CONFORMITY INDEX: A REVIEW
LOÏC FEUVRET, M.D.,* GEORGES NOËL, M.D.,* JEAN-JACQUES MAZERON, M.D., PH.D.,*†
AND PIERRE BEY, M.D.*‡
*Institut Curie, Orsay, France;†Pitié Salpêtrière Hospital, Paris, France;‡Institut Curie, Paris, France
We present a critical analysis of the conformity indices described in the literature and an evaluation of their
field of application. Three-dimensional conformal radiotherapy, with or without intensity modulation, is based
on medical imaging techniques, three-dimensional dosimetry software, compression accessories, and verification
procedures. It consists of delineating target volumes and critical healthy tissues to select the best combination of
beams. This approach allows better adaptation of the isodose to the tumor volume, while limiting irradiation of
healthy tissues. Tools must be developed to evaluate the quality of proposed treatment plans. Dosimetry software
provides the dose distribution in each CT section and dose–volume histograms without really indicating the
degree of conformity. The conformity index is a complementary tool that attributes a score to a treatment plan
or that can compare several treatment plans for the same patient. The future of conformal index in everyday
practice therefore remains unclear.© 2006 Elsevier Inc.
Conformal radiotherapy, Conformity index, Review.
Throughout the history of radiotherapy, with successive
technological progress and various methods of irradiation,
the objective has always been to homogeneously deliver
100% of the prescribed dose to 100% of the target volume
containing the identifiable tumor and/or tumor cells poten-
tially present while limiting the dose to the adjacent healthy
tissues. The human brain has a limited spontaneous three-
dimensional conceptual capacity and therefore requires a
support to allow three-dimensional visualization. Spatial
representation of dose distribution has been progressively
facilitated with progress in medical imaging and dosimetric
software, thereby helping the radiotherapist to achieve these
Chronologically, the first method used to represent dose
distribution was CT for two-dimensional dosimetry. Evalu-
ation of the quality of the treatment plan was based on
visual dosimetric analysis, section by section. However,
detailed comparison between several treatment plans is dif-
ficult and imprecise. Improvements in dosimetry software
have progressively allowed visualization of the spatial ar-
rangement of the tumor, critical organs, and isodoses in the
form of a single three-dimensional representation that can
be observed from all angles. Quantification of this three-
dimensional dose distribution is represented in the form of
dose–volume histograms that can be used to define the
maximum, minimum, mean, and modal dose values deliv-
ered to each volume of interest, as well as the dose delivered
per unit or percentage volume of these structures. This dose
distribution modeling is easy to interpret for the tumor
volume, because it defines the isodose that covers a given
percentage of the tumor volume. This modeling also indi-
cates the doses delivered to the critical organs delineated
(particularly the maximum doses and doses delivered per
unit or percentage of volume of critical organs), and allows
comparison of these doses to theoretical doses considered to
be the maximum tolerated doses. In contrast, a number of
healthy tissues crossed by the beam (brain, gastrointestinal
tract, muscles, etc.) cannot be taken into account, because of
difficulties of delineation and absence of sufficient data
concerning the tolerance of these tissues to the absolute
dose received, or the magnitude of the volume irradiated
(1). For example, not all of the spinal cord is systematically
contoured on dosimetric imaging; dosimetric imaging be-
fore irradiation of the cervical region does not extend as far
as the first lumbar vertebra, corresponding to the termina-
tion of the spinal cord, and only part of the contoured spinal
cord is actually irradiated. This therefore raises the question
of the acceptable mean dose for this portion of the con-
toured spinal cord, which represents only a fraction of the
As a result of these developments, various treatment
plans can now be easily and relatively rapidly obtained for
a same patient. The choice between several options ensuring
the same tumor coverage and the same protection of critical
Reprint requests to: Loïc Feuvret, M.D., Institut Curie-Centre de
protonthérapie d’Orsay (CPO), Bâtiment 101, Campus universita-
ire, 91898 Orsay cedex, France. Tel: (?33) 01-69-29-87-29; Fax:
(?33) 01-69-29-87-19; E-mail: firstname.lastname@example.org
Received June 30, 2005, and in revised form Sept 7, 2005.
Accepted for publication Sept 9, 2005.
Int. J. Radiation Oncology Biol. Phys., Vol. 64, No. 2, pp. 333–342, 2006
Copyright © 2006 Elsevier Inc.
Printed in the USA. All rights reserved
0360-3016/06/$–see front matter
organs is generally in favor of the option that most effec-
tively protects the other healthy tissues, with the simplest
ballistics or the less expensive treatment. However, this
choice nevertheless remains difficult, despite all of the do-
simetric parameters routinely available. An additional tool
is required for integration and analysis of all these data. The
ideal solution would therefore be a system that integrates all
data and presents them in a simple and quantitative form,
i.e., provides a score expressing the relationship between
irradiated tumor tissue and nonirradiated healthy tissues.
Various treatment plans could then be compared on the
basis of the scores attributed to each option. The conformity
index could constitute this ideal tool. Although the confor-
mity index was first proposed in 1993 by the Radiation
Therapy Oncology Group (RTOG) and described in Report
62 of the International Commission on Radiation Units and
Measurements (ICRU), it has not become part of routine
practice (2, 3). With the growth of conformal radiotherapy,
the conformity index could logically be supposed to play an
important role in the future. However, this role has not yet
been defined, probably because the value of conformal
radiotherapy is just beginning to be demonstrated in terms
of prevention of adverse effects and tumor control (4–7).
The clinical value of conformal radiotherapy techniques
must be clearly confirmed before the conformity index can
be used and evaluated as a complementary tool.
In the first part of this article, we will report the clinical
value of conformal radiotherapy to define the context in
which the conformity index could be used. In the second
part, we will describe data from the literature concerning the
history of the conformity index in radiotherapy and its
possible future development (8).
