Editorial - Prof Kalla

COXIBs and cariac safety

Arthritis and the Internet

Glucocorticoid osteoporosis

Return to SARAA website

Editorial

It has been some time since the last issue of Rheumatology News. Fortunately, we were able to find sponsorship for our Journal and you can expect three editions in 2002. We are grateful to MSD for their ongoing support of Rheumatology in South Africa.
It is a pleasure for the members of the Editorial Board to congratulate Dr David Gotlieb on being elected as the President of SARAA for the next 2 years (2001 - 2002). He has clearly earned the support and respect of his colleagues and we are confident that he will take our association forward during his tenure. His internet expertise will see many new developments of the SARAA website, and his personal drdoc website will naturally grow from strength to strength.
It is not often that we have non-rheumatologists contributing to our journal, but in this issue the two main articles are both by non-rheumatologists. The first paper is by Dr Shirley Middlemost, a cardiologist from Johannesburg. She provides an excellent overview of the non-GIT side-effect profile of the COXIBs, which seems to resemble those seen with the non-selective NSAIDs. The kidneys appear to have a physiological activity of COX II, resulting in sodium and water retention when this isoenzyme is inhibited in its production of prostaglandins. While the COXIBs have certainly improved the gastrointestinal risks associated with conventional NSAIDs, caution is necessary when these new agents are used in patients with cardiac or renal disease.
Another important issue arising from the use of these new anti-inflammatory relates to the absence of an anti-platelet effect. The concurrent use of aspirin removes the GIT protection, while failing to use aspirin may predispose to myocardial infarction. The data seems to suggest that there is no intrinsic thrombogenic effect due to the COXIBs, but this is an area that is likely to receive increasing interest in the future. This aspect grows in importance when we recognise that the target population is particularly predisposed to ischaemic heart disease. These are also the group of patients more likely to have comorbid diseases such as hypertension, diabetes mellitus or cardiac failure.
Corticosteroid therapy is often required for the treatment of a variety of rheumatic diseases, whose mortality and morbidity has improved significantly over the years. As a result, many patients are surviving into the menopause and suffer the consequences of osteoporosis and fragility fractures. The paper by Dr Steven Soule, an endocrinologist at Groote Schuur Hospital, Cape Town provides an excellent overview of the mechanisms of bone loss due to corticosteroid therapy. He discusses the conditions favouring bone loss in these patients and addresses some of the controversies. He also provides a pactical approach to the treatment and management of patients who are at risk of developing osteoporosis while taking corticosteroids. The algorithm provides a useful stepwise approach to evaluating the at-risk patient. It is interesting that the therapeutic approach closely resembles that used for idiopathic osteoporosis.
The SARAA congress held in Pretoria in March, 2001, maintained the excellent standard which has become a regular feature of our biennial meeting. Delegates attended from all parts of South Africa, North and West Africa as well as shores as far as the UK and USA. Dr Gotlieb has done an excellent job of recording the highlights of the meeting. Due to limitation of space, the report has been extensively edited, but it is hoped that the content will provide a useful reflection of the academic content of the meeting.
Wishing one and all everything of the best for 2002.
Asgar Ali Kalla.
Editor.

Go to top

By Dr S Middlemost Cardiologist

The effects of nonsteroidal anti-inflammatory drugs (NSAIDS) are mediated through their ability to inhibit cyclooxygenase (COX)-catalyzed prostaglandin production. Two isoforms of COX, COX-1 and COX-2, have been identified. The nonselective NSAIDS, such as ibuprofen, diclofenac and naproxen, inhibit both of the COX isoenzymes. The selective COX-2 inhibitors have little effect on the COX-1 isoenzyme and are therefore sometimes referred to as 'COX-1 sparing agents'.

Cyclooxygenase-1 is constitutively expressed throughout the body and is thought to play an essential role in normal gastrointestinal and platelet function, whereas COX-2 is induced in the presence of inflammation. The anti-inflammatory effects of nonselective NSAIDS (those that inhibit both isoforms of the COX enzyme) therefore appear to be mediated through inhibition of COX-2. Their harmful effects in the gastrointestinal tract as well as their antiplatelet effects occur primarily through inhibition of COX-1.The newer agents, that selectively inhibit COX-2 (COX-2 inhibitors) have anti-inflammatory and analgesic effects that are similar to those of nonselective NSAIDS.

Differences in the adverse effects between nonselective NSAIDS and the COX-2 inhibitors.

1. Gastrointestinal Toxicity The gastropathy induced by NSAIDS has been shown to pose a substantial health hazard: 1 in 70 will develop a symptomatic ulcer, 1 patient in 150 will develop a bleeding ulcer and 1 patient in 1200 will die of a bleeding ulcer. (1) However, rofecoxib has been shown to exert analgesic and anti-inflammatory effects comparable to the high-dose NSAIDS, but with a markedly superior gastrointestinal safety profile. (2) In the VIOXX Gastrointestinal Outcomes Research (VIGOR) trial, 8076 patients with rheumatoid arthritis were treated with rofecoxib 50 mg once daily or naproxen 500 mg twice daily. Patients were not allowed to use aspirin in this trial. After a median follow up of 9 months, there was an 88% risk reduction in patients at low risk for serious gastrointestinal events. In those at high risk (> 65 years age, prior history of an upper GI event, positive test for H. pylori, or long-term use of corticosteroids) there was a 51% risk reduction. In the ADVANTAGE (Assessment of Difference between VIOXX and Naproxen to Ascertain Gastrointestinal Tolerability and Effectiveness) trial, 5556 patients with osteoarthritis (OA) were randomized to rofecoxib 25 mg once daily or naproxen 1 g/day for 12 weeks. Unlike VIGOR, patients taking low-dose aspirin (< 81 mg/day) were eligible for ADVANTAGE. Rofecoxib use was associated with a 25% reduction in the risk of discontinuing therapy due to a GI adverse effect (p<0.006), and a 20% reduction in the number of patients using concomitant GI medications to alleviate GI adverse effects (p=0.012). (3)

2. Renal Effects The role of the COX-isoenzymes in the kidney The most important clinical effects of the NSAIDS on the kidney are decreased sodium excretion, decreased potassium excretion, and decreased renal perfusion. This is mediated by inhibiting the synthesis of renal prostaglandins that are important for solute homeostasis and for maintaining renal blood flow. Recent evidence has confirmed that COX-2 is widely distributed in the kidney and that it is present in the constitutive form in vasculature, glomerulus, tubule segments and interstitium. As a result one could predict similar effects on renal function with COX-2 inhibitors as with nonselective NSAIDS. The COX-1 isoenzyme, in part, controls renal hemodynamics and the glomerular filtration rate (GFR), while COX-2 affects salt and water metabolism.

Studies have evaluated the effects of COX-2 inhibitors on renal prostaglandins. Although there have been no studies directly comparing the effects of these agents, it appears that both the COX-2 inhibitors and the nonselective NSAIDS similarly reduce renal excretion of PGE2 and prostacyclin. The former prostaglandin reduces sodium reabsorption, and the latter increases potassium secretion and maintains renal blood flow by causing vasodilatation.

