The month in research
Epidemiology
Smoking and social networks
N Engl J Med 2008;358:2249-58
When researchers first began the Framingham heart study back in 1948, they asked participants for the contact details of close
friends to help them keep track of the cohort. More than half a century later, other researchers are using these incidental
data to study the spread of behaviours such as smoking through social networks.
The latest study, analysing data between 1971 (the children of the original cohort) and 2000, makes it clear that as the prevalence
of smoking falls smokers are becoming increasingly marginalised. By 2000, they tended to cluster at the edge of social networks
and associate mainly with other smokers. The same network analysis shows that groups of smokers seem to quit roughly at the
same time, and that spouses, siblings, friends, and colleagues can all influence a person to quit smoking. Having a spouse
who quit decreased a person’s chance of smoking by 67% (95% CI 59% to 73%), for example. Friends and close colleagues in small
firms helped reduce the chance of someone smoking by around a third.
Social pressure to quit may be a good thing, but driving hardcore smokers to the edge of society probably isn’t, says a linked
editorial (p 2284). The isolation and stigma that go with marginalisation could make smokers even harder to reach. (This item
was first published in the BMJ, 2008;336:1212.)1
Influence of contacts on subject’s smoking
Haematology
Blood donation risk in teens
JAMA 2008;299:2279-86
Teenagers aged 16 or 17, who donate almost a tenth of the whole blood collected by the US Red Cross, have a significantly
higher risk of complications related to donation than older teenagers and adults. In blood centres run by the Red Cross, 10.7%
of 16 and 17 year olds, 8.3% of 18 and 19 year olds, and 2.8% of adults had some kind of complication in 2006, usually minor
bruising, pallor, or light headedness. Far fewer lost consciousness and fewer still sustained injuries or needed outside medical
help, but the risks were still highest in the youngest donors, in an analysis of data from nine donation centres. Rates of
injury secondary to fainting were 5.9 per 10 000 in 16 and 17 year olds, 2.4 per 10 000 in 18 and 19 year olds, and 0.4 per
10 000 in adults aged at least 20. The odds of injury were 14 times higher for the youngest donors relative to adults (odds
ratio 14.46, 95% confidence interval 10.43 to 20.04).
In this study, 16 year olds who had even a minor complication were less likely than others to donate again that year (52%
v 73%; 0.40, 0.36 to 0.44). (This item was first published in the BMJ, 2008;336:1212.)
Renal medicine
Vascular access for haemodialysis
JAMA 2008;299:2164-71
People having haemodialysis need vascular access, and arteriovenous fistulas are the most popular option. It is relatively
straightforward to create such a fistula, but a substantial minority fail to “mature” into useful access for dialysis. To
find out if the antithrombotic agent clopidogrel might help, US researchers did a placebo controlled trial in 877 men and
women with advanced or end stage renal disease.
The results were disappointing. Clopidogrel (75 mg a day for six weeks after a 300 mg loading dose) helped prevent early thromboses
but made no difference to the number of fistulas that remodelled into useful access. About 60% of the fistulas in both groups
were ultimately unsuitable for haemodialysis (relative risk 1.05, 95% confidence interval 0.94 to 1.17), which was a much
higher failure rate than the authors expected.
Although the trial was stopped early, it was powerful enough to show with some confidence that clopidogrel isn’t an effective
treatment for patients with new fistulas, says a linked editorial (p 2205). Nephrologists probably shouldn’t use it. Thrombosis
may not be the key factor determining whether a new fistula will eventually work. (This item was first published in the BMJ, 2008;336:1156-7.)
Molecular oncology
Understanding trastuzumab
J Clin Oncol 2008;26:1789
Trastuzumab (Herceptin) has been heralded as a breakthrough in molecular oncology because its development was based on a detailed
understanding of a signalling pathway that promotes growth of tumour cells.
The drug disrupts the proliferation of cancerous cells. It is a humanised monoclonal antibody. Its antigen binding domain
(Fab) binds to the famous HER2/erbB2 tyrosine kinase receptor, which is overexpressed in some breast cancers. Although some
HER2/erbB2 positive patients improve when treated with trastuzumab, as many as 70% do not, and the reasons have been unknown
till now.
Researchers have provided evidence that the drug’s oncolytic properties are a result, at least in part, to a different mechanism
than was previously suggested. Known as antibody dependent cell mediated cytotoxicity (ADCC), natural killer cells recognise
cells with antibodies bound to them and ingest and eventually kill them.
Fcγ receptors are important for normal ADCC. Some variations in the gametes in the sequences of genes that encode these receptors
were correlated with patients’ responses to trastuzumab. These results came from a study of 54 patients with HER2/erbB2 positive
metastatic breast cancer.
These results indicate, the authors argue, that not only can a patient’s response to trastuzumab be predicted, but also that
drugs increasing the body’s capacity to induce ADCC might improve the outcome of treatment.
Molecular oncology
Treating acute promyelocytic leukaemia
Nat Cell Biol 2008;10:538-46; 547-55
Arsenic trioxide, although toxic in general, has been used to treat acute promyelocytic leukaemia for more than a decade—but
with no understanding of how it worked. What was known however was that the drug destroys PML-RARα, an oncogenic fusion protein
seen in patients who have promyelocytic leukaemia. Two studies offer important clues about how this happens.
Researchers have found that arsenic trioxide attaches several copies of SUMO (small ubiquitin related modifier), a signalling
molecule, to PML-RARα, and that this addition prompts the protein’s destruction, resulting in oncolysis. This in turn happens
through another signalling molecule called ring finger protein 4 (RNF4), which directs the protein complex formed between
SUMO and PML-RARα to the proteasome, an intracellular organelle where enzymes are normally destroyed.
The search is now on for a less toxic molecule that might cause aggregation of SUMO and tag the oncogenic molecule for destruction.
Student BMJ 2008;16:235 | 18
