The overlap syndrome

There is more than one 'overlap syndrome' in medicine.  For respiratory and sleep physicians, however, the term generally refers to patients with the combination of chronic obstructive pulmonary disease (COPD) and obstructive sleep apnoea (OSA).

Such patients will often have profound oxygen desaturation associated with their OSA.  They are also more likely than patients with either syndrome alone to develop pulmonary hypertension.

A joint Spanish-American study in the blue journal this month has looked at this condition in quite a helpful way.  They wanted to find out if patients with the two of these conditions together did worse than patients with just COPD.  And if they did worse, did treatment with CPAP seem to make any difference?  They took three groups of around 200 patients each.  One group had patients with the 'overlap syndrome' on no treatment.  Another group had patients with the 'overlap syndrome' on CPAP.  The third group was patients with COPD alone.

They recruited patients for 5 years from 1996 to 2001 and then followed them. The patients were selected from patients referred to a sleep clinic - so the patients with COPD alone were chosen from amongst a population of snorers who had no OSA.

Amongst those patients with OSA (defined as an apnoea-hypopnoea index over 5) and COPD, CPAP was recommended to all.  So the 'not treated' group were those who ended up being non-compliant with CPAP therapy.  As always, this raises the question of whether they make other good health-related choices (ie is non-compliance with CPAP a marker of bad health behaviours rather than in itself an unhealthy choice?).  

The bottom line was that 213 patients died.  Death from any cause, and particularly death from cardiovascular causes, were more frequent in untreated overlap patients (RR 2.23, C=95%CI 1.59-3.14)than in the overlap/CPAP treated group.  There was no significant difference between the overlap/CPAP group and the COPD -alone group.

Likewise, untreated overlap patients were more likely to have a first admission to hospital with an exacerbation (OR 2.13, 95% CI 1.61-2.80) than COPD only patients.

So, what have we learned?  In patients with COPD, the coexistence of OSA is a poor prognostic marker, indicating higher risk of hospitalisation with exacerbation and higher risk of death.  That increased risk is not evident in patients with COPD and OSA who use CPAP. 

This means that prescription of CPAP for COPD patients who have OSA is one of the very few interventions which may make a difference to mortality outcomes and is, therefore, VERY IMPORTANT.

Antibiotics in COPD exacerbations? Show me the sputum!

Are antibiotics needed in treatment of acute exacerbations of COPD (AECOPD)?

I have been reading a paper in the Blue journal from mid January which tries to address this question. As it was deemed editorial – worthy, it caught my eye.

To date, we administer antibiotics to patients with exacerbations of chronic obstructive airways disease based on the Anthonisen criteria. These were defined for a paper published in 1987 – now some years ago – and take into account increasing breathlessness, sputum volume and sputum purulence. There were limitations to that trial, not least among which was the absence of corticosteroids, such as prednisolone, from treatment regimes. We know that corticosteroids hasten clinical and physiological recovery and shorten hospital stays in patients with COPD. We don’t know that about antibiotics.

This recent, Dutch, study randomized 223 patients who between them suffered 265 episodes of exacerbation to receive either doxycycline 200mg daily for 7 days or placebo, in addition to a standardized treatment regime which included intraveinous and then oral prednisolone. Patients likely to have pneumonia (xray signs, fever) were excluded reasonably well. The primary endpoint was clinical response on Day 30, meaning that their symptoms and signs had either completely resolved or improved without getting worse again.

They found the following:
• At day 30 there was no significant difference between the groups
• At day 10, however (before which time you would generally hope to be sending someone home from hospital – the norm being around 7 to 9 days in our sort of institution), there was a slight difference in favour of the treatment group in terms of clinical response (80 vs 69% clinically responded. P=0.03). 67% of the doxy group were ‘cured’ at day 10, vs 51% of the placebo group (P=0.01). The difference was not statistically significant at day 30.
• More patients were changed from placebo to open-label antibiotic because of lack of efficacy than from doxy to open-label antibiotic for the same reason, at day 10 (15vs28%, P=0.01) and day 30 (33vs45%, P=0.13)
• Symptom scores were more improved at day 10 for the doxy group than for the placebo group, but not at day 30
• Microbiological outcomes varied at day 10. Common bugs were H.influenzae (41%), Pneumococcus (24%), Moraxella catarrhalis (22%). 52 of 78 patients in the doxy group who grew bacteria were cleared (67% ) vs 25 of 73 patients (34%) in the placebo group.
• 46 of the doxy group and 62 of the placebo group (37 vs 46%) had treatment failure.
• Withdrawal from the trial was similar in both groups, but 23 (17% ) of the placebo group withdrew for lack of efficacy, as opposed to only 8 (6%) of the doxy group withdrawing for that reason

The doxy seemed, in sub-group analysis, particularly beneficial in patients with increased sputum volume and purulence and a C-reactive protein of over 50.

