Dose-Response

Study

of

N,N-Dimethyltryptamine

in

Humans

I.

Neuroendocrine,

Autonomic,

and Cardiovascular

Effects

Rick

J. Strassman,

MD,

Clifford

R.

Qualls,

PhD

Background:

To

begin applying

basic

neuropharmaco-

logical hypotheses

of

hallucinogenic drug

actions

to

hu-

mans,

we

generated

dose-response

data for

intravenously

administered

dimethyltryptamine

fumarate's

(DMT)

neu-

roendocrine,

cardiovascular,

autonomic,

and

subjective

effects

in

a

group of

experienced

hallucinogen

users.

Methods:

Dimethyltryptamine,

an

endogenous

mam-

malian

hallucinogen

and

drug

of

abuse,

was

adminis-

tered

intravenously

at

0.05,0.1,0.2,

and

0.4

mg/kg

to

II

experienced

hallucinogen

users,

in

a

double-blind,

sa-

line

placebo\p=m-\controlled,

randomized

design.

Treat-

ments

were

separated by

at

least

1

week.

Results:

Peak

DMT

blood levels and

subjective

effects

were seen

within

2 minutes

after

drug

administration,

and

were

negligible

at

30

minutes.

Dimethyltryptamine

dose

dependently

elevated blood

pressure,

heart

rate, pu-

pil

diameter,

and rectal

temperature,

in

addition

to

el-

evating

blood

concentrations

of

\g=b\-endorphin,

cortico-

tropin,

cortisol,

and

prolactin.

Growth hormone blood

levels

rose

equally

in

response

to

all

doses of

DMT,

and

melatonin levels

were

unaffected. Threshold doses for

sig-

nificant

effects relative

to

placebo

were

also hallucino-

genic

(0.2

mg/kg

and

higher). Subjects

with five

or

more

exposures

to

3,4-methylenedioxymethamphetamine

dem-

onstrated

less robust

pupil

diameter

effects than those

with

two

or

fewer

exposures.

Conclusions:

Dimethyltryptamine

can

be

adminis-

tered

safely

to

experienced

hallucinogen

users

and

dose-response

data

generated

for several

measures

hy-

pothesized

under

serotonergic

modulatory

control.

Additional

studies

characterizing

the

specific

mecha-

nisms

mediating

DMT's

biological

effects

may prove

useful

in

psychopharmacological

investigations

of

drug-induced

and

endogenous

alterations

in

brain

function.

(Arch

Gen

Psychiatry.

1994;51:85-97)

HALLUCINOGENIC

drugs

re¬

liably

induce

a

unique

constellation of

subjec¬

tive

effects

in

humans.

They

elicit

perceptual

il¬

lusions and

hallucinations,

primarily

vi¬

sual but often

auditory.

Mood effects

range

from

euphoria

to

panic

to

bland indiffer¬

ence.

Somatic

effects

include

dissocia¬

tion,

relaxation,

or

tension.1

"Classic" hal¬

lucinogens

include

phenethylamines

(eg,

mescaline),

indolealkylamines (eg, psilo-

cybin

and

dimethyltryptamine

[DMT]),

and the

lysergamides

(eg,

lysergic

acid di-

ethylamide

[LSD]).2

These

unique

compounds

were

the fo-

cusof

in

tensive

clinical research

in

the 1950s

and

1960s.

The

discovery

of

LSD may

ar¬

guably

have

been

as

important

to

the de¬

velopment

of

a

"biological" psychiatry

as

was

the

contemporaneous

discovery

of

chlorpromazine.

Hallucinogens'

use

in

clinical research

was

associated with

ac¬

ceptable

risks when

subjects,

both

psy¬

chiatric

patients

and

normal

controls,

were

carefully

screened,

prepared,

and fol¬

lowed

up.3

Clinical research

with

hallu¬

cinogens,

however,

became

increasingly

difficult with their

widespread

abuse

by

young adults.

Highly

publicized

"adverse

reactions"

to

hallucinogens, primarily

LSD,4

resulted

in

their

placement

into

the

restrictive

schedule

I

of

the

Controlled

From

the

Departments

of

Psychiatry

(Dr

Strassman)

and

Medicine

(Dr

Qualls),

School

of

Medicine,

and

the

Department

of

Mathematics

and Statistics

(Dr

Qualls),

University

of

New

Mexico,

Albuquerque.

Downloaded From: http://archpsyc.jamanetwork.com/ by John Chavez on 07/19/2016

SUBfECTS

AND

METHODS

The

protocol

was

approved by

the Scientific

Advisory

Com¬

mittee

of the General Clinical Research Center

(GCRC)

and

the

Human

Research

Review Committee

of the

University

of

New Mexico

School of

Medicine,

Albuquerque,

the

New

Mexico State

Pharmacy

Board,

the

US

Drug

Enforcement

Administration,

and the US

Food

and

Drug

Administra¬

tion.23 Witnessed

written

informed

consent

was

obtained

from all

subjects.

Confidentiality

and

anonymity

were

main¬

tained

throughout

the

study.

SUBJECTS

We

recruited

only

experienced hallucinogen

users.

All sub¬

jects

but

one were

recruited

by

word

of

mouth;

this last

subject responded

to

an

announcement sent to

a

graduate

student

training

office

at

a

local

university.

Subjects

were

initially

interviewed

by

one

of

us

(R.J.S.)

to

screen

for

medical and/or

psychiatric

conditions and

as¬

sess

level of

experience with,

and

ability

to

manage

dysphoric

reactions to,

hallucinogenic drugs. Subjects

were

dropped

from further consideration if

they

were

taking

any medica¬

tions

long-term,

had

a

history

of

psychosis

not

secondary

to

drugs

or

fever,

or

had

a

chronic medical disorder.

Prospective

subjects

then received

a

Structured Clini¬

cal

Interview

for

DSM-IÍI-R

(SCID-R),

Outpatient Version,24

administered

by

a

trained

psychiatric

research

nurse.

First-

degree

relatives'

psychiatric

histories

were

obtained

during

the

SCID-R.

Diagnoses

were

confirmed

by

discussion with

two

research

psychiatrists.