Conformal radiotherapy is designed to achieve the best
adaptation of the shape of a high isodose envelope to the
exact shape of the target volume (9). In 1995, Carrie et al.
reported that conformal radiotherapy could be the next
major revolution in the field of radiotherapy (10). In 2001,
the same authors concluded that, at the dawn of the third
millennium, conformal radiotherapy had already become
the standard radiotherapy modality (11). The practical mo-
dalities of this type of irradiation vary considerably in the
various centers that perform this technique. It is therefore
important to clearly define what we mean by conformal
radiotherapy in terms of the minimum and necessary stan-
dards, such as those consensually proposed by about 30
European institutions in 1997 (DYNARAD: development
and standardization of new DYNamic RADiotherapy tech-
nics) (9, 12). These standards are composed of a series of
mandatory steps: compression, acquisition of anatomic data
by imaging combined with definition of target volumes
(ICRU 50) and critical organs, planimetry, and, finally,
verification and implementation of therapy (13–16). This
progress has been achieved at the price of more complex
patient management with major human and material invest-
ments, for an increasingly well demonstrated clinical ben-
efit. Analysis of the first published results demonstrates a
benefit in terms of quality of life, prevention of adverse
effects, and local control, particularly when dose escalation
Risk factors for healthy tissue complications have now
been clearly defined: irradiated volume, total dose, and dose
per fraction (1, 17). The incidence of these complications
can be decreased by modifying some of these parameters.
Prostate cancer was the first model to be studied in terms of
prevention of late adverse effects by reduction of the irra-
diated volume (7). In 1999, Dearnaley et al. published the
first randomized study comparing the incidence of late
adverse effects after conventional radiotherapy or confor-
mal radiotherapy delivering the same total dose. They de-
scribed a significant reduction of the incidence of proctitis
and rectal bleeding with conformal radiotherapy (6). In a
general review of the place of conformal radiotherapy in the
treatment of non–small-cell lung cancer, Giraud et al. con-
cluded that it has already been clearly demonstrated that
conformal radiotherapy significantly decreases toxicity to
healthy tissues (18). At least for certain tumor sites, con-
formal radiotherapy therefore seems to significantly de-
crease the incidence of late adverse effects and probably
improves the patient’s quality of life by reducing the vol-
ume of healthy tissue irradiated.
Dose escalation delivered by nonconformal radiotherapy
seemed to improve the local control rate of certain tumors,
but at the price of increased toxicity (4, 5). The concept of
the improved tissue tolerance obtained by conformal radio-
therapy and the concept of the dose-effect for certain tumors
were therefore combined in dose escalation protocols to
improve local control and survival rates, while maintaining
the same toxicity levels. Encouraging results have been
reported for non–small-cell lung cancer and prostate cancer
Conformal radiotherapy therefore constitutes a useful
approach that will achieve its full potential only when
radiotherapists have more clearly defined their treatment
options. In other words, for a given treatment plan, what
risks of tumor underdosage are radiotherapists willing to
accept to avoid exceeding a certain level of toxicity, or
what risks of toxicity are they willing to accept to ensure
optimal treatment of the tumor? The rapidly growing
applications of conformal radiotherapy can provide the
physician with a panel of various irradiation technical
options for the same patient. This enormous volume of
data, along with increasingly sophisticated systems, has
not simplified treatment decisions: The radiotherapist is
overwhelmed by clinical, radiologic, geometric, dosimet-
ric, and radiobiologic parameters. Analysis of each of
these parameters is essential but complex and time-con-
suming, and it has become essential to develop a rapid
tool to integrate all of these data. Ideally, after analysis of
each parameter, this tool should provide unique quanti-
tative information and should be able to quantitatively
assess the quality of a treatment option. At the present
334 I. J. Radiation Oncology ● Biology ● Physics Volume 64, Number 2, 2006
time, apart from clinical and radiologic parameters, only
geometric and dosimetric data can be incorporated into
such a tool, because the correlation between clinical and
radiobiologic data is still insufficient and cannot be used
in routine clinical practice. The conformity index could
correspond to this synthetic tool, because it simply quan-
tifies the degree of congruence between isodoses, tumor
contours, and healthy tissue contours by geometric inter-
DEVELOPMENT OF THE CONCEPT OF
The conformity index was developed as an extension of
section-by-section dosimetric analysis and dose–volume histo-
grams and can be defined as an absolute value resulting from
the relationship between tumor volume or a fraction of this
volume and the volume delineated by an isodose or a fraction
of this volume. It can also be defined by the ratio of an isodose
with another isodose (prescription isodose, reference isodose,
minimum isodose, maximum isodose). As we have seen
above, the use of this type of tool could facilitate the choice of
treatment and comparisons of various treatment plans for con-
formal radiotherapy, stereotactic radiotherapy, and brachyther-
apy, as recommended by the American Brachytherapy Society
(22). However, the multiple indices proposed, as well as the
difficulty in interpreting them, raise a number of problems. We
will therefore describe each of the various indices reported in
the literature (Table 1).
Conformity index and stereotactic radiotherapy
Radiation Therapy Oncology Group criteria. In 1993, the
RTOG proposed routine evaluation of stereotactic radio-
therapy treatment plans based on several parameters: refer-
Table 1. Comparison of the various volume-based conformity indices in various clinical settings
RTOG (1, 24)
Quality of coverageRTOG ?
I min: Minimal isodose surrounding
RI: Reference isodose
I max: Maximum isodose in the
RI: Reference isodose
RI: Volume of the reference isodose
TV: Target volume
Homogeneity indexRTOG ?
SALT (30, 31)
Lesion coverage factor ?CVF? ?
