T m: JOURNA L O F I NVEST IGATIVE DEHMATOLOGY, 65:472- 475, 1975
Copy rig ht © 1975 by The Williams & Wilk ins Co.
Vo l. 65, No.5
Printed in U.S.A.
FURTHER STUDIES OF PHOTO AUGMENTATION IN HUMANS: PHOTOTOXIC
REACTIONS
K AYS H . KAIDBEY, M.D ., AND ALBERT M. KLIGMAN, M .D ., P H.D .
Department of Dermatology, University of Pennsylvania , Sch ool of M edicine, Philade lphia, Pennsylv ania
T h e a ddi t io n of m od est a m o un ts of lo n g ul t rav iolet li ght (UV- A) to Y2 minim a l e ryt h e m a
dose (M E D ) of s unburnin g irrad iatio n (UV-B) produ ced e ryth e m a v is ible at 24 hr, t h e
p rototy pe of t he photoau g m e n tation phe no m e n on . Th e results we re t he sa m e wheth er UV-A
was given befor e or afte r UV -B . Photoau gm e n ta tio n co uld be d e m o ns tr ate d afte r a n in ter va l
of 6 hr between d oses, b u t n ot a fter 1 d ay. Phot oau g m en ta ti on h as a lso bee n d e m o n strated
clini cally a nd histologicall y wit h two to pi cal photose n s itize rs, coa l tar a nd 8- m etho xypso ra le n .
Pi gm e n tat io n d e velo ps sh ort ly a fte r huma n s kin
is ex p osed to lon g ul trav io let li gh t (UV-AJ. This
phe n om e n on , known as im media t e p ig m e n t d a rk e nin g, occ urs as a r es ul t o f t h e oxid ation of
prefo rmed colorless m e la nin a nd p ossi bl y redistr ibut io n of mela n os m es [1] . T h e d a rkenin g was
p rev ious ly regard ed as a useful a d aptation , provid ing a te m pora ry protection aga inst s unburnin g
ra di atio n (UV-B) un t il m ela n oge n esis co uld get
und er way [2 ). Willis et a l teste d t hi s h y pothes is
b y d ete rminin g t h e minim a l e ryth e m a d ose (MED )
in s ites d a rke n ed by prior expos ure t o UV-A [31.
The resul t was opposite to th eir e xpectation s.
S unburn rays (UV-B) ca u sed great e r redn ess in
d a rke n ed sites a nd led to gr eater s u sce pt ibili ty
rath er t h a n protection . They ca lled this poten t iat io n of t h e s unburnin g effects of UV-B a nd UV-A
ph otoa u g m e ntat ion . E a rli er , v an d e r Le un a nd
Stoop , w hile t ry ing to d e m o nstrate photo recove ry
from eryth em ic r a di ation a lso ob se rved t h at p r ior
irradi ation wi t h vis ibl e li gh t in creased s usce pt ibil ity to UV-8 [4 ). T hey t hou g ht t hat e nh a nce m e n t
could b e ex pl a ined b y a ddition of e n e rgies .
Th e curre nt work is a n exten s ion 0(' Willis et a I's
obse rvatio n s o n photoa u gm e nta ti o n . W e wis h ed t o
find ou t th e minimum d oses of UV-A b y UV-B
t hat wo uld elicit photoa u gm e ntatio n ; wheth er
t he re wo uld be a differe n ce if UV-A we re given first
or second ; a n d how lon g a n in te rva l b etwee n d oses
woul d still b e a ugm en tative. Fin a lly, we a tte mpted
to establ ish wh ether photoa u gm e n tat ion co uld be
de m onstr ate d in phototoxic reactio ns as well.
L ight So urces
Long ultrav iolet light (UV-A) was prov ided by t he
xe non sola r simul ato r with t he Schott WG345 fil te r to
eliminate erythemi c radi ation (in te nsity 25 X 10' /1w /
cm ') (5 ]. Erythemic radi ation (UV- B) was obta ined from
a bank of 5, closely set, !1uorescent Westinghouse F S20
sunl a mps moun ted in a refl ecto r housing. The la mp-tos kin distance was 20 cm.
