
|
Unspecified
|
RR: 1.02 (0.89 to 1.18)
|
0
|
0.76
|
No
|
No
|
No
|
No
|
Serious
|
Moderate
|
|
Risk of new non-traumatic fracture
|
Ca+Vit D vs. Ca (or Placebo)
|
Unspecified
|
OR: 0.55 (0.12 to 2.44)
|
0
|
0.43
|
No
|
No
|
No
|
No
|
Serious
|
Moderate
|
|
Denosumab vs. BPs
|
Unspecified
|
RR: 1.16 (0.68 to 1.98)
|
0
|
0.59
|
No
|
No
|
No
|
No
|
Serious
|
Moderate
|
|
Incidence of VF
|
Teriparatide vs. Alendronate
|
Unspecified
|
RR: 0.13 (0.05 to 0.34)
|
0
|
<0.001
|
No
|
No
|
No
|
No
|
Serious
|
Moderate
|
|
Alendronate+Ca+Vit D vs.
Ca+Vit D
|
Unspecified
|
OR: 0.46 (0.21 to 1.02)
|
0
|
0.06
|
Serious
|
No
|
No
|
No
|
Serious
|
Low
|
|
Incidence of NVF
|
Teriparatide vs. Alendronate
|
Unspecified
|
RR: 1.28 (0.81 to 2.02)
|
0
|
0.29
|
No
|
No
|
No
|
No
|
Serious
|
Moderate
|
|
Alendronate+Ca+Vit D vs. Ca+Vit D
|
Unspecified
|
OR: 1.48 (0.50 to 4.37)
|
0
|
0.47
|
Serious
|
No
|
No
|
No
|
Serious
|
Low
|
|
PINP
|
Teriparatide vs. Alendronate
|
1 months
|
SMD: 3.51% (3.15% to 3.87%)
|
0
|
<0.001
|
No
|
No
|
No
|
No
|
Serious
|
Moderate
|
|
6 months
|
SMD: 5.02% (3.35% to 6.69%)
|
91
|
<0.001
|
No
|
Serious
|
No
|
No
|
Serious
|
Low
|
|
18 months
|
SMD: 4.97% (4.48% to 5.46%)
|
0
|
<0.001
|
No
|
No
|
No
|
No
|
Serious
|
Moderate
|
|
CTX
|
Teriparatide vs. Alendronate
|
1 months
|
SMD: 4.83% (2.87% to 6.79%)
|
96
|
<0.001
|
No
|
Serious
|
No
|
No
|
Serious
|
Low
|
|
6 months
|
SMD: 5.77% (2.19% to 9.34%)
|
97
|
0.002
|
No
|
Serious
|
No
|
Serious
|
Serious
|
Very Low
|
|
18 months
|
SMD: 5.33% (4.23% to 6.43%)
|
80
|
<0.001
|
No
|
Serious
|
No
|
No
|
Serious
|
Low
|
|
BPs vs. Vit D
|
Unspecified
|
MD: –72.27% (–85.19% to
–59.34%)
|
0
|
<0.001
|
No
|
No
|
No
|
Serious
|
Serious
|
Low
|
BPs: Bisphosphonates; Ca: Calcium; Vit D: Vitamin D; NTF: Non-traumatic
fracture; AE: Adverse events; VF: Vertebral fractures; NVF:
Non-vertebral fractures; PINP: N-terminal propeptide of type I collagen;
CTX: C-telopeptide of type I collagen; MD: Weighted mean difference;
SMD: Standard mean difference; OR: Odds ratio; RR: Risk ratio; CI:
Confidence intervals.
Effects of pharmacological treatments for GIOP
Primary outcome
BMD of the lumbar spine (LSBMD)
LSBMD was reported in six SRs [17]
[18]
[19]
[20]
[21]
[22]. Compared with low-dose BPs, standard-dose BPs improved the
LSBMD (MD: 0.95%, 95% CI: 0.37% to 1.53%, p
<0.001). The doses of different classes of BPs are shown in
Supplementary Material 4. Compared with alendronate, teriparatide
had better efficacy in increasing LSBMD, and its evidence level is high.
Compared with Vit D alone, BPs, risedronate and alendronate also showed
better effects. There was no significant difference between ibandronate and
Vit D in increasing LSBMD (MD: 3.77%, 95% CI: 0.05%
to 7.49%, p=0.05). Compared with Vit D alone, the combined
application of BPs was more effective in increasing LSBMD. Compared with Ca
(or placebo), Ca+Vit D was more effective in increasing LSBMD.
Compared with BPs, denosumab had better clinical efficacy for increasing
LSBMD (MD: 2.32%, 95% CI: 1.72% to 2.91%, p
<0.001) ([Table 3] and [Fig. 2]).
