If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Address for correspondence: Dr. Kevin Boczar, MD, FRCPC, University of Ottawa Heart Institute, 40 Ruskin Street, Ottawa ON, K1Y 4W7, Tel: +1-6136967334, Fax: +1-613697247,
Division of Cardiology, Department of Medicine, University of Ottawa Heart Institute, Ottawa, ON, CanadaSchool of Epidemiology and Public Health, University of Ottawa, Ottawa, ON, CanadaDepartment of Echocardiography, Division of Cardiology, University of Pennsylvania, Philadelphia, PA, USA
Colchicine is an anti-inflammatory therapy with a low associated cost that has been shown in two large studies to reduce cardiovascular (CV) events but is associated with side effects. The main objective for this analysis is to determine whether colchicine is cost effective for the prevention of recurrent-CV events in patients who have suffered a myocardial infarction (MI).
Methods
A decision model was developed to estimate the healthcare costs in Canadian dollars and clinical outcomes amongst patients who have suffered a MI and are treated with colchicine. Probabilistic Markov modeling was used in combination with Monte Carlo simulation to derive expected lifetime costs and quality-adjusted life years (QALYs), which permitted the calculation of incremental cost-effectiveness ratios (ICERs). Models were derived for both short-term (20-months) as well as long-term (lifelong) colchicine use in this population.
Results
Long-term colchicine use was dominant over standard of care with lower average lifetime costs per patient ($91,552.80 versus $97,085.84) and higher average QALYs per patient (19.92 versus 19.80). Short-term colchicine use also dominated over standard of care. Results were consistent over a range of scenario analyses.
Conclusion
Based on two large randomized controlled trials, treatment of patients post MI with colchicine appears cost-effective when compared to standard of care at the current price. Based on these studies and currently accepted willingness to pay thresholds in Canada, health-care payers could consider funding long-term colchicine therapy for CV secondary prevention while we await results from ongoing trials.
Introduction
Atherosclerotic cardiovascular disease is a leading cause of morbidity and mortality around the world.
One-year costs associated with cardiovascular disease in Canada: Insights from the REduction of Atherothrombosis for Continued Health (REACH) registry.
Causes of Death C. Global, regional, and national age-sex specific all-cause and cause-specific mortality for 240 causes of death, 1990-2013: a systematic analysis for the Global Burden of Disease Study 2013.
Inflammation has long been known to play a key role in the progression of atherosclerosis and its complications, such as myocardial infarction (MI), stroke and cardiovascular (CV) death; however, until recently effective inflammatory treatment targets have remained elusive. Large trials in the area have been disappointing and have failed to show CV outcome benefits from therapies that target specific inflammation pathways: including low-density lipoprotein (LDL) oxidation;
Effects of the P-selectin antagonist inclacumab on myocardial damage after percutaneous coronary intervention for non-ST-segment elevation myocardial infarction: results of the SELECT-ACS trial.
Colchicine is an anti-inflammatory agent that has been around for centuries and is used for inflammatory conditions like gout, pericarditis, and juvenile idiopathic arthritis. It has been shown in several large trials to have CV outcome benefit.
Thus, while colchicine represents a promising novel therapeutic agent in the treatment of atherosclerotic cardiovascular disease, its use does have potential negative consequences including increased overall cost to the healthcare system. Decision analysis combined with economic analysis allows us to determine the net benefit of a new technology by weighing the potential benefits of a novel therapy (in terms of improved health outcomes) against some of the drawbacks associated with its use (including potential side effects or increased health care system costs). Furthermore, it allows for the determination of whether a new technology will lead to an improvement in the overall health of the wider population (i.e., whether it is cost-effective). The main objective of this study was to assess the cost-effectiveness associated with using colchicine for the prevention of recurrent-CV events.
Methods
Decision Problem
The specific decision problem that this study addresses is whether a Canadian health care payer should reimburse treatment with colchicine for the reduction of recurrent cardiovascular events in adult patients in Canada with a prior MI. To address this decision problem, a cost effectiveness analysis was performed to compare colchicine to current standard of care therapies (i.e. which do not include anti-inflammatory therapies), from the perspective of the Canadian public healthcare payer over a lifetime horizon (specifically from the perspective of the Ontario Ministry of Health & Long-Term Care). Both long-term (lifelong) and short-term (20-months) treatment options with colchicine were considered.