Experimental Procedures
Mi nim al ery thema dose . The MED for eac h indi vidual
was determin ed by exposing l -c m squ a res on t he back for
60, 75, 90, 105, a nd 120 sec of ery t hemic radia ti on . The
sites were read 24 hr late r.
Min. imum doses of U V-B and UV-A for photoaugm en tation. T he lowest combinat ion of UV-A a nd UV-B
required to eli cit redn ess in 24 hr was established in t he
followin g way : a grid consisting of 4 hori zontal a nd 4 '
ve rt ical rows of l -cm squ ares was outlined on the
un tann ed mid -bac k by mea ns of t hin strips of adh esive
ta pe. Vert ical rows one to three received 10, 15, a nd 20
min of UV-A, res pective ly. Wi t h filtered xe non radi at ion
it ta kes ove r 60 min of ex posure or a bout 90 joul es/c m 2 to
cause redn ess in the average subject. Thus t he hi ghest
dose of UV -A was one-third t hat required to produ ce
erythema. Ho ri zo ntal sq ua res were irradi a ted with ~ l4
a nd
of MED from the !1uarescent sunl a mps. In thi~
way, eac h sq ua re received different overl ap ping doses of
UV-A and UV -B. One hori zontal a nd one verti cal column
of cont rol squ a res received either UV -B or UV -A alone.
The reactions were evaluated 24 hours la ter.
Sequ.en ce of UV administra tion . First t he minimal
co mbined doses of UV -A and UV-B t hat would produce
photoaugmentation were dete rmined for eac h of 10 indi viduals; t hen in one site UV -B was given before UV -A
and in the ot her t he sequ ence was reve rsed .
M ATERIAL S AND METH OD S
Interval between ex posu.res. Five, l -cm squ a res on t he
Subjects
mid bac k of 8 subj ec ts were irradi ated for 15 min wi t h
T hese we re adu lt, white, healt hy, male prisoner volun UV -A. Single sites t hen rece ived V2 MED of erythema
teers between the ages of 21 and 40 yea rs, from whom inradi at ion imm edi ate ly a nd 2, 4, 6, and 24 hr late r.
for med conse nt was obta ined .
Photoau.gm entation an.d photot oxicity. Phototoxic responses are ty pically medi ated by UV -A. We exa mi ned
two phototox ic mate ri als, coal tar a nd 8- methoxy psorReprin t requests to: Dr. A. M. K lig man , Depa rt ment
alen (8-MOP) .
of Dermatology, Duhrin g Labo ratories, Hos pital of the
U nive rsity of Pennsylvania, P hil adelphia, Pe nnsylva ni a
Coal tar phototoxicity: On the back, 1.5-cm squares
19104.
outlined with adh esive ta pe received 40 /11 of a 25 %
*
472
Nov. 1975
PH OTOAUGMENTATION I N HUMANS
eth an olic solu t ion of coal tar dist illate (Doa k P ha rm aca l
C
Jamaica, New York ). An eq ua lly sized sq uare of
So., n Wrap was placed over t he site a nd sealed to t he
:.ra with overlapping strips of impermeable plastic tape
s BII:nderm , 3M Co) for. 2 hr. Pai red sites were t hen
~ diated with 2.5 mm of UV -A a lone or t he same dose to
Ir~ h was added :x MED of erythema radiation. Un; I~ed s ites received the comb inat ion of 2.5 min of UV-A
re: f.I MED . Reactions were graded 24 hr later.
an 8_Methoxypsora len (8-MOP): Forty-microliter port· 5 of a 0.1 % 8-MOP solu t ion in eq ual parts of 95 %
~~n no l a nd propylene glycol were pipetted in to 1.5-cm
=qu~res of skin a nd occ~ude? for 2 hr as a~ove. Pa!red
sites received eithe r 2 mm of UV-A a lone or m co mbma tion with 'Vol MED, the controls bemg the sa me as above.
The reactions were ~ ra d ed at 48 hr.
.
Phototoxic reactIOns were scored as follows: 0, no
reaction ; I, minima l eryth ema; 2, moderate erythema; 3,
intense erythema a nd edema. . .
.
