AJR 2001; 177:521-524
© American Roentgen Ray Society
Is Diagnostic Review of Radiotherapy-Planning CT Scans Important in the Conformal Therapy Era?
Melanie C. Smitt1 and
Vivek K. Mehta
1
Both authors: Department of Radiation Oncology, Stanford Hospital, Stanford
University Medical Center, 300 Pasteur Dr., Stanford, CA 94305.
Received December 21, 2000;
accepted after revision March 8, 2001.
Address correspondence to M. C. Smitt.
Abstract
OBJECTIVE. Detailed CT scans are often acquired during the
radiotherapy planning process. This study was performed to determine the
incidence of important benign and cancer-related CT findings on these
scans.
SUBJECTS AND METHODS. From December 1998 to December 2000, 162
radiotherapy patients who were to be treated curatively underwent treatment
planning CT scans on a helical scanner in the radiology department at
Washington Hospital, Fremont, CA. All CT scans were prospectively interpreted
relative to diagnoses, and reports were dictated for the medical records. The
diagnostic reports and records on all patients were reviewed to determine the
incidence of previously unknown benign or cancer-related findings, the impacts
of such findings on treatment, and the need for additional radiologic studies
or procedures on the basis of the CT interpretations.
RESULTS. Incidental benign findings were noted for 32 patients
(20%). Potentially important benign findings were noted for three patients:
two with aneurysms and one with a possible deep vein thrombosis. Potentially
cancer-related findings were reported in 20 patients: a single liver lesion
(four patients), multiple liver lesions (two patients), possible or probable
lymphadenopathy (11 patients), abnormal soft tissue (one patient), a
small-bowel obstruction (one patient), and a breast mass (one patient). After
reviewing prior diagnostic studies and obtaining additional recommended
studies, the physicians found that only three of the previously unknown
findings required further investigation: two aneurysms, which did not require
near-term treatment, and one metastatic neck node.
CONCLUSION. Routine diagnostic interpretation of radiotherapy
planning scans resulted in few important medical findings and changed patient
care for less than 1% of the patients.
Introduction
Recent advances in radiation treatment planning software have allowed the
direct incorporation of CT images into routine radiation oncology practice.
Several studies have documented improved tumor targeting and normal tissue
dose distributions over various tumor sites with conformal radiotherapy fields
designed using CT-based plans. Conformal radiotherapy fields adhere tightly to
the three-dimensional reconstruction of tumor target volumes obtained from CT
data. Current trends suggest that most patients receiving curative treatment
will ultimately undergo such three-dimensional planning procedures. Optimal
incorporation of this technology, however, requires CT images to be
reproducibly obtained in the treatment position. Several companies now have
commercial CT scanners designed for use in radiation oncology departments.
These scanners feature flat tabletops registered to the linear accelerator
couch, lasers adapted for radiotherapy marking needs, and computer
workstations allowing rapid contouring of relevant structures with creation of
digitally reconstructed radiographs in beam's eye-view projections. The
placement of CT scanners in the radiation oncology department generally
results in the loss of expert diagnostic interpretation of the images by
radiologists. In addition to the possible loss of oncologic information, the
presence of medically important benign disease could be overlooked. The
purpose of this study was to examine the results of routine diagnostic
interpretation of CT images obtained for radiation planning at a
community-based radiation oncology center.
Subjects and Methods
In August 1998, the Washington—Stanford Radiation Oncology Center (a
joint venture of Washington Hospital and Stanford University) acquired a
computer capable of three-dimensional treatment planning (Cadplan; Varian,
Palo Alto, CA). For patients undergoing CT-based planning, an initial
simulation was performed on a conventional simulator (Ximatron; Varian).
Patients underwent site-specific immobilization with placement of reference
marks, and appropriate simulator films were taken. Treatment planning CT scans
were then performed at Washington Hospital on a helical scanner (General
Electric Medical Systems, Milwaukee, WI) using a flat tabletop insert. Slice
intervals were similar to site-specific protocols already in place at the
hospital throughout the radiotherapy field, ranging from 3 to 7 mm. IV
contrast medium was used at the discretion of the managing radiation
oncologist. Scans were generally performed without use of contrast agents,
with the exception of the scans of 19 patients; their scans were primarily of
the brain or head and neck, areas in which contrast administration was
required to achieve optimal tumor visualization. CT images were transferred
directly to the planning computer via an integrated services digital network
(ISDN) line, and hard-copy films were also created.
From December 1998 through December 2000, treatment planning CT scans
obtained at Washington Hospital were prospectively interpreted by the
Washington Hospital Radiology Group to diagnose any medically important benign
or cancer-related conditions. During this period, 162 patients who received
curative treatment underwent CT planning and became the subjects of this
study. The indications for conformal therapy planning were consistent with
established benefits and policies at Stanford University. The site and stage
distributions for these patients are shown in Tables
1 and
2. Initial staging protocols
were site-specific, in keeping with the National Comprehensive Cancer Network
guidelines [2]. In particular,
no breast and only four prostate patients underwent pretreatment CT or MR
imaging, whereas all patients with other diagnoses had received routine
cross-sectional imaging at presentation.
All treatment planning CT scans were prospectively interpreted relative to
diagnosis, and reports were dictated for the medical records. Any previous
diagnostic CT reports and available medical records on all patients were
reviewed and assessed to determine the incidence of previously unknown benign
or cancer-related findings, impacts of such findings on treatment, and the
need for additional radiologic studies or procedures on the basis of the
treatment planning CT interpretations. Findings were classified as incidental
benign findings if medical follow-up or further studies were not required,
important benign findings if medical follow-up or studies were required, and
cancer-related findings if the diagnostic report suggested possible
unsuspected sites of carcinoma requiring medical follow-up or additional
studies.