BMD of the femoral neck (FNBMD)
In terms of increasing FNBMD, risedronate was more effective in increasing
FNBMD than Vit D alone (MD: 2.20%, 95% CI: 0.56% to
3.84%, p=0.008). Compared with alendronate, teriparatide had
better efficacy in increasing FNBMD. Vit D had better efficacy than BPs in
increasing FNBMD. Compared with the combined application of BPs, Vit D alone
was more effective in increasing the efficacy of FNBMD (MD: 36.20%,
95% CI: 26.87% to 45.52%, p <0.001).
Compared with Vit D alone, risedronate was more effective in increasing
FNBMD.
BMD of total hip (THBMD)
THBMD was reported in a total of 2 SRs [19]
[22]. The existing
evidence indicates that teriparatide has better efficacy in increasing THBMD
than alendronate (SMD: 0.17%, 95% CI: 0.05% to
0.28%, p=0.004). Denosumab was more effective in increasing
THBMD than BPs (MD: 1.52%, 95% CI: 1.10% to
1.94%), and the difference was statistically significant (p
<0.001).
BMD of the distal radius (DRBMD)
One SR showed changes in DRBMD [18].
Compared with Ca (or placebo), Ca+Vit D significantly increased
DRBMD (MD: 2.49%, 95% CI: 0.62% to 4.36%),
and the difference was statistically significant (p=0.0092).
Secondary outcome
Risk of infection
Compared with BPs, denosumab in GIOP patients did not increase the risk of
infection (RR: 2.16, 95% CI: 0.38 to 12.34), and the difference was
not statistically significant (p=0.39) ([Table 4] and [Fig. 2]).
AEs
AEs were reported in two SRs [19]
[20]. The existing evidence indicates
that the combination of alendronate with Ca+Vit D does not
significantly increase the incidence of AE compared with Ca+Vit D
treatment alone (OR: 1.04, 95% CI: 0.72 to 1.51, p=0.84).
There was no significant difference in the incidence of AE between
teriparatide and alendronate (RR: 1.02, 95% CI: 0.89 to 1.18,
p=0.76).
NTFs
There was no significant difference in the incidence of new NTFs between
denosumab and BPs (RR: 1.16, 95% CI: 0.68 to 1.98, p=0.59).
Compared with Ca (or placebo), Ca+Vit D did not significantly
increase or decrease the incidence of new NTFs (OR: 0.55, 95% CI:
0.12 to 2.44, p=0.43).
Incidence of VFs
In terms of reducing the incidence of VFs, teriparatide significantly reduced
the risk of fracture compared with alendronate (RR: 0.13, 95% CI:
0.05 to 0.34), and the difference was statistically significant (p
<0.001). There was no significant difference in the application of
alendronate whether combined or not with Ca+Vit D (OR: 0.46,
95% CI: 0.21 to 1.02, p=0.06).
Incidence of NVFs
There was no significant difference in the incidence of NVFs between
teriparatide and alendronate (RR: 1.28, 95, 95% CI: 0.81 to 2.02,
p=0.29). There was no significant difference between Ca+Vit
D and alendronate+Ca+Vit D (OR: 1.48, 95% CI: 0.50
to 4.37, p=0.47).
PINP
After 1 (SMD: 3.51%, 95% CI: 3.15% to 3.87%),
6 (SMD: 5.02%, 95% CI: 3.35% to 6.69%), and
18 (SMD: 4.97%, 95% CI: 4.48% to 5.46%)
months of follow-up, teriparatide was more effective in increasing PINP
levels than alendronate, and the difference was statistically significant (p
<0.001).
CTX
In terms of the influence on CTX, teriparatide was more effective in
increasing the content in serum than alendronate, and the difference was
statistically significant. Compared with Vit D, BPs reduced the level of CTX
in serum (MD: –72.27% 95% CI:
–85.19% to –59.34%), and the difference was
statistically significant (p <0.001).
Discussion
In this umbrella review, we evaluated 6 SRs of pharmacological interventions for
GIOP, including calcium, Vit D, BPs, denosumab, teriparatide and their combined
applications, which provided a stronger evidence-based foundation for us to further
understand the efficacy of drug therapy for GIOP. In combination with the GIOP
treatment guidelines published by the ACR [12], we found that due to the lack of a sufficient evidence-based study, the
recommendation strength of many drug applications was low, or the application of
drugs was restricted by certain conditions. In this study, we have summarized the
latest SRs on drug therapy for GIOP, which can provide the latest and best
evidence-based recommendation for patients and medical personnel to select drugs for
GIOP. The findings of this study are the latest supporting reference and can be used
to help revise the guidelines. In addition, due to the limitation of the level of
clinical evidence, we recommend that users carefully consider low-level and very
low-level evidence in this umbrella review or select appropriate drug prescriptions
according to the comorbidities, advantages and disadvantages of GIOP patients.
