.1 3 

3/10/87?~ D. Barnes
HOLMESBURG PRISON-BASED ASSESSMENT

Facts
Doses of 0.4 - 10.6 ug of 2.3.7.8-TCDD did not elicit a

chloracnegenic response in men to whose skin the
dose was applied.

Dose of 7500 ug applied in the same manner did elicit
a chloracnegenic response.

Questions
How much was absorbed?
Use Poiger and Schlatter as a basis for a rough estimate: 50%

I. Reality check What do these data "predict" about the
likelihood of seeing chloracne in the general population as
a consequence of "background" doses?

The chloracnegenic NOAEL is in the range of
50% of (16 7500 ug/ person)
(8 - 3750 ug/person) 70 kg/person I
.1 - 54 ug/kg
Considered as a LADD. this becomes
.1 - 54 ug/kg (70 yr 365 d/yr)
.000004 - .002 ug/kg-d 4 - 2000 pg/kg-d

This suggests that the daily doses to which folks are
currently exposed (1?10 pg/kg-d. a la Commoner analysis and
Japanese/Swedish diet studies) are not likely to be associated
with chloracne: they are close to the low end of the range
(4 pg/kg-d) which did not elicit chloracne.

 

Another way.to looking at these data to reach that same
conclusion is to consider the lifetime cumulative total of
2.3.7.8-TCDD from outside sources and. assuming a worst case of
infinite halflife. compare that dose with the Holmesburg doses.
In this case,

1?10 pg/kg-d 70 yr 365 d/yr 26,000 260,000 pg/kg
.026 - .26 ug/kg

These doses span a range that from well-below to a bit above the
level which did not elicit a chloracnegenic response. Therefore.
we can conclude that it is not surprising that these "background"
dose levels of 1-10 pg/kg-d are not associated with chloracne in
the general population.

II. Reality check What do these data suggest about the 
likelihood of cancer in the Holmesburg population?
- 1
There were 10 persons who received the 7500 ug dose, which. as
shown above. equated to 2000 pg/kg-d. Plugging this into
the UCL estimate of LMS'risk, we find:

 Slope se

1.6 10' 2 103 pg/kg-d

.3 (This is an overestimate since we are well
out of the linear range by this time --
however. the point is made: the risk is

high.)

Since the population is so small and the issue of confounding
.epxosure so confusing. however. that it is unlikely that a
defintive statement could be made on the basis of cancer
incidence figures for this group. even if we could track them
down. -

Reality check What can these data suggest to us about
exposure in other epidemiological investigations?

A number of studies have been conducted on populations. some
of whose members have had chloracne. The Holmesburg data suggest
that study subjects with chloracne received doses above .1 ungg
(LADD 4 Those without chloracne received doses below
54 ug/kg (2000 An effect level cannot be definitely
determined from the these data.

However. as a "geduncken experiment?. let's take the effect
level to be the midpoint of the range: 1000 pg/kg-d. This
would imply that members of a cohort who had chloracne (assuming
it was generated as a result of exposure to 2.3.7.8-TCDD) would
have an UCL of risk of

UCL Risk 1.5 10?4 1000 .1
Again, this is an overestimate because we are outside of

the linear range: but the risk is still projected to be
quite high.
Reality check: Does the incidence of cancer seen amongst the
"chloracne cohort" in the various studies support this idea?

IV. Reality check Can the Holmesburg data be used in some way
to deduce the relative sensitivity of humans and animals?

measure of relative sensitivity could be generated from the
Holmeaburg data vs. data on chloracnegenic experiments in
animals: nude mice and rabbits.