Adverse effects on renal function In elderly patients on a sodium-replete diet, the COX-2 inhibitors (rofecoxib, celecoxib) reduce sodium excretion in the urine to a similar extent as non-selective NSAIDS (indomethacin 50 mg three times daily, naproxen 500 mg twice daily). In these healthy subjects, there was no reduction in glomerular filtration rate (GFR). However, in salt-depleted subjects, the COX-2 inhibitors like the nonselective NSAIDS, promote sodium and potassium retention and produce a significant reduction in GFR. Furthermore, in high-risk patients, the COX-2 inhibitors have been shown to precipitate renal insufficiency, with no clinically significant differences in the renal effects between the COX-2 inhibitors. These drugs should not be used at all in patients with very reduced renal function (creatinine clearance <30ml/minute). NSAID-related renal effects are clinically significant in patients with risk factors such as diabetes and poorly controlled hypertension. Situations causing a decrease in the actual or effective circulating volume such as dehydration, overzealous use of diuretics, blood loss, congestive heart failure, cirrhosis or renal insufficiency, also place patients at increased risk for adverse renal effects. (4,5) Acute renal failure may develop after the first dose of the NSAID. Both a high dose and the use of NSAIDS with a longer-acting half-life, increase the risk of this complication. (6) It is thus important to monitor renal function in these high-risk patients within days, rather than weeks, after the use of these drugs.

Hyperkalemia Patients at risk for the development of hyperkalemia include those with renal insufficiency, diabetes, and those receiving drugs that decrease potassium secretion, such as the ACE inhibitors. In these situations, it is advisable to monitor serum potassium within days of commencing treatment, rather than after several weeks.

Peripheral edema Peripheral edema occurs with both classes of drugs when compared to placebo. And the effects of celecoxib and rofecoxib are similar to nonselective agents. Although trials have reported differences between the COX-2 inhibitors, comparisons of efficacy or toxicity between members of this class should only be made using standardized techniques and equivalent doses over the entire dosing interval. Lower extremity water displacement evaluation may be necessary to truly differentiate between these drugs.

3. Blood Pressure Effects Pooling data from randomized trials conducted in younger adults (mean age 46 years), use of nonselective NSAIDS results in an increase in mean blood pressure of 5 mmHg. The only patients in whom a statistically significant rise in mean arterial pressure could be found, however, were those with preexisting hypertension, either untreated or controlled by medication. Recent onset of use of NSAIDS in the elderly is associated with a 1.7-fold higher risk of requiring antihypertensive therapy compared with nonusers. Although there is less information concerning the effects of the COX-2 inhibitors on blood pressure, studies have confirmed that hypertension may be aggravated to a similar extent as the nonselective agents. For example, the effects of rofecoxib 50 mg daily on mean diastolic and systolic BP were similar to that of ibuprofen 800 mg three times daily and diclofenac 50 mg three times daily. (7) In another study conducted over a 6-month period, 0.1% of patients receiving either 12.5 or 25 mg rofecoxib discontinued therapy due to hypertension as compared with 0.4% of the ibuprofen-treated patients. (8)

In a placebo-controlled arthritis trial, worsening of hypertension did occur more frequently in older patients taking celecoxib (1.1%) as compared with placebo (0.3%; p<0.05), and the percentage of patients initiating antihypertensive medication was more common on celecoxib (3.7%) than on placebo (2.4%) (p<0.05). (9) In the VACT trial, no difference was found in the incidence of hypertension between rofecoxib (12.5 or 25 mg per day), celecoxib (200 mg per day) and acetaminophen. (10) Overall, the rates of discontinuation for hypertension and leg edema were low (~ 0.1% and ~ 0.5%, respectively) for the COX-2 inhibitors. These data reinforce that the COX-2 inhibitors are similar to the nonselective NSAIDS in terms of effects on blood pressure.

While the mechanism(s) remain speculative, salt and water retention together with an increased peripheral vascular resistance may play an important role. The class of antihypertensive medication a patient is taking does play a role in the NSAID-mediated increases in blood pressure. With the exception of calcium channel blockers, COX-2 inhibitors and nonselective NSAIDS affect all classes of antihypertensive agents. (11)

From a practical point of view, clinicians should avoid the excess use of NSAIDS or COX-2 inhibitors and try simpler analgesics. Moreover, patients should be warned about the danger of unsupervised use of over-the-counter NSAIDS. Blood pressure should be monitored once therapy is initiated. An adjustment in the dose of antihypertensive medication may be required to maintain blood pressure at goal level. High-risk patients include the elderly, persons with congestive heart failure, cirrhosis or renal insufficiency, individuals with hypertension, volume-depleted patients, or diabetics. Patients with heart failure may need additional diuretics to keep them edema free. Daily weights and BP monitoring should be done in these high-risk patients, with measurement of serum creatinine and potassium within a week of initiating treatment.

4. Precipitation / Worsening of Heart Failure By decreasing tubular prostaglandin E2 (PGE2), the NSAIDS increase sodium reabsorption. This manifests clinically as a 15 - 20% reduction in the response to loop diuretics. (12) Whilst some patients may gain weight or develop edema; congestive heart failure (CHF) develops rarely in those who are not at risk for this complication. An increase in systemic vascular resistance and a decline in cardiac output have been shown to occur following the use of NSAID therapy in patients with established or incipient heart failure. (13) A recent paper estimated the relative risk of first admission to a hospital with CHF in recent users of nonselective NSAIDS, compared with nonusers. The increased relative risk for this complication was also evaluated in patients with a history of heart disease but without CHF. (14) Use of NSAIDS in patients with a history of heart disease was associated with a marked increase in risk for first admission for heart failure compared to those without such a history. Both high-dose and long plasma half-life of the NSAID increased this risk. Thus, in such high risk-patients elderly patients, recent use of NSAIDS doubled the odds of being admitted with an episode of CHF. (14) Others have confirmed similar outcomes, with a doubling of the risk of hospitalization with CHF when patients receiving diuretics were given NSAIDS, the risk being greatest in those with pre-existing CHF. (15) Based on these observations, the indiscriminate use of NSAIDS in patients with compensated left ventricular dysfunction should be discouraged. If, however, it is deemed necessary to prescribe these agents, the lowest dose should be used. Ideally, the NSAIDS with a long plasma half-life should be avoided.

5. Do COX-2 Inhibitors Increase Thrombotic Cardiovascular Events?

The cardioprotective effect of aspirin is potentially mediated through its inhibition of platelet function. Selective COX-2 inhibitors such as rofecoxib differ from the nonselective NSAIDS in that they do not inhibit platelet function. (16,17) However, the selective COX-2 inhibitors are similar to the nonselective NSAIDS in their ability to inhibit the production of systemic prostacyclin (PGI2), a vasodilatory prostanoid that is an inhibitor of platelet aggregation. These data raise the concern that COX-2 selective inhibitors may be pro-thrombotic. (16,18) Or alternatively, the nonselective NSAIDS with potent and sustained antiplatelet activity might be cardioprotective (naproxen).

A recent meta-analysis concluded that the COX-2 inhibitors may increase the risk of thrombotic cardiovascular events. (19) There were important limitations in this meta-analysis. Cardiovascular complications were not pre-specified endpoints in any of the trials cited. The authors compared the incidence of cardiovascular events in the COX-2 inhibitor-treated patients (older with cardiovascular risk factors) to a younger and healthier placebo group with a low cardiovascular risk profile. It would have been more meaningful to compare the patients treated with the COX-2 inhibitors to the patients in the placebo groups of the arthritis studies.