We

then

performed

a

medical his¬

tory,

physical

examination,

and

laboratory

screening

tests

(including

a

complete

blood

count

with differential cell

count,

24-item

chemistry panel, thyroid

functions

[including thy

-

rotropin], urinalysis,

and

electrocardiogram).

The

demographic

characteristics of

our

subjectsare

given

in

the Table.

Subject

11

was

dropped

after

receiving

the

non-

blind

(0.04-

and-0.4-mg/kg)

doses and the

0.05-, 0.1-,

and

0.2-mg/kg

doses because he

had

a

major

depressive episode

precipitated by multiple personal

Stressors

that

appeared

un¬

related

to

his

participation

in

the

study.

He

responded

to

de-

sipraminehydrochloride. Theaverage

(±SEM)

age

(exclud¬

ing

subject

11)

was41.5± 1.5

years. The group

was

high

func¬

tioning,

with

only

one

subject

not

beinga professional

or

student

in

a

professional

training

program. There

was a

high

incidence

of

prior major

depression

in

the group.

Only

one

subject

was

suffering

any

current

Axis I

disorder;

this

was an

adjustment

disorder

resulting

from

an

impending

divorce. There also

was

a

high

incidence

of

first-degree

relatives

suffering

from de¬

pression

and/or

alcohol

abuse.

The

number of exposures

to

hallucinogens ranged

from

six

to

"hundreds."

Two

subjects

described

a

history

of

cocaine

dependence/abuse,

currently

in

remission;

the remainder had little

or no

exposure

to

co¬

caine

or

amphetamines.

Not

all

subjects

had used

marijuana

equally, although

all had

some

experience

with

it.

METHYLENEDIOXYMETHAMPHETAMINE USE

Of

interest

was

the

breakdown

of

subjects by

amount

of

methylenedioxymethamphetamine

(MDMA,

"Ecstasy")

use.

All but

two

had used

MDMA

at

least

once.

Six

of

the

11

sub¬

jects

had used

MDMA

five

or more

times;

five had used

it

twice

or

less.

Methylenedioxymethamphetamine

is

a mam¬

malian

serotonergic

neurotoxin25

with

equivocal long-term

effects

on

human neuroendocrine and/or

behavioral func¬

tion.26

Thus,

we were

interested

in

assessing

differential ef¬

fects

of the

serotonergic agonist,

DMT,

in

these

two

groups

of

users.

For

the sake of

approximating

equal

cell

sizes,

we

described the

six

former

subjects

as

having

had

high

expo¬

sure

and the latter five

as

having

had low

or no

exposure.

DRUG

Dimethyltryptamine

fumara

te

was

prepared

by

David

E. Ni¬

chols,

PhD,

Purdue

University,

West

Lafayette,

Ind. Pu¬

rity,

determined

by

gas

chromatography-mass

spectrom-

etry

and

high-performance liquid chromatography,

was

greater

than 99.9%.

It

was

prepared

for clinical adminis¬

tration

by

the

Inpatient

Pharmacy

of

the

University

of

 

ew

Mexico

Hospital

in

vials

containing

40

mg/mL.

Sterility

and

pyrogenicity testing

revealed

no

abnormalities

in

a

repre¬

sentative

sample

of vials. The

dimethyltryptamine

fuma-

rate

solution

was

drawn

into

a

sterile tuberculin

syringe

at

the

appropriate

dose and diluted with sterile saline

to

1.0

mL

before administration.

PROCEDURES

Nonblind

Drug

Administration

First,

all

subjects

received nonblind administrations of

0.04

and

0.4

mg/kg

of

intravenous

(IV)

dimethyltryptamine

fu-

marate

in

the

GCRC,

usually

on

consecutive

days.

Previ¬

ous

work22

had demonstrated

no

tolerance

to

the cardio¬

vascular and

subjective

effects of

a

full

IM

dose

of

DMT

given

twice

daily

for

5 consecutive

days.

Thus,

we

did

not

believe tolerance

to

DMT's

effects would

occur

at

this

in¬

terval.

We

decided

to

administer

the

drug

IV

rather than

IM

as

had been

reported previously.'6

This

was

because

of

our

initial

dose-finding

studies

using

a

subject

with

expe¬

rience

smoking

DMT

free

base,

the usual form and

route

of administration

by

recreational

users.27

The

onset

and

in¬

tensity

of effect of

1.0

mg/kg

of

IM

dimethyltryptamine

fu-

marate

was

described

by

this

subject

as

significantly

less

rapidly developing

and

hallucinogenic

than his

previous

experience

with the smoked

drug.

Further

dose-finding

stud¬

ies

with him and another

experienced

DMT

smoker estab¬

lished that 0.4

mg/kg

of

IV

dimethyltryptamine

fumarate

produced

a

"rush" and

hallucinogenic

effect

comparable

with those

seen

with

a

"full

dose"

of

smoked

free

base.

The nonblind

days

served several purposes.

They

al¬

lowed

us

to

assess

the cardiovascular and

subjective

Downloaded From: http://archpsyc.jamanetwork.com/ by John Chavez on 07/19/2016

effects of

a

subclinical and

fully hallucinogenic

dose of di¬

methyltryptamine

in

a

novel and

potentially

anxiety-

provoking

setting.

These

days

also

provided

the research

team

and

subjects

an

opportunity

to

become

acquainted

and familiar

with each other

in

the research

center

setting

without the

extensive

blood

drawing

and

use

of

the rectal

thermistor that could

exaggerate

any

potentially

paranoid

reactions.

Subjects

also

were

able

to

"calibrate" them¬

selves

to

what

to

expect

regarding

minimal and maximal

effects of

the

drug,

one

of

the purposes of the double-blind

study being

whether

they

could

distinguish

among

incre¬

mental doses of

DMT.

Finally,

these nonblind

days

pro¬

vided

subjects

with the

opportunity

to

drop

out

of the

study

before

we

had collected

extensive

data

on

them.

Subjects

were

admitted

to

the

GCRC

by

9

am

after hav¬

ing

fasted

since

midnight.

An IV

line

was

inserted

into

a

forearm

vein

and

kept

patent

with

heparinized

saline.

Sub¬

jects

were

allowed

to

relax for

30 minutes

before the

drug

was

given.

