In,. J. Radralron O ncology Bml. Phy s., Vol. 28. pp. 523-526 0360~3016/94 $6.

00 + 00
Printed in the U.S.A. All rights reserved. Copyright 0 1993 Pergamon Press Ltd.

??Special Feature

CURRENT RADIOSURGERY PRACTICE: RESULTS OF AN ASTRO SURVEY

DAVID A. LARSON, M.D., PH.D., FRANK BOVA, PH. D., DONALD EISERT, M.D.,
ROBERT KLINE, PH.D., JAY LOEFFLER, M.D., WENDELL LUTZ, PH.D., MINESH MEHTA, M.D.,
JATINDER PALTA, PH.D., KEVIN SCHEWE, M.D., CHRISTOPHER SCHULTZ, M.D.,
ED SHAW, M.D. AND J. FRANK WILSON, M.D.
Task Force on Stereotactic Radiosurgery,American Society for Therapeutic Radiologyand Oncology

Purpose: Although there is increasing interest in radiosurgery, little quantitative data regarding current patterns
ofurgery practice are available. We developed a radiosurgery questionnaire to obtain information on radio-
surgery practice.
Methods and Materials: We distributed the questionnaire to the entire membership of the American Society of
Therapeutic Radcand Oncology in early 1993. Responses were obtained from 74 facilities that practice
radiosurgery, corresponding to over 6000 treatments carried out since 1983 by 135 radiation oncologists and 130
physicists.
Results: Most res,pondents were found to work within a multidisciplinary team, consisting of the following specialists
(average hours dtevoted per patient on day of treatment in parentheses): radiation oncologist (3.8), neurosurgeon
(3.2), physicist (6.1), radiologist (0.7), nurse (2.7), other (3.0). On average, neurosurgeons and nurses who perform
Gamma Knife radiosurgery devote significantly more time-per-patient on the day of treatment than their peers who
perform linac radiosurgery. On average, less experienced radiation oncologists and physicists (I 24 months ex-
perience, or I 50 patients treated) devote significantly more time-per-patient on the day of treatment than their
more experienced peers. Although there are many more linac radiosurgery facilities than Gamma Knife facilities,
on average the number of patients treated per month per facility is significantly larger at the latter. On average,
follow-up responsibilities are nearly equally shared by radiation oncologists and neurosurgeons, except at Gamma
Knife facilities, where neurosurgeons assume a larger percentage of follow-up responsibility. The percentages of
patients treated at linac facilities for metastases or primary CNS malignancy are larger than the corresponding
percentages at G.amma Knife facilities; the opposite is true for arteriovenous malformation, acoustic neuroma, and
meningioma.
Conclusion: Current radiosurgery practice usually involves a team approach, with participation of specialists from
radiation oncology, neurosurgery, physics, radiology, and nursing. The average number of M.D. and Ph.D. hours
required per treatment on the day of radiosurgery is high.

Radiosurgery.

INTRODUCTION types of techniques are used, what indications are most

common, and what degree of effort is usually required,
Although concepts of stereotactic radiosurgery were de-
among other things. Therefore, we developed a radiosur-
veloped over 40 years ago (Leksell, 195 1), widespread use
gery questionnaire, which was distributed to the entire
of the technique did nlot occur until the 1980’s. Along
membership of the American Society of Therapeutic Ra-
with proliferation of facilities there has been an increase
diology and Oncology (ASTRO). To the best of our
in number of clinical and scientific publications, symposia,
knowledge, no formal survey of radiosurgery practice pat-
conferences, and workshops on radiosurgery. A growing
terns had been carried out previously.
number of radiosurgery hardware and software products
have become available commercially. In addition, the
public has been exposed1 to radiosurgery through frequent METHODS AND MATERIALS
newspaper and television reports.
Since radiosurgery is obviously of interest to physicians, A simple questionnaire containing nine questions was
patients, and the public, it seems worthwhile to attempt distributed to the entire ASTRO membership in early
to find out how many facilities practice radiosurgery, what 1993. Questions were designed to elicit information on