Model Overview
A probabilistic Markov cohort model was used to derive the estimated direct costs (i.e. healthcare-related costs) and health outcomes (life years and quality adjusted life years (QALYs)) amongst patients who have suffered a MI and are treated with colchicine. The dosing of colchicine 0.5mg orally taken daily was used as the treatment dose of choice, as per the results of the COLCOT and LoDoCo 2 studies
We used a lifetime study horizon (400 cycles of 1-month duration, equating to 33.33 years) to analyze the base-case costs (in Canadian dollars (CAD)) and utility values from the perspective of the Canadian publicly funded health care system, and we incorporated certain model assumptions that were based on previous published literature.
The net present value of future costs (in CAD) and QALYs was determined using a 1.5% discount rate as per guidelines outlined by the Canadian Agency for Drugs and Technologies in Health (CADTH) .
A Markov model combined with probabilistic analysis allowed estimation of costs (2022 CAD) and quality-adjusted life years (QALYs) which then allowed for the calculation of incremental cost-effectiveness ratios (ICERs). In the reference case, uncertainty regarding the value of each parameter was incorporated within the probabilistic analysis. Methodological uncertainty was explored by comparing the reference case results to those from non-reference case analyses using discount rates of 0% and 3% as per CADTH directives.
The model was developed in Microsoft Excel and we completed the Impact Inventory from the Second Panel on Cost-effectiveness in Health and Medicine (Supplemental Table S1).
Recommendations for Conduct, Methodological Practices, and Reporting of Cost-effectiveness Analyses: Second Panel on Cost-Effectiveness in Health and Medicine.
We developed a Markov model to follow a hypothetical cohort of patients who had suffered a MI (Figure 1). Base states in the model included a stable state, a stable after recurrent MI state, a stable after stroke state, and a death state. With each cycle, patients entered the event model which included the event states of: stable, recurrent MI, stroke, non-MI revascularization, pneumonia (potential side-effect of colchicine), gout flare, and death. From all events except gout, patients could either recover or experience death, and patients would then return to the appropriate base state following the event. Patients experiencing a gout flare all returned to their appropriate base state following the event. The model generated cycle specific estimates of costs and QALYs thereby allowing for the calculation of lifetime costs and QALYs.
Figure 1Markov Model. A survivor of myocardial infarction can be assigned to treatment with colchicine 0.5mg oral daily and standard of care, or standard of care alone (Decision Tree). Within each cycle, a survivor of myocardial infarction can then each cycle either remain stable, suffer a recurrent myocardial infarction, suffer a stroke, have a coronary revascularization procedure performed, suffer a pneumonia infection, suffer a gout attack, or remain in the death state. Following non-fatal recurrent MI, stroke, or pneumonia events, patients then cycle back to their stable states. Standard of care therapies following a myocardial infarction include antiplatelet agents, statins, and a beta-blocker. Abbreviations: MI: myocardial infarction, PO: oral
The cohort of patients were assumed to mirror the patient population within the COLCOT and LoDoCo 2 trials evaluating the impact of colchicine on cardiovascular outcomes.
The mean age of participants was 62, and patients had a history of MI with treatment including antiplatelet agents, statins, beta-blockers, and angiotensin-converting enzyme inhibitors. This is similar to the treatment regimens in the COLCOT and LoDoCo2 trials and what is done in clinical practice.
Model inputs
Short- and long-term colchicine was evaluated against current standard of care therapies post MI. There are several perceivable disadvantages to colchicine therapy including the upfront cost, an increased risk of pneumonia for patients, and a potential signal for increased mortality. As per the drug-regimens used in the COLCOT and LoDoCo2 trials, all patients received standard of care therapies, with patients receiving colchicine being administered 0.5mg orally every day. Strategies of both short-term (20 months) as well as long-term colchicine therapy were considered, separately, in the model.
Survival probabilities associated with standard of care therapy as well as colchicine treatment are summarized in Table 1, and were derived from the COLCOT and LoDoCo 2 studies where a total of 10,267 patients were randomly assigned to either placebo or colchicine. The uncertainty around these probabilities were incorporated into the probabilistic analysis.