Dimethyltryptamine

was

infused

over

30

sec¬

onds and flushed with

5 mL

of sterile saline

over

the

next

15

seconds. Blood pressure and heart

rate

(HR)

were

taken

with

an

automatic

cuff several

times

before and

frequently

during

the

30 minutes

following

drug

administration. The

research

team

(a

registered

nurse

and

psychiatrist)

sat

qui¬

etly

on

either side of the

subject,

attentive

to

verbal and

nonverbal

cues,

but did

not

offer

any

direction

or

advice

unless

absolutely

necessary

during

the

first

several

min¬

utes,

the

time

of the

peak

effect of

the

drug. Sublingual

ni-

troglycerin,

and

IV

diphenhydramine

and

diazepam,

were

available

at

the bedside for sustained

hypertension,

aller¬

gic

reactions,

or

excessive

anxiety,

respectively.

As

drug

ef¬

fects

resolved,

discussion focused

on

the

nature

of the

subject's

experiences.

After

subjects

ate

a

snack

or

meal,

they

were

discharged

home.

No

subj

ect

withdrew

voluntarily

after the nonblind

days.

One

subj

ect

was

withdrawn because of

a

marked diastolic

blood

pressure

response

(to

>

105

mm

Hg)

to

the

0.04-mg/kgdose.

Double-blind,

Placebo-Controlled,

Randomized

Drug

Administration

A

computer-generated

randomization sequence

was

coded

and

kept by

the

Inpatient

Pharmacy

of

the

University

of

New

Mexico

Hospital. Subjects

were

admitted

by

9

am

to

a

dimly

lit

(<

100

lux)

room

on

the

GCRC. Two IV

lines

were

started:

a

small-gauge

metal needle

in

one arm

for

drug

administra¬

tion,

and

a

large-gauge,

three-holed

plastic

angiocatheter

for

blood withdrawal.

A

reusable rectal thermistor

probe

(YS-

400,

Yellow

Springs

Instruments,

Yellow

Springs,

Ohio),

con¬

figured

to

a

physiological

monitor

(Vita-Log

PMS-8,

CPS

Ine,

Palo

Alto, Calif),

was

inserted. The

monitor

sampled

rectal

temperature

every

minute,

and the thermistor

probe

has

an

accuracy of

±0.03°C.

Blood pressure and

HR

(vital

signs),

and

pupil

diameter

(measured

by

placing

a

card with black

circles

differing by

1-mm increments

juxtaposed

10

to

15

cm

from the

subject's eyes),

were

taken

at

30, 15,

and

2 minutes

before

drug

administration.

Blood

samples

for

cor¬

ticotropin,

ß-endorphin,

prolactin,

GH,

cortisol,

and mela-

tonin

levels

were

drawn

30

and

2 minutes

before

drug

ad¬

ministration.

Drug

and saline

flush

were

administered

over

45

seconds,

and

time 0

was

when

the saline

flush

was com¬

pleted.

Doses

were

0.05,0.1,0.2,

and

0.4

mg/kg

of dimeth¬

yl

tryptamine

fumarate

and sterile

saline

placebo.

Blood

samples

were

drawn,

and

pupil

diameter and vital

signs

measured

2,

5,10,15,30,

and

60 minutes

after

drug

administration.

Study

sessions

occurred

at

an

interval of

at

least

1

week and took

place

from

January

to

September

1991.

The

one

woman

sub-

ject

was

studied

during

the

early

follicular

phase

of her

cycle.

A

negative

serum

pregnancy

test

drawn the

night

before her

study days

was

always

available

before

drug

administration.

ASSAYS

Dimethyltryptamine

(free base)

blood levels

were

assayed

using

the gas

chromatography-mass

spectrometry

technique

of Walker

et

al.28

Limit

of

detectability

was

1

ng/mL.

Cor¬

ticotropin

and

ß-endorphin

levels

were

assayed by

immu-

noradiometric assays

(Nichols

Diagnostic

Laboratories,

San

Juan

Capis

trano,

Calif),

and lower limits of

detectability

were

0.2

and

2.2

pmol/L,

respectively.

Limits

of

detectability

for

cortisol,

GH,

and

prolactin

radioimmunoassays

(all

from

Di¬

agnostic Products,

Los

Angeles,

Calif)

were

6

nmol/L,

1

µ /L,

andl

µg/L, respectively.

Growth hormone and

prolactin

val¬

ues

below

1

µg/L

were

entered

as

0.9

µ£/ .

The melatonin

radioimmunoassay29

limit of

detectability

was

5

pmol/L.

All

assays

were

performed

in

duplicate.

STATISTICS

Variable

values

are

given

as mean

±

SEM.

All

analyses

were

per¬

formed

using

PC SAS Version

6.03

(Cary,

NC).

Several

derived

variables

were

computed

for

analyses.

Baseline values for blood

pressure,

HR,

and

pupil

diameter

were

derived

by

averaging

the three values

obtained before

drug

administration

(ie,

—30,

—15,

and

—2

minutes).

Baseline

temperature

was

computed

by

averaging

values for

5 minutes

immediately

preceding drug

administration.

Hormone

baseline

values

were

computed by

averagingresults

from

the

—30-

and

2-minu

te

samples.

Maxi¬

mum

rise

above baseline

is

indicated

AMax;

AAUC represents

the

area

under the

curve

above

baseline,

calculated

by

the

trap¬

ezoidal rule.

Our

primary

statistical tool

was

the

univariate

analysis

of

variance

(ANOVA)

with

repeated

measures,

with dose

as

the

repeated

factor.

Pairwise

comparisons,

when the overall

F

value

generated

a

P<

.05,

were

performed

using

the

"contrast"

state¬

ment

in

the

general

linear

models

procedure

in SAS. To

assess

a

possible modulatory

role of

methylenedioxymethamphetamine

exposure

on

dimethyltryptamine's

effects

on

a

particular

vari¬

able,

we

performed

a

two-way

univariate

repeated-measures

ANOVA

using

MDMA

history

as

the

independent

variable.

Continued

on

next

page

Downloaded From: http://archpsyc.jamanetwork.com/ by John Chavez on 07/19/2016

MISSING DATA

We

did

not

analyze

blood hormone level results

for

the

one

female

subject

in

the group because her

veins

collapsed

at

some

point

for almost every

session,

and

her

blood data

were

incomplete.