LVRI: Lesion volumes covered by the
LV: Lesion volumes
TVRI: Target volume covered by the
reference isodose (Lomax and
TV: Target volume (Lomax and
LV?RI: Lesion volume receiving
isodoses ? reference isodose
LV: Lesion volume
HTVRI: Healthy tissue volume
covered by the reference isodose
LV: Lesion volume
LUF: Lesion underdosage factor
HTOF: Healthy tissue overdosage
TVRI: Target volume covered by the
VRI: Volume of the reference isodose
TVRI: Target volume covered by the
TV: Target volume
VRI: Volume of the reference isodose
CN: Conformation number proposed
by van’t Riet
NCO: Number of critical organs (CO)
VCOref,i: Critical organ volume
receiving at least the reference
VCO,i: Critical organ volume
Lesion underdosage factor ?LUF? ?
Healthy tissue overdosage factor ?HTOF? ?
Geometric conformity index: g ? LUF ? HTOF
Lomax and Scheib (32)
Healthy tissue conformity index ?
van’t Riet et al. (33)
Conformation number ?CN? ?
Baltas et al. (34)
COnformal INdex (COIN)?CN??
Abbreviations: RTOG ? Radiation Therapy Oncology Group; SALT ? Saint-Anne, Lariboisière, Tenon.
335 Conformity index ● L. FEUVRET et al.
ence isodose values of the treatment plan, reference isodose
volume (VRI) (defined by the RTOG as the prescription
isodose), and the target volume (TV) (3, 23, 24).
Quality of coverageRTOG?Imin
where I min ? minimum isodose around the target, and RI
? reference isodose.
where I max ? maximum isodose in the target, and RI ?
where VRI? reference isodose volume, and TV ? target
The values of each of these parameters have been defined
to determine the quality of irradiation.
For “quality of coverage” (Eq. 1a), if the 90% isodose
covers all of the clinical and pathologic target volume,
treatment is considered to comply with the protocol. If the
80% isodose covers all of the clinical and pathologic target
volume, the protocol violation is considered to be minor. On
the other hand, if the 80% isodose does not cover all of the
clinical and pathologic target volume, the protocol violation
is considered to be major.
If the “homogeneity index” (Eq. 1b) is ?2, treatment is
considered to comply with the protocol. If this index is
between 2 and 2.5, the protocol violation is considered to be
minor, but when the index exceeds 2.5, the protocol viola-
tion is considered to be major, but may nevertheless be
considered to be acceptable.
The RTOG “conformity index” (Eq.1c) is easy to inter-
pret. A conformity index equal to 1 corresponds to ideal
conformation. A conformity index greater than 1 indicates
that the irradiated volume is greater than the target volume
and includes healthy tissues. If the conformity index is less
than 1, the target volume is only partially irradiated. Ac-
cording to RTOG guidelines, ranges of conformity index
values have been defined to determine the quality of con-
formation, because a value of 1 is rarely obtained. If the
conformity index is situated between 1 and 2, treatment is
considered to comply with the treatment plan; an index
between 2 and 2.5, or 0.9 and 1, is considered to be a minor
violation, and an index less than 0.9 or more than 2.5 is
considered to be a major violation. However, this index
presents a major drawback: It can never take into account
the degree of spatial intersection of two volumes or their
shapes. In extreme cases, it may be equal to 1 while these
two volumes are situated away from each other and present
entirely different shapes (Fig. 1). The conformity index
alone therefore cannot provide any practical information.
With this index, compliance with the treatment plan can
only be assessed by visualization of CT sections and dose–
volume histograms. However, the extreme case of noncon-
cordance of the target and the isodoses seems unlikely,
because this index was initially proposed for radiotherapy
under stereotactic conditions, a modality ensuring the most
rigorous and most precise treatment planning.
These indices raise two particular comments. First of all,
cutoff values were defined to determine whether the treat-
ment complied with the protocol, but values indicating that
a treatment plan is unacceptable are not defined. What
criteria are used to determine the choice of dosimetry with
suboptimal parameters? It is difficult to answer this question
in view of the limited information concerning a possible
correlation between clinical data and these theoretical pa-
rameters (25). The second comment, in relation to external
beam radiotherapy, was raised by Knoos et al. and con-
cerned the limitations of the RTOG conformity index (26),
because the definition of VRImay differ from one center to
another. This parameter can correspond to either the mini-
mum isodose volume containing the target volume, as is
often the case in radiosurgery, or the 95% isodose volume
according to ICRU 50 guidelines. Consequently, volumes
and conformity indices may vary according to the isodoses
selected. For example, during stereotactic radiotherapy of
brain metastases, the use of conformal beams can improve
the conformity index but can also lower the conformity by
decreasing the reference isodose level, and therefore in-
creasing VRI. Consequently, and to standardize techniques
and description of the results of external classical irradia-
Fig. 1. Four possibilities for which the VRI/TV ratio is equal to 1
(index proposed by the RTOG) (1) (target volume, shaded; volume
of reference isodose, enclosed in black dashes).
336 I. J. Radiation Oncology ● Biology ● PhysicsVolume 64, Number 2, 2006
tion, it seems more appropriate to systematically use ICRU
50 guidelines (95% isodose volume), because they corre-
spond to the parameters used for treatment planning.
Saint-Anne, Lariboisière, Tenon (SALT) criteria. For ste-
reotactic radiotherapy of arteriovenous malformations, the
SALT group has quantified the global quality of treatment
planning by the standard deviation of the differential dose–
volume histogram calculated for each vascular lesion (Eq.
2a) (27-31). The standard deviation represents the scatter of
dose data around a mean dose and constitutes a way of
representing the homogeneity of the dose delivered to the
lesion. Furthermore, to facilitate the choice of a prescription
isodose, this team uses a quantitative protocol of automatic
definition of this isodose derived from indices based on
coverage of the lesion (Eqs. 2b-e). The parameters are as
where D(j) ? relative dose in the lesion voxel j, Dmean ?
relative mean dose in the lesion, and LV ? lesion volume in
The lesion coverage volume factor (CVF):
where LVRI? lesion volume covered by the reference
isodose, and LV ? lesion volume.
The lesion underdosage volume factor (LUF):
where LV?RI? lesion volume receiving an isodose ?
reference isodose, and LV ? lesion volume.