Rofecoxib in Rheumatoid Arthritis: Increased risk of thrombotic cardiovascular events? In the VIGOR Trial, where patients with rheumatoid arthritis were treated with either naproxen or rofecoxib, the rates of death from gastrointestinal events and from cardiovascular causes were similar. Ischemic cerebrovascular events occurred in 0.2% of the patients in each group. The difference in thrombotic events was mostly attributable to a difference in the incidence of myocardial infarction between the groups. The rate of myocardial infarction was significantly lower in the naproxen group than in the rofecoxib group (0.1% versus 0.4%). Although patients requiring aspirin for secondary cardiovascular prophylaxis were excluded from the study, 4% of such patients were included and were not taking aspirin. These patients accounted for 38% of the patients in the study who had myocardial infarctions. In the other patients the difference in the rate of myocardial infarction was not significant (0.2% in the rofecoxib group and 0.1% in the naproxen group). In addition, there was no association between hypertension and myocardial infarction. Interestingly, it has been shown that patients with rheumatoid arthritis have a higher risk if myocardial infarction. (20) So the question remains as to whether these patients should be given aspirin for cardiovascular prophylaxis.

It should be noted that, compared to the other NSAIDS that inhibit platelet function to a much lesser extent (ibuprofen 80%, diclofenac 40%), the antiplatelet action of naproxen is similar to low-dose (81mg) aspirin. (21) Moreover, this potent antiplatelet effect of naproxen is sustained throughout the dosing interval. (17) It is thus conceivable that naproxen conferred aspirin-like protection from thrombotic cardiovascular events. However, it could not be determined whether a cardioprotective effect of naproxen, or a pro-thrombotic effect of rofecoxib, accounted for the difference in the incidence of the thrombotic vascular events in the VIGOR trial.

Due to the cardiovascular results from the VIGOR study, an interim combined analysis was undertaken to provide a global assessment of cardiovascular outcomes in the rofecoxib clinical program to date. (22) All Phase IIb to Phase V clinical trials of 4 weeks or longer with rofecoxib were evaluated. The relative risk of cardiovascular thrombotic serious adverse experiences in patients taking rofecoxib as compared to placebo, naproxen and other non-naproxen, nonselective NSAIDS was the focus of this analysis. The naproxen studies were examined separately because of pharmacology data, which demonstrated that unlike the other NSAIDS (ibuprofen, diclofenac), naproxen therapy is associated with a sustained potent antiplatelet which is similar to aspirin.

The APTC (Antiplatelet Trialists' Collaboration) endpoint was used in this analysis. (23) It is the most common and widely accepted endpoint to quantify the overall cardiovascular impact of antithrombotic compounds in cardiovascular trials. The APTC endpoint measures fatal and irreversible morbid cardiovascular events and estimates the combined incidence cardiovascular, hemorrhagic, and unknown cause of death, myocardial infarctions, and cerebrovascular accident. The meta-analysis examined the incidence of the APTC endpoint using data from over 28,000 patients. More than half the patients studied had at least one risk factor for heart disease.

The results are summarized in Figure 1. The analysis showed that the relative risks of serious cardiovascular events were similar with rofecoxib and placebo or nonselective NSAIDS other than naproxen. In the placebo-controlled trials of rofecoxib, the majority of patients were elderly (mean age 75 years) with the highest baseline incidence of traditional risk factors in the rofecoxib clinical program. This elderly cohort was therefor an excellent population in which to affirm the cardiovascular safety of rofecoxib.

The relative risk of 1.69 seen in the comparison of rofecoxib to naproxen was higher than the relative risk of 0.79 seen in the comparison of rofecoxib to nonselective NSAIDS other than naproxen. This outcome supports the contention that the imbalanced rates of thrombotic cardiovascular adverse events observed in the clinical trials that compared rofecoxib with naproxen, may have been due to a reduced rate of events in the naproxen group, rather than an increased rate in the rofecoxib group. The 'aspirin-like' antiplatelet action of naproxen could account for this difference.

COX-2 inhibitors and Osteoarthritis (OA): Increased thrombotic risk?

Celecoxib in OA Trials CLASS was a double-blind, randomized controlled trial in which 8059 patients were randomized to receive 400 mg of celecoxib twice per day, or 75 mg of diclofenac twice per day or 800 mg of ibuprofen 3 times per day. (24) Unlike the VIGOR trial, aspirin use (<325 mg/day) was permitted. The CLASS trial, which included patients with either rheumatoid arthritis (30%) or OA (70%), demonstrated no significant difference in cardiovascular events compared with the NSAIDS.

Rofecoxib in OA Trials Daniels performed a meta-analysis of 6000 patients with OA treated in nine double-blind, placebo and active comparator controlled studies with rofecoxib, placebo or nonselective NDSAIDS (ibuprofen 2400 mg, diclofenac 150 mg, nabumetone 1500 mg). (25) The patients (mean age 63 years; women 70%) had a mean duration of disease of 10 years. The baseline cardiovascular risk factors included hypercholesterolemia, hypertension, coronary artery disease, diabetes, and current tobacco use. Due to the variable duration of treatment (6 to 86 weeks) in the studies, results are expressed as events per 100 patient-years. Results are shown in Tables 1-3.

MI= myocardial infarction; CVA=cerebrovascular accident; TIA= transient ischemic attack; UA=unstable angina.

These data confirm that the incidence of thrombotic cardiovascular events was similar in patients taking rofecoxib, placebo, or the nonselective NSAIDS ibuprofen, diclofenac, and nambumetone. (26)

Conclusions In patients with arthritis, rofecoxib causes significantly less gastrointestinal side effects than the nonselective NSAIDS, even when used in combination with low-dose aspirin. However, with regards to adverse effects on blood pressure, renal function and the precipitation of peripheral edema or congestive heart failure, the same precautions that apply to the use of nonselective NSAIDS, are applicable to the COX-2 inhibitors.

Finally, the incidence of cardiovascular thrombotic adverse events with the COX-2 inhibitors is similar to placebo and to the nonselective NSAIDS, with the exception of naproxen. The latter does appear to confer 'aspirin-like' cardiovascular protection.