However,

we

did

ana¬

lyze

her

autonomie

and cardiovascular responses.

Pupil

diameter

was

difficult

to

measure

consis¬

tently,

as

subjects

varied

widely

in

their

ability

and/or

desire

to

open their eyes

during

acute

drug

intoxi¬

cation.

Thus,

these

measurements

were

obtained

at

the

subjects'

discretion.

Additionally,

several sub¬

jects

had

technically

poor rectal

temperature

trac¬

ings

because

of difficulties

with

probe placement,

and

these data could

not

be used.

Two

subjects

had

un¬

satisfactory

temperature

data

for

the 0.05- and 0.2-

mg/kg

sessions;

one

had

no

0.1-mg/kg

session

data;

and

one

had

no

temperature

data from his

placebo

session.

Finally,

one

subject

refused the rectal therm¬

istor

probe,

and

no

temperature

data

were

collected

from him.

As SAS

cannot

perform repeated-

measures

ANOVA

on

subjects

with

missing data,

we

used

a

less robust one-way ANOVA

for

the

tempera¬

ture

and

pupil

diameter

data,

with dose

as

the inde¬

pendent

variable. Fisher's

Least

Significant

Differ¬

ence

Test

for

pairwise

comparisons

was

performed

if

the

overall

F

value

generated

a

significance

of

P<.05.

To

assess

the effect of

DMT

exposure

on

MDMA's

ef¬

fects

on

pupil

diameter and rectal

temperature,

we

performed

a

two-way

ANOVA

with dose and

MDMA

history

as

independent

factors.

Substances

Act. Human

studies

with these

drugs

effec¬

tively

ended

in

the

mid-1970s.

Despite

legal

restrictions

and

severe

penalties

for

manufacture, possession,

and dis¬

tribution of

hallucinogens,

their

use

among

college

stu¬

dents has

held

relatively steady

over

the

past

20

years.5

Basic

research

into

the mechanisms of

action

of hal¬

lucinogens

continued, however,

and contributed

to

re¬

cent

advances

in serotonin

(5-HT)

receptor

pharmacol¬

ogy.6

Recent

behavioral,

in

vivo,

and

ligand binding

data

suggest

agonist

or

partial

agonist

effects

at

the

5-HT2,

5-HT1A,

and

5-HTiC

receptors.7"10 Dopaminergic11

and

nor¬

adrenergic12

effects

have been

described but

not

studied

recently.

Hallucinogenic drugs

have diffuse and

relatively

well-

documented

biological

effects,

presumably

related

to

their

serotoninergic properties.

Growth hormone

(GH),

pro¬

lactin,

ß-endorphin,

corticotropin,

and cortisol levels

are

all affected

by

serotonergic

stimuli.13

Consistent

with the

presumed

serotonergic properties

of the classic halluci¬

nogens,

levels of

several

of

these hormones

increase

with

their

administration.

For

example,

human

prolactin

blood

levels

rise in

response

to

dimethyltryptamine,14

and

rat

plasma

corticoid levels

are

stimulated

by

LSD.15

Cardio-

vascular

variables,

core

temperature,

and

pupillary

di¬

ameter

also

are

affected

by

serotonergic

hallucinogens.

Examples

of these effects include

DMT's

hypertensive

and

mydriatic,16

LSD's

pyretogenic

and

mydriatic,17'18

and

psilocybin's hypertensive

and

tachycardie19

properties.

Dimethyltryptamine

is

a

short-acting

hallucino¬

gen, first discovered

in

hallucinogenic

Amazonian

snuffs.20

It

was

later

synthesized

and

determined

to

be

active

by

parenteral

(intramuscular [IM])

administration

only.16

Its

discovery

in

human

body

fluids

prompted

a

flurry

of

investigations

into its

possible biosynthesis

in

man

and

its

role

as a

putative

endogenous

psychotogen.21

Diffi¬

culties

distinguishing

peripheral

levels of

DMT

between

normal controls and

patients

with

endogenous psycho¬

ses

led

to

a

loss of

interest in

the function of this hallu¬

cinogenic

tryptamine.22

It

has been

impossible

to test

in

humans

hypoth¬

eses

of

hallucinogen

mechanisms of

action

developed

from

basic research.

Systematic

psychopharmacological

in¬

vestigations

are

the necessary

interface

bridging

recent

neuropharmacological findings

to

human effects.

Dose-

response

investigations

applying

current

psychiatric

re¬

search methods

to

assess

hallucinogens'

effects

can

gen¬

erate

data

providing

the bases for

more

experimental

studies,

such

as

selective blockade

with

receptor

sub¬

type

antagonists.

Several

properties

recommend

DMT

as an

appro¬

priate

drug

with which

to

renew

such

investigations.

Its

short duration

minimizes

prolonged dysphoric

reac¬

tions in

the

potentially

stressful

environment

of

a

mod¬

ern

clinical

research

center.

The

lack

of

a

widespread

"folk¬

lore"

concerning

its

effects

provides

less bias and

expectations

from

subjects. Finally,

DMT's

role

in

hu¬

man

brain function has

yet

to

be

explicated

in

normal

and

pathological

states.

That

is,

understanding

effects and

mechanisms of

action

of

DMT

may

shed

light

on

the

eti¬

ology

and

treatment

of

endogenous hallucinatory

con¬

ditions.

We

have

performed

a

dose-response

study

of the

neu¬

roendocrine, cardiovascular,

autonomie,

and

subjective

effects of

dimethyltryptamine

fumarate

using

a

double-

blind,

placebo-controlled,

randomized

design.

This

ar¬

ticle focuses

on

biological

data

generated

by

this

study.

RESULTS

The

major

findings

of this

study

are

the

following:

(1)

Intravenous

dimethyltryptamine

fumarate

was

fully

hal¬

lucinogenic

in

doses of

0.2

and

0.4

mg/kg.

Effects

were

felt

nearly instantaneously, peaked

within

2 minutes

af¬

ter

injection,

and resolved

within 20

to

30 minutes.

The

time

course

of

DMT

blood levels matched the march of

subjective

effects.

