Rapid Responses to:
|
|
Rapid Responses: Rapid Responses published:
-
![[Read Rapid Response]](/icons/misc/sectionDOWN.gif)
High level of hs-CRP should be treated with statin irrespective of lipid profile.
- Gauranga Dhar (16 November 2008)
-
Maybe JUPITER will change something
- Mikael Rabaeus (16 November 2008)
-
Re: Maybe JUPITER will change something
- Alexander Jablánczy (18 November 2008)
-
Early Interruption of Jupiter Trial
- Ken Leeper (19 November 2008)
-
Three more “fatal myocardial infarctions” in rosuvastatin group than placebo in JUPITER
- Malvinder S Parmar (20 November 2008)
-
JUPITER: no cardiovascular deaths prevented
- Eddie Vos, Colin P. Rose, MD (27 November 2008)
|
|
|||
|
Gauranga Dhar, Physician and teacher Bangladesh Institute of Family medicine and Research. Dhaka, Bangladesh-01209
Send response to journal:
|
Chronic inflammation is the most important factor for atherosclerosis. Among the inflammatory biomarkers, hs-CRP plays crucial role. Bogalusa heart study shows that atherosclerosis starts in early childhood. Cigarette smoking, HTN, hyperglycemia, obesity, hyperlipidemia cause endothelial dysfunction and subsequent increased production of hs- CRP with the help of IL-6. Increased level of hs-CRP is directly proportional to cardiovascular outcomes. Besides metabolic factors written above there are so many other reasons which cause elevated level of hs- CRP. For example bacteria C. pneumoniae, H. pylori, Herpes simplex infection can also stimulate atherosclerosis. Recent study in Denmark showed decreased rate of mortality in adults hospitalized due to pneumonia who were under prior statin therapy. Main cause of death of patients with rheumatoid arthritis is atherosclerotic disease e.g. heart attack. This is due to high rise of inflammatory markers like hs-CRP and TNF alpha as well. Resent study in Russia (Ref:PMID: 18991814 [PubMed - in process]) has shown that HMG-CoA- reductase inhibitors, statins reduce hs-CRP level more rapidly in patients with RA than in patients with IHD which subsequently may reduce cardiovascular outcomes in patients with RA. According to me, statin is an unique drug, use of which will be extended in other indications other than lowering lipid parameters. Reserches now under way to use statins plus antibiotics in acute and chronic infections. I think irrespective of lipid profile, use of statins will give beneficial effects in adults in terms of cardiovascular outcomes. In my practice I will start to screen my suspected adult patients for hs-CRP level and if high, I will go for statin. Competing interests: None declared |
|||
|
|
|||
|
Mikael Rabaeus, Medical Director Health Management Center, Clinique de Genolier, 1272 Switzerland
Send response to journal:
|
Credit should be given to the BMJ and Drs Donner-Banzhoff and Sönnichsen for tempering the unpleasant hype that followed the publication of the JUPITER trial, hailed as a "ground-breaking trial that will change the paradigm of cardiovascular prevention"... In addition to the very pertinent remarks by the authors of the BMJ editorial, I would like to point out that the methodology in itself of the study is far from stain-free : - it is hard to find any good reason to interrupt the study early, considering the very low absolute benefit - the mortality figures (the only that cannot be discussed) show an even lower absolute reduction of 0.25 % (NNT = 400). Curiously, a short time before the study was interrupted, the reduction was only around 0.1 % (NNT = 1000), according to the curves shown. To say the least, the study was interrupted at a very convenient moment... - the competing interests are staggering. In particular, the first author and chair of the steering committee is a co-inventor of the hs-CRP test which has been licensed to Astra-Zeneca. JUPITER should simply be discarded, being irrelevant and to the least doubtful in its conclusions. In fact, one might speculate as to whether it actually doesn't add evidence to the fact that statins are probably much less efficient than we want to believe. I would like to remind everybody of the CORONA trial (N. Engl. J. Med. 2007;357:2248-61.)in which rosuvastatin as compared to placebo had no benefit whatsoever on clinical outcomes in HIGH-risk patients (CAD patients with heart failure) Maybe JUPITER will change something in our daily practice : we should consider closely if our patients really are in need of statins. M. Rabaeus MD Competing interests: None declared |
|||
|
|
|||
|
Alexander Jablánczy, MD private office, Sault Ste Marie
Send response to journal:
|
Of course I cannot read your article as evidently your interest is not spreading light but the darkness of Mammon. My wife tells me while we supposedly pay $25 dollars a month for this accursed internet actually somehow it works out to be $800 for three months. So I never pay for anything they must be stealing this somehow. So I cannot comment on the actual figures of this study, but when the whole kerfuffle about the hormone replacement therapy came out in JAMA a few years ago it was accessible on the internet and since then any study I can get my hands on is pretty much the same. I use four coloured pens for easy explanation to myself and patients. It refers to that study but is applicable to this and any other. I draw two lines in blue ten thousand units long assuming the study had 10,000 subjects. Then I draw inside that line a segment 9900 units long in green. This of course means that whether or not the whole group got this or that treatment placebo or drug 9900 people or subjects didnt die or suffer any ill effects. Of the remaining 100 unit length I colour at the end black the last fifty. This of course means that whichever they belonged to fifty died anyway in either group. So we are left with fifty to colour red who died only in one group. Now the reporters of these studies have the audacity to say that there was say fifty % reduction in death rate because they compare the fifty red plus fifty black with the fifty red. But they fail to mention the glaringly obvious that we started with 10,000 not 100 which is a bit of a difference. So my humble contention is that a much less than one percent difference is simply inconsequential. You have this massive majority of over 9900 to whom any ministrations of a pharmaceutical simply made no difference wahtsoever. If I gave you ten thousand talents but keep ten or five as retainer fee do you really give a damn? Now it's true if you are one of the five survivors it makes all the difference in the world, but for the others it's immaterial. So half a percent has a different ring to it than fifty percent. Competing interests: nil except that I am an MD |
