of PTSD

Identification

L. ALTER,

CAROL

ALAN

The

authors

with

measured

in a group

HARRIS,

GROBOIS,

M.D.,

at least

the rate

3 years

apy or radiation, PTSD field trial,

SEPTIMUS,

of cancer since

DAVID

M.S.W.,

HELENE

ALIZA

(PTSD)

MD.,

AXELROD,

BRIAN

in Cancer

and

were interviewed were compared

MEYERS

FRANCINE

PH.D.

MANDEL,

M.D.

KAPLAN,

of posttraumatic

Patients

diagnosis,

GOLDENBERG

BARBARA

determinants

survivors.

PH.D.

PELCOVITZ,

BARBARA

SANDRA

Survivors

who

receiving

had

no active

(N = 27). Patients, with a community-based

stress

a history treatment,

who

disorder

of cancer such

diagnosis

as chemother-

part of the DSM-JV group matched for

were

control

age and socioeconomic status. One member of the survivor group (4%) and no members of the control group met criteria for current PTSD (NS). Six of the survivors (22%) and no control subjects met lifetime criteria (P <0.02). Cancer patients have higher rate individuals

of PTSD than found who have experienced

in the community. other traumatic

Symptoms events.

closely

resemble

(Psychosomatics

O

ver the past several years, our treat patients with cancer has steadily. As a result of this progress, percentage increase

ability to increased a greater

of these patients is surviving. has lead to numerous reports

psychological sequelae of patients cancer. Symptoms experienced have increased levels tress, difficulties

of global in returning

This on the

surviving involved

psychological to vocational

tioning, psychosexual concerns, an increased rate of psychiatric Also, specific psychological

and for some, morbidity.’’#{176} symptoms re-

lated to the cancer experience, such as reexperience, avoidance, and heightened arousal, which are typical of posttraumatic stress dis-

of

37: 137-143)

ness. However, the DSM-IV PTSD Committee is currently considering modification of the definition of Criterion A.’2 who

This study investigates whether patients have a history of cancer diagnosis and

treatment, nosis and

disfunc-

1996;

those

a

such

who were

were at least 3 years receiving no active

as chemotherapy

the criteria

or radiation,

for PTSD.

the DSM-IV PTSD with a community-based

Patients field

for age and socioeconomic though studies of cancer

Received

January

meet

who

trial control

all of

were

part

of

were compared group matched

status survivors

26, 1994;

postdiagtreatment,

revised

(SES). have

June

Albeen

6,

1994;

order (PTSD), have also been reported.’” DSM-III-R criteria for PTSD specifically

ex-

accepted June 29, 1994. From the Department of Psychiatry, North Shore University Hospital-Cornell University Medi-

clude

for

cal College.

chronic

diagnosis definition, disorder

VOLUME

illness

as a stressor

qualifying

of PTSD (Criterion A); therefore, by patients do not meet criteria for the if the trauma identified is medical ill-

37.

NUMBER

2

#{149} MARCH

-

APRIL

1996

University

Address Cancer

reprint Center,

requests 3322

N.

to Dr. Alter, Broad

St.,

Temple P.O.

Box

38346, Philadelphia, PA 19140. Copyright Medicine.

©

1996

The

Academy

of Psychosomatic

137

PTSD

in Cancer

descriptive, the

Patients

none

presence

have

systematically

of PTSD.

tematically

assessing

The

assessed

advantage

of sys-

the prevalence

of PTSD

in

this population is that, unlike more general approaches, this methodology allows for specific questions about the impact of the illness and examines symptoms or reactions characteristically

seen

in individuals

affected

experience.

Our

site

individuals

with

chronic

completed

study

have

been

to severe

also

examined

associated

stress

and

with

focus events

on a patient’s or prior trauma

stressor (illness SES, education, or), PTSD,

which

termined magnitude

also

of

PTSD

Severity

been

with

of illness

threat of the illness important predictors demonstrated

poor

adjustment

that

ceiving after

of PTSD.

factors, illness

such as (stress-

associated

with

settings

have

to one’s of PTSD.

de-

of the

life may Kilpatrick

life-threatening

is particularly

collected

also be et al.’5

events

are

the

victim

is

were

and was then

by the research team A total of 65 women

ceived as unexpected, greater incidence of symptoms.

patients intrusive

was

per-

developed and avoidant

a

The

in participating

data

for

these

analyses

were

collected

The current is an event

138

as

central type of enough PTSD.

definition in Criterion A of PTSD outside the range of normal human

200). from

by phone

(n

=

24),

not

having

to participate at another wishing to discuss illness

time; (n =

2). Refusers were comparable in age, SES, disease type, severity, type of treatment, and other parameters studied. A control comparison group, demographically comparable to the survivor sample and living in the same geographic area, was recruited by means of a random-digit procedure.