(2)

Multiple

biological

variables

rose

dose

dependently

in

response

to

DMT.

The

time

course

of elevations

in

blood levels of

pro-opiomelanocortin

Downloaded From: http://archpsyc.jamanetwork.com/ by John Chavez on 07/19/2016

*N0S indicates

not

otherwise

specified.

Subject

11

suffered

a

major depression during

the

study

and

was

withdrawn before

completing

it.

His data

are

included for illustrative purposes.

(POMC)

peptides (corticotropin

and

ß-endorphin),

HR,

blood

pressure,

and

pupil

diameter

paralleled subjective

effects and

DMT

blood

concentrations.

Elevations of

pro¬

lactin and cortisol blood

levels

lagged

5

to

10

minutes

behind these

changes.

Core

temperature

and

GH

effects

did

not

begin

until the

first

set

of

perturbations

had al¬

most

resolved

(15

to

30

minutes)

and

may

not

have

peaked

by

60

minutes

after

DMT

administration.

DMT

BLOOD LEVELS

There

was no

measurable

DMT in

blood

samples

before

drug

administration

or

after saline

placebo

administra-

tion

for

any

subject

except

for

subject

4. He

had

mea¬

surable

DMT

levels

at

one

point

before

active

drug

ad¬

ministration

and

twice

after

placebo

injection.

His

low

levels

in

these

circumstances

are

consistent

with

endog¬

enous

levels

in

a

study

of

normal volunteers

using

a

simi¬

lar

assay.30

However,

he

regularly

took "Chinese herbs"

and

"homeopathic

tinctures"

as

tonics,

unlike any

of

our

other

subjects.

Plasma levels of

DMT

peaked

at

the

first

blood

draw,

2 minutes

after

the

injection

was

completed, closely

cor¬

responding

to

peak

psychological

effects.

Generally,

a

dou¬

bling

of dose resulted

in

a

doubling

of AMax values. These

data

are

displayed

in

Figure

1.

As

previously

de-

Downloaded From: http://archpsyc.jamanetwork.com/ by John Chavez on 07/19/2016

Figure

1.

Mean

dimethyltryptamine

(free base)

values

in 10

subjects

after

four doses of intravenous

dimethyltryptamine

fumarate

(0.05

mg/kg

[squares],

0.1

mg/kg [triangles],

0.2

mg/kg

[diamonds],

and 0.4

mg/kg

[closed circles])

and saline

placebo

(open circles).

scribed,3'

DMT

levels varied

widely

among

subjects, peak

values

ranging

from

32

to

204

ng/mL

after the

0.4-

mg/kg

injection.

BASELINE EFFECTS

There

were no

differences among baseline values for any

biological

variable

among the

five

treatment

conditions.

NEUROENDOCRINE

EFFECTS

POMC PEPTIDES

The

AAUC

and

AMax

for both

corticotropin

and

ß-endorphin

rose

dose

dependently

in

response

to

di¬

methyltryptamine

fumarate,

with

the

0.4-mg/kg

dose's

effects

being statistically

greater than all

other doses

for

both of these variables.

The

effects of

placebo

and the

0.05-

and

0.1-mg/kg

doses

could

not

be

separated statistically.

Of

interest

was

the stimulation of

corticotropin

and

ß-en¬

dorphin

levels

by

saline

placebo.

This has the appear¬

ance

of

a

conditioned

response

to

placebo

injection

in

all

these

subjects

who had

previously

received the

some¬

what

anxiety-provoking 0.4-mg/kg

dose

nonblind. These

data

are

displayed graphically

in

Figure

2

(corticotro¬

pin)

and

Figure

3

(ß-endorphin).

Prolactin

The

AMax and

AAUC

for

prolactin

demonstrated

a

sig¬

nificant

dose-dependent stimulatory

effect of

dimethyl¬

tryptamine

fumarate. The

0.4-mg/kg

dose raised both of

these

variables

significantly

more

than

placebo

and the

0.05-

and

0.1-mg/kg

doses. The effects of

0.4

and

0.2

mg/kg

on

AMax

prolactin

could

not

be

distinguished

from each

other,

although

the

0.4-mg/kg

dose's effects

on

AAUC

were

significantly

greater

than those of

0.2

mg/kg.

The

effects of

0.2,

0.1,

and

0.05

mg/kg

of

dimethyltryp¬

tamine

fumarate and of

placebo

on

AAUC

were

not sta¬

tistically

separable.

These data

are

graphically

presented

in

Figure

4.

Note

that

prolactin

effects

were

slightly

de¬

layed

relative

to

the

POMC

peptides.

Cortisol

The

AAUC and AMax for

cortisol both

demonstrated dose-

dependent stimulatory

effects of

dimethyltryptamine

fu¬

marate.

The

0.4-mg/kg

dose raised

these

two

variables

more

than all other doses.

The

effects of

0.4

mg/kg

on

AAUC for cortisol could

not

be

statistically

distin¬

guished

from the

effects

of

0.2

mg/kg,

however. Neither

could

the

0.4-mg/kg

dose's effects

on

AMax

for cortisol

be

separated

statistically

from

0.2 and 0.05

mg/kg.

Ef¬

fects of

the

0.05-,

0.1-,

and

0.2-mg/kg

doses

on

AAUC

and AMax

were

not

statistically

different,

although

their

effects

were

numerically

greater

than

those of saline.

These

data

are

graphically

presented

in

Figure

5.

Growth

Hormone

There

was a

trend

for

DMT

to

produce

a

rise in

AAUC

for

GH.

The

AMax for

GH

rose

in

response

to

DMT

ad¬

ministration.

Although

all doses of

DMT

raised

GH

lev-

Figure

2.

Mean blood

corticotropin

values

in

10

subjects

after four

doses

of intravenous

dimethyltryptamine

fumarate

(0.05 mg/kg [squares],

0.1

mg/kg

[triangles],

0.2

mg/kg

[diamonds],

and 0.4

mg/kg

[closed circles])

and saline

placebo (open circles).

Downloaded From: http://archpsyc.jamanetwork.com/ by John Chavez on 07/19/2016

els relative

to

saline

placebo,

effects of different doses could

not

be

distinguished

from each other. These data

are

graphically

displayed

in

Figure

6.