Table 1Distributions for probabilistic analysis
Expected ValueReference
Probability Distribution
Event Probabilities
Probability of recurrent MI in acute phase (Standard Care)
In the base case analysis we assume that patients continue treatment beyond the 5 year time horizon of the clinical trials and the benefit of treatment is maintained long-term. This decision was made as within the LoDoCo2 trial, the survival benefit was stable at 5 years and showed no evidence of treatment waning.
Scenario analyses addressed the impact of these assumptions. In a scenario analysis, we adopted a 20-month time horizon, which allowed estimation of the proportion of the forecasted QALY gains for colchicine which were generated after the period covered by the COLCOT trial (short-term colchicine use model) as well as a 49-month time horizon, approximately covering the follow-up period of the COLCOT and LoDoCo2 trials combined. These analyses which examine the impact of extrapolation are recommended within the CADTH guidelines.
As per CADTH guidelines, in a further scenario analysis, we assessed the impact of treatment discontinuation by assuming that after both 20 months and 49 months, respectively, patients would discontinue treatment with colchicine with no continued effect after discontinuation.
The costs (in CAD) associated with each treatment strategy are summarized in Table 2. The monthly costs associated with colchicine administration strategies and the costs of cardiovascular events, complications, and admissions were obtained from the literature and converted to 2022 CAD based on the Bank of Canada inflation calculator.
Cohen DM, D.G. Tugwell, P. Sanmartin, C. Ramsay, T. Direct healthcare costs of acute myocardial infarction in Canada’s elderly across the continuum of care. The Journal of the Economics of Ageing. 2014:44-49.
Kang JSB, M.C. Qiu, F. Knudtson, M.L. Austin, P.C. Ko, D.T. Wijeysundera, H.C. Relation between initial treatment strategy in stable coronary artery disease and 1-year costs in Ontario: a population-based cohort study. CMAJ Open. 2016;4(3):E409-E416.
Cohen DM, D.G. Tugwell, P. Sanmartin, C. Ramsay, T. Direct healthcare costs of acute myocardial infarction in Canada’s elderly across the continuum of care. The Journal of the Economics of Ageing. 2014:44-49.
Kang JSB, M.C. Qiu, F. Knudtson, M.L. Austin, P.C. Ko, D.T. Wijeysundera, H.C. Relation between initial treatment strategy in stable coronary artery disease and 1-year costs in Ontario: a population-based cohort study. CMAJ Open. 2016;4(3):E409-E416.
Utility values were derived from a variety of sources. Utility values associated with the base state of post-MI status were taken from an analysis and comparison of the MONICA/KORA registry to the general population.
Utility values for the recurrent MI state were obtained from literature which investigated the stability of time-tradeoff utilities in survivors of MI.
Utility values for the recurrent MI and post-stroke state were applied to patients indefinitely following their respective event (i.e. stroke or recurrent MI). Finally, disutility values associated with adverse events were determined from expert opinion, and were applied only to the specific cycle during which the adverse event occurred.
Markov modeling was used to calculate mean total health care costs and total QALYs gained for each strategy. Cycle specific estimates were summed and discounted over the lifetime horizon to provide estimates of total costs and QALYs. Probabilistic analysis using Monte Carlo simulation with 5000 replications was performed to account for uncertainty in the input values providing an estimate of the expected values for costs and QALYs. Cost effectiveness was assessed by estimating the incremental cost per QALY gained against a base willingness-to-pay threshold of $50,000 CAD/QALY. This threshold was chosen based on a previous statement by the Assistant Deputy Minister for Ontario.
Sapsford R. Deputy Minister’s response to the investigation into the Ministry of Health and Long-Term Care’s decision-making concerning the funding of Avastin for colorectal cancer patients. In:Letter dated August 26, 2009.
The threshold value for a QALY should reflect the marginal efficiency of the health care system that the result is related to. Empirical work related to this for Canada has not been conducted, although an analysis for the UK health care system suggests an appropriate threshold may be between $20,000 to $30,000 per QALY.
Claxton K, Martin S, Soares M, et al. Methods for the estimation of the National Institute for Health and Care Excellence cost-effectiveness threshold. Health Technol Assess. 2015;19(14):1-503, v-vi.
A higher cost per QALY threshold would be suggestive of the Canadian system either being less efficient than the UK system, or better funded.