Reprint requeststo: David A. Larson, M.D., Ph.D., Dept. of board and administration of the American Society for Thera-
Radiation Oncology, University of California, San Francisco, peutic Radiology and Oncology in the execution of this survey.
CA 94143. Accepted for publication 13 August 1993.
Acknowledgements-The authors appreciate the support of the
523
524 1. J. Radiation Oncology 0 Biology 0 Physics Volume 28, Number 2, 1994

Table I. Year of first treatment (N = 69) Table 3. Linac energy used (N = 64)

Year Facilities Energy (MeV) Facilities

1983-1984 1 I5 2
1985-1986 2 10 5
1987-1988 6 6 53
1989-1990 21 5 1
1990-present 39 4 3
Total 69 Total 64

technique, hardware, software, personnel, date of first
treatment, patients treated, indications, and effort on day
of procedure as well as in follow-up. We assumed that
several hundred physician members of ASTRO might
claim to perform radiosurgery, and we hoped that a sub-
stantial fraction of those physicians would respond.
Questionnaires were returned in January and February
1993, and analyzed in March 1993. Only a single ques-
tionnaire was distributed, without repeat mailings, and
follow-up with individual respondents to clarify or com-
plete answers was not performed. Results were presented
to the ASTRO executive board in May 1993, at which
time the executive board recommended their publication.
For results in the form of continuous variables, such
as number of hours devoted per treatment day, number
of patients treated per month, and distribution of follow-
up responsibility, unweighted averages were obtained, i.e.,
results from individual facilities were equally weighted,
without regard to number of patients treated. Compari-
sons of means of continuous variables were carried out
by calculating the appropriate t-statistic and using tables
of critical values oft to determine the level of significance.
Results were considered to differ significantly ifp < 0.05.
RESULTS
A total of 98 survey forms were returned, from 90 fa-
cilities (2 forms were returned from each of 8 facilities).
The following countries were represented: Australia (I),
Belgium (l), Republic of China (l), Germany (2), Canada
(3), and USA (82). Radiosurgery was being performed at
74 of the 90 facilities (radiosurgery had not yet been per-
formed at 16 facilities from which responses were ob-
tained). These 74 facilities reported over 6000 radiosurgery
treatments carried out by 135 participating radiation on-
cologists and 130 participating physicists.
Table 4. Linac technique(s) used (N = 63)
Technique Facilities
Arc 61
Static beams 5
Dynamic 3
Total 69*
* At some facilities more than one technique is used.
Summaries of results are presented in tabular form.
The number (N) of facilities for which sufficient infor-
mation was provided is listed in each case. Table 1 shows
the number of facilities starting radiosurgery treatment
in various years, and demonstrates the rate of growth in
facilities.
Table 2 shows that the number of radiation oncologists
per facility who perform radiosurgery ranges from I-5,
with 60% of facilities claiming more than one participating
radiation oncologist. On average there are approximately
two radiation oncologists per facility who perform radio-
surgery. The table also shows that the number of physicists
per facility who perform radiosurgery ranges from 1-4,
with nearly 60% of facilities claiming more than one par-
ticipating physicist. On average there are nearly two phys-
icists per facility who perform radiosurgery.
Linac radiosurgery is used by 64 responding facilities
whereas Gamma Knife radiosurgery is used by 6 (one
facility used both techniques). Linac facilities use five dif-
ferent linac companies, usually with 6 MV X rays (Table
3), and usually with arcs (Table 4).
Commercial frames are used in most cases (Table 5),
provided by five different companies. More than one-
third of facilities use noncommercial planning software
(Table 6).