Note the

delay

in

GH

effects relative

to

the other

neuroendocrine results.

Melatonin

We

only assayed

six

subjects'

blood

samples

for melatonin

from their

0.4-mg/kg

sessions.

Utilizing one-sample

t

tests,

there

was no

effect of

0.4

mg/kg

of

dimethyltryptamine

fu¬

marate

on

either AMax

or

AAUC for melatonin

(data

not

given),

so

the

remaining

samples

were

not

assayed.

AUTONOMIC EFFECTS

Rectal

Temperature

The AMax and AAUC for

temperature

demonstrated

a

significant

effect of

dimethyltryptamine

fumarate.

For

AAUC,

the

0.4-mg/kg

dose's effects

were

greater

than those

of

0.1

and

0.05

mg/kg

and

saline,

but could

not

be dis¬

tinguished

from those of

0.2

mg/kg.

The

effects of

0.2,

0.1,

and

0.05

mg/kg

of

drug

and of

saline

could

not

be

statistically separated.

For

AMax

in

temperature,

the

0.4-

and

0.2-mg/kg

doses' effects

were

greater

than

those

pro¬

duced

by

0.1

or

0.05

mg/kg

or

placebo.

The

AMax ef¬

fects of

0.4

and

0.2

mg/kg

could

not

be

statistically

sepa¬

rated

from each

other;

neither could the

effects of

the three

latter

treatments

be

separated.

These data

are

graphi¬

cally displayed

in

Figure

7,

where the

delayed

effect rela¬

tive

to

other

variables

can

be

seen.

Figure

3. Mean blood

ß-endorphin

values

in

10

subjects

after four doses

of intravenous

dimethyltryptamine

fumarate

(0.05

mg/kg

[squares],

0.1

mg/kg

[triangles],

0.2

mg/kg

[diamonds],

and

0.4

mg/kg [closed

circles])

and saline

placebo (open

circles).

r-

P^

~"--V,_

 ..

V

 

 

 

2

5

10

15

 

30

Time,

min

60

Figure

4.

Mean blood

prolactin

values

in

10

subjects

after four doses of

intravenous

dimethyltryptamine

fumarate

(0.05

mg/kg [squares],

0.1

mg/kg

[triangles],

0.2

mg/kg

[diamonds],

and 0.4

mg/kg ¡closed circles])

and saline

placebo

(open circles).

Pupil

Diameter

As

discussed

in

the

"Procedures"

section,

our

data

for

pu¬

pil

diameter contained the

most

missing

points

and should

be

interpreted cautiously.

Missing

values made

a

calcu¬

lation of AAUC

pupil

diameter

impossible

for the

more

disruptive

0.2-

and

0.4-mg/kg

doses.

Furthermore,

these

data

were

analyzed

using

ANOVA

without

repeated

mea¬

sures,

a

less

robust

analytic

tool.

The

AMax

pupil

diameter

demonstrated

significant

effects of

DMT.

The

0.4-mg/kg

dose

increased

pupil

di¬

ameter

more

than

all other

treatments.

Effects of

pla¬

cebo and

0.05, 0.1,

and

0.2

mg/kg

could

not

be distin¬

guished

from each other. These data

are

displayed

graphically

in

Figure

8.

CARDIOVASCULAR EFFECTS

Dimethyltryptamine

dose

dependently

elevated AMax

and

AAUC values for both

HR

and

mean

arterial blood pres¬

sure

(MAP).

For

AMax

HR,

effects of

0.4

and

0.2

mg/kg

were

significantly

greater

than those of

0.1

and

0.05

mg/kg

and

saline;

these latter three

treatments'

effects

were

not

distinguishable

from each other. Neither

were

the ef¬

fects of

0.4

and

0.2

mg/kg

separable

from each other. Ef¬

fects

on

AAUC

HR

were

less

clearly separable by

dose

but

followed

the

same

pattern.

The

0.4-mg/kg dimethyltryptamine

fumarate dose

raised AAUC

MAP

significantly

more

than

0.1 and 0.05

mg/kg

and saline. These latter three

treatments'

effects could

not

be

distinguished

from each other. Neither

were

the

Downloaded From: http://archpsyc.jamanetwork.com/ by John Chavez on 07/19/2016

Figure

5.

Mean blood cortisol values

in

10

subjects

after four

doses of

intravenous

dimethyltryptamine

fumarate

(0.05

mg/kg

[squares],

0.1

mg/kg

[triangles],

0.2

mg/kg ¡diamonds],

and 0.4

mg/kg

[closed circles])

and saline

placebo

(open circles).

effects of

0.2

and

0.4

mg/kg statistically separable.

The

0.4-

and

0.2-mg/kg

doses caused

significantly

greater

elevations

of

AMax

MAP

than

0.1

and

0.05

mg/kg

and

saline;

the

two

former

doses'

effects could

not

be

distinguished

from each

other. These

data

are

graphically displayed

in

Figure

9

(HR)

and

Figure

IO

(MAP).

MDMA EFFECTS

For

all of the

biological

variables under

investigation,

only

AMax

pupil

diameter values

were

smaller

across

all doses

of

DMT in

the

high-exposure

compared

with the low-

or

no-exposure MDMA

group

(high

exposure, 0.6±0.2

mm;

low

or no

exposure, 1.5±0.2

mm;

F=11.3,

P=,002).

SUBJECTIVE

EFFECTS

A

detailed

description

of

subjective

effects,

and the devel¬

opment of and

preliminary

data

generated by

our new

rat¬

ing

scale,

the

Hallucinogen

Rating

Scale

(HRS),

are

presented

in

a

companion

article.32

Briefly, subjects

felt the

onset

of

effects before

the 45-second

infusion

was

completed.

Peak

hallucinogenic

effects

(with

0.2

and

0.4

mg/kg)

occurred

usually

by

the first

blood draw and vital

signs

check,

at

2

minutes

after

injection,

a

useful

temporal

reference

point

for

subjects. Subjects

remained

moderately

intoxicated for

another

10

to

20 minutes.

They

felt

relatively

normal

at

20

to

30

minutes

after

injection.

Several

described

a

relaxed,

"at ease"

feeling

for

an

additional

30

minutes.