Informed

consent

was

ob-

tained for all patients. The adult cancer survivor and control groups were composed entirely of white, suburban women of comparable age and took

place

of structured also completed

psychological

part of the DSM-IV PTSD field trials. A question of the PTSD field trials was the event that constitutes a stressor of high magnitude to qualify for a diagnosis of

in our =

to participate in the study in order to obtain 27 interviews (the number necessary for the DSMIV field trial). Reasons for refusal were lack of

series group

METHODS

identified

contacted

during a regularly scheduled the hospital. Each participant vor and control groups was

of cancer

re3 years

to arrange for an interview. were sequentially invited

found

recurrence

follow-up,

of Oncology (N a letter of invitation

SES.’7 Interviews

when

at our site

and at least

of cancer

her oncologist

dialing

factors of PTSD. when a traumatic

unexpected,

diagnosis

in active

more likely to develop PTSD. Cella and colleagues,’6 in a study of adult cancer survivors, that

subjects

of questionnaires

treatment,

hospital’s Division Each woman received

interest

stress-

the extent

currently

no active

time (but willing n = 12), and not

of higha higher

subsequent and

one of the major predictive There is also evidence that event

factors

that patients with a history trauma or life stress have

incidence ors.’2”4

risk

history of stressful life and the severity of the

also measured. of PTSD in other

battery

Sample

have been identiof development of study we chose to

severity). Other and time since

have

were Studies

two

the development

Although numerous factors fied as increasing the risk PTSD,’3 in this preliminary

to evaluate All

by extreme Women

This

chosen illness.

described below. Only data are presented in this report.

stress. that

the entire

was

distress

(the

in Table

control group and SES. The

or

The survivor measure of

Symptom

of the sample 1. The

homes

follow-up visit to in both the surviinterviewed with a

interviews. a self-report

[SCL-90-R]). The characteristics marized

at patients’

study

group

Checklist are sumand

the

were comparable in terms of age ages of the survivors ranged from

31 to 69 years. The average age at the time of interview was 54± 11 years old. The age of women in the community-based sample ranged

PSYCHOSOMATICS

Alter

1.CharacterIstics tion (N = 27)

TABLE

of the study

last at least 1 month. A DSM-III-R diagnosis current disorder is met if the patient meets

popula-

Prevalence

(%)

teria

n

in each

area,

with

symptoms

I month and being present A diagnosis of a lifetime

Diagnosis Breast

81

22

Other

19

5

Chemotherapy

44

12

linked

Radiation

II

3

trauma.

Chemo

30

8

all patients

(ongoing)

33

9

only

11

3

experienced. Research interviewers dents in doctoral programs

symptoms months

Treatment

+ Radiation

Tamoxifen Tamoxifen Recurrence

risk

Low

26

7

Intermediate

63

17

High

11

3

from

40 to 63 years;

the average

age

was

ogy,

signs

completion None of

of active

and 5.4 years the participants

48 ±

since initial had any of interview.

at any time more than 6 interview. Questions are

patient’s

In the case

experience

“cancer”

survivor

as the

social

Also,

interviewers

the

and

the standards for DSM-III-R

necessary diagnosis

training

sessions

were

jointly

to listen

to audiotapes

interviewers

from

the

for the of

The interthe reliability authors’

site

0.72

both high-magnitude ors. This measure

and low-magnitude stressis a structured interview de-

Survivors also completed the SCL-90-R,’8 a 90-item self-report measure, to assess the

signed

systematically

presence

the

of current

current

DSM-IV

as

sites was views.

for

four

held

to of

The Potential Stressful Life Events Interview’5 was used to obtain information regarding

whether

the other

a

consisted manual that

previous SCID-PTSD interviews. rater kappa coefficient measuring of

event

worker,

Training of interviewers a SCID-PTSD training

clearly delineated meet the criteria

group,

worst

were graduate stuin clinical psychol-

a master’s-level

PTSD.

of a specific

of the cancer

cited

at least 6 months. is met if

with

assess

at the time

to the

lasting

in the past disorder

for cri-

compared

to

disease

occurred before the

psychiatrist. of reading

6 years old. Community-based control subjects had no history of cancer. Women were an average of 4.6 years posttreatment diagnosis.

et a!.

and

field

lifetime

psychological

trial inter-

distress.