Model Validity
Face validity was confirmed with experts in the field. Internal validation was conducted by the senior health economist reviewing the model structure and coding. Validity of our model was tested by comparing event rates from the COLCOT and LoDoCo2 trials with the risk from our simulated colchicine and standard of care cohorts at 23 and 49 months of elapsed follow-up time, respectively. To assess the external validity of our model, we compared events in the standard of care arm from the CANTOS trial (a similar study population) to results obtained from our standard of care model cohort at 22.6 months of follow-up (which was the median follow-up time from the CANTOS trial).
Results
With respect to validity of our model, cumulative mortality estimates from our model were similar to the results seen in both the COLCOT and LoDoCo2 trials results.
Results of our probabilistic analysis found that compared to standard of care, long-term colchicine was dominant. (Table 3, Figure 2) This strategy was associated with a slightly lower discounted life expectancy (23.62 years versus 23.75 years), but with a reduced lifetime incidence of MIs, revascularisation, stroke, and gout flares (Table 4). Long-term colchicine was associated with lower health care costs than the standard of care group $91,552.80 versus $97,085.84 (Table 3). The incremental reduction in costs with long-term colchicine use ($5,533.04) were almost exclusively due to the lifetime costs of colchicine ($2,640.36), being more than offset by the reduction in MI and revascularization events. Cumulative QALYs were 19.92 in the long-term colchicine group, and 19.80 in the standard of care group, respectively.
Table 3Cost Effectiveness Results for Base Analysis and Scenario Analyses
Lifetime costs
Lifetime QALYs
ICER: Colchicine versus Standard of Care
Base Case
Standard of care
$97,086
19.80
Long-term colchicine
$91,553
19.92
Dominant
Short-term colchicine
$96,636
19.86
Dominant
Discount rate 0%
Standard of care
$121,938
24.83
Long-term colchicine
$115,082
24.96
Dominant
Short-term colchicine
$113,736
24.91
Dominant
Discount rate 3%
Standard of care
$78,389
16.16
Long-term colchicine
$74,171
16.24
Dominant
Short-term colchicine
$73,173
16.21
Dominant
48-month time horizon
Standard of care
$15,274
3.36
Long-term colchicine
$14,582
3.37
Dominant
Short-term colchicine
$14,952
3.37
Dominant
20-month time horizon
Standard of care
$6,121.38
1.41
Colchicine
$5,865.58
1.42
Dominant
Allowance for discontinuation with treatment
Standard of care
$97,087
19.79
Long-term colchicine
$96,161
19.86
Dominant
Short-term colchicine
$96,394
19.84
Dominant
ICER – incremental cost per QALY gained; MI - = myocardial infarction; QALY – quality adjusted life year
Figure 2Scatterplot of incremental costs against incremental QALYs gained. Standard of care therapies following a myocardial infarction include antiplatelet agents, statins, and a beta-blocker. Dashed line represents a cost-effectiveness threshold of $50,000/QALY. Abbreviations: CAD: Canadian dollar, QALY: Quality-adjusted life years
In our probabilistic analysis we found that compared to standard of care, short-term colchicine was also dominant (Table 3, Figure 2). It was associated with slightly higher discounted life expectancy (23.77 years versus 23.75 years). Short-term colchicine was associated with lower health care costs compared to the long-term colchicine group $96,636.33 versus $97,085.84 (Table 3). Cumulative QALYs were 19.86 in the short-term colchicine group, and 19.80 in the standard of care group, respectively.(Table 3).
In a separate comparison, long-term colchicine therapy dominated over short-term use with lower lifetime costs and higher lifetime QALY gains, as above.
At a willingness-to pay threshold for a QALY of $50,000, long-term colchicine had a 72.2% probability of being cost-effective (i.e. the ICER was greater than $50,000 in 72.2% of the 5,000 simulations). Short-term colchicine had a 25.8% probability of being the most cost-effective strategy (Figure 3).