All

subjects

but

one

experienced

visual

hallucinatory

phenomena

at

0.2

and

0.4

mg/kg

of

dimethyltryptamine

fu-

marate.

Brief,

poorly

formed

auditory

hallucinations

were

experienced by

most

subjects.

Somatic

effects consisted of

an

intense

rush,

felt

throughout

the

body

or

specifically

in

the

head,

and

bodily

dissociation,

usually

associated with

a

transient anxious

or

fearful emotional

reaction.

Cogni¬

tive

effects

were

not

particularly

profound;

rather,

an

"ob¬

serving ego"

was

maintained,

alert and

attentive

to

the

per¬

ceptual

and

somatic

effects of the

drug.

The

0.2-mg/kg

dose's

perceptual

perturbations,

rush,

and

disorganizing

effects

were

less

intense

than for

0.4

mg/kg,

and

0.2

mg/kg

was

the

dose

preferred by

many

subjects.

The

0.1-mg/kg

dose of di¬

methyltryptamine

fumarate

produced hallucinatory phe¬

nomena

in

only

one

subject

(the

woman

in

the

study).

So¬

matic

effects

predominated,

and

many

subjects

disliked this

dose the

most.

The

0.05-mg/kg

dose

was

mistaken for

pla¬

cebo

by

several

subjects.

If effects

were

noted,

they

were usu¬

ally

somatic,

with

a

warm,

relaxed,

and

"floating" bodily

sensation.

COMMENT

We

safely

administered

graded

doses of

IV

DMT,

an en¬

dogenous hallucinogen

and schedule

I

drug

of

abuse,

to

a

group

of

experienced hallucinogen

users.

Dose-

dependent

responses

were

observed for several

neuro¬

endocrine,

cardiovascular,

and

autonomie

variables

in

a

randomized, double-blind,

saline

placebo-controlled

de¬

sign.

Peak

psychological

effects

paralleled

the

time

course

of

DMT

blood

levels,

which

appeared, peaked,

and fell

rapidly.

Blood levels of the POMC

peptides,

corticotro¬

pin

and

ß-endorphin,

closely

followed the

course

of

DMT

TT

2 5

 — 

10

15

30

Time,

min

60

Figure

6. Mean blood

growth

hormone values

in 10

subjects

after

four

doses of intravenous

dimethyltryptamine

fumarate

(0.05

mg/kg

[squares],

0.1

mg/kg

[triangles],

0.2

mg/kg

¡diamonds],

and 0.4

mg/kg [closed

circles])

and saline

placebo

(open circles).

Downloaded From: http://archpsyc.jamanetwork.com/ by John Chavez on 07/19/2016

levels and

hallucinogenic

effects,

as

did

pupil

diameter

and cardiovascular

responses.

Cortisol and

prolactin

lev¬

els

rose

in

a

dose-dependent

fashion and followed

the

first

group

of variables'

course

by

5

to

10

minutes.

Core

tem¬

perature

and

GH

stimulation

were

most

delayed,

al¬

though

GH

effects

were

not

separable by

dose. Stimula¬

tion

of

daytime

melatonin levels

was

not

seen.

The threshold for

significant biological

effects of

di¬

methyltryptamine

relative

to

placebo

was

also that for

its

hallucinogenic

properties:

0.2

mg/kg

and

higher.

The

0.4-

mg/kg

dose of

IV

dimethyltryptamine

fumarate

pro¬

duced blood

levels and

psychological

effects

compa¬

rable with those

seen

with administration of

more

than

twice

higher

doses of

IM

dimethyltryptamine hydrochlo¬

ride

or

creatine

sulfate.31 However,

the

IV

route

gener¬

ated

a

more

rapid

onset

(nearly

instantaneous)

and off¬

set

for these

effects

than

the

IM

route.

Dimethyltryptamine,

found

in

Amazonian

halluci¬

nogenic

plant

mixtures

and human

body

fluids,

meets

electrophysiological,33

pharmacological,34

and behav¬

ioral35

criteria

for

hallucinogenicity

in

lower animals.

It

had been used

safely

in

humans

in

this

country

and

Eu¬

rope.14·16·31

Its

short duration of

action

was

believed

ca¬

pable

of

providing

a

discrete,

more

manageable

halluci¬

nogenic

state

in

a

clinical research

setting

than

might

the

8-

to

12-hour effects of

lysergic

acid

diethylamide.

Hallucinogenic

drugs

are

serotonergic

agonists

or

par¬

tial

agonists,

in

addition

to

having adrenergic

and

dopa¬

minergic properties.

The

5-HT

receptor

subtypes

of

inter-

Figure

7.

Mean rectal

temperature

values

relative

to

baseline

temperature

in

10

subjects

after four doses of intravenous

dimethyltryptamine

fumarate

(0.05 mg/kg

[squares],

0.1

mg/kg

[triangles],

0.2

mg/kg

[diamonds],

and

0.4

mg/kg

[closed circles])

and saline

placebo

(open circles).

Baseline

temperature

values

were

calculated

by averaging

the

values of the 5

minutes

immediately

preceding dimethyltryptamine

administration.

Figure

8. Mean

pupil

diameter

values in 11

subjects

after four doses of

intravenous

dimethyltryptamine

fumarate

(0.05

mg/kg

[squares],

0.1

mg/kg

[triangles],

0.2

mg/kg

[diamonds],

and 0.4

mg/kg [closed

circles])

and saline

placebo (open

circles).

est

include the

5-HT2/lc

and

5-HT1A.36·37

We

therefore chose

our

measured variables because

they

are

believed

regulated

or

modulated

to

a

significant

degree by

serotonergic

neu¬

rotransmission.

These variables included

multiple

neuro¬

endocrine

markers,

blood

pressure,

HR,

rectal

temperature,

and

pupil

diameter.

As DMT

has

nearly

equal

affinities for

both the

5-HT2

and

5-HT1A

subtypes,34

many

of the effects

to

be discussed

may

devolve from

activation

of

either;

how¬

ever,

functional

interactions

in

humans also have been de¬

scribed38

and

may

be

important.