The

subject has been exposed to any traumatic events that would qualify under the DSM-III-R

Global indices

Criterion

A guidelines for PTSD. High-magniare those traditionally meeting

portray the individual’s level ing information on the number

A for PTSD; these include been a victim of or witness

the intensity of distress. Items are rated on a five-point scale of distress and result in nine

tude

stressors

Criterion as having

such things to combat,

physical or sexual assault or abuse, crime or natural disaster, or other events ered outside of the range of “normal” experience. Low-magnitude stressors loss

of job,

ration

financial

or divorce,

difficulties,

major

illness

violent considhuman include

marital

sepa-

in self or signifi-

cant

others, and death of a significant other. The Structured Clinical Interview for DSM-III-R (SCID), PTSD section, was used to ascertain tion of presence ence, VOLUME

the presence of PTSD. the SCID systematically of symptoms

arousal,

and

37 #{149} NUMBER

in three avoidance;

2

#{149} MARCH

The

PTSD assesses

areas: symptoms

-

APRIL

1996

por-

the

reexpenmust

Severity Index (OS!), one of three global of the SCL-90-R, can most sensitively of distress by usof symptoms and

primary symptom dimensions and three indices of distress. Internal consistency nine symptom dimensions as measured efficient

alpha

retest

ranges

reliability

Derogatis’8 the concurrent measure. Statistical of the survivor

from ranges

cites

numerous and

to 0.90.

from

0.8

studies

construct

validity

Testto

0.9.

supporting of

this

analysis included a comparison group with the control group for

the following high-magnitude

categorical events,

tude

PTSD-lifetime,

events,

0.77

global for the by co-

variables:

number

number

of

of low-magniPTSD-current,

139

PTSD

in Cancer

Patients

and SES, using Fisher’s whether the proportions approximately time PTSD

exact test, in each

equal. Also, were compared

PTSD

by means

nosis,

type

TABLE

exact

test

ongoing

for diag-

tamoxifen

control

counterpart the survivor

subjects.

The

Mann-Whitney

of time

treatment out PTSD

from

test

diagnosis

for those with were compared

and

time

PTSD and by means

Survivor

Control

P

<

PTSD

to the f-test) was subjects with the

also used to compare the survivors with to those without PTSD for the OS!. The butions

n(%)

use,

and severity of illness. For continuous variables, such as age, the Mann-Whitney U test (a nonparametric used to compare

PTSD and PTSD subscale prevalence in survivor group and control group,

2.

survivors with lifewith those without

of Fisher’s

of treatment,

to determine group were

Current

1 (4)

0

NS

Lifetime

6 (22)

0

0.02

Reexperience Current

10(37)

1 (4)

0.002

Lifetime

13 (48)

6 (22)

0.046

Arousal

was

Current

3(11)

1(4)

NS

Lifetime

8 (30)

2 (7)

0.03

PTSD distri-

Avoidance

from

Current

2(7)

0

NS

Lifetime

8 (30)

0

0.004

those withof the Sav-

Note:

NS

age test.

=

P’I’SD = posttraumatic not significant.

stress disorder;

RESULTS time Thirty-three percent of survivors and subjects reported a history of experience high-magnitude

stressor

(not

control with a

significant

[NS]).

of diagnosis,

or history

low-magnitude events sults from the SCL-90-R meeting

criteria

for

of other

high-

or

(Table 3). However, reindicated that women

lifetime

PTSD

had

scores

Although, by definition, 100% of survivors had experience with a low-magnitude event, 11% (3/27) had an additional low-magnitude event,

on the GSI that were significantly higher than those not meeting criteria for PTSD (P <0.02). Also, we sampled a group of women who will

and 33%

remain

(9/27)

of the control

subjects

reported

on tamoxifen

a history of low-magnitude events. One member of the survivor group (4%) and no members of the control group met crite-

indefinitely met criteria

ria for current PTSD (NS). However, six (22%) of the survivors met lifetime criteria, whereas none of the control subjects did (P < 0.02). Al-

proaches significance cal significance with

though

three

of the

high-magnitude the endorsement only relative the results avoidance, ble 2 and

six

survivors

finding

that

although

such

Although diagnosed

symptoms

140

time

of treatment

since

treatment

received,

or diagnosis

severity

of illness

cancer

significant,

it ap-

and might reach a larger sample.