The healthy tissue overdose volume factor (HTOF):
where HTVRI? healthy tissue volume covered by reference
isodoses, and LV ? lesion volume.
Geometric conformity index g:
The standard deviation is dependent on the shape of the
differential dose–volume histogram: The narrower the dif-
ferential dose–volume histogram, the smaller the standard
deviation and the better the quality of treatment. The last
factors (Eqs. 2b–e), dependent on the reference isodose,
present the advantages of being relatively easy to under-
stand and are independent of the anatomy of the lesion.
Comparison between SALT and RTOG criteria. In 2001,
the SALT team published a comparison of two quantitative
protocols of automatic definition of the prescription isodose
from RTOG indices and the SALT group indices described
above (Eqs. 1a–c and 2a–e) for irradiated arteriovenous
malformations (The prescription isodose was selected em-
pirically based on superimposition of the isodoses and le-
sion contours) (28, 31). The following constraints were
imposed for each of the two protocols: For the SALT
protocol, the lesion volume covered was equal to 90% of the
total lesion volume, and for the RTOG protocol, the ratio of
the minimum isodose to the lesion over the prescription
isodose was equal to 0.9. Note that the concept of coverage
of the lesions was based on volumes for the SALT group,
but was based on isodoses for the RTOG. This difference
has a number of consequences. The RTOG protocol gener-
ates geometric conformity indices situated out of the range
of protocol values, whereas the values of the SALT protocol
mostly remain satisfactory (i.e., with a standard deviation
?12%). When the criteria of the RTOG protocol are used,
there is a tendency toward overdosage of healthy tissues but
no underdosage of lesions, in contrast with the SALT pro-
tocol, which accepts a 10% underdosage of the lesions and
therefore minimizes overdosage of healthy tissues. With
this study, the authors demonstrated the limits of a model
based exclusively on isodoses, which does not explicitly
comprise specific evaluation of irradiation of healthy tissues
(Eqs. 1a–c). To improve treatment evaluation, the authors
proposed the use of these two complementary protocols to
take into account all volume parameters of the lesion and
healthy tissues (28, 31). They therefore proposed a modifi-
cation of the RTOG conformity index (Eq. 1c) according to
the following formula:
that is, Eq. 1c ? Eq. 2b ? Eq. 2d. This revised formula
allows quantification of healthy tissue irradiation (by
HTOF) based on the various criteria of the RTOG protocol.
Conformity index and target volume
In view of the advantages of the SALT group indices and
the disadvantages of the RTOG indices, Lomax and Scheib
decided to apply the SALT group CVF, developed for
arteriovenous malformations, to brain tumors (Eq. 3) ? (Eq.
2b). This index represents the percentage of target volume
receiving at least the prescribed dose (32). This concept
therefore constitutes one of the recurring questions faced by
radiotherapists: “What proportion of the target volume re-
ceives the prescribed dose?” The target volume, also called
LV (lesion volume) by the SALT group, and the target
volume covered by the reference isodose (TVRI) can be
calculated automatically by computer (Eq. 3).
337 Conformity index ● L. FEUVRET et al.
where TVRI? target volume covered by the reference
isodose, and TV ? target volume.
This index ranges from 0 (All of the target volume is
situated outside of the prescription isodose) to 1 (All of the
target volume is irradiated at the prescribed dose). The
quality of irradiation of the target volume can be correctly
determined with this index, but it does not provide sufficient
information about the overall treatment plan. Let us take the
typical example of whole-brain irradiation. The index is
equal to 1, because the target volume is covered by the
reference isodose. This example illustrates the main disad-
vantage of this index: The volume of adjacent healthy
tissues is not taken into account. Overall, this index can help
the radiotherapist to achieve only one of his objectives:
treatment of the target.
Conformity index and healthy tissues
Lomax and Scheib proposed an index taking into account
exclusively the irradiation of healthy tissues (Eq. 4) (32).
This index is as follows:
Healthy tissues conformity index?TVRI
where TVRI? target volume covered by the reference
isodose, and VRI? volume of the reference isodose.
It measures the proportion of the volume of the reference
isodose comprising the target volume, i.e., indirectly the
volume of healthy tissue included in the reference isodose.
It ranges from 0 (no spatial concordance between the two
volumes and no protection of healthy tissues in the isodose
concerned) to 1 (perfect conformation). According to Lo-
max and Scheib (32), irradiation is considered to be con-
formal if and only if this index is ?0.6. This index is fairly
difficult to interpret. First of all, the real significance of an
index equal to 1 may be very different from that of perfect
conformation, because the reference isodose can be totally
included in the target volume, but part of the target volume
may not be irradiated at the prescribed dose. A value of 0.5
can correspond to one of the following two possibilities:
Either the tumor is correctly treated with a surrounding
volume of irradiated healthy tissues equal to that of the
target, or the target volume and the volume of the reference
isodose have the same shape and the same value, but only
half overlap (Table 2). Overall, this index could help to
achieve the radiotherapist’s other objective: protection of
Global conformity index (target volumes
and healthy tissues)
The previous two indices (Eqs. 3 and 4) provide indisso-
ciable, complementary information (irradiation of the target
volume and irradiation of healthy tissues). To compensate
for the defects of these two indices, van’t Riet et al. pro-
posed an index called conformation number (CN) (33).
Calculation of this CN simultaneously takes into account
irradiation of the target volume and irradiation of healthy
tissues (Eq. 5). This number is defined as follows:
where CN ? conformation number, TVRI? target volume
covered by the reference isodose, TV ? target volume, and
VRI? volume of the reference isodose.
