inhaled steroids for COPD may increase pneumonia

The data are pretty mixed on the efficacy of inhaled corticosteroids (ICS) in patients with COPD, yet in some studies 85% of COPD patients are on them. in this light, there a recent large case-control study found that weaning some patients off ICS found a dramatic decrease in pneumonia cases !!! (see copd inhaled steroid PNA chest2015in dropbox, or CHEST2015; 148(5): 1177). details:

--case-control study from Quebec, looking at a new-user cohort of patients with COPD (mean age 78.6, 50% male) put on ICS during 1990-2005, followed thru 2007 or until a serious pneumonia event (first hospitalization or death from pneumonia).

--they compared those who discontinued ICS, vs those who continued, to assess development of a serious pneumonia event, adjusting for age, sex, comorbidities, and respiratory disease severity (which they defined as: number of prescriptions for b-agonists, ipratropium and tiotropium, theophylline, oral steroids, and antibiotics; and if there were hospitalizations with primary diagnosis of COPD; all measured in year before index date)

--results:

    --103,386 users of ICS, of whom 14,020 had a serious pneumonia event during 4.9 yrs of follow-up (2.8/100/yr); those who developed pneumonia  were sicker overall (more COPD meds, antibiotics, COPD admissions, recent steroid use)

    --discontinuation of ICS was associated with 37% decrease in serious pneumonia events [RR 0.63 (0.60-0.66)]

    --risk reduction was rapid: 20% in first month to 50% by 4th month after discontinuation (the vast majority of benefit was in the first 3 months)

    ​--risk reduction was most pronounced with stopping fluticasone [RR 0.58 (0.54-0.61)]; less with stopping budesonide [RR 0.87 (0.78-0.97)]

so, what does this all suggest?

--pneumonia may be a serious complication from continued ICS usage. There are other likely or known adverse effects, often related to suppression of the hypothalamic-pituitary axis: some ocular effects including cataracts and increased intraocular pressures, bone density may be adversely affected, and there are local problems with candida and irritation.

--this is an observational study, so we really need a formal intervention study to show benefit of stopping ICS. But this study is really large and  found quite dramatic differences in pneumonia, so should be taken seriously until a large enough intervention trial is done. There have been a couple of smaller trials done with unclear benefit of stopping ICS: the INSTEAD trial found an insignificant decrease in pneumonia in switching patients with moderate COPD from a combo long-acting b-agonist (LABA)/fluticasone to just a different LABA, but was a very small study with only 6-month follow-up. The WISDOM trial was larger with 2456 patients with moderate to severe COPD (GOLD 3-4 and at least one COPD exacerbation in past year), randomizing patients to  step-wise decreases in fluticasone over 12 weeks vs continued use, and found no difference in COPD exacerbations in those off fluticasone, but they did find a pretty small 43 ml decrease in FEV1.0 at 52 weeks​ (see NEJM 2014 Oct 2;371:1285). There was again a non-significant decrease in pneumonia in those in the ICS withdrawal arm.

--there has been a potential conceptual error that those who respond to oral steroids should respond to the less-toxic ICS, but there are no data to support this contention

--my sense, and my clinical experience, does suggest that there are many patients, including some with very severe COPD, who do not respond to ICS. It is certainly appropriate to try ICS on patients with relatively frequent COPD exacerbations, or if  inadequate symptom control with tiotropium augmented with LABA, but those who are pretty stable and without exacerbations probably should not be on them, and those on ICS without evident relief might deserve a trial of titrating them off ICS.

--so, the benefit of ICS is hardly uniform (the individual studies have very mixed results), ICS do have potentially significant adverse effects, and, per the WISDOM trial, there does not seem to be major clinical problems by trying to wean people off them. But, nonetheless, many people even with stable mild to moderate COPD are taking ICS​.

--Bottom line: it is probably not indicated to use ICS except if there seems to be an asthma component to the COPD, inadequate symptom control, or COPD exacerbations. but ICS are undoubtedly not required in all of the 50-85% of COPDers on them. and this study adds the potential for a very serious adverse event to the list of their complications. so, best to avoid them if possible and consider discontinuing them when the patient is stable to see how they do.