In 5 subj ects, 4-mm punch bIOpSies were ob ta med from
hototox ic reactions provoked by l!.V -A ~ I one and UV -.A
PI
UV-B. The spec im ens were fI xe d m 10% formalm
pus sta ined with
. h ematoxy I·m an d eos m
. .
and
RESULTS
Minimum doses to elicit photoaugmentation. In
11 of 20 s ubjects, 1/ 2 MED. was e.levated t.o a n.
erythema level hwhen co.mbdmed :WIth
I 130 mb~n of
UV -A while 6 ot ers reqUire 15 mm . n su Jects,
thema occurred wit h 1/4 MED and 15 min of
ery -A. Erythema d I·d not appear 111
. any 0 t· t h e
UV
control sites.
Sequence of administration. The res ul ts were t he
same whether UV -B was given before or after
UV-A.
Int erval between doses. In 6 of 8 subj ects,
hotoaugmentation could be demonstrated when
~12 MED was given 2, 4, a nd 6 hr after 15 min of
UV -A. With a gap of 24 hr between ~xposures,
erythema did n ot ap pear m any of t he eight.
TABLE.
Subj ect
Potentiation of photo toxic reactions with
suberythemal doses of UV-Bn
8-MOP
COlli tar
UV -A UV ·A + " , MED
UV ·A UV-A + "., MED
3
2
1
1
4
2
5
6
7
8
9
1
2
2
1
2
3
1
2
10
3
2
2
2
2
2
3
2
2
3
1
1
2
2
2
2
1
1
1
1
2
1
2
2
2
2
2
1
1
1
0 = No reaction
1 = Minimal erythema
2 = Moderate erythema
3 = Intense erythema a nd edema
For detail s of experimenta l desi gn see Materials and
M ethods.
a
473
FIG. 1. Coal tar phototoxicity (UV -A). Note perivascular edema a nd minimal mononuclear cell infiltrate (H &
E, x 94).
Phototoxic reactions to tar and S-MOP. Clinical:
The add it ion of ~ MED to t he UV -A- irradiated site
resulted in intensified reactions of one grade point
in at least 6 of t he 10 subjects with both coal tar
and 8-MOP (Tab.). All of course showed eryt hema, usu a lly mild, with UV-A a lone.
Histopathologic appraisal: Coal ta r : With UV-A
a lone t he epidermis was largely unaffected except
for occasional intercellular edema. An infiltrate of
mononuclear cell s formed a round t he venules in
t he upper dermis along with mild edema (Fig. 1). A
few n eutroph il s were someti m es interspersed. With
UV-A a nd -% MED, t he epidermis showed some
spongiotic foci with vacuoli zation of malpighian
cells. The uppermost cells displayed increased
eos inophilia and blurring of cellular outlin e. Intracellular edema was prominent, especia lly in the
basal cell layer. S unburn cells were distinctly rare.
In t he dermis, t he perivascular infiltrates were
larger and extended down to t he deeper vessels
(Fig. 2). Neutrophil were more ab undant in t he
predominantly mononuclear reaction.
8-MOP: In con trast to coal tar, epidermal
chan ges were far more evident with UV -A a lone . In
some portions, cellular necrosis was evident. Vacuoli zation and spongiosis were quite pronounced
(Fig. 3). In t h e dermis, a slight mononuclear
infilt rate developed a round t he superficia l vessels.
When ~ MED was added to the UV -A dose the
epidermal damage was harply accentuated. Necrosis was t he dominant effect with the cells
unusu ally swollen (Fig. 4). Sunburn cells were rare.
T he dermal perivasc ular infi ltrate was a lso intensi fied and involved deeper vessels. The cells were
predominantly mononuclear, admixed with some
n eutrop hil s. The vessels were dilated and occasiona lly t he endoth elia l cells were swollen.
DISCUSSION
UV -A can induce erythema but on ly with energies several orders of magnitude greater than with
474
Vol. 65, No.5
KAIDBEV AND KLIGMAN
UV -B. In prev ious work from t his laboratory and
now aga in we have shown that skin exposed to s mall
doses of UV -A beco mes more vulnerable to t he
effects of erythem ic radiation (290- 320 nm) . Willis
et al [3] used relatively large doses of UV -A (30 min
or 45 J/cm 2 ) from t he same lamp source. We found
that the sunburn reaction could be potentiated
with as little as 10 min (15 J/cm 2) of UV -A in more
than half the subjects receiving 1/2 MED.
