Comment and reply on: Efficacy and safety of the single pill combination of amlodipine 10 mg plus valsartan 160 mg in hypertensive patients not controlled by amlodipine 10 mg plus olmesartan 20 mg in free combination

Dear Sir,

In a recent issue of the journal, Braun et al.1 published the findings from a multicentre, open-label, non-randomised, two-phase study examining the efficacy and safety of a fixed dose combination of valsartan (VAL)/amlodipine (AML) 160/10 mg as a second-line treatment in patients with moderate hypertension whose blood pressure (BP) was not adequately controlled by a free combination of olmesartan (OLM)/AML 20/10 mg. When discussing the results from this study1, the authors acknowledged and briefly exposed a number of limitations associated with this study, including the open-label, non-controlled design and the short duration of treatment, prior to concluding that ‘once-daily treatment with the single-pill combination of amlodipine 10 mg plus valsartan 160 mg in those patients not controlled by previous treatment with the free combination of amlodipine 10 mg plus olmesartan 20 mg leads to a substantial and statistically significant reduction of elevated BP’. Having read this publication, it is clear that there is an inconsistency regarding the data reported in the abstract and the results section. The value reported for the mean BP at week 4 in the results section (page 425) is quoted as 165.4/103.7 mmHg whereas in the abstract (page 421) it is quoted as 149.6/93.4 mmHg. Given that the primary and secondary efficacy variables are changes in DBP and SBP, respectively, between week 4 and week 8, this inconsistency could lead to the study results being misinterpreted.

Furthermore, a number of additional study limitations that could have biased the results have been identified. One major limitation is the dosages of VAL/AML and OLM/AML used. The VAL/AML 160/10 mg dose is the maximum approved dose of this combination in Europe. In contrast, a fixed combination of OLM/AML 20/10 mg is not licensed in Europe and is not the maximum dose available. It is therefore proposed that in order to provide a fair comparison, the maximum licensed dose of OLM/AML, 40/10 mg, should have been used instead. Furthermore, there was no attempt to disguise the fact that patients received two pills in phase 1 followed by one pill in phase 2, and this may have influenced the results. The reason for this, as stated in the paper, is that the only single-pill angiotensin receptor blocker/calcium channel blocker combination available in Germany is VAL/AML. In order to have a valid comparison of both regimens, this trial should have either been performed with both VAL/AML and OLM/AML fixed-dose combinations or been done in a blinded fashion using a dummy in the VAL/AML phase.

Finally, the authors conclude that the short duration of phases 1 and 2 would not impair the conclusions drawn. However, it is questionable whether it is correct to claim that the improvements in BP seen at the end of phase 2 were due to the VAL/AML combination given the linear nature of the change in DBP, and close to linear change in SBP seen in Figure 1. Indeed, although many antihypertensive agents demonstrate good efficacy within the first few weeks of administration, clinical trials often assess efficacy over 6–8 weeks, and even longer in some instances, to evaluate the full effect of the drug. Furthermore, the best method of comparing whether an agent has an effect is to include a comparator arm as commonly used in other studies comparing the benefits of switching from one antihypertensive agent to another2–4.

Declaration of funding

None.

Declaration of financial/other relationships

V.B. has disclosed that he has been in receipt of research sponsorship from AstraZeneca and Recordati; that he has served as a consultant/advisor to Daiichi Sankyo Novartis; AstraZeneca; Sanofi-Aventis; Bristol Myers Squibb; Boehringer Ingelheim and that he is on the speakers bureau for Novartis, AstraZeneca, Pfizer, Servier, Recordati and Boehringer Ingelheim.

Authors'reply

Dear Sir,

We thank Dr Barrios for his comments, also pointing out an inconsistency in the Results section of the publication1.

Whereas the numbers given in the abstract are correct, the quoted BP values given on page 425 of our article1 represent the blood pressure levels at week 0 instead of week 4. We apologize for this inconsistency and would like to provide the correct values below. At week 0 mean systolic blood pressure in the ITT population was 165.4 ± 9.5 mmHg. It was reduced during treatment phase 1 to 149.6 ± 11.1 mmHg at week 4 and during treatment phase 2 to 141.7 ± 12.3 mmHg. This represented an additional BP reduction of 7.9 ± 11.8 mmHg after phase 2. Mean diastolic blood pressure in the ITT population was 103.7 ± 2.0 mmHg at week 0. It was reduced during treatment phase 1 to 93.4 ± 3.9 mmHg at week 4 and during treatment phase 2 to 84.3 ± 7.4 mmHg. This represented an additional BP reduction of 9.1 ± 7.1 mmHg after phase 2. The additional systolic/diastolic blood pressure reduction observed from week 4 to week 8 were statistically significant (p < 0.0001 for both). As the additional reductions observed were presented correctly, this should not impair the conclusions drawn.

The trial was conducted between July and December 2007. At that point of time, a single-pill combination of olmesartan (OLM) and amlodipine (AML) was not commercially available in Europe, and so could not have been used in the trial. The choice of dosage was based on the most commonly used ones. In Europe and many other countries, the starting doses of OLM and VAL in monotherapy are 10 and 80 mg and the most common dosages used are 20 and 160 mg, respectively (p. 427, Ref. 1). Today, a single-pill combination containing OLM 40 mg and AML 10 mg is available in Europe. However, in a randomized, double-blind, factorial trial the differences in mean SBP/DBP between OLM/AML 20/10 mg and OLM/AML 40/10 mg were 0.9/2 mmHg and the differences in least squares means were 0.4/1.7 mmHg. It can only be speculated if our trial would have come to different results using the dosage of 40/10 mg based on these small differences2.

It is true, that especially in the long-term treatment of hypertension in daily practice, single-pill combinations are associated with a better drug adherence and better outcomes in comparison to free combinations (p. 422, Ref 1). However, in this trial with short treatment periods we observed comparable compliance and drug exposure (p. 425, Ref 1) in both treatment phases which makes a substantial bias unlikely.

The design was chosen to mimic daily clinical practice and decision making. We agree that, beyond the alternative designs included in the discussion, other potential designs could have been used associated with different strengths and weaknesses. However, based on the data available for OLM and AML it is unlikely that the extent of blood pressure reduction observed in the trial is mainly caused by biasing factors.

Declaration of funding

The authors were not sponsored by Novartis or by any other commercial organization for their response to this letter. The original study commented on1 was sponsored by Novartis.

Declaration of financial/other relationships

N.B. declares that he is not currently sponsored by any pharmaceutical company and that he has no relevant financial relationships. S.K. is an employee of Novartis Pharma GmbH. Both N.B and S.K. share co-authorship of the response to the letter, on behalf of all the authors of the original article1.

Acknowledgements

None.

Acknowledgements

None.