| We read with interest Farin Kamangar
and colleagues’ Review1 in which they
suggest a link between opium use
and cancer risk. The authors argue
that genetic polymorphisms are
unlikely to confounded their fi ndings.
However, many of the studies
they refer to do draw attention to
associations between polymorphisms
and cancers. For instance, Abnet and
colleagues2 concluded that variants
at 10q23 in PLCE1 showed genomewide
associations for gastric cardia
cancer and esophageal squamouscell
carcinoma.Moreover, most
of the cited studies report odds
ratios (ORs) typically between 1·2
and 1·4 per allele. Kamangar and
colleagues’ suggest that these ORs
are indicative of a weak association
between polymorphisms and
cancer. Nevertheless, many diff erent
alleles have been shown to have a
signifi cant association with cancer,3,4
and we believe that the association
of polymorphisms and cancer should
not be overlooked, even though
associations could be weak with some
alleles and strong with others.
Many studies in diff erent countries,
and especially Iran, have confirmed
that opium can be contaminated with
lead.5–8 Such contamination could
be a contributing factor for the high
prevalence of cancer in opium users.
All but two of the 18 studies reviewed
by Kamangar and colleagues were
reports from Iran, but no adjustment
was made for this potential
confounder. Given that environmental
factors are an important contributor
to cancer, the results of this Review
may not be generalisable to whole
populations.
We declare that we have no competing interests. |