REFERENCES

1. Tramer et al. Pain 2000;85:169-182

2. Bombardier C, Laine L, Reicin A, et al. Comparison of upper gastrointestinal toxicity of rofecoxib and naproxen in patients with rheumatoid arthritis. N Engl J Med 2000;343:1520-28

3. Geba GP, Polis AB, Dixon ME, et al. Gastrointestinal tolerability in primary care patients treated with naproxen or rofecoxib for osteoarthritis (OA): the Advantage trial. Ann Rheum Dis 2000; 59(Suppl. 1):134-135

4. Brater DC, Anderson SA, Baird B, Campbell WB. Effects of ibuprofen, naproxen, and sulindac on prostaglandins in man. Kidney Int. 1985;27:66-73

5. Toto RD, Anderson SA, Brown-Cartwright D, Kokko JP, Brater DC. Effects of acute and chronic dosing of nonsteroidal anti-inflammatory drugs in patients with renal insufficiency. Kidney Int. 1986;30:760-768

6. Henry D, Page J, Whyte I, Nanra R, Hall C. Consumption of nonsteroidal anti-inflammatory drugs and the development of functional renal impairment in elderly subjects: results of a case-control study. Br J Pharmacol. 1997;44:85-90

7. Daniels B, Gertz B, Morrison B, Seidenberg B. Renal safety profile of rofecoxib, a specific inhibitor of COX-2, in controlled clinical trials. (Abstract). Arthritis Rheum 1999;(Suppl)42(9):S143

8. Brater DC, Harris C, Redfern JS, Gertz BJ. Renal effects of COX-2 selective inhibitors. Am J Nephrol 2001:21:1-15.

9. Whelton A, Maurath CJ, Verburg KM, Geis GS. Renal safety and tolerability of celecoxib, a novel cyclooxygenase inhibitor. Am J Ther 2000;7:159-175

10. Geba GP, Weaver AL, Schnitzer TJ, et al. A clinical trial comparing rofecoxib to celecoxib and acetaminophen in the treatment of osteoarthritis (OA): early efficacy results. Ann Rheum Dis 2000;59(Suppl 1)

11. Houston M, Weir M, Gray J, et al. The effects of nonsteroidal anti-inflammatory drugs on blood pressures of patients with hypertension controlled by verapamil. Arch Int Med 1995;1949-54

12. Patrono C, Dunn MJ. The clinical significance of inhibition of renal prostaglandin synthesis. Kidney Int. 1987;32:1-12

13. Dzau VJ, Packer M, Lilly LS, et al. Prostaglandins in severe congestive heart failure. N Engl J Med 1984;321;347-352

14. Page J, Henry D. Consumption of NSAIDS and the development of congestive heart failure in elderly subjects. Arch Int Med 2000;160:777-784

15. Heerdink ER, Leufkens HG, Herings RM, et al. NSAIDS associated with increased risk of congestive heart failure in elderly subjects taking diuretics. Arch Int Med 1998;158:1108-1112

16. McAdam BF, Catella-Lawson F, Mardini IA, et al. Systemic biosynthesis of prostacyclin by clyclooxygenase (COX)-2: the human pharmacology of a selective inhibitor of COX-2. Proc Natl Acad Sci USA 1999;96:272-7

17. Van Hecken A, schwartz JI, Depre M, et al. Comparative inhibitory activity of rofecoxib, meloxicam, diclofenac, ibuprofen and naproxen on COX-2 versus COX-1 in healthy volunteers. J Clin Pharmacol 2000;40:1-12

18. Schedtje JF, Ji YS, Liu WL, DuBois RN, Runge MS. Hypoxia induces cyclooxygenase-2 via the NF-kappaB p65 transcription factor in human vascular endothelial cells. J Biol Chem 1997;272:601-608

19. Mukherjee D, Nissen S, Topol E. Risk of cardiovascular events associated with selective COX-2 inhibitors. JAMA 2001;286:954-959

20. Wallberg-Jonsson S, Johansson H, Ohman ML, Rantapaa-Dahlqvist S. Extent of inflammation predicts cardiovascular disease and overall mortality in seropositive patients with rheumatoid arthritis: a retrospective cohort study from disease onset. J Rheumatol 1999;26:2562-2571

21. FDA Advisory Committee. Cardiovascular Safety Review of Rofecoxib. Rockville, Md: Food and Drug Administration;2001

22. Shapiro DR, et al. Cardiovascular safety profile of rofecoxib: A meta-analysis. (Abstract). European League Against Rheumatism Congress 2001; Jun 13-16: Prague, Czech Republic

23. Antiplatelet Trialists' Collaboration. Collaborative overview of randomized trials of antiplatelet therapy-I: prevention of death, myocardial infarction, and stroke by prolonged antiplatelet therapy in various categories of patients. BMJ 1994;308:81-106

24. Silverstein FE, Faich G, Goldstein JL, et al, for the Celecoxib Long-term Arthritis Safety Study. Gastrointestinal toxicity with celecoxib vs nonsteroidal anti-inflammatory drugs for osteoarthritis and rheumatoid arthritis: the CLASS study: a randomized controlled trial. JAMA 2000;284:1247-1255

25. Daniels B, Seidenberg B. Cardiovascular safety profile of rofecoxib in controlled clinical trials. (Abstract) Arthritis Rheum 1999;42(9 suppl):S143

26. Daniels B, Seidenberg B. Cardiovascular Safety Profile of Rofecoxib in Controlled Clinical Trials (Abstract). Arthritis Rheum 1999;Suppl (435):S143.

Go to top

The world has changed. It is becoming smaller.

Whilst borders exist between nations, communications have broken down barriers, and the gaps that separate us are just a phone call away. We are now able to research and acquire knowledge beyond our expectations and dreams. Because of the massive linking and networking of computers, people are able to write and be read, by millions of others worldwide, in an instant.

With this has come the capacity for the ordinary man or woman in the street to research anything and everything possibly imaginable. All that is required is a simple computer and communication device, a modem, connected to a telephone line.

Somewhere out there for various reasons, people with expertise or personal opinion, have recorded their knowledge on computers in a way that is accessible by all who connect to the web of information, now known to all as the internet.

The Internet has allowed browsing to infinite computer sites worldwide. The network of sites is known as the World Wide Web ("WWW"). In fact, the Internet is more than just a web.

It has enabled communication through electronic mail between individuals or groups in a format known as email. It has allowed groups of people with common discussion points to meet in "chat rooms" and newsgroups. It has allowed transfer of information and data with a click of a button. It has almost infinite power. This includes both positive and negative aspects, but the balance is undoubtedly beneficial to the world at large.

The net is expanding every day. There are now over 2 billion web-pages, increasing at 7 million per day, and is likely to double within 2 years. However, this clearly produces problems. It is increasingly difficult to access 'good' information. How does one sort out the poorer quality website, from the sites with good information. It must be noted, that if the public cannot find a particular site, then no matter how good it is, it may as well not exist. The search for information returns hundreds of thousands of response. As a result one needs to sift out the 'deadwood amongst the trees'.

There are no limits who can put information on the web. There are no standards, and no controls whatsoever. Therefore, many parents have been alarmed about access of minor children to the web. However good far outweighs the bad, but one requires to be critical about one reads, bearing in mind that it represents an opinion, and not always fact.

This applies especially in the field of medicine, as people are bombarded with claims of cures and magical remedies. Beware of the trap. Many people prey on the sick for financial gain.

Several general "do's and don'ts are included in figure 1.

The medical field has lay and professional websites and information available. Both may be highly unreliable, and caution is required in using the information. Medical websites may in addition be too complex, but once the public gets used to the Internet and understands it, then the process gets much easier. See figure 2.

A number of good general search engines are available to search for information. These include, Excite, Yahoo, WebCrawler, Infoseek (Fig 3.) but programs are available for download that will integrate all of these in one search function. My favourite is to download and install a program called copernic, located at www.copernic.com .

Professional medical literature searching is now easy. In the past, what took a doctor weeks to research, is now available online in minutes. There are several search engines for medical articles. My favourite is healthgate. However the American national institute of health offers huge medical reference search engines. There is no harm in the general public researching their disease, although clearly this information, for the lay public, is likely to be difficult to understand.