In this umbrella review, we found that many drug treatments, such as standard-dose
BPs, Ca+Vit D, teriparatide, alendronate+Ca+Vit D, BPs, Vit
D+BPs, risedronate, alendronate, and denosumab, showed better efficacy for
increasing LSBMD compared with that in the control group. Notably, since the control
groups included in this umbrella review were all positive drug controls, users need
to choose the best drug prescription according to the corresponding reference drug
and the patient’s tolerance to the drug when selecting the above single-drug
or combination therapies. In terms of dose application of BPs, our study showed that
low-dose BPs were not more effective in increasing BMD than standard-dose BPs [17]. However, there was no significant
difference in the increase in FNBMD between the two doses. We believe that these
findings may be due to the different responses achieved with different doses at
different anatomical sites and to the fact that each site has a different blood
supply [23]
[24]. Drug metabolism may be affected because the blood flow in the lumbar
spine is rich and the blood flow in the total hip joint and the femoral neck is poor
[23]
[24]. Therefore, the effects of higher doses of BPs on BMD of the total
hip and femoral neck deserve further study, but the effects of higher doses of BPs
on metabolic organ function should also be observed. A clinical study with a
follow-up time of 16 weeks showed that alendronate combined with Vit D could
significantly improve osteoporosis without obvious side effects [25]. We found that compared with the
application of Vit D or BPs alone, Vit D+BPs had better efficacy in
increasing LSBMD, which suggests that the combination of Vit D and BPs is an obvious
option for the treatment of lumbar osteoporosis in GIOP patients, rather than the
application of Vit D or BPs alone. However, the patient’s tolerance to the
combination should also be considered.
In terms of improving FNBMD, teriparatide, alendronate+Ca+Vit D, Vit
D and risedronate all have better effects on increasing BMD. We found that the
application of teriparatide has a better impact on increasing FNBMD than alendronate
by synthesizing the existing evidence. In addition, we believe that the combined
application of teriparatide and alendronate is not recommended because bone
formation markers such as osteocalcin can be significantly decreased after the
application of alendronate, which will reduce the role of teriparatide in promoting
bone formation [26]. Therefore, while
considering the severity of osteoporosis in the femoral neck of GIOP patients, if
the patients have good tolerance to teriparatide and alendronate, there is moderate
evidence that supports the recommendation that teriparatide be selected
preferentially. Valenti et al. found that risedronate can affect bone metabolism by
upregulating the expression of cyclooxygenase-2 (COX-2) [27], and the inhibition of COX is associated
with reduced bone formation and delayed fracture healing in vivo. In this review,
moderate-strength evidence indicates that risedronate has a better effect on
increasing FNBMD than Vit D, which provides an option for GIOP patients who cannot
tolerate Vit D.
In addition, we also reviewed the evidence of adverse reactions, fracture risk, and
infection risk of different drug therapies. We found that most of the included drug
therapies had no difference in the above indicators, which indicates that there was
no significant difference in the increase or decrease in AEs between the existing
commonly used drugs. Notably, compared with alendronate, teriparatide can reduce the
incidence of VF, which suggests that teriparatide is an optimal choice for GIOP
patients with a high risk of VF and no drug contraindications. The study by Bouxein
et al. [28] showed that compared to placebo,
teriparatide reduced the rates of new VFs, adjacent VFs, and nonadjacent VFs in
patients with vertebral fractures and osteoporosis by 72%, 75%, and
70%, respectively, which indicates that teriparatide has a significant
advantage in reducing vertebral fractures.
Although there are still other drugs used in the treatment of GIOP, there is still a
lack of high-level evidence-based recommendations, and more pharmaceutical
researchers are needed to design and implement higher quality RCTs or SRs to
evaluate the efficacy and safety of these drugs in the treatment of GIOP. In this
umbrella review, it is encouraging that we found some moderate- to high-intensity
evidence that teriparatide, BPs and denosumab have better clinical efficacy in
increasing the BMD of patients with GIOP.
In addition to the above findings, this umbrella review also has the following
shortcomings. First, since this study did not include SRs involving non-RCTs, there
may be a lack of new drug therapies in this umbrella review. Second, the control
group was not limited to blank controls or placebo in the included SR, which is not
conducive to our horizontal comparison of the efficacy of different drug treatments
in the same outcome index. Third, although our research findings suggest that
teriparatide, BPs, and denosumab are drug choices for improving BMD in GIOP
patients, there is still a lack of high-level evidence to compare the efficacy
differences between these drugs.
Conclusions
In this umbrella review, we have summarized and compared the SRs of drug therapy for
GIOP, and the existing evidence indicates that teriparatide, BPs, and denosumab have
better clinical efficacy in increasing the BMD of patients with GIOP. These findings
can be used to provide evidence-based care to patients and to assist clinical
medical personnel in selecting the best drug prescription.
Authors’ Contributions
HD Liang: participation in study design, execution, analysis, article drafting and
critical discussion; JL Zhao: participation in study design, critical discussion; TZ
Tian: participation in study design, article drafting and critical discussion. All
authors read and approved the final manuscript.