The question raised by the article, and also by the editorial, is whether antibiotics really need to be added in to treatment regimes for patients with acute exacerbations of COPD. I suppose there are two ways of looking at this.

From my perspective, as doctor and holder of the keys to the pharmacy, it may well be correct to say that absence of benefit at 30 days is reason to withhold antibiotics from patients with acute exacerbations of COPD. It should be pointed out, however, that we all use corticosteroids in AECOPD in the absence of precisely the same benefit.

Why do we do this? Well, perhaps here we consider the patients perspective. We use corticosteroids because patients feel better faster, and because they stay in hospital for less time. We like it when our patients feel better faster, because psychologically any improvement is linked to our intervention, which in turn encourages future adherence to suggested treatment regimes.

If my doctor tells me that they are withholding a treatment with a narrow-spectrum antibiotic that will probably make me feel better quicker because it won’t still be having that effect in a month, I would probably be a little annoyed.

I would like to see a study setting out to prove whether what we already know about steroids. applies to antibiotics too: Do some patients feel better quicker, and get out of hospital sooner, if they take antibiotics –regardless of the effect at 30 days? Perhaps the study has been done. Time to get on PubMed.

Of course the argument is more complex than this. Antibiotic resistance is a major public health problem. However, the very real clinical issue of how we maintain our patients’ confidence by introducing treatments that work to help them achieve tangible health improvements, didn’t make it into the discussion about this paper in the AJRCCM.

Andrew

Can Spiriva® save lives?

Tiotropium (Spiriva®) is almost universally prescribed for treatment of chronic obstructive pulmonary disease. It has been demonstrated to improve lung function (better spirometry, less air trapping) and health-related quality of life, and to reduce the frequency of exacerbations and hospitalizations in studies that compared use of tiotropium to placebo.

But can it save lives?

To date, several interventions in particular clinical settings in patients with COPD have been demonstrated to improve survival. These include; smoking cessation; use of supplemental oxygen in patients with persistent and profound hypoxaemia (low blood levels of oxygen); use of non-invasive ventilation in patients with ventilatory failure (high levels of carbon-dioxide in the blood; and even lung volume reduction in a very select group of patients. An analysis of data collected in the previously-published UPLIFT trial was released in the Blue Journal in mid-November, and suggests that tiotropium used regularly may be added to this list.

This was an interesting study. Two groups of patients (totaling 5993 patients) with COPD were randomized to receive either tiotropium or placebo. They were allowed to take any other inhalers except anticholinergics, and were even allowed to smoke (30% of them were smokers). Many patients discontinued the drug (44.6% in the placebo arm and 36.2% in the treatment arm) but continued in the study over its four year duration, and then a 30 day follow up period, during which tiotropium was stopped and ipratroprium (Atrovent ®) administered. .

During the 1440 days of the treatment period, 411 patients died while receiving placebo and 381 while receiving tiotropium (Hazards ratio (95% CI) 0.84 (0.73-0.97) p-0.016). If those in the two groups who discontinued treatment were included there was still a significant difference ( 491 vs 430, HR 0.87 (0.76-0.99), p-0.034). However, the follow-up period was a particularly bad one for those in the tiotropium group. During that period, 16 patients from the tiotropium group died, as opposed to only four from the placebo group. Six of those ‘tiotropium group’ deaths were patients who had remained on the medication right up until the end of the study – raising the question of whether withdrawal of the medication may have been harmful. By the end of the follow-up period, the survival difference was not statistically significant.

In a sub-group analysis it seemed as if smoking status at baseline might be a significant differentiator. Not too much was made of this in the discussion, but certainly for those patients still smoking there was no difference in risk of death between the two groups (medication vs placebo).

I think that the study design was a bit messy, including smokers and non-smokers and with a real hotch-potch of other inhalers in play. In this regard, however, it can certainly claim to be like real life. This recently-published new analysis of old data from the UPLIFT study gives us even more reason to prescribe tiotropium, and encourage our patients to remain on it.

Andrew

Symbicort catching up with Seretide in COPD.

An article published in the blue journal last month, accompanied by an editorial, draws to light two related phenomena. Firstly, the combination of an inhaled corticosteroid, long-acting beta-agonist and tiotropium is beneficial in chronic obstructive pulmonary disease. (This, at long last, gives respiratory physicians the opportunity to use 'triple therapy', a phrase which has an inexplicable allure for medicos). Secondly, for Astra-Zeneca and Symbicort (budesonide/eformoterol) it's all about catching up with Glaxo and Seretide (fluticasone/salmeterol). And they just might do it yet.