The first fraction of this equation defines the quality of
coverage of the target; the second fraction defines the vol-
ume of healthy tissue receiving a dose greater than or equal
to the prescribed reference dose. The CN ranges from 0 to
1, where 1 is the ideal value. A value close to 0 indicates
either total absence of conformation, i.e., the target volume
is not irradiated (Fig. 2b) or a very large volume of irradi-
ation compared to the target volume. The last situation
corresponds to the typical case of whole-brain irradiation
Table 2. Comparison of the various volume-based conformity
indices in various clinical settings
Abbreviations: TV ? Target Volume (gray); VRI? Volume of
the Reference Isodose (dotted line); TVRI? Target volume cov-
ered by the Reference Isodose ? intersection of TV and VRI.
338 I. J. Radiation Oncology ● Biology ● PhysicsVolume 64, Number 2, 2006
for a single, small metastasis, where the index tends toward
0, because the VRIis much greater than the TVRI(Fig. 2c).
Although this index takes into account irradiation of the
target volume and healthy tissues, it is not perfect and can
be difficult to interpret, because a CN of 0.6 can correspond
to two different situations (lines 2 and 3, Table 2): The
target volume is partly irradiated with complete protection
of healthy tissues, or both the target volume and the healthy
tissues are only partially irradiated. van’t Riet et al. also
emphasized the importance of clearly specifying the target
volume, defined according to ICRU 50, to which the con-
formity index corresponds (33). When comparing different
treatment plans or different irradiation techniques, it is
important to use the same terms, while bearing in mind that
analysis of an index is usually a relative comparison for the
same patient. For example, in this study for which the ratio
between the planning target volume and the clinical target
volume was 1.9, for a same treatment plan, the conformity
index was equal to 0.65 for the planning target volume and
decreased to 0.37 for the clinical target volume. To illustrate
this variation, refer to example 1 of Table 2, noting that the
planning target volume is equal to 5 cm3. By applying the
above ratio (1.9), the clinical target volume is 2.63 cm3, and
the clinical target volume situated in the reference isodose is
therefore 2.63 cm3. Calculation of the CN shows a reduction
from 0.5 to 0.26. Depending on the choice of tumor volume
(CTV, planning target volume, etc.) and the margins used,
the results can therefore vary considerably, leading to erro-
neous conclusions. This comment can be applied also to the
reference isodose selected. All of the formulae described
above refer to the reference isodose. In practice, this isodose
usually corresponds to the ICRU isodose 95%. An analysis
of conformity indices for isodoses greater than the 95%
isodose would provide information about the degree of
homogeneity of irradiation of the target. For isodoses less
than the 95% isodose, the indices could be used to evaluate
the volume of healthy tissue irradiated. However, between
two treatment plans ensuring the same tumor coverage,
what role is played by the volume of healthy tissue irradi-
ated or the doses received by these tissues in the main
criteria of choice which are complexity of treatment, cost,
etc.? No consensus has been reached concerning the concept
of irradiated healthy tissues; i.e., what isodose or absolute
dose to be determined must be defined to predict the risk of
toxicity in an irradiated volume?
A summary table (Table 2) compares the various indices
described above in relation to various treatment plans. The
first two indices could lead to false-positive results; i.e., they
may give a value equal to 1 according to the treatment plan
and be falsely reassuring in terms of the quality of irradia-
tion. A conformity index must reach a value of at least 1 if
and only if conformation is perfect. Only the index proposed
by van’t Riet et al. satisfies this criterion, although the two
parameters of Eq. 5 must be analyzed separately (33).
Conformity index taking into account critical organs
Baltas et al. applied the CN to brachytherapy by adding
a supplementary parameter, the concept of critical organs
(CO) (34). This index is called COIN (COnformal INdex)
(Eq. 6) (Fig. 3) and is defined as follows:
VCO, i??, (6)
where NCO? number of critical organs, VCOref,i? critical
organ volume receiving at least the reference dose, VCO,i?
critical organ volume, and CN ? van’t Riet’s conformation
The COIN takes into account the quality of tumor irra-
diation, irradiation of noncritical healthy tissues, and irra-
diation of critical organs. The first two parameters corre-
spond to the CN described above (Eq. 5). It is multiplied by
other indices correlated with the various critical organs.
Each of these indices
tends toward 1 with improvement of the degree of protec-
tion of critical organs compared to the reference dose.
Overall, each component of the COIN ideally tends toward
1, and this index therefore ranges from 0 to 1. Initially
Fig. 3. Diagram of the various volumes required to calculate the
COnformity INdex (COIN) (34)
Fig. 2. Illustration of extreme cases for the conformation number
(CN) proposed by van’t Riet et al. (33): (a) ideal conformation, (b)
no concordance, (c) extreme case of whole-brain irradiation.
339Conformity index ● L. FEUVRET et al.
proposed for brachytherapy, the COIN can also be applied
to external beam radiotherapy, because, by analogy, the
COIN seems to be a relevant index for certain high-preci-
sion irradiation techniques (proton therapy, ion therapy,
intensity modulation, radiosurgery) associated with a very
high dose gradient. During treatment planning, it is impor-
tant to ensure optimal coverage of the target, to reduce doses
to critical organs to a minimum and to optimally spare
adjacent healthy tissues. Computers can be used to deter-
mine the volume of each critical organ (VCO) and the
proportion of its volume covered by the reference isodose
VCOref. The final index is based on a global score attributed
to the proposed treatment plan by taking into account irra-
diation of the tumor, noncritical healthy tissues, and critical
organs. Note that the parameters of Eq. 5 can be determined
by dose–volume histograms. However, two comments must
be taken into account. First of all, when comparing various
treatment plans, it is difficult to guide one’s choice accord-
ing to the degree of protection of critical organs estimated
one by one, because the index provides only global infor-
mation. However, it is possible to independently analyze the
index attributed to each critical organ and give preference to
a solitary vital organ, rather than a paired organ. This
approach is already applied routinely in a more rudimentary
fashion with the maximum dose received and the dose
expressed per volume. Second, the acceptable maximum
doses for critical structures are known, but differ from one
organ to another, and these doses rarely reach the dose level
required to sterilize the tumor. As shown in the formula, the
entire calculation is based on the volume of the reference
isodose, usually the 95% isodose in external beam radio-
therapy. This reference isodose is related to the isodose
surrounding the target. It is not envisaged that a different
acceptable isodose is selected for each critical organ. The
volume of the critical organ contained in the reference
isodose (95%) is probably very low when the prescribed
dose at the ICRU point is much higher than the maximum
tolerated dose. Consequently, the parameter (Eq. 7) is ap-
proximately equal to 1 and does not affect the COIN. This
parameter does not seem to be very useful apart from cases
in which the dose to the target volume is approximately
equal to the acceptable doses to critical organs. However,
the principle of combining several indices (ideally tending
toward 1) can, by extension, be applied to external beam
radiotherapy by calculating an index for each critical organ
based on the maximum tolerated dose for each organ. Thus,
in the presence of tumor irradiation in a zone close to many
critical organs, the COIN would allow more detailed com-
parison of various treatment plans with a global coefficient
and a coefficient attributed to each critical organ according
to its specific tolerance.