Table 2. Number of radiation oncologists and physicists per facility

Radiation oncologists per facility 0 1 2 3 4 5 Total

Number of facilities (N = 69) 0 27 25 11 5 1 69
Total number of radiation oncologists 0 27 50 33 20 5 135
Physicists per facility 0 1 2 3 4 5 Total
Number of facilities (N = 68) 0 28 21 16 3 0 68
Total number of physicists 0 28 42 48 12 0 130

Note: Refers only to radiation oncologists and physicists who perform radiosurgery.
 Current radiosurgery practice 0 D. A.LARSONefal. 525

Table 5. Type of steretotactic frame(s) used (N = 53) who perform linac radiosurgery. Less experienced radia-
tion oncologists (I 24 months experience or I 50 patients
Type Facilities
treated) and less experienced physicists (I 24 months ex-
Commercial 51 perience or I 50 patients treated or I 2 patients treated-
Noncommercial 2 per-month) devote significantly more time than their more
Frameless 2 experienced peers, presumably because radiosurgery is
Total 55*
performed more efficiently as one gains experience. Fi-
* At some facilities mor’e than one type of frame is used. nally, and possibly paradoxically, less experienced nurses
(I 50 patients treated) devote significantly less time than
their more experienced peers, possibly because larger,
Table 6. Type(s) of software used (N = 69) more established programs have a better appreciation for
the role of nursing in radiosurgery, or because more es-
tablished programs have more resources for providing
Commercial 45 nursing care.
Noncommercial 21 The average number of patients treated-per-month per
Total 12*
facility since the starting date is presented in Table 8.
* At some facilities more than one type of software is used. Although there are substantially more linac radiosurgery
facilities than gamma knife facilities, the overall number
of patients treated-per-month per facility is significantly
The number of hours devoted to radiosurgery on the greater at gamma knife facilities (9.1 vs. 2.7), as is the
treatment day by various specialists is presented in Table number-per-facility per month among subcategories of
7. On average, radiation oncologists, neurosurgeons, and patients with arteriovenous malformation (3.2 vs. 0.6),
physicists devote 3.8, 3.2, and 6. I hours, respectively, on acoustic neuroma (I .3 vs. 0.2), and meningioma (1.1 vs.
the day of the procedure. Nurses and others also devote 0.3); in other subcategories differences are not significant.
substantial time (2.7 and 3.0 hours, respectively). Neu- Monthly treatment rates did not differ significantly based
rosurgeons and nurses who perform gamma knife radio- on length of program operation except among patients
surgery devote significantly more time than their peers treated for arteriovenous malformation (AVM). However,

Table 7. Number of hours spent per treatment on day of treatment (average)

Radiation
Category N oncologist Neurosurgeon Radiologist Physicist Nurse Other*

Ail 67 3.8” 3.2” 0.7 6.1” 2.1 3.0

Linac 61 3.8 3.od 0.6 6.1 2.2h 2.8
Gamma knife 6 3.3 5.5d 1.1 5.6 7.3h 4.1
Start > 24 mo ago 33 3.4b 3.2 0.1 4.9e 2.9 2.8
Start 5 24 mo ago 34 4.lb 3.2 0.6 7.1” 2.4 3.1
Treat > 2 pts/mo 33 3.7 0.8 5.2’ 3.3 3.4
Treat 5 2 pts/mo 34 3.8 3’:: 0.6 6.9’ 2.0 2.5
> 50 pts treated 27 3.1’ 3.2 0.6 4.99 3.7’ 3.1
5 50 pts treated 40 4.2’ 3.2 0.7 6.8g 1.9’ 2.8

* Tech (RTT, electronics, computer, CT), engineer, specialist, manager, dosimetrist, programmer, coordinator, or not described.
a-’ Numbers with the salme superscript differ significantly (p < 0.05) from one another.

Table 8. Number of patients treated per month per facility (average), since listed starting date, by diagnosis

Acoustic 1” CNS
Category IV AVM neuroma Meningioma malignancy Metastases Other* Total

All 65 0.8 0.3 0.4 0.6 1.0 0.1 3.2

Linac 60 0.6” 0.2d 0.3’ 0.6 0.9 0.1 2.7’
Gamma knife 5 3.2” 1.3d l.lf 1.3 1.8 0.4 9.1’
Start > 24 mo ago 32 l.lb 0.3 0.3 0.5 0.9 0.1 3.3
Start<24moago 33 0.4b 0.2 0.4 0.7 1.1 0.2 3.1
> 50 pts treated 2.6 1.4” 0.4’ 0.5 0.98 1.3h 0.2 4.8’
i 50 pts treated 39 0.4’ 0.1’ 0.3 0.49 0.8h 0.1 2.1’