T he seq uence in which UV-A and UV-B was
adm inistered did not affect the outcome. Photoaugmentation occurred with a 6- hr interval betwee n expos ures but not with a 24- hr in terval.
This phenomenon h as to be taken into account
in certain common situations. For example, exposure to midday summer erythemic radiation will
render the skin more susceptibl e to long ul t rav iolet
FIG. 3. 8-MOP phototox icity elicited by exposure to
rays for t he rest of that day . Formerly, it was said UV-A alone. Note intercellul ar edema and vac uoli zation
that late afternoon sun was harmless, t he ery- of keratinocytes (H & E, x 190).
them ic rays having been filtered out by t he atmosphere. However, afternoon sun is rich in UV -A. It
can only worsen a sunburning dose received earlier.
It should be pointed out t hat many sunscreens do
not absorb in t he long ul t raviolet region. One
should beware of afternoon sun when sunbathing
for the first time or two in summ er.
P hotoaugmentation can also be demonstrated in
phototoxic reactions. When a suberythem ic dose of
UV -B was added, the reaction produced by UV-A
alone was intensified. With 8-MOP and coal tar
the augmented response was particularly evident
at the microscopic level.
Photoaugmentation may be a poss ible exp lan ation for the difficu lties in vestigators have had in
demonstrating the phototoxic capab ilities of photosensitizing drugs. Tetracycline phototoxicity is a
case in point. As with a ll we ll -know n p hototoxic
FIG. 4. Spec imen from same subject as in Figure 3.
agents, t he action spectrum is in t he long ultravio- 8-MOP phototoxic ity produced by same dose of UV-A
let region [6]. Irradiation of tetracycline-treated plus Y. MED of UV -B. More pronou nced epidermal injury
albin o an im als with UV -A from artificial sources (H & E, x 94).
.. ;
::'
\
) .
I '
."
F IG. 2. Spec im en from sa me subject as in Figure 1
exposed to an eq ual dose of UV -A to whic h r.I MED of
UV -B was added. Note marked acce ntuation of perivascu lar infiltrate and some vacuo li zat ion of malp ighi an
cells ( H & E, x 94).
will rather regu larly elic it phototoxicity [7]. In
hum ans, however, we and others have not been
ab le to cons istently demonstrate phototoxicity
with a variety of artificia l li ght sources of appropriate spectral quality [8 ]. Many persons are unaffected despite hi gh doses of drug a nd UV -A. By
cont rast no difficulty seems to be experienced by
workers who use hours of exposure to natural
sunli ght in both a nim als [9 ] a nd hum a ns [10,11].
We suggest that it is because of photoaugmentation t hat natural sunli ght is so mu ch more effective
in triggeri ng phototoxic reactions to drugs . We
t hink it likely t hat photoaugmentation is a general
biologic respo nse t hat will a ppl y to all species of
ph otosensitiz in g drugs whether given ora lly or
topically .
Recently, Yi ng et al [12] argued that photoa ugmentat ion is due to ph otoadd ition of energies.
Thus, when energies of UV-A a nd UV-B were
ex pressed as a fraction of t he dose of eac h required
Nov.1975
to ca use minimal erythema, t he combinations of
UV -A and UV -B t hat just caused erythema added
up to 1. Evidence for t he unity theory has also been
presented by Sayre et al for the two wavelengths,
254 a nd 297 nm (13] . These photoaddition theories
all state t hat combinations of suberythemic doses
of different spectra l ranges will yield redness at
that point when t he sum of t he fract ions is 1. For
example, erythema would occur with V.I and ~
l\1EDs ofUV-B and UV-A or the reverse. Similarly,
tWO V2 MEDs will give redness. The data compiled
by Findlay (14] for t hreshold erythema responses
to polychromatic li ght over t he 250 to 320 nm
range , however, s howed no ev idence of simple
summation of energies. In our hands, too, t he
administration of 1/2 MED of UV -C (cold quartz,
254 nm) a nd V2 MED of UV-B did not cause
erythema (unpub lis hed observatio ns). In our
xenon system with the WG345 filter , doses of UV-A
in excess of 90 J/cm 2 (60 m in) a re required to
produce eryt.hema. We d~monstr ate d photoaugmentation with on ly one-sixt h t hat dose (15 J) of
UV-A and V2 MED of UV-B. It is also worthwhile
mentioning t hat the time-course of UV -A erythema is entirely different from that of UV -C and
UV -B. T here is no latency period. Erythema devel ops immediately after irrad iation and shows considerable variability , reaching a maximum in a few
minutes to an hour and often disappearin g en tirely by 24 hr (15]. This is in marked contrast to
UV-B erythema, which reached a maximum in 24
hr . Much , t herefore, depends on t he time of
grading.
