Statin treatment and onset of cough without pulmonary involvement
Three cases of cough in patients treated with atorvastatin
Within the Viger project (Vigilance in geriatrics) three cases of cough induced by atorvastatin therapy have drawn our attention.
In all cases the drug withdrawal determined symptomatology remission and in one case of rechallange the cough reappeared leading to the definitive interruption of the pharmacological treatment.
All patients were in good health and the temporality of the events induced the physician who reported the cases to impute the adverse reaction to the statin exclusively.
Here there are the three cases in brief.
Case 1. A 54 years old man who had been administered atorvastatin (40 mg/day) for cardiovascular prevention and manifested cough 5 days after the treatment began. Persistent cough continued for 7 days during the therapy and disappeared only when the statin was withdrawn. The objective examination did not reveal either pulmonary compromise or respiratory problems, and the patient anamnesis did not suggest alternative explanations. One month later, atorvastatin was administered again with consequent reappearance of the cough; the therapy interruption led to complete remission of the drug adverse reaction also in this case.
Case 2. A 70 years old woman, with hypercholesterolemia, treated with 40 mg/day of atorvastatin; two days after the beginning of the therapy she manifested dry cough. The persistence of the symptom for three more days led to the drug withdrawal with consequent definitive remission of the cough. The patient was then successively treated with fibrates, without presenting similar conditions anymore. Also in this case the patient anamnesis did not suggest any alternative explanation. The iatrogenic nature of the event was therefore hypothesised and atorvastatin ascribed as only suspect drug for the onset of the adverse reaction.
Case 3. A 71 years old man in whom the cough appeared four days after he began the treatment with atorvastatin (10 mg/day) for hypercholesterolemia. The patient was also in therapy with losartan and hydrochlorothiazide, notoriously associated to possible cough onset. The duration of the sartan therapy had not been specified; anyway the physician who reported the adverse reaction ascribed it only to atorvastatin, since the patient had never manifested cough before its introduction. The statin was immediately withdrawn, with consequent definitive remission of the cough. Also in this case, the objective examination and the clinical history of the patient suggested the iatrogenic nature of the event.
Analysis of the reports relative to Lombardy Region collected by the AIFA database
We consulted the AIFA database from January 2004 to July 2013, analysing only the cases regarding Lombardy Region. We found five reports, including the three cases previously discussed, all correlating cough and statins. In all these reports the statin was considered the only suspect drug (Table 1).
Our analysis considered as potential confounding factors pulmonary pathologies and concomitant drugs known for causing cough as possibly implicated in the onset of the adverse reaction.
Table 1. Cases of statin-induced cough in the Italian National Network for Pharmacovigilance relative to Lombardy Region (January 2004 – July 2013)
￼￼Analysis of the international databases
We consulted the following international databases that provide free online access and are referential for this kind of analyses, spanning from January 1, 2004 to December 31, 2012: Adverse Event Reporting System – AERS of the FDA (US); Canada Vigilance Adverse Reaction Online Database; Australian Database of Adverse Event Notifications – DAEN.
Suspect reports of “cough” by statins have been collected based on the MedDRA codex (Medical Dictionary for Regulatory Activities).1 Only for the Canadian and Australian reports it was possible to exclude the patients affected by pulmonary pathologies or in therapy with other suspect drugs known for causing cough and possibly implicated in the onset of the adverse reaction, therefore the total result might be overestimated (Table 2).
The table highlights the cases of iatrogenic cough relative to the administration of atorvastatin, simvastatin, rosuvastatin, fluvastatin, cerivastatin, mevastatin and pitavastatin. Within the considered period, about 9 years, we found 977 spontaneous reports of cough associated with the use of statins.
The highest number of reports for iatrogenic cough is associated to rosuvastatin (425), followed by atorvastatin (344) and by simvastatin (144), probably caused by higher prescription rate for these drugs; lovastatin (9 cases) and cerivastatin (2 cases) have been associated to the lowest number of cases. For mevastatin and pivastatin, on the other side, we did not find similar reports.
Even though the data collected by the different databases are not homogeneous - because they reflect prescribing and reporting approaches characteristic and different for every country - the elevate number of reports we found, despite possibly overestimated, indicates a plausible relation with the statin treatment. Besides, the involvement of different molecules of statins does not exclude a possible class effect.