* Pituitary, functional, benign, nasopharynx ca, or not described.

aq Numbers with the same superscript differ significantly (p < 0.05) from one another.
526 1. J. Radiation Oncology 0 Biology 0 Physics Volume 28, Number 2, I994

Table 9. Patients treated, by diagnosis

Acoustic 1” CNS
Category N AVM neuroma Meningioma malignancy Metastases Othe? Total

All 65 2017 549 605 1028 1871 252 6322

Linac 60 1172 236 357 777 1479 203 4225
Gamma knife 5 845 313 248 251 392 49 2097
Linac (W) 60 28 6 8 18 35 5 100
Gamma knife (%) 5 40 15 12 12 19 2 100

* Pituitary, functional, benign, nasopharynx ca, or not described.

programs which have treated more than 50 patients had oncologists (54% vs. 44%), except at gamma knife facilities
higher rates than less experienced programs both overall (76% vs. 24%).
(4.8 vs. 2.1) and for various subcategories: AVM (1.4 vs.
0.4), acoustic neuroma (0.4 vs. 0.1), primary CNS malig-
DISCUSSION
nancy (0.9 vs. 0.4), and metastases ( 1.3 vs. 0.8).
The total number of patients treated, according to di- The results of the 1993 ASTRO radiosurgery question-
agnosis, is listed in Table 9. These number are further naire provide information regarding current radiosurgery
subcategorized based on treatment technique. Approxi- practice patterns, predominantly in the United States.
mately twice as many patients were treated with linac However, caution must be exercised when interpreting
radiosurgery compared to gamma knife, although this in- the results. First, the questionnaire was purposely designed
volved more than ten times as many facilities. It is inter- to be brief, to enhance the return rate. As a result, the
esting to note that the percentages of patients treated at information obtained, though broad, may be somewhat
linac facilities for metastases or primary CNS malignancy superficial. Second, although responses were received from
are larger than the corresponding percentages at gamma 74 radiosurgery facilities with 135 radiation oncologists,
knife facilities; the opposite is true for arteriovenous mal- in most cases only one radiation oncologist responded,
formation, acoustic neuroma, or meningioma. This may and other team members might well have responded dif-
reflect selection biases of team members (possible stronger ferently. Third, some answers-such as how many hours
oncologic orientation at linac facilities vs. stronger neu- a particular team member devotes on the day of radio-
rosurgical orientation at gamma knife facilities) or patterns surgery-may reflect (imprecise) perceptions, rather than
of referral. hard data. Finally, the true number of facilities to which
Finally, the distribution of follow-up responsibilities is this data might apply (or which might have provided more
presented in Table 10. On average, these responsibilities complete information) may be higher by a factor of 2-3.
are nearly equally shared by neurosurgeons and radiation Since responses were obtained from fewer than half the
known North American gamma knife facilities, it may be
that responses were also returned from fewer than half
the linac radiosurgery facilities. Therefore, the total num-
Table 10. Distribution of follow-up responsibility (average) ber of facilities performing radiosurgery in North America
might be 150-200, and the number of participating ra-
Radiation diation oncologists might be well over 200, as might the
oncologist Neurosurgeon Other*
Category N
number of participating physicists.
@I @) (%)
Nevertheless, two main conclusions seem warranted:
All 67 44 54 (1) a team approach, involving specialists from radiation
Linac 61 46 52 oncology, neurosurgery, physics, radiology, and nursing,
Gamma knife 6 24 76 is widely used, (2) the average number of specialist-hours
Start > 24 mo ago 33 43 54 (M.D. and Ph.D. hours) required per treatment on the
Start I 24 mo ago 34 45 55
day of radiosurgery is high (> 13.)
Treat > 2 pts/mo 33 44 56
Treat < 2 pts/mo 34 44 53
> 50 pts treated 27 40 58
I 50 uts treated 40 46 52 REFERENCE
1. Leksell, L. The stereotaxic method and radiosurgery of the
* Medical oncologist. brain. Acta Chir. Stand. 102:3 16-3 19; 195 1.