The study was a randomised, placebo controlled trial of Spiriva (Tiotropium) alone vs Spiriva and Symbicort (to use the local trade names) in combination. The Symbicort dose was 320/9mcg, which is not a dose marketted in Australia (where we have 100 or 200/6 or 400/12). Around 600 subjects were recruited, 300 or so in each arm. Subjects had to have pretty severe COPD, with an FEV1 of less than 50% predicted on their initial pre-bronchodilator spirometry, although 25% ended up categorised as moderate if the GOLD - compliant definition of post-bronchodilator FEV1 was used.

The goal of the study was to demonstrate improvement in lung function - which they did. There was a 6% improvement in pre-bronchodilator FEV1 at follow up in the 'triple therapy' (TT) group compared with the Spiriva-alone group. There were also more subjects in the TT group with a significantly improved quality of life. Most strikingly, there was a very significant (62%) reduction in exacerbations in the TT group (7.6% of patients vs 18.5% of patients) over the short 12 weeks of the trial. (It's worth noting that the definition of exacerbations here required a hospital attendance, not just an escalation of therapy with glucocorticoids directed by a doctor).

The benefits of this combination of inhaled medications in COPD now seems beyond dispute. There seems to be a class effect (ie Symbicort and Seretide are each beneficial) and we can expect Symbicort to get PBS listing in Australia at some point for patients with moderate COPD, although it does not yet currently have this.

The accompanying editorial in the blue journal does point out the following important points:

1. Although Seretide has not been demonstrated to reduce the frequency of exacerbations, the trials of Seretide in COPD had a more inclusive definition of exacerbation and the drugs can't be compared on this point on the data already to hand;
2. 40% of patients in this trial were already on 'triple therapy' before the trial, and the inhaled steroid and LABA had to be stopped. This may just be an example of trials catching up with clinical practice;
3. Not enough patients of moderate disease severity have been enrolled in the trials. Aparently a recent post-hoc analysis of the TORCH data (one of the Seretide trials) showed that patients categorised as moderate by GOLD criteria (post-bronchodilator FEV1 50-80%) had quite poor prognosis generally in terms of rate of loss of lung function, deterioration of quality of life and rate of exacerbations. If COPD is now being modelled as an inflammatory disorder and patients with relatively early disease do progress why are we not investigating the impact of antiinflammatory inhalers on earlier-stage disease?

So why do I think that Symbicort is working hard to catch up to Seretide, and that the unwritten primary-endpoint here was to win market share in patients with COPD? It strikes me that Astra-Zeneca learnt from the failure of Glaxo to demonstrate a survival-benefit for the Seretide/Spiriva combination in a big trial published in the New England Journal in 2007. Although a lot of useful secondary endpoints were hit in that trial, it is remembered as the trial that missed the jackpot. The current trial has notably not tried to aim so high. It is a short trial, with relatively few patients, and looks to me very much as if is has been put together to demonstrate that Symbicort is also 'good enough' to use in patients with COPD.

The trial authors list the following as goals of treatment in COPD:

• reducing symptoms
• improving health status and exercise tolerance
• reducing the risk of exacerbations.

These management goals are actually limited by the efficacy of the treatments we have available. They were designed for clinicians, not researchers. So having stuck to the established management goals, Astra-Zeneca have now proved that Symbicort helps to do all of what we are already doing better. It strikes me as circular thinking, and a little disappointing. Give me something new to offer my patients! Please.

Notably absent from the list of treatment goals is to 'reduce disease associated mortality'. It's not there as a goal because, apart from smoking cessation and oxygen in select patients, we don't know how to do it yet. Mortality reduction is the most meaningful of all possible treatment goals in established medical research traditions, but I suspect it is beyond the potential of our current medications in COPD.

Andrew

COPD and stiff arteries

COPD is increasingly prevalent, and an increasingly common cause of death. It is currently the fourth leading cause of death in the USA. Patients with COPD often die of other diseases, such as cardiac disease - which is currently the leading cause of death in that country and ours. In the last 10 years there has been increasing recognition that COPD is a disease with systemic effects, beyond the lungs. Death from ischaemic heart disease has been associated with lowered FEV1 (a fivefold in death from AMI as FEV1 fell from 109 to 88% of predicted in a US population). However, establishing how COPD may be a discrete risk factor in cardiovascular disease, distinct from the common risk factor of cigarette smoking, is proving difficult.