The truth is that using the huge amount of information becomes much easier the more you make use of it. Everyone finds great sites, and it is advisable that you add these to your "web-browser favourites"

But what about meeting others online? There are many support groups out there who are forming "web communities". Many are available on a free subscriber or registration basis, such as the Microsoft network, or on America on-line. But the most accessible support groups are on the Internet newsgroups. There are two main arthritis Internet newsgroups. These are misc.health.arthritis and alt.support.arthritis

The newsgroups are accessible from your browser or email program and the requirements to access the newsgroups can be simply obtained from your Internet service provider who will guide you how to do this at no charge within minutes. The process is similar to how you set up your email addresses to access email.

On the newsgroup, people place messages and these are readable by all, and answerable by all who visit that particular newsgroup. Hence any question will result in responses from many people who have had that experience and can give appropriate advice. However, again the responses are opinions and not necessarily fact, and therefore should be treated with caution, and discussed with your doctor.

In South Africa, most doctors are now confronted by a host of Internet documents. The practitioner should not see this as a threat, and most do not. However, the public should appreciate that the doctor has limited time, and cannot read everything. A short summary of the questions you want to pose is appropriate rather than placing reams and reams of notes on the desk. I personally encourage my patients to read and research their problem. There is no harm in obtaining a second opinion, and there is no such thing as a stupid question. As practitioners, it is our obligation to answer the concerns of our patients. The Internet represents an enormous opportunity for the practitioner. It keeps the doctor up-to-date, as we now have to be able to answer the latest information, so it is incumbent on the doctor to be as interested as the patient in continuing medical education. Patient empowerment must be encouraged, and the Internet has revolutionized this.

In my own personal experience, I have found that ¾ of my patients have either direct or indirect access to the Internet, either through home, work or relatives on-line.

The World Wide Web is expanding and getting faster and more accessible to all.

I strongly recommend people to research their own disease and to ask questions.

What makes a site a good site?

1. Content - preferably original, relevant and up to date. 2. Reliability of source and established bona fides / references. 3. No claims of cure, guarantees or false promises 4. No demands for money for information 5. Stimulating presentation - but remember - content is king.

A table of some of my favourite arthritis links is noted below in Figure 5. There are too many to list, so only a key few are listed. Enjoy your web browsing.

Dr David Gotlieb 
Rheumatologist Constantiaberg Medi-Clinic 
drdoc on-line

Figure 1.

1.	Don't believe everything you read.
2.	Regard medical information with caution.
3.	Discuss your findings with your doctor.
4.	Do not change your therapies unilaterally.
5.	Don't give out personal details to everyone you meet on the Internet.
6.	Do not give out your credit card information, unless you are sure whom you are dealing with and that the process is secure.

Figure 2.
Software on-line

Software is easily available and usually free on the Internet.
Several basic requirements are suggested before downloading software programs. The main requirement is a good antiviral package to 
prevent damage to your own computer by malicious programmers, who insert and activate their computer code when you execute their 
programs. This can be done entirely without your knowledge and disseminated by you every time you send information such as email 
or programs to others.I suggest that you buy either Norton antivirus, or McAfee antivirus, and keep them updated by downloading updates on-line.

Software sites

www.download.com
www.tucows.com

Figure 3.

Good general search engines


Yahoo:		www.yahoo.com
Excite:		www.excite.com
Webcrawler:	www.webcrawler.com
Alta vista:	www.altavista.com
Ask Jeeves:	www.askjeeves.com


Good medical reference medline sites

	
NIH:
http://www.ncbi.nlm.nih.gov/pubmed/


It is a recommendation that you write your search criteria as simply as possible. 
For example, "rheumatoid arthritis and the lung" or "methotrexate and rheumatoid arthritis". Avoid using capital letters.

Fig 4. 
The background to the drdoc on-line website

Computers have always been my main hobby and I first became connected to the Internet in 1996. 
The medium allowed a natural combination of my love of Rheumatology, and my passion for computers and medical graphics.
One of the first things that I did was to become active in on-line interaction with and between patients on the Internet newsgroups.

 I became the rheumatologist on-line, consulting on the newsgroups and responding to email questions 
around the world. Because of the volumes of questions, many of which were repetitive, I started my own
 website in 1997. The aim was to provide the lay public with information that was practical, readable, 
understandable, and that answered every question out there related to arthritis. 
The articles and website design were all written and organised by myself. 
This grew to a large and busy site, all done after hours, at no cost. The website currently is one of the
largest websites in the world on arthritis, and has won several awards. It was the top website on the 
SAWebchart for several months and is still in the top 10. It received the Mail and Guardian newspaper 
award for top physician website in 1998. It was featured in an Internet symposium at the 
American College of Rheumatology annual meeting in Washington in 1997, and was in 
fact the only physician site demonstrated at that meeting.
The website receives thousands of visitors, and now has approximately 38000 page impressions 
(one of the largest of its kind WORLDWIDE) per month. It is unique as a resource of its kind, in arthritis. 
Letters stemming from the newsgroups now approximate 30 per day, all answered at no cost and all 
researched from on-line sources, as required.
The only limitation is time, since I still maintain an active full time rheumatology practice.

Fig 5. 
Essential links to information on rheumatology:

Societies and foundations:

1.	American college of rheumatology	http://www.rheumatology.org
2.	Arthritis foundation USA			http://www.arthritis.org
3.	Arthritis Society Canada			http://www.arthritis.ca
4.	Ankylosing spondylitis society		http://www.spondylitis.org
5.	Osteoporosis foundation			http://www.nof.org
6.	Lupus association			http://www.lupus.org
7.	Psoriasis association			http://www.psoriasis.org		
8.	SARAA website				http://www.arthritis.co.za/saraa.html


Patient based websites / Public arthritis websites:

1.	The arthritis pages of drdoc on-line			http://www.arthritis.co.za
2.	Arthritis at about.com			http://arthritis.about.com/health/arthritis/

Other interesting resources

1.	The arthritis better living spa		http://www.arthritisconnection.com
2.	National institute of health USA		http://www.nih.gov/health/
3.	Int'l league of association of rheumatologyhttp://www.ilar.org

Newsgroups

1.	Misc.health.arthritis
2.	Alt.support.arthritis

Go to top

GLUCOCORTICOID-INDUCED OSTEOPOROSIS - MECHANISMS AND MANAGEMENT
Dr Steve Soule

Department of Medicine
Endocrine-Diabetes Unit
University of Cape Town

Ph 021-4045007

Fax 021-4486815

ssoule@uctgsh1.uct.ac.za

"Spontaneous fractures occurred from time to time, involving sternum, clavicle and ribs. The autopsy showed skeletal osteoporosis with spinal curvature, the bones being easily cut with a knife … the spongy part of the bone having largely disappeared"

Harvey Cushing 1932

Therapeutic doses of glucocorticoids are widely used in the treatment of a variety of rheumatological conditions. Until relatively recently, however, the adverse effects of high dose glucocorticoids on the integrity of the skeleton have been inadequately appreciated, and patients have suffered as a consequence. Thus the 10-20% loss of trabecular bone in the first 3-6 months of high dose steroid use together with the subsequent slower loss of 2% per annum results in a vertebral fracture risk increased by 3-5 fold as well as a 50-60% increase in the risk of femoral neck fracture. Current estimates of the overall fracture prevalence in long term users of corticosteroids range from 30-50%.1 The end result may be a patient whose primary rheumatological condition is quiescent, but who is incapacitated by multiple vertebral fractures with associated degenerative spondylosis. In this review of glucocorticoid osteoporosis we will focus on recent advances in osteoclast and osteoblast biology which have enhanced our understanding of the pathogenesis of steroid osteoporosis and opened up exciting new options for therapy. We will also examine several well-designed therapeutic studies which have allowing us insight into the relative merits of available therapies.