Overall, improvements of the various conformity indices
presented above have led to the COIN index, which strives
to be ideal by taking into account all irradiated tissues:
tumor, healthy tissues, and critical and noncritical organs.
These indices are based exclusively on “global intersection
in a Euclidean space” geometric models. However, it is
important to be able to apply these very theoretical values to
the clinical context (35, 36). Lomax and Scheib quantified
the degree of conformation obtained by radiosurgery for
551 benign (arteriovenous malformations, acoustic neuro-
mas, meningiomas, pituitary adenomas) and malignant tar-
gets (metastases) and found that it is not always justified to
try to achieve the ideal value of the conformity index at any
price (32). Dosimetric priorities are not the same for benign
and malignant lesions. For benign lesions, priority is given
to the conformation (Eq. 4), sparing critical organs at the
price of decreased tumor coverage (Eq. 3). For metastases,
tumor coverage takes precedence over conformation. For
example, for pituitary adenomas, the median healthy tissues
conformity index (TVRI/VRI) was equal to 0.78 (Eq. 4) for a
median target volume conformity index (TVRI/TV) of 0.91
(Eq. 3). In contrast, for metastases, the median healthy
tissues conformity index was 0.58 for a median target vol-
ume conformity index of 1. Lomax and Scheib (32) there-
fore preferred to analyze the healthy tissues conformity
index and the target volume conformity index indepen-
dently, without combining them, as in the case of the CN
(Eq. 5). In any case, interpretation of conformation data
alone is not sufficient, and certain technical parameters
related to the machine itself must also be taken into account.
It is difficult to adjust a spherical dose distribution (Leksell
Gamma-Knife) to the irregular contours of small targets less
than 0.25 cm3. In extreme cases, the target volume could be
smaller than that produced by the smallest collimators. For
the 74 acoustic neuromas analyzed, the healthy tissues con-
formity index (TVRI/VRI) was stable and independent of the
size of the lesion for lesions measuring more than 1 cm3. In
contrast, for lesions measuring less than 1 cm3, the healthy
tissues conformity index was lower as the volume de-
creased. For these small tumors, a lower conformation is
generally less critical, because the volume of healthy tissue
receiving the total dose is inevitably small. However, this
lower conformation is not necessarily less critical in terms
of toxicity simply on the basis of the small tumor volume.
Nakamura et al. evaluated the risk factors of toxicity, in-
cluding the conformity index, after gamma knife radiosur-
gery of 1,612 arteriovenous malformations (25). On multi-
factorial analysis, only the maximum diameter of the largest
lesions was identified as an independent risk factor for
complications, although the conformity index paradoxically
improved with increasing target volume, as described by
Lomax and Scheib (32). These various findings clearly
indicate that the ideal index does not exist. Complementary
clinical and technical information remains essential to in-
terpreting any conformity index.
In the future
The choice of the best dose distribution and the best
irradiation modalities could therefore be based on the use of
the conformity index, but also dose–volume histograms and
analysis of CT sections. It must be remembered that the
conformity index is calculated without any distinction be-
tween the various types of target volumes and the various
340 I. J. Radiation Oncology ● Biology ● Physics Volume 64, Number 2, 2006
types of healthy tissues, although many data remain uncer-
tain or unknown, because radiobiologic parameters vary
from one tumor to another and within the same tumor. The
sensitivity of tumor cells to irradiation depends on cell
characteristics (cell repair, cell cycle), the environment (ox-
ygen, pH, temperature), and the type of irradiation used
(relative biologic efficacy) (37, 38). This problem remains
unresolved for healthy tissues and critical organs, because
clinical data on the relationship between delivered doses
and the short-term and long-term morbidity of noncritical
and critical healthy organs are rare as a result of the absence
of systematic prospective reporting of these data in the past
and the absence of systematic dosimetric data on these
anatomic structures. Late adverse effects have been taken
into account only more recently, and follow-up is therefore
still limited (39–42). Probabilistic mathematical formulae
have been developed to predict tumor control probability
and normal tissue complication probability. These formulae
take the volume of the organ considered into account, its
theoretical capacity for repair of radiation-induced lesions,
and the dose delivered (43). These are only mathematical
formulae established from retrospective observations allow-
ing theoretical comparisons, but they do not provide real
probability figures. Rosenzweig et al. presented the prelim-
inary results of a phase I dose escalation study for locally
advanced non–small-cell lung cancers (44). Because of
Grade 3 pulmonary toxicity and Grade 5 toxicity for 3 of the
14 patients irradiated at the first dose level (70.2 Gy), the
protocol was modified, and only patients with a normal
tissue complication probability less than 25% were subse-
quently eligible for dose escalation. Eleven patients re-
ceived a dose of 70.2 Gy, and another 10 patients received
a dose of 75.6 Gy. Three patients of the first dose level and
only one patient of the second dose level developed Grade
3 pneumonia. This study illustrates the potential value of
this type of radiobiologic model, but it is not yet used in
routine clinical practice.