It seems to us exceed in gly improbable that t he
biologic effects of different erythema-producing
waveba nd s can be added arit hm et ically. The responses e li ~it~d by the vari?us bands differ mar~ ­
edly in their t ime patterns, 111 color and hue, a nd In
histologic changes. It would be remarkab le indeed
if t h ese distinctive reactions would follow t he rule
of unity in regard to on e feature, erythema.
PIiOTOAUGMENTATION I N HUMAN S
475
REFERENCES
1. Pathak MA , Stratton K : Effects of ultraviolet and
visible rad iation and the production of free radicals
in skin, T he Biologic Effects of Ultraviolet Radiat ion (w ith Emphas is on the Sk in ). Edited by F
Urbach. Oxford, Pergamon, 1969, pp 207-222
2. Monash S: Immed iate p igme ntation in su nli ght an d
a rtifi cia l li ght. Arch Derm atol 87:686- 690, 1963
3. Willis I, Kligman AM , Epstein J: Effects of long
ultraviolet rays on hum a n sk in: photoprotective or
photoaugmentative? J Invest Dermato l 59:4 16420, 1973
4. van der Leun, JC, Stoop TH: Photorecovery of
ultrav iolet eryth ema, The Biologic Effects of U ltravio let Rad iat ion (with Emp hasis on the Skin).
Edited by F Urbach. Oxford, Pergamon, 1969, pp
215-254
5. Berger DS: Specification and design of sola r ultravio let s imulators . J Invest DermatoI53:192- 199, 1969
6. Stratigos JD, M agnus IA: Photosensitivity by
demethy lchlortetracycline and sulfanilam ide. Br J
Dermatol 80:391-405, 1968
7. Ison, A, Bla nk H: Testing drug phototox icity in mi ce.
J Invest Dermatol 49:508- 511, 1967
8. Maibach HI, Sams WM , Epstein JH: Screening for
drug tox icity by wavel engths grea te r than 3100 A.
Arch Dermatol 95:12- 17, 1967
9. Sams WM Jr, Epste in JH: T he ex peri m enta l production of drug phototoxici ty in guin ea pigs. I. Using
s unli ght. J In vest Derm atol 48:89- 94 , 1967
10. Harber LC, Tromovitch TA, Baer RL: Stud ies on
photose ns itivity due to demethyl chl o rtet racycline. J Invest Derm atol 37:189- 193, 1961
11. Frost P , Weinstein GD, Gomez EC: Methacycline
a nd demeclocy cline in relation to sunlight. JAMA
216:326-329, 1971
12. Yin g CY, Parrish JA, Pathak MA: Additive eryt hemoge ni c effects of middl e-(280- 320 nm) a nd
long-(320- 400 nm) wave ultraviolet li ght. J Invest
Dermatol 63:273- 278, 1974
13. Sayre RM , Olso n RL, Everett MA: Quantitative
stud ies on eryt hema. J Invest D erm ato l
46:240- 244, 1966
14. Findlay GH: An automatic fractionator for li ght
dosage on t he sk in . Its application to t he polychromatic minim a l erythema dose . Br J Dermatol
79: 148-152, 1967
15. Bachem A: Time fa ctors of erythema a nd pigme ntation produced by ultrav iolet rays of different
wavelengths. J Invest Dermatol 25:215- 218, 1955