PATHOGENESIS OF GLUCOCORTICOID-INDUCED OSTEOPOROSIS

The glucocorticoid receptor is widely distributed and there are undoubtedly many interacting factors contributing to steroid osteoporosis.
Osteoblasts
The recognition that osteoblast number is critically dependent on the balance between osteoblast maturation and apoptosis has particular relevance to steroid osteoporosis. In vitro studies have clearly demonstrated that corticosteroids not only inhibit osteoblast proliferation and matrix synthesis (by regulating a variety of target genes including type I collagen, osteocalcin, insulin-like growth factors and their binding proteins and the critical transcrition factor cbfa-1), but also promote osteoblast apoptosis by shifting the balance between pro-and antiapoptotic factors (BAX and bcl-2 respectively).2 It is thus not surprising that, following the initiation of steroid therapy, biochemical markers of bone formation decline substantially and histomorphometric studies reveal an impressive reduction in bone formation. The reduction in osteoblastic new bone formation is now regarded as the principle cause of the osteopaenia associated with steroid use.

Osteoclast
A word about osteoclast physiology is relevant at this point. Osteoclastogenesis and bone resorption are critically important to the integrity of the skeleton and tight homeostatic control is therefore essential. The receptor activator of nuclear factor kB (RANK) is a newly discovered member of the TNF family that is expressed on the surface of haematopoietic osteoclast precursor cells.3 The ligand is RANK ligand (RANKL) which is a transmembrane protein expressed on osteoblasts/stromal cells. The interaction between RANKL and RANK (i.e. osteoblast and precursor osteoclast respectively) induces differentiation to the mature osteoclast phenotype and thus enhances bone resorption.4 Of particular interest is that the osteoblast/stromal cell is capable of synthesising and secreting a soluble decoy receptor named osteoprotegerin (OPG) when stimulated by, for example, oestrogens. The secreted OPG binds in an auto-or paracrine fashion to RANKL on the osteoblast surface, thereby limiting interaction of RANKL with RANK and inhibiting osteoclastogenesis and bone resorption.

The relevance to steroid osteoporosis is as follows: in vitro studies have revealed a classic biphasic response of osteoclasts following steroid exposure. Initially a rapid and transient increase in the RANK ligand/OPG ratio prolongs osteoclast lifespan.5 However, more prolonged exposure to glucocorticoids causes both a profound reduction in the differentiation of osteoclast precursors and an increase in osteoclast apoptosis.6 The clinical correlate is that the majority of studies in humans on high dose steroids reveal reduced bone resorption both histologically and as measured by biochemical markers. Parenthetically, activated T cells also express RANK ligand: the resulting imbalance between OPG and excess RANK ligand could play an important role in the induction of excessive bone resorption in chronic arthritis, such as RA.7

Calcium homeostasis
The traditional view of the impact of steroids on skeletal mineral metabolism argues that glucocorticoids inhibit intestinal calcium absorption via a direct effect on the calcium transport system in the small intestine. Furthermore, calciuresis occurs secondary to a direct inhibitory effect of steroids on tubular reabsorption of calcium. The resultant negative calcium balance is postulated to cause secondary hyperparathyroidism and increased osteoclastic resorption. However, the evidence for secondary hyperparathyroidism in patients on long term steroids is at best inconsistent and histomorphometric studies in these patients usually indicate reduced osteoclastic activity (see above).8,9 It is thus unlikely that disturbed calcium homeostasis is the dominant mediator of the bone loss associated with high dose steroid use. This view accords with the limited beneficial effects of calcium or vitamin D in the prevention and treatment of steroid osteoporosis.

Hypogonadism
Both males and females exposed to high dose glucocorticoids may develop hypogonadism which may have a deleterious impact on the maintenance of bone mineral density. Approximately 50% of males on prolonged high dose steroids have a subnormal serum testosterone concentration, while oligo-amenorrhoea is frequently noted in females.10 Sex steroid deficiency is associated with both a reduction in osteoblastic OPG secretion and an increase in cytokine activity which together stimulate osteoclastic absorption. Assessment of gonadal function by monitoring the menstrual cycle in females and serum testosterone in males is thus important in patients on high dose steroids.

IMPACT OF GLUCOCORTICOIDS AND DISEASE ACTIVITY ON BONE MINERAL DENSITY

The interpretation of changes in bone mineral density (BMD) in patients with inflammatory arthritis is complicated by the adverse effects of the inflammatory process on bone density. Numerous studies indicate that in active RA, for example, bone resorption is increased and bone formation is normal or reduced when compared with normal controls.11 The pathogenesis of this altered bone metabolism is complex, involving non-disease specific factors (age, female sex, postmenopausal status) and disease-specific factors. The latter include disease activity (inflammatory cells, cytokines and hypogonadism), disease outcome (reduced mobility) and disease medication (especially corticosteroids). The resultant uncoupling of bone formation and resorption leads to generalised bone loss which may confound the assessment of the independent effects of glucocorticoids on BMD. Thus a prospective study by Gough et al of 148 patients with early RA followed without corticosteroid treatment emphasised several important points.12 First, disease activity was independently associated with loss of BMD in the lumbar spine, amounting to 5-10% over 2 years. Second, suppression of disease activity using modalities other than corticosteroids stabilised the bone loss.

Despite the proven adverse effects of active inflammatory disease on BMD, multiple studies now convincingly demonstrate the independent deleterious effects of corticosteroids. Within hours of exposure to glucorticoids, there is a fall in circulating osteocalcin, a bone matrix protein synthesised by osteoblasts. This is followed by a rapid fall in bone mass that is maximal over the first 3-6 months of therapy. The study of Laan et al in patients with early RA demonstrated an average loss of 8% of trabecular bone and 2% of cortical bone over a 20 week period in response to treatment with a mean dose of prednisone of only 7.5mg/day.13 Although the reduction in BMD is maximal within the first 3-6 months of glucocorticoid therapy, the evidence suggests that there is continued bone loss in most patients after several years of therapy, albeit at a reduced rate. Thus Saito et al were able to show losses 2-3 times greater than matched controls in elderly patients who had been using glucocorticoids for a mean of 2 years.14 Consequently, BMD in corticosteroid treated subjects studied cross-sectionally relates to multiple variables including age and body weight (determining baseline BMD) together with duration and dose of glucocorticoid treatment (determining rate of loss of BMD). Unfortunately, alternate day administration of steroids does not diminish bone loss. Furthermore, there is data suggesting that high dose inhaled steroids may also impact negatively on bone density - a doubling of the cumulative dose of inhaled steroid in a large cross-sectional study was associated with a 0.16 SD reduction in BMD. The cumulative effect of inhaled steroids may thus have important consequences on fracture risk later in life.15