Integration of these probabilistic data into clinical re-
search could become necessary and even mandatory in the
same way as ICRU guidelines, because they can complete
geometric data provided by calculation of the conformity
index and therefore allow selection of a treatment plan on
the basis of dosimetric, geometric, and functional parame-
ters (for example, in lung cancer they could be used to select
the treatment plan with the best probability of sparing
respiratory function based on pulmonary function tests or
ventilation-perfusion scintigraphy) and biologic parameters
obtained by new, rapidly developing functional imaging
techniques (positron emission tomography [PET]) (45). For
example, PET can be used to visualize the hypoxic tumor
fraction, a factor of radioresistance, and obtain so-called
anatomometabolic images resulting from fusion of PET data
and morphologic imaging data (MRI or CT) (46).
The conformity index constitutes an attractive tool, be-
cause it could facilitate decisions during analysis of various
treatment plans proposed for conformal radiotherapy. Its
advantages are its simplicity and the integration of multiple
parameters. It is not widely used in routine clinical practice
at the present time, except for stereotactic radiotherapy, for
which the conformity index was created. However, apart
from its contribution to everyday practice, it would also
facilitate comparison between various available techniques
and could be used to evaluate new technologies.
An ideal tool does not exist at the present time. The
conformity index, or rather conformity indices, are too
diverse to achieve the desired objective, i.e., to quantify the
quality of a treatment with 100% sensitivity and specificity.
Attempts to reduce conformation to a single index could
lead to omission of essential information. Furthermore, be-
cause these indices are based on purely geometric parame-
ters, many qualitative aspects of irradiation are overlooked
(tissue characterization, effect of fractionation, etc.) related
to the living neoplastic and healthy tissues irradiated, com-
prising far too many uncertainties. Furthermore, other rea-
sons argue to the usefulness of no complex and unencom-
passing CI: the inaccuracy of dose prescription from current
treatment planning systems, the inconsistency of organ and
target contouring among practitioners, and the radiation
damage to “parallel” vs. “serial” organs. All these increase
the difficulties of applying a simple index or indices to a
general treatment site.
The future of conformity indices in everyday practice
therefore remains unclear.
1. Huchet A, Caudry M, Belkacemi Y, et al. [Volume-effect and
radiotherapy part two: Volume-effect and normal tissue]. Can-
cer Radiother 2003;7:353–362.
2. Anonymous. Prescribing, recording and reporting photon
beam therapy (supplement to ICRU Report 50). Report 62,
International Commission on Radiation Units and Measure-
ments, Washington, DC 1999.
3. Shaw E, Kline R, Gillin M, et al. Radiation Therapy Oncology
Group: Radiosurgery quality assurance guidelines. Int J Ra-
diat Oncol Biol Phys 1993;27:1231–1239.
4. Armstrong J, Raben A, Zelefsky M, et al. Promising survival
with three-dimensional conformal radiation therapy for non-
small cell lung cancer. Radiother Oncol 1997;44:17–22.
5. Armstrong J, McGibney C. The impact of three-dimensional
radiation on the treatment of non-small cell lung cancer.
Radiother Oncol 2000;56:157–167.
6. Dearnaley DP, Khoo VS, Norman AR, et al. Comparison of
radiation side-effects of conformal and conventional radio-
therapy in prostate cancer: A randomised trial. Lancet 1999;
7. Lee WR, Hanks GE, Hanlon AL, et al. Lateral rectal shielding
reduces late rectal morbidity following high dose three-dimen-
sional conformal radiation therapy for clinically localized
prostate cancer: Further evidence for a significant dose effect.
Int J Radiat Oncol Biol Phys 1996;35:251–257.
8. Paddick I. A simple scoring ratio to index the conformity of
341 Conformity index ● L. FEUVRET et al.
radiosurgical treatment plans. Technical note. J Neurosurg Download full-text
9. Rosenwald JC, Gaboriaud G, Pontvert D. [Conformal radio-
therapy: Principles and classification]. Cancer Radiother
10. Carrie C, Ginestet C, Bey P, et al. [Conformal radiation
therapy. Federation nationale des centres de lutte contre le
cancer (FNCLCC)]. Bull Cancer 1995;82:325–330.
11. Carrie C, Ginestet C. [New techniques and methods in trans-
cutaneous radiotherapy]. Bull Cancer 2001;88:59–65.
12. Kolitsi Z, Dahl O, Van Loon R, et al. Quality assurance in
conformal radiotherapy: DYNARAD consensus report on
practice guidelines. Radiother Oncol 1997;45:217–223.
13. Bey P, Gerard JP. [Practical determination of volume and
doses in conformal radiotherapy]. Cancer Radiother 1998;2:
14. Chavaudra J. [Last ICRU recommendations for the prescrip-
tion, recording and reporting of external bean therapy]. Can-
cer Radiother 1998;2:607–614.
15. Garcia R, Oozeer R, Le Thanh H, et al. [Are there physical,
technical and morphological limits to conformational radio-
therapy?]. Cancer Radiother 2001;5:53s–56s.
16. Pommier P, Ginestet C, Carrie C. [Is conformational radio-
therapy progressing? What are its difficulties and limits?].
Cancer Radiother 2001;5:57s–67s.
17. Dorr W, Hendry JH. Consequential late effects in normal
tissues. Radiother Oncol 2001;61:223–231.
18. Giraud P, Helfre S, Lavole A, et al. [Non-small-cell bronchial
cancers: Improvement of survival probability by conformal
radiotherapy]. Cancer Radiother 2002;6:125s–134s.
19. Corn BW, Hanks GE, Schultheiss TE, et al. Conformal treat-
ment of prostate cancer with improved targeting: Superior
prostate-specific antigen response compared to standard treat-
ment. Int J Radiat Oncol Biol Phys 1995;32:325–330.