statisti-

are major researchers

that being for cancer

and often traumatic events in one’s life, have not agreed on exactly what symptoms

are to be expected

for

someone surviving cancer. Studies examining both adult and child survivors of cancer exten-

in the cancer-survivor group. No differences were found between subjects meeting criteria for lifetime PTSD and those not meeting criteria in the areas of age, type

statistically

it is widely acknowledged and receiving treatment

psychological

survivor greater

levels

diagnosis,

for breast

DISCUSSION

of the subscales (reexperience, and arousal) are summarized in Taindicate

is not

experienced

events in addition to the cancer, of PTSD symptoms was made to the cancer. Data summarizing

are present in both the control and samples, they are found at significantly

treatment

(n = 9) and found that none of them for PTSD (P < 0.07). Although this

or at

sively

describe

the heightened

presence

of psy-

chological distress; increased concerns about sexuality, intimacy, and physical well-being; and concerns about the future.’’#{176}’9 However, researchers have not been able to agree on whether the presence of psychiatric diagnoses

PSYCHOSOMATICS

Alter

TABLE

Survivor

3.

group

Age Years

since

Years

since diagnosis

treatment

Global Severity score Diagnosis,

Index

Researchers

predic tors of PTSD

PTSDLifetime (n=6)

(n=21)

52 ± 10.8

54 ± 11.1

P

thinking

NS

4.8 ± 1.5

4.5 ± 1.7

NS

6.3±

5.2±

NS

59 ± 9.6

1.6

48 ±.6

0.05

Stage

I

0

Stage

Breast,

II

Other Treatment,

NS

5(24)

4(67)

13(62)

NS

2(33)

3(14)

NS

4(67)

8 (40)

NS

n (%)

Radiation +

Chemotherapy Radiation

0

4(19)

NS

2(33)

6(30)

NS NS

only

0

3(15)

Tamoxifen,

ongoing

0

9 (43)

<0.07

0

7(33)

NS

Severity, n (%)

Low risk Intermediate risk

5(83)

High

1(17)

2(10)

NS

3 (50)

7 (33)

NS

Note:

dicated. stress

risk n (%)

Values are means NS = not significant; disorder.

is truly elevated with the general focused

12(57)

NS

in-

± SD unless otherwise PTSD = posttraumatic

in this population, population. Such

on the presence

as compared studies have

of generalized

the DSM-III-R specifically illness as meeting Criterion

anxiety

rules out A. There-

fore, the presence of reexperience, arousal, avoidance symptomatology has never been tematically studied in this population. knowledge, this is the first controlled report on prevalence of PTSD using

and sys-

To our study to structured

#{149} NUMBER

-

APRIL

1996

with

Scale

(IES)

and

other

at assessing other condiand have found increased

these

measures

this

remain

linked

also been highly of psychological study

is the

aimed

to acute correlated distress.

first

at arriving

to

use

a

at a diag-

GSI, indicating increased levels of acute psychological distress. However, the distress is found in women who do not presently meet

not be the current status, predict

a history continued A number

gested PTSD

stress-

basis

for their

psychological

of PTSD at an earlier “distress” in these

time may patients.

of risk

been

factors

have

to explain why one person and another may not. The

life threat is often cited as a predictor Although any diagnosis of cancer ceived as life-threatening, severity prognostic indicators at the time may

offer

specific

further

elucidation

perceived

life

sug-

may develop perception of of PTSD.’2 may be perof disease or of diagnosis

of an individual’s

threat.