Figure 3Cost-effectiveness acceptability curve for the probability that each treatment strategy is cost effective against the threshold value for a QALY. Standard of care therapies following a myocardial infarction include antiplatelet agents, statins, and a beta-blocker. Abbreviations: CAD: Canadian dollar, QALY: Quality-adjusted life years
Scenario analyses relating to discount rates demonstrated consistency with our primary analysis (Table 4). Our results remained stable when using alternative discount rates of 0% and 3%. In the scenario analysis with reduced time horizons of 20 and 49 months, estimated incremental QALY gains were only 0.01, suggesting that less than 17% of the forecasted QALY gains from colchicine occur during the initial time periods representative of the COLCOT and LoDoCo2 trials, respectively. In a scenario analysis where patients stopped colchicine at 20 or 49 months, respectively, results remained unchanged with both short-term and long-term colchicine therapy continuing to show dominance over standard of care.
Discussion
In this cost-effectiveness analysis, we evaluated the cost-effectiveness of short-term and long-term colchicine against the current standard of care therapies for the secondary prevention of CV events, from the perspective of the Ontario public healthcare system. Our results demonstrate that in comparison, both short-term and long-term colchicine usage are cost-effective strategies in this area, as both were dominant over current standard of care. The predominant driver of lower costs with colchicine therapy was the reduction in CV events, which made up for the marginal increase in costs associated with the drug itself. In our pre-specified willingness to pay threshold of $50,000/QALY, long-term colchicine also emerged as dominant option over short-term use, with lower overall costs and higher QALY gains.
In recent years, targeting inflammation for the reduction of atherosclerotic cardiovascular disease has become a topic of great interest.
much attention has been focused, with varying success, on trying to identify potential anti-inflammatory agents that could be utilized to reduce CV risk for patients. The anti-inflammatory medication canakinumab represented a breakthrough in this domain, as the CANTOS trial demonstrated benefit for the reduction of CV events in patients who had suffered a previous MI and who had a high residual inflammatory burden.
This was largely due to the high cost associated with the medication. Following this, much attention focused around trying to find alternative anti-inflammatory agents that could fill the void. Colchicine, a medication with a low-associated cost and long track-record as a commonly used anti-inflammatory agent, seemed to hold promise in this regard.
Several pivotal trials released following CANTOS demonstrated clinical benefit for colchicine in reducing CV events. In the COLCOT trial, colchicine use in patients with recent MI led to a significantly lower risk of ischemic cardiovascular events than placebo.
Separately, the LoDoCo2 study showed that in patients with chronic coronary disease, the risk of cardiovascular events was significantly lower among those who received 0.5 mg of colchicine once daily than among those who received placebo.
Colchicine is not without side effects, however. Some studies, including LoDoCo2 have found a small and potentially non-statistically significant increase in all-cause mortality.
Thus, while the current medications used for secondary prevention in patients who have suffered a MI is well-established, whether or not colchicine should enter this domain is certainly not. Additionally, if colchicine were to enter guideline directed medical therapy, it is not clear whether colchicine should be used for a short-term course after a CV event, or for long-term use. The cost-effectiveness of the drug will certainly need to be considered in this clinical decision making, particularly in publicly funded health care systems. Further research is currently being conducted (NCT03048825) looking at colchicine for cardiovascular indications, and will further explore the potential increase in all-cause mortality associated with colchicine use. This will certainly help to shed further light on this area to determine if it truly remains a dominant treatment option in this domain.
By conducting probabilistic analysis, we were able to incorporate a wide range of uncertainty into our model, thereby improving the robustness of our results. Our analyses incorporated the potential increase in all cause mortality but still concluded that the use of colchicine was worthwhile. As evidenced by our study, there is an overwhelming probability that colchicine is cost-effective over standard care using our pre-specified willingness to pay threshold of $50,000/QALY, and this conclusion remains consistent with much lower willingness-to-pay thresholds. However, whether a short-term or long-term strategy of colchicine should be employed is a much more nuanced decision-problem. Long-term colchicine provided slightly higher cumulative QALYs with a higher cost, largely related to the cost of long-term use of the medication. However, long-term use was still deemed to be a cost-effective strategy at the accepted willingness to pay threshold of $50,000/QALY.