The question often arises as to whether a subset of corticosteroid treated patients is immune to the adverse effects of steroids on BMD. Reported polymorphisms in the glucocorticoid receptor and the observation that in some cross-sectional studies certain subjects treated with long term steroids have a normal BMD would provide some support for this possibility. However, there are several convincing arguments suggesting that corticosteroid-induced bone loss is a universal phenomenon. Firstly, one should recall that for DEXA of the lumbar spine, for example, the normal range extends from 80 to 120% of the mean. Therefore an average loss of 20% following steroid use will result in the range in this population extending from 60 to 100% of the mean. It is therefore not surprising that certain patients treated with steroids are found to have a normal BMD, since the mean value has been reduced by an amount less than the width of the normal range. Secondly, biochemical markers of osteoblastic function are universally reduced after exposure to high dose steroids, arguing against the existence of a cohort of patients protected from the adverse skeletal effects of steroids.16 Finally, the fact that BMD in corticosteroid treated patients remains normally distributed with a standard deviation identical to that in control subjects would support the concept of universal reduction in BMD following long term exposure to corticosteroids.17

A further important issue relates to recovery of BMD after cessation of steroids. There is, in general, substantial reversibility of steroid-induced bone loss. For example, after surgical cure of Cushing's syndrome there is a marked increase in bone density, which normalises in subjects cured for a mean of 9 years.18 Likewise Laan et al, in a prospective study of bone density in RA patients receiving treatment with low dose steroids, demonstrated that lumbar BMD increased significantly after withdrawal of steroids, while hip BMD normalised.12 There are, however, data suggesting that 5 years after ceasing steroids the non-vertebral fracture risk remains 20% greater than in controls - this may merely reflect the fact that one needs longer term follow up for BMD and fracture risk to normalise.19

We can conclude that bone loss is a virtually universal feature in patients exposed to high dose steroids, relates predominantly to reduced osteoblastic bone formation and may amount to a 20% loss of bone density in the lumbar spine and proximal femur after 5 years of use. Moreover, studies suggesting that fracture rates in steroid treated patients at any given BMD are higher than in control subjects hints at unmeasured adverse effects of steroids on bone architecture, and further strengthens the argument for detailed evaluation of bone density in patients about to embark on treatment with high dose steroids.

INTERVENTIONS TO MAINTAIN OR INCREASE BONE DENSITY

Calcium and vitamin D supplementation
In several randomised studies of the impact of glucocorticoids on BMD the control group received calcium supplementation and yet continued to suffer a substantial reduction in bone density.20 This clearly indicates that considerable bone loss occurs despite calcium supplementation, despite data demonstrating a reduction in biochemical markers of bone turnover in calcium treated patients exposed to high dose steroids. For example, Sambrook et al studied changes in BMD in over 100 patients commencing steroids (mean prednisone dose 13.5 mg) and randomised to one of three treatment arms (1) calcium 1000 mg alone (2) calcium and calcitriol 0.5-1 µg daily (3) calcium, calcitriol and calcitonin 400 IU intranasally daily. The patients who received calcium prophylaxis alone showed a mean loss of BMD at the lumbar spine and femoral neck of 4.3 and 2.9% respectively after the first year, proving that calcium supplementation is inadequate to prevent clinically significant reductions in bone density in corticosteroid treated patients.21

The prophylactic role of native vitamin D (cholecalciferol) in combination with calcium remains rather contentious after two contradictory studies were recently published. Buckley et al studied patients with RA treated with low dose prednisone and randomised to receive either placebo or a combination of calcium 500 mg per day plus vitamin D 500 IU per day. The active treatment arm showed a 2% more positive change in BMD than the placebo arm.22 However, in the study of Adachi et al the use of cholecalciferol 50 000 IU once weekly plus calcium 1000 mg daily had no impact on the reduction of lumbar spine BMD in a cohort of steroid-treated patients as compared with placebo.23

We can conclude that the data to support calcium and vitamin D supplementation is contradictory. However, the documented potential for steroids to induce a negative calcium balance together with the positive effect of calcium on limiting bone turnover in steroid treated patients suggests that all steroid treated patients should have a calcium intake of 1-1.5 g daily - this can be achieved either by 4-6 dairy servings daily or alternatively by a variety of available calcium supplements. Vitamin D supplementation should probably not be universal but should be reserved for those steroid-treated patients with documented vitamin D deficiency (low 25-hydroxyvitamin D) or at high risk of vitamin D deficiency (frail, housebound, resident in Northern hemisphere).

Active vitamin D metabolites
The variety of different vitamin D metabolites available (25-hydroxyvitamin D, calcitriol and alphacalcidol), and the different doses used for the prophylaxis or treatment of glucocorticoid-induced osteoporosis renders interpretation of the available studies difficult. One of the most useful studies is that of Sambrook et al mentioned above, where patients about to commence steroids were randomised to receive calcium, calcium plus calcitriol (mean 0.6 µg daily) or these two agents combined with nasal calcitonin.21 Bone loss from the lumbar spine was 4.3%, 1.3% and 0.2% respectively, although no protective effect of any therapy was evident in the proximal femur. There was no significant benefit of adding calcitonin to the calcitriol. In contrast, a trial comparing calcitriol 0.5 µg daily with hormone replacement in hypogonadal young women with SLE revealed a progressive reduction in BMD in those receiving the vitamin D analog, while there was an increase in BMD in those randomised to oestrogen replacement (resultant 3.7% between group difference at the spine at 2 years).24

The data for alphacalcidol (1-alpha-hydroxycholecalciferol) is rather more consistent with positive effects on bone density at the lumbar spine demonstrated in patients post cardiac transplant, in non-transplant patients commencing corticosteroids and in a population of patients with established osetoporosis.25-27 There are, however, no reports of measurable benefits of alphacalcidol on femoral bone density and no positive fracture data are available for either native vitamin D or any of the several available vitamin D metabolites. These agents are thus best reserved as adjunctive therapy in patients treated with oestrogen replacement or bisphosphonates or alternatively as second line therapy in patients for whom these other agents are unacceptable.