Levels of the

POMC

peptide

hormones,

corticotro¬

pin

and

ß-endorphin,

both

rose

dose

dependently

in

our

subjects. They

rose

quite

quickly, peak

values

being

seen

at

5 minutes

after

injection.

With doses of

0.1

mg/kg

and

higher,

AAUC

and

AMax values for these

two

hormones

were

nearly

identical,

confirming

their

corelease.39

These

rises in

ß-endorphin

and

corticotropin

extend hu¬

man40·41

data

demonstrating

serotonergic

stimulatory

ef¬

fects. Either

5-HT1A42

or

5-HT243

subtypes

may

be

in¬

volved.

Prolactin levels

rose

in

our

subjects

in

a

dose-

dependent

manner,

confirming

and

extending

the data

of Meltzer

et

al,14

wherein

cyproheptadine

inconsis¬

tently

blocked

DMT's

stimulatory

effect.

Multiple

ex¬

perimental

procedures44"47

have

suggested

a

stimulatory

effect of

serotonergic

neurotransmission

on

prolactin

lev¬

els.

Subtype

delineation

is

not

clearly

understood,

but both

5-HT1A48

and

5-HT249

receptors

appear

involved.

Cortisol levels

rose

dose

dependently

in

our

sub¬

jects,

consistent

with

the

robust

elevation of

corticotro¬

pin

described

above,

and

with earlier

human

data.14

Se-

Downloaded From: http://archpsyc.jamanetwork.com/ by John Chavez on 07/19/2016

  —I- 

2

5

10

15

30

Time,

min

 

60

Figure

9. Mean heart

rate

values

(in

beats

per

minute

[bpm])

in

11

subjects

after four doses

of

intravenous

dimethyltryptamine

fumarate

(0.05

mg/kg

[squares]

0.1

mg/kg [triangles],

0.2

mg/kg [diamonds],

and

0.4

 

[closed

circles])

and saline

placebo

(open

circles).

rotonin

receptor

activation

stimulates

peripheral

cortisol

levels,

most

likely

via

corticotropin release,30

with rel¬

evant

subtypes

having

been

discussed

previously.

The

cor¬

tisol effects

in

our

subjects

did

not

provide

as

useful

a

differentiation

among

doses

as

the

POMC

hormones.

Hallucinogen-induced hyperthermia

in

humans51

may

occur

by

means

of

central52

and/or

peripheral33

5-HT2

activation.

Dimethyltryptamine's hyperthermic

effects

are

probably

mediated

by

the

5-HT2

subtype,

as

5-HT1A

ago¬

nists

are

hypothermie

in

man.42 The

delayed

time

course

of the

temperature

rise

seen

with

DMT

is

of

interest

and

may

reflect

more

complex

homeostatic mechanisms

con¬

trolling

this variable than those with

a more

rapid

time

course.

Previous

human studies have

described variable ef¬

fects of

longer-acting hallucinogenic drugs

on

cardio¬

vascular

variables,

including hypotension

and

hy¬

pertension,

and

tachycardia

and

bradycardia.54

Dimethyltryptamine's

excitatory

cardiovascular

prop¬

erties

confirm and extend

previous

human data55

and

suggest 5-HT2

agonism.

The

5-HT2

agonists

usually

elicit

a

rise

in

blood

pressure,

without reliable

in¬

creases

in

HR.56·57

In

contrast,

5-HT1A

agonists

gener¬

ally produce

a

drop

in

blood

pressure,

again

without

consistent HR

effects.38

Dimethyltryptamine

resembles

other classic hallu¬

cinogens

in

its

mydriatic

properties59;

this

pupillary

di¬

latation

may

be

5-HT2 mediated.60

As

the

pupillary

di¬

ameter

data

were

both the least

reliably

measured

(ie,

we

did

not

use

automated

equipment,

which also could

have

measured

more

dynamic

features of

pupillary

function)

and contained the

most

missing

values,

they

must

be

con¬

sidered

preliminary.

We

found

no

dose

dependency

of

DMT's

stimula¬

tion

of blood

GH

levels. Meltzer

et

al14

demonstrated

a

cyproheptadine-sensitive

stimulatory

effect of

one

DMT

dose

on

peripheral

GH

levels

in

humans.

Perhaps

if

we

had continued

drawing samples

for

2

hours,

as

did Melt¬

zer

et

al,

we

might

have found better

separation

of ef¬

fects

by

dose.

Subjective

effects of

IV

dimethyltryptamine

were

con¬

sistent

with

previous

reports

of

parenterally

adminis¬

tered DMT16·31·61 and

were

marked

by perceptual,

disso¬

ciative,

affective,

and

cognitive

alterations

typical

of the

classic

hallucinogens.

Onset

from the

IV

route

was

nearly

instantaneous,

rather than

taking

several

minutes

to

de¬

velop,

and effects

were

shorter than

previously pub¬

lished

IM

data.

However,

this

time

course

is

quite

simi¬

lar

to

reports

of

DMT

smoked free base.27 The three

subjects

in

the group with

experience

smoking

DMT

free

base all volunteered that the

IV

route

was

somewhat

more

rapid

and

intense in

onset

and effects than

smoking

the

compound.

In

this

regard,

DMT is

actively transported

across

the blood-brain

barrier

in

the

rat.62

A

more

de¬

tailed

account

of these effects and their

quantification by

a new

rating

scale

are

described

in

a

companion

article.32

Dimethyltryptamine,

as

a

putative

directly

acting

5-HT

agonist

in

humans,

can

be

most

appropriately

compared

with

6-chloro-2-(l-piperazinyl)

pyrazine

(MK-212)

and

m-chlorophenylpiperazine

(m-CPP),

although only

m-CPP

has been

given

intravenously

to

humans. Neither

drug

pro¬

duces the

typical hallucinogenic

syndrome

in

normal

sub-

Figure

10. Mean arterial blood

pressure

(MAP)

values in 11

subjects

after

four doses of intravenous

dimethyltryptamine

fumarate

(0.05 mg/kg

[squares],

0.1

mg/kg

[triangles],

0.2

mg/kg

[diamonds],

and 0.4

mg/kg

[closed circles])

and saline

placebo (open

circles).

Downloaded From: http://archpsyc.jamanetwork.com/ by John Chavez on 07/19/2016