20. Giraud P, Jaulerry C, Brunin F, et al. [Upper aerodigestive
tract cancers: Clinical benefits of conformal radiotherapy and
intensity modulation]. Cancer Radiother 2002;6:37s–48s.
21. Hayman JA, Martel MK, Ten Haken RK, et al. Dose escala-
tion in non-small-cell lung cancer using three-dimensional
conformal radiation therapy: Update of a phase I trial. J Clin
22. Nag S, Bice W, DeWyngaert K, et al. The American Brachy-
therapy Society recommendations for permanent prostate
brachytherapy postimplant dosimetric analysis. Int J Radiat
Oncol Biol Phys 2000;46:221–230.
23. Nedzi LA, Kooy HM, Alexander E III, et al. Dynamic field
shaping for stereotactic radiosurgery: A modeling study. Int J
Radiat Oncol Biol Phys 1993;25:859–869.
24. Shaw E, Scott C, Souhami L, et al. Single dose radiosurgical
treatment of recurrent previously irradiated primary brain tu-
mors and brain metastases: Final report of RTOG protocol
90–05. Int J Radiat Oncol Biol Phys 2000;47:291–298.
25. Nakamura JL, Verhey LJ, Smith V, et al. Dose conformity of
gamma knife radiosurgery and risk factors for complications.
Int J Radiat Oncol Biol Phys 2001;51:1313–1319.
26. Knoos T, Kristensen I, Nilsson P. Volumetric and dosimetric
evaluation of radiation treatment plans: Radiation conformity
index. Int J Radiat Oncol Biol Phys 1998;42:1169–1176.
27. Dejean C, Lefkopoulos D, Nataf F, et al. [Radiosurgery of
cerebral arteriovenous malformations: Methodology of the
dose prescription and its implications for dosimetric results.
Retrospective study of 408 cases]. Neurochirurgie 2001;47:
28. Dejean C, Lefkopoulos D, Foulquier JN, et al. [Automatic defi-
nition of prescription isodose for stereotaxic radiation of arterio-
venous malformations]. Cancer Radiother 2001;5:138–149.
29. Grandjean P, Platoni K, Lefkopoulos D, et al. Use of a general
inverse technique for the conformational stereotactic treatment
of complex intracranial lesions. Int J Radiat Oncol Biol Phys
30. Lefkopoulos D, Schlienger M, Touboul E, et al. Quantitative
evaluation of treatment planning for linac multi-isocentric
radiosurgery. Proc. of the XIth Intern. Conf. on: “Computers
in Radiation Therapy,” XIth ICCR, Manchester, UK. 1994;
31. Lefkopoulos D, Dejean C, El-Balma H, et al. Determination of
dose-volumes parameters to characterise the conformity of
stereotactic treatment plans. Proc. of the XIIIth Intern. Conf.
on: “Computers in Radiation Therapy,” XIIIth ICCR, Heidel-
berg, Germany. 2000;356–358.
32. Lomax NJ, Scheib SG. Quantifying the degree of conformity
in radiosurgery treatment planning. Int J Radiat Oncol Biol
33. van’t Riet A, Mak AC, Moerland MA, et al. A conformation
number to quantify the degree of conformality in brachyther-
apy and external beam irradiation: Application to the prostate.
Int J Radiat Oncol Biol Phys 1997;37:731–736.
34. Baltas D, Kolotas C, Geramani K, et al. A conformal index
(COIN) to evaluate implant quality and dose specification in
35. Dibiase SJ, Kwok Y, Yovino S, et al. Factors predicting local
tumor control after gamma knife stereotactic radiosurgery for
benign intracranial meningiomas. Int J Radiat Oncol Biol
36. Oozeer R, Chauvet B, Garcia R, et al. [Dosimetric evaluation
of conformal radiotherapy: Conformity factor]. Cancer Ra-
37. Huchet A, Caudry M, Belkacemi Y, et al. [Volume-effect in
radiation therapy part one: Volume-effect and tumour]. Can-
cer Radiother 2003;7:79–89.
38. Kulik C, Mazurier J, Lartigau E. [Probabilities of controlling
tumors and complications (TCP/NTCP) after radiotherapy:
Methodologic, physical, and biological aspects]. Cancer Ra-
39. Lartigau E, Dubray B, Mornex F. [Biological mechanisms of
late effects of ionizing radiations]. Cancer Radiother 1997;1:
40. Mornex F, Beauvois S, Van Houtte P. [Late effects of ionizing
radiations on central nervous system, spinal cord, and periph-
eral nerves]. Cancer Radiother 1997;1:677–682.
41. Mornex F, Bolla M, Pavy JJ, et al. [Late effects of ionizing
radiations on normal tissues]. Cancer Radiother 1997;1:621-
42. Mornex F, Pavy JJ, Denekamp J, et al. [Scoring system of late
effects of radiations on normal tissues: The SOMA-LENT
scale]. Cancer Radiother 1997;1:622–668.
43. Jackson A, Ten Haken RK, Robertson JM, et al. Analysis of
clinical complication data for radiation hepatitis using a par-
allel architecture model. Int J Radiat Oncol Biol Phys 1995;
44. Rosenzweig KE, Mychalczak B, Fuks Z, et al. Final report of
the 70.2-Gy and 75.6-Gy dose levels of a phase I dose esca-
lation study using three-dimensional conformal radiotherapy
in the treatment of inoperable non-small cell lung cancer.
Cancer J 2000;6:82–87.
45. Chapman JD, Bradley JD, Eary JF, et al. Molecular (func-
tional) imaging for radiotherapy applications: An RTOG sym-
posium. Int J Radiat Oncol Biol Phys 2003;55:294–301.
46. Lonneux M, Sibomana M, Pauwels S, et al. [Current data and
perspectives on positron emission tomography oncology-ra-
diotherapy]. Cancer Radiother 1999;3:275–288.
342I. J. Radiation Oncology ● Biology ● PhysicsVolume 64, Number 2, 2006