Hypotheses on the onset of cough induced by statins
Statins are involved in the production of prostacyclin (PGI2) and prostaglandins (PG) and in the upregulation of enzyme cyclooxygenase-2 (COX-2). Birbaum and colleagues showed that atorvastatin cardioprotective effects are mediated by the induction of PG synthetase activity, following A2 cytosolic phospholipase (cPLA2) and COX-2 upregulation, PGI2 synthetase and, to a small extent, PGE2 synthetase.2 Similarly, Degraeve et al. documented COX-2 induction and higher PGI2 production in human aortic smooth muscle cells following metavastatin and lovastatin administration.3
Prostanoids like PGD2, PGE2, PGF2alpha and PGI2 mediate a wide range of physiological effects by means of five receptor classes named: DP, EP, FP, IP and TP.
In particular, inhaled PGE2 can induce cough as a result of an important increase of cough reflex sensitivity itself, by binding to specific receptor molecules.4,5
A second hypothesised mechanism, probably involved in the potentiation of the prostaglandin-mediated action, is correlated to the effect caused by these drugs on the transmission way of nitric oxide (NO).
The activation of this way mediates protective effects against vascular damages; statins’ beneficial role in atherosclerotic vascular diseases is in fact well known.
In particular, statins stabilise messenger RNA in endothelial nitric oxide synthase (eNOS) leading to increased enzyme expression and NO production.6,7
Besides, increased inducible nitric oxide synthase (iNOS) expression, associated to stabilisation and potentiation of statin-mediated COX-2 activity, promotes significant cardioprotective effects.8 Recent results indicates that NO produced by NOS is an important stimulator in the normal cough reflex circuit, determining increased sensitivity of this reflex.
In conclusion, the possible mechanisms by which statins stimulate cough are linked to increased induction of COX-2 and NOS expression followed by increased nitric oxide and prostaglandins production, which in turn enhance cough reflex sensitivity.9
Exceptional cases of interstitial pulmonary disease have been reported for some statins, especially in long-term therapies; nevertheless, cough without pulmonary involvement is not yet included in the data-sheet for statins.
The clinical cases presented and the analysis of spontaneous reports collected by AIFA (for Lombardy Region) and international databases suggest a plausible involvement in the onset of iatrogenic cough. The most likely mechanism might be linked to the capability of increasing prostaglandins and nitric oxide production, which in turn induce an increase of cough reflex sensitivity.
Awareness about this correlation can be useful both to the physician – for providing a correct differential diagnosis and rationally establishing the possible iatrogenic nature of the cough – and the patient, who - knowing this possible adverse effect - could report its onset to the physician avoiding useless ulterior prescriptions.
Table 2 – Reports of statin-induced cough on international databases
|Drug||Adverse Event Reporting|
System Database FDA (AERS)
|Australian Database of Adverse|
Event Notifications (DAEN-TGA)
|Rosuvastatin||399||26 (of which 3 with pulmonary problems) = 23||3|
|Atorvastatin||315||24 (of which 2 under ACE inhibitor therapy, indicated as dubious, 2 with pulmonary problems) = 20||11 (of which 1 with pulmonary problems plus use of ACE inhibitors, indicated as dubious, and 1 under ACE inhibitor therapy indicated as dubious) = 9|
1Clinical Pharmacological Unit, CNR
Neuroscience Institute, Biomedical and Clinical Science Dept., University Hospital L. Sacco, Milan
2IRCCS Fondazione Ca’ Granda Ospedale Maggiore Policlinico, Milan
3Pharmaceutical Service, Local Health Service, Mantova
- Drug Saf 2002;25:445-52. CDI
- Cardiovasc Res 2005;65:345-55. CDI NS
- J Biol Chem 2001;276:46849-55. CDI NS
- Cough 2007;3:2. CDI NS
- Ciba Found Symp 1980;78:333-50. CDI NS
- J Cardiovasc Pharmacol 2003;1:474-80. CDI NS
- 7. Stroke 2000;31:2442-9. CDI NS
- Am J Physiol Heart Circ Physiol 2006;290:1960-8. CDI NS
- Cough 2011;7:5. CDI