Results
The incidence and type of previously unknown findings on the treatment
planning CT scans are summarized in Tables
3 and
4. Incidental benign findings
without clinical importance were noted in 32 patients (20%). Important benign
findings were noted in three patients (2%). Two of these findings were
aneurysms that did not meet size criteria for urgent intervention, and one was
a possible deep vein thrombosis. In the latter patient, further evaluation
with Doppler sonography yielded negative results.
Potential cancer-related findings were recorded in 20 patients (12%). These
findings included a single liver lesion (n = 4), multiple liver
lesions (n = 2), possible or probable lymphadenopathy (n =
11), abnormal soft tissue (n = 1), a small-bowel obstruction
(n = 1), and a breast mass (n = 1). After clinical review,
appropriate physical examination, and further radiologic studies, one patient
was found to have a new metastatic deposit. This patient had apparently
developed neck adenopathy in the interval between the initial consultation and
simulation. CT findings led to a decision to perform a neck dissection before
radiation. The findings for remaining patients were believed, on further
investigation, to represent liver cysts, postoperative change or seroma, or
vascular structures.
Diagnostic review of the treatment planning CT scans led to the performance
of 10 additional radiologic studies or procedures. These included:
contrast-enhanced CT (n = 4), liver sonography (n = 1),
fine-needle aspiration (n = 1), MR imaging (n = 2), Doppler
sonography (n = 1), and small-bowel series (n = 1). Most
(n = 6) of these studies were ordered in breast cancer patients to
evaluate multiple liver lesions or possible lymphadenopathy. Histologic
diagnosis was pursued only when recommended by the diagnostic physicians after
additional imaging studies. To date, no patient has had a recurrence in a
nonbiopsied area identified by the treatment planning CT scans.
Discussion
CT-based conformal radiotherapy has shown advantages in determining tumor
coverage, normal tissue doses, and safe escalation of radiotherapy doses
[3,
4]. For certain malignancies,
these advantages appear to have translated into improved cure rates and
decreased long-term morbidity
[5,
6]. Accurate acquisition of CT
images in the radiotherapy treatment position is critical to optimal
implementation of conformal radiotherapy techniques. CT images can be obtained
on diagnostic scanners and then transferred electronically to a treatment
planning computer. However, there are practical benefits to having a CT
scanner in the radiotherapy department, such as the ease of scheduling
patients as well as an increase in patient and staff convenience. In addition,
commercial CT simulators include features of importance to radiation
oncologists: a flat tabletop registered to the linear accelerator couch,
generation of digitally reconstructed radiographs for field verification, and
computer workstations adapted to radiotherapy needs. The 1994 Patterns of Care
facilities survey had already found an increase in the use of dedicated CT
scanners in radiotherapy departments
[7]. This study was performed
in anticipation of our department's acquisition of a CT simulator to determine
whether continued diagnostic interpretation of our treatment planning CT scans
would be of value in patient care.
A few previous site-specific studies have examined the incidence of
unsuspected comorbid disease discovered on diagnostic or treatment planning CT
scans
[8,9,10,11]
(Table 5). Although a fair
percentage of patients' scans do reveal some benign finding, only 1-10% reveal
findings requiring medical follow-up or near-term treatment. The likelihood of
relevant benign disease undoubtedly depends on characteristics of the studied
patient population. Elderly patients and those with other risk factors for
comorbid disease would be expected have a higher incidence. Among prostate
cancer patients in the prior studies, 1-9% of their scans revealed medically
important comorbid disease, primarily aneurysms. Second malignancies were also
noted. Although the overall rate of medically important comorbid disease in
our study was low, the rate among our prostate patients was 4%, consistent
with other reports. Among our breast cancer patients, no benign findings that
required treatment were found; a rate of less than 1% was reported in another
study [11].
View this table:
[in this window]
[in a new window]
|
TABLE 5 Review of Previous Studies on Incidence of Unsuspected Comorbidity
Discovered on Diagnostic or Treatment Planning CT Scans
|
|
Detecting additional foci of cancer on treatment planning scans has obvious
importance in terms of prognosis and potential effects on treatment programs
or radiotherapy field design. Prior studies showed a 3-7%
[9,
11] incidence of unexpected
cancer foci as compared with less than 1% in this study. However, these prior
studies included patients with metastatic disease and higher numbers of
patients with locally advanced disease. Looking only at patients receiving
curative treatment in those studies, the incidence of new cancer findings was
4% and 2%, respectively [9,
11]. We did not routinely
examine the results of diagnostic scan interpretation in patients who were not
treated curatively because any such findings would be unlikely to affect
overall patient outcomes. Histologic diagnosis of all CT findings was not
obtained in this study, and the true frequency of pathologic findings could be
higher. However, this study primarily addressed the likelihood of changes in
clinical treatment with diagnostic interpretation using reasonable clinical
judgment. To date, no patient has had a clinical recurrence in nonbiopsied
areas.
Evaluating the impact of diagnostic interpretation of treatment planning
scans requires recognition of the technical shortcomings of such scans.
Radiotherapy planning scans may be obtained with markers or immobilization
devices that create artifacts. Patient positioning may result in unusual
fields of view. Optimal amounts of contrast media, whether IV or oral, are
rarely administered in the radiotherapy department. For patients undergoing
diagnsotic cross-sectional imaging as part of their routine staging workup,
treatment scans obtained with such limitations seem unlikely to generate new
findings. For patients who do not undergo routine diagnostic CT scans,
planning scans could understimate