However,

in our

sample, severity of disease did not in any way predict development of PTSD symptoms. The suggestion that ongoing treatment with for PTSD may offer an

perceived 2 #{149} MARCH

especially

One interesting finding in our study was that women meeting lifetime but not current criteria for PTSD had elevated scores on the

over

37

of Events

structured interview nosis of disorder.

inclusion of chronic or for PTSD.2#{176}

VOLUME

arousal,

Several studies have docuof such symptoms by use

conditions and have with current levels

tamoxifen

as an eligible

and

scores in cancer survivors.24’5 ‘ ‘ .16.2 I .22 Although many have asserted that these scales indicate stress-related symptoms that may be indicative

diagnostic interviews in a sample of chronically ill patients. Our findings are consistent with proposed changes in DSM-IV that allow for illness

of stress

criteria for PTSD, but have a history of meeting criteria at some other point in their lifetime. This finding suggests that although PTSD may

and depressive disorders.49’#{176} Furthermore, although some have completed extensive psychiatric interviews, such as the SCID and DIS, patients, by definition, would not meet criteria for PTSD on the basis of Criterion A alone because chronic

Impact

Furthermore,

Tamoxifen

trauma,

of the

of PTSD,

Chemotherapy

Prior

or increased

conditioned stimuli. mented the presence instruments aimed tioned responses

n (%)

Breast,

a model

coping have determined that in cancer, as in other life stresses, there may be some characteristic responses, such as avoidant or intrusive

No PTSD

1.8

using

et a!.

was

associated

with

a decreased

is intriguing. Tamoxifen patients an enhanced sense uncontrollable

process

prevention

of life threat.

as

risk

treatment of control well

as

It is known

a

PTSD

in Cancer

that while they

Patients

patients are undergoing

report

less

distress

than

period just after

completing

Similarly,

currently

people

they active

treatment

the process

of dealing

do

perspective

is that

in the

treatment.

experiencing

standing

trauma

often after

do not manifest symptoms of PTSD until the traumatic incident is ended.23 Another risk factor identified as predictive has been the history of experience of prior trauma or another

severe

stressor.

Whereas

patients meeting criteria for lifetime a history of a prior high-magnitude

50%

33% of those without PTSD had such Although this difference did not meet significance results ated

criteria

may

in our limited

be indicative

these

when

evalu-

of a trend

diagnosis

to this

experience a constellation impairs their functioning.

the any-

is a syndrome

that

has

a major

tioning. The intrusiveness, and avoidance phenomena

impact

that PTSD on func-

symptoms that can severely limit an individual’s ability to function adequately. This study of PTSD in cancer survivors

to clarify of trauma.

the experience and may also

our understanding The advantage

under-

and treating

the psychiat-

to account reactions nightmares,

for the whole spectrum of typical to overwhelming trauma, such as social isolation, and a sense of a

those fail

foreshortened future. The failure of many studies to document an increased prevalence of depressive or anxiety disorders in cancer survisupports

prove

the notion

that

a trauma-specific

be more inclusive. that have proven

beneficial

Finally, helpful

treatin al-

to this

of help

of the experience of conceptualizing

that

population.24

these

data

support

the

large

body of work describing significant psychological sequelae in this patient population, and these results underscore the need for both further descriptive study and interventions aimed at the specific

physiologic arousal, are a cluster of

begins to better characterize the patient surviving cancer

a PTSD

of the illness. Approaches like only on conditioned responses”

It is clear

Our findsurvivors

of symptoms By definition,

from

for a broader

lowing trauma survivors to cope with the effects of exposure to overwhelming events should also

adds

thing to clinical practice and research. ings suggest that some chronic-illness

in predicting

diagnosis may ment strategies

applying

population

cancer

ric effects that focus

vors

a history. statistical

sample,

with a larger sample. A central question is whether

PTSD

of

PTSD had stressor,

with

it allows

ence

symptoms of cancer

This

associated

with

the

experi-

survival.

work

was presented

at the Academy

of

Psychosomatic Medicine Annual Meeting, San Diego, CA, October 30, 1992. This research was supported in part through the NIMH-funded and NJMH Grant JROJ

DSM-JV MH4372-04.

field

trials

References I. Morris

T, Greer

adjustment Cancer

1977;

2. CelIa

D, Tross

from

Hodgkin’s

54:616-622 3. Mumma G, chosexual pact

of

disease.

study.

of

advanced

ABVD,

adjustment

J Consult Lesko

of acute

Anderson

survivors

D,

Hosp

Clin

to survival

6. Fobair

Psychol

1986;

L:

Long-term

psy-

leukemia vs.