Bisphosphonates
The bisphosphonates, analogs of pyrophosphate, are rapidly incorporated into bone matrix and are the most extensively researched forms of therapy available for the prevention and treatment of steroid osteoporosis. These agents act principally to inhibit osteoclastic resorption (by inhibiting both the recruitment of osteoclast precursors and the function of mature osteoclasts) and hence reduce bone turnover, allowing filling in of existing resorption cavities. Interestingly, recent data suggests that the bisphosphonates also prevent steroid-induced osteoblast and osteocyte apoptosis.3 There is a wealth of clinical data supporting the positive effects of these agents on bone density both in the setting of the prevention and treatment of glucocorticoid-induced osteoporosis.28 Groundbreaking studies with oral Pamidronate (Aredia®) demonstrated a dramatic 19% increase in lumbar spine BMD after 12 months in steroid treated patients, as compared with a 9% decrease in those receiving placebo.29 The oral formulation is not available locally, although three monthly infusions of Pamidronate (90 mg ivi initially, then 30 mg 3 monthly) appear to produce broadly comparable increases in bone density.30 Alendronate (Fosamax®) has similar proven efficacy in the setting of steroid-induced osteoporosis. The study of Saag et al randomised 477 patients either commencing or already established on oral glucocorticoids (median dose 10 mg prednisone daily) to calcium plus vitamin D or to these agents in combination with alendronate 5 or 10 mg daily.31 Bone mineral density increased at the spine (2.1 and 2.9% in the alendronate 5 and 10 mg groups respectively), trochanter (1.1 and 2.7%), femoral neck (1.2 and 1.0%) and total body (0.4 and 0.7%) in those treated with alendronate. Benefit was evident in men, premenopausal and postmenopausal women and occurred irrespective of the duration of previous glucocorticoid use. It is worth emphasising that the efficacy of the 5 and 10 mg doses of alendronate was similar, except in postmenopausal women not receiving oestrogen therapy in whom the higher dose was more effective. Furthermore, there was a 50% reduction in new vertebral fractures, although this only reached significance in the subgroup of postmenopausal women. The newer bisphosphonate risedronate (Actonel®) has also proven effective in the prevention and treatment of glucocorticoid osteoporosis. The changes in bone density when compared to placebo were broadly similar to those noted with alendronate, although data obtained from pooling the two studies indicated over 50% reduction in new fractures in the first year of treatment.32,33 One may conclude that the bisphosphonates are not only effective in maintaining and increasing bone density in glucocorticoid treated patients, but are the only agents conclusively proven to reduce vertebral fracture rate.

Sex hormone replacement
Despite the wealth of information delineating the hypogonadism induced by glucocorticoids in both men and women, there is a paucity of controlled data on the impact of hormone replacement on bone density in these patients. Moreover, there are no studies reporting fracture data. This dose not, of course, imply that hormone replacement is ineffective in the prevention and treatment of glucocorticoid osteoporosis, but rather reflects the lack of pharmaceutical company interest in funding such studies. Indeed, the in vitro observation that oestradiol reverses the osteoblast apoptosis induced by dexamethasone suggests that sex steroids may have a specific anti-glucocorticoid effect.34

There is certainly no evidence to support a role for HRT in premenopausal women with a regular menstrual cycle or in men with a normal serum testosterone concentration. However, in three randomised trials in postmenopausal women, HRT produced an increase in BMD at the lumbar spine ranging from 0.7 to 13.7%, while no benefit was seen at the hip. Similarly, in the only two randomised studies in hypogonadal men, testosterone replacement resulted in BMD increase at the lumbar spine of 5 and 17%.35 Thus although the replacement of sex steroids in hypogonadal patients appears to produce consistent increases in lumbar spine bone density, there are presently no sufficiently large studies to document anti-fracture efficacy.

Calcitonin
Calcitonin binds to receptors on osteoclasts and inhibits bone resorption. Several randomised studies of parenteral calcitonin have demonstrated that bone loss is slowed in the lumbar spine in the active treatment group when compared with controls. The work of Luengo and Rizzato, for instance, documented a difference of 6.5 and 12.5% respectively in lumbar BMD after 12-15 months in steroid treated patients receiving parenteral calcitonin when compared to control subjects.36,37 Several problems have, however, limited the widespread use of parenteral calcitonin. These include the high prevalence of side effects, the poor patient acceptability and the fact that the only randomised placebo-controlled trial failed to demonstrate a significant benefit in the active treatment group. Furthermore, only one study reports a protective effect at the hip and there are no fracture data.38

The availability of the intranasal preparation has stimulated a re-evaluation of the role of calcitonin in the prevention and treatment of steroid osteoporosis. Adachi et al studied 31 patients starting glucocorticoid therapy for polymyalgia rheumatica randomised to receive nasal calcitonin 200 IU daily or placebo over a 12 month period. Calcitonin was marginally more effective than placebo in reducing the rate of bone loss from the lumbar spine since mean spinal BMD changed from 1.06 to 1.04 g/cm2 (-1.29 ± 6.76%) in the calcitonin group and from 1.11 to 1.08 g/cm2 (-4.95 ± 3.5%) in the placebo group over 12 months.39 Although this between group difference was statistically different, the clinical relevance of these small changes in bone density remains questionable. In the proximal femur the intergroup differences were of similar magnitude, but favoured placebo. It therefore seems reasonable to conclude that there are inconsistent data supporting the prophylactic and therapeutic use of calcitonin, both parenteral and intranasal, in the setting of glucocorticoid osteoporosis and that this agent should not be used as first line therapy.

Flouride
Flouride is theoretically an attractive therapeutic agent for glucocorticoid osteoporosis since it specifically stimulates osteoblast proliferation by inhibiting phosphotyrosine phosphatases (PTPases), hence promoting tyrosine phosphorylation of the MAP-kinase signal transduction pathway. Although impressive increases in bone density may result, reports of increased vertebral fracture rate in non-steroid exposed patients treated with high dose flouride have dampened enthusiasm for this agent.40 Provided, however, one uses the correct dosage (25 mg twice daily) and formulation (monoflourophosphate), there are data supporting the use of flouride for both the prevention and treatment of glucocorticoid osteoporosis. Lems and co-workers studied 44 patients commencing glucocorticoids randomised in a blinded fashion to flouride 25 mg bd or placebo.41 After 2 years the difference in bone density at the spine was +5.2% (P<0.01), and at the hips +0.8% (P>0.05), prompting the conclusion that in steroid treated patients without osteoporosis, flouride prevents bone loss at the lumbar spine. A second study in patients with established steroid osteoporosis revealed that flouride produced additional increases in lumbar spine BMD when added to Etidronate.42 Concerns about dosing and safety, the lack of fracture data and the fact that there is no increase in hip bone density following the use of flouride suggest that this agent should only be used cautiously by experienced practitioners as adjunctive therapy in patients with severe bone loss. Suggested strategy for evaluation and treatment
Clearly the best method to prevent steroid osteoporosis is to use glucocorticoids as sparingly as possible - one should therefore aim to use the lowest possible dose for the shortest possible period and consideration should also be given to the use of steroid sparing agents. Unfortunately the reported bone-sparing properties of deflazacort have not been substantiated by recent data which suggest that the potency of this agent has been overestimated, resulting in the use of dosages which were subtherapeutic.43 The rapid and substantial reductions in bone density that occur within the first 3-6 months of steroid therapy (prednisone>7.5mg/day) emphasise that the optimal approach is preventive rather than therapeutic. Figure 1 illustrates a suggested outline for the prevention and treatment of steroid osteoporosis based upon currently available data. It should be realised that after commencing glucocorticoids the bone density will usually drop a further 1-2 standard deviations in the first year, and this needs to be factored into the decision making process. The precise cutoff of BMD for initiating specific therapy to maintain bone density is arbitrary and needs to be individualised depending on, amongst other factors, patient age, expected dose and duration of steroid use, existing fractures and life expectancy. In the absence of clear evidence, however, most would accept that a T-score (number of SD's above or below young normal mean value) less than minus 1 or certainly minus 2 would warrant intervention.

Conclusion
Glucocorticoid-induced osteoporosis is a common, debilitating and largely preventable condition, now understood to predominantly reflect imbalances in the birth and death rate of osteoblasts. The availability of several effective preventive and therapeutic strategies places a responsibility on the physician prescribing glucocorticoids to consider the possibility of bone loss and fracture and to intervene both timeously and appropriately.

Go to top