Limitations
Our model had several important limitations, and our results should therefore be interpreted within the context of the data inputs and modelling assumptions. First, the efficacy of colchicine was based on two single randomized controlled trials, which limits the generalizability of our results. However, at this point in time, these trials encompass the largest body of Phase 3 evidence on the subject, and are currently what is available for decision-makers. The results of this analysis strongly suggest that given the current data on relative effectiveness, colchicine is highly likely to be cost effective in this context at its current price. However, with the concerning signal regarding mortality associated with colchicine, the results of ongoing randomized-controlled trials are needed to help to clarify this issue further before adopting colchicine into guidelines. The uncertainty regarding the potential signal for mortality is reflected within the scatterplot of incremental costs against incremental QALYs gained shown in Figure 2. As further data emerges, this will clarify the risks (or lack thereof) associated with colchicine-use, and will subsequently reduce the amount of uncertainty within the model. Our scenario analyses adopted assumptions related to discontinuance of treatment and of waning treatment effect which were not favourable towards colchicine, and despite this, colchicine remained a cost-effective option for this indication. The median follow-up for COLCOT and LoDoCo2 studies were 23 and 29 months, respectively, but we assumed therapy would be effective for a much longer period. This was a potentially optimistic estimate of the long-term effects of colchicine therapy. We addressed this by performing a scenario analysis looking at waning of treatment effect as well as discontinuation of therapy, and our results remained consistent. Finally, as the trials did not present sex-stratified results, we were unable to conduct sex-stratified models within our analysis. However, given the small sex differences seen in the incidence of cardiovascular events we expect that our conclusions would not change with the addition of sex-stratified analyses.
Conclusions
Treatment with colchicine for secondary prevention of CV events is likely cost-effective in the Canadian healthcare system. Both short- and long-term therapy with colchicine was dominant over standard care in the reduction of CV events. In our analysis, long-term colchicine therapy emerged as a cost-effective option over short-term use, with an ICER of $46,241.17. Ongoing trials (NCT03048825, NCT04848857, NCT04181996) are certainly needed and will shed additional light on the potential harms associated with colchicine use. Given this uncertainty, it is currently difficult to strongly recommend colchicine use for cardiovascular indications at present until further data are available. Future investigations of anti-inflammatory drugs targeted at atherosclerosis must continue to consider the balance of effectiveness and economic impact on patients and the health-care system.
Contemporary mortality risk prediction for percutaneous coronary intervention: results from 588,398 procedures in the National Cardiovascular Data Registry.
One-year costs associated with cardiovascular disease in Canada: Insights from the REduction of Atherothrombosis for Continued Health (REACH) registry.
Causes of Death C. Global, regional, and national age-sex specific all-cause and cause-specific mortality for 240 causes of death, 1990-2013: a systematic analysis for the Global Burden of Disease Study 2013.
Effects of the P-selectin antagonist inclacumab on myocardial damage after percutaneous coronary intervention for non-ST-segment elevation myocardial infarction: results of the SELECT-ACS trial.
Recommendations for Conduct, Methodological Practices, and Reporting of Cost-effectiveness Analyses: Second Panel on Cost-Effectiveness in Health and Medicine.
Cohen DM, D.G. Tugwell, P. Sanmartin, C. Ramsay, T. Direct healthcare costs of acute myocardial infarction in Canada’s elderly across the continuum of care. The Journal of the Economics of Ageing. 2014:44-49.
Kang JSB, M.C. Qiu, F. Knudtson, M.L. Austin, P.C. Ko, D.T. Wijeysundera, H.C. Relation between initial treatment strategy in stable coronary artery disease and 1-year costs in Ontario: a population-based cohort study. CMAJ Open. 2016;4(3):E409-E416.
Sapsford R. Deputy Minister’s response to the investigation into the Ministry of Health and Long-Term Care’s decision-making concerning the funding of Avastin for colorectal cancer patients. In:Letter dated August 26, 2009.
Claxton K, Martin S, Soares M, et al. Methods for the estimation of the National Institute for Health and Care Excellence cost-effectiveness threshold. Health Technol Assess. 2015;19(14):1-503, v-vi.
Contemporary mortality risk prediction for percutaneous coronary intervention: results from 588,398 procedures in the National Cardiovascular Data Registry.
KEB – supported by a CIHR Fellowship Award (FRN: 171284)
RSB – was supported in part by the University of Ottawa Heart Institute’s Vered Chair in Cardiology and holds a University of Ottawa Distinguished Chair in Cardiovascular Research. He has industry grant support and has received Honoria from GEHC, JDI and Lantheus Medical Imaging, not related to this work.
00029-X&id=gr1.jpg){kind=link}
00029-X&id=gr2.jpg){kind=link}
00029-X&id=gr3.jpg){kind=link}