Psychiatry I,

at increased

survivors:

CelIa risk

conventional

1992; D,

im-

14:43-55

et al:

disease

alternating

for problems

in psy-

and sexual

function

of survivors

7. Bloom

treated

with

R, Bloom

survivors

ABVD.

by

MOPP,

Cancer

1992;

J: Psychological

disease.

I Clin

probOncol

response

adaptation

among

of acute

leukemia

transplantation

Psychosom

a pro-

survivors

1989; 63:406-413 I, Mumma 0, et al: Long-term

adjustment

of bone marrow

to mastectomy:

study. Cancer 1987; 59:189-196 R, Morvay T, et al: A longitudinal

of psychological

cancer. Cancer 9. Lesko L, Ostroff

therapy.

J, et al: Psychosocial

of Hodgkin’s

4:805-814

spective comparison 8. ElI K, Nishimoto

logical of

P. Hoppe

lems among

analysis

Hodgkin’s

adaptation. Cancer 1992: 70:2214-2224 A, Anderson I, CelIa D, et al: Comparison adaptation

Hodgkin’s

or MOPP

70:2508-2516

transplantation Gen

A,

psychosocial

and social

follow-up

1986; Mashberg

marrow

chosocial 5. Komblith

a two-year

5: Psychological

adjustment

4. Komblith

142

P: Psychological

40:2381-2387

chemotherapy. disease

H, White

to mastectomy:

Med

survivors:

vs. conventional 1992;

of

psychoimpact chemo-

54:30-47

PSYCHOSOMATICS

Alter

10.

Schmale

A, Morrow

cancer

survivors.

11. CelIa

D, Pratt

A, Holland

sea, vomiting, patients atry

0, Schmitt Psychosom

1986;

in cured

disorder:

for

Working

disease

18. Derogatis

A.

Vol

Press

by Oldham

DC,

J Ab-

posttraumatic

1989;

15. Kilpatrick

DO,

American

MR.

Psychiatric

of posttraumatic

morbidity: factors. Br

HS,

Freedy

JR:

results.

Meeting,

16. CelIa

D, Mahon

VOLUME

37.

American Washington,

Psychiatric DC,

S, Donovan

NUMBER

2

#{149} MARCH

Scoring,

MD,

and

Clinical

Psy-

1977

5: Death

anxiety

in cancer

survival:

cross-validation study. J Pers Assess

1993 21. Manuel 22. Reiker factors,

Association

review An-

- APRIL

GM,

Roth

of Human

recurrence

1996

Association,

5, Keefe Stress

P. Fitzgerald curative 1989;

a

1987;

F, et al: Coping 1987;

E, Kalish therapies,

Washington,

and

DC,

with cancer.

13:149-158 L, et al:

Psychosocial

behavioral

outcomes.

64:2399-2407

tional University 24. Foy DW (ed): Strategies. New

1992

M: Cancer

Psychiatric

23. Kristal H: Massive

Posttraumatic

stress disorder field trialreport: a comprehensive of initial

Administration,

Draft Criteria, 3/1/93: Task Force on DSM-IV.

American

Cancer

154:221-228

Resnick

Status.

CT, Yale University, 1975

I. Towson

-

Research,

preliminary

Journal A: The etiology

SCL-90-R:

Manual

D, Tross

20. DSM-IV

Review

J, Riba

predisposing, precipitating, and perpetuating J Psychiatry

LR:

Haven,

51:45 1-561 of

Psychiatric

12. Edited

Washington,

DSM-IV.

1993

14. McFarlane

Paper. New

Procedures 19. CelIa

in postiraumatic

the

100:346-355 in the diagnosis

stress disorder, in American

nual

nau-

Am I Psychi-

issues

considerations

of Psychiatry,

traumatic event. Behav Med 1990: 16:15-22 17. Hollingshead AB: Four Factor Index of Social

chometric

norm Psychol 1991; 13. Davidson J: Issues

Press,

Hodgkin’s

of chemotherapy.

JR. Foa EB: Diagnostic

Tasman

anticipatory

143:641-643

12. Davidson stress

of

1983; 45:163-169

J: Persistent

and anxiety

after completion

M, et al: Well-being Med

et a!.

Psychic Trauma.

Press, Treating

1968 PTSD:

New

York,

Interna-

Cognitive-Behavioral

York, Guilford, 1992

as a

143