|||
|
|
|||
|
Ken Leeper, General Practitioner Billinghay Medical Practice Billinghay Lincoln LN4 4AU UK
Send response to journal:
|
I believe the scientific community must unite behind the basic principle that it is unacceptable to interrupt a clinical trial early - unless it is clearly shown that the study arm is more dangerous to patients than the currently accepted standard treatment. This is especially important for trials like Jupiter where the proposed intervention is to be promoted for long term use. Clearly drug manufacturers and authors seeking maximum kudos will want trials stopped at a time that shows the intervention in its best light - as others have already commented. It is perfectly feasible that adverse effects of continuing the intervention may have outweighed benefits before the planned end of the trial. This trial has been stopped before providing any evidence of long- term benefit. It must be robustly criticized for this. However it could become a useful trigger for an internationally accepted standard that future trials could be expected to adhere to. I would propose that all trials seeking to demonstrate long-term benefits from new interventions should be expected to run for at least 5 years in most cases. Participants in trials are treated shabbily, possibly unethically, when trials are stopped early in this manner. They deserve better - and future generations depend on better standards of research. Let us all now unite to bring to an end such unprofessional behaviour. Competing interests: None declared |
|||
|
|
|||
|
Malvinder S Parmar, Associate professor, Northern Ontario School of Medicine 640 Ross Ave. East, Suite E, Timmins, ON. P4N 8P2, Canada
Send response to journal:
|
There is no doubt that results of JUPITER [1] are intriguing and interesting. However, one needs to study the information in detail, before changing practice drastically. The way the data is presented raises some questions and concerns. 1. hs-CRP was used as entry criteria and many different sub-group analyses were presented in the paper BUT an important one, if not the most important, about what happened to patients in this group whose CRP didn't come down. What was the difference in event rates in this group of patient’s. This analysis should have been included. 2. In Table 3 [page 2201] of the original paper in NEJM, the authors report results of non-fatal myocardial infarction [22 in active group versus 62 in control group p<0.00001] and in the line below they report “any myocardial infarction” [31 in active group and 68 in control group p0.0002]. Ideally, the categories should have been 'any MI' and "fatal MI' but here some thinking did go into presenting the data, that should favour active agent where possible. They didn’t define what they mean by ‘any myocardial infarction’ and I deduce from the data that, “Any myocardial infarction” = “fatal myocardial infarction” + “non- fatal myocardial infarction,” and, that could be STEMI or NSTEMI. So, if the above deduction is correct, which likely is, then rosuvastatin group had 9 ‘fatal MIs’ [31-22 =9] whereas the control group had 6 [68-62=6]. So Active group had three more “fatal myocardial infarctions” and this indicates that the results were likely camouflaged knowingly [likely] or unknowingly [let's give a benefit of doubt] and this should be taken into consideration. Although, only 3 more fatal MI’s but these are the real ‘hard-end” points. If a 0.01% difference in HbA1C could be statistically significant, then this hard end point with a difference of 3 bodies should not only be statistically but clinically significant. 3. Interestingly, there were total 142 primary end points in rosuvastatin group compared to 251 in placebo BUT any death ‘198’ were significantly higher than the primary end points in rosuvastatin group, compared to control group. So, if the patients didn't die from cardiovascular disease, they did die for other reasons. The fact that the numbers don't match here is somewhat interesting and requires further clarification. 1. Ridker PM et al. Rosuvastatin to prevent vascular events in Men and women with elevated C-reactive protein. NEJM 2008; 359:2195-207 Competing interests: None declared |
|||
|
|
|||
|
Eddie Vos, maintains health-heart-org Sutton (Qc) Canada J0E 2K0, Colin P. Rose, MD
Send response to journal:
|
In support of this excellent Editorial, we add the following: 4 of the 5 combined primary endpoints were non fatal: two were "disease" [myocardial infarction and (probably ischemic) stroke], the 3rd was a treatment option [revascularization, a procedure without evident mortality benefit and being 90% determined by the first hospital visited in an acute coronary syndrome: ref. http://www.bmj.com/cgi/content/full/330/7489/441 ], a 4th was hospitalization for unstable angina [a non-significant 41% on -statin reduction] and, finally, “confirmed death from cardiovascular causes”, an endpoint that was not reported separately. However, calculating back from NEJM's Table 3, there would have been a TOTAL of 31 vascular fatalities in the rosuvastatin vs. 37 in the placebo groups, a highly non-significant difference. As in many statin studies, could the 3 non-fatal endpoints in which benefit was found be from statin's nitroglycerin mimicking action [by promoting NO-synthase], in which case using a drug like nitroglycerin directly, and thus reducing angina, would have shown the same non-fatal primary endpoint benefit? Something that should finally be investigated. Competing interests: None declared |
|||