A paper in the September 15th edition of the AJRCCM (180 pp513-520) has shed some light on this subject, and scored an accompanying editorial. In this study 18 men with COPD were matched (smoking history and age) with 17 controls who did not have COPD. Pulse wave velocity and pulse wave analysis at carotid, radial and femoral arteries indicate significantly increased arterial stiffness in the subjects with COPD. A complicated array of tests of endothelial function did not indicate any difference between the groups.

Central arterial stiffness is, according to the article and the editorial, a predictor of cardiovascular mortality. It is usually due to either smooth muscle or endothelial dysfunction or elastin loss - or a combination of all three. Given the equivalent endothelial function between the groups in this study, the authors surmised that COPD may lead to elastin degradation in arterial walls. This may be a means by which COPD leads to increased cardiovascular risk independent of other risk factors.

Other studies have suggested that there is endothelial dysfunction in COPD, and that it is worse in patients with lower FEV1 or more extensive emphysema - evidence which is not in agreement with that presented in the recent study. Obviously more work - much more work -needs to be done (I hate hearing that line at conferences, but there it is!)

Anyway, the most tantalising part of the picture is the question of clinical relevance. All of this cardiovascular science is news to me, but it turns out that simvastatin has been shown to reduce arterial stiffness in patients with rheumatoid arthritis. There is also some evidence that statins reduce lung parenchymal destruction and pulmonary vascular remodelling in smoking-related lung disease. So although this is far from being evidence based medicine, we might have some reason to anticipate statins becoming part of routine respiratory medicine in the future.

Andrew

Pulmonary rehab

In the last decade pulmonary rehabilitation has become a mainstay of management of COPD. Improvements in quality of life and exercise tolerance are generally the goals of pulmonary rehab programs, which should run for at least 8 weeks at a time. In our small country communities, pulmonary rehabilitation programs have resulted in apparently significant reductions in utilisation of hospital resources over winter.

These programs generally include supervised exercise and education. I have not, much as I love to sing, been involved with choral pulmonary rehab.

A friend recently brought to my attention this video clip - a moving rendition of the Coldplay song 'Fix You'. The song is sung by an aged gent with a rich baritone - and an oxygen concentrator at his side - backed up by the 'Young at heart' choir. I found it moving, and inspirational.

Andrew

Remembering theophylline

My registrar recently asked me why some of my patients with chronic obstructive pulmonary disease (COPD) were on theophylline - the implication being that this is an outmoded drug, and I am an anachronistic practitioner.

Actually, it was put very politely - but the question hung in the air for a moment before I answered.

Theophylline is an old drug - having been used first for the treatment of asthma in the 1950s - which has long been recognised to relax airway smooth muscle. It has retained its place in the treatment of chronic obstructive pulmonary disease over decades. Every consensus statement on the management of COPD includes a place for methylxanthines, which in Australia means either oral theophylline or intraveinous aminophylline. Its use predates the era of 'evidence based medicine' and, as the medication is long out of patent, the resources being put into research of its effects are limited. Nevertheless there is ongoing research into the impact of theophylline in the airways (and the kidneys).

Most of the recent research focuses on the potential role of theophylline to reverse steroid resistance in COPD. A recent review in the Lancet (2009,373;1905-17) by Professor Peter Barnes covers this ground in great detail. A study published in Thorax journal in May (64(5);424-9) randomised a small number of patients (35) to receive either steroids and placebo or steroids and low dose theophylline during an exacerbation. All patients demonstrated evidence of relatively reduced airway inflammation over time - out to 3 months - but there was evidence of a greater reduction in some inflammatory markers (particularly histone deacetylases) in the theophylline group. The study did not evaluate clinically significant parameters.

Incidentally, the European Journal of Clinical Investigation has a forthcoming article (2009 Sep;39(9):793-9) reporting a study in which 217 patients with impaired renal function (GFR 30-60 ml/min) undergoing coronary angiography were randomised to receive either IV saline, IV saline + n-acetyl cysteine or IV saline + NAC + theophylline 200mg twice daily for the day prior and on the day of the angiogram. Contrast induced nephropathy occured in 5, 7 and zero patients in each of the groups respectively.

So, if theophylline protects the kidneys and reduces inflammation in the airways, why are we so slow to use it?

Side effects! Theophylline needs to be kept within a therapeutic range. Caffeine-like side effects occur in many people (nausiea, vomiting, headache, insomnia) at initiation of treatment - and persist in up to 10% . In overdose, side effects can include intractable vomiting, cardiac arrhythmias and seizure. So it's not an entirely safe drug.

Nevertheless, it's a drug with a place - and an old drug that may end up with some real evidence to support our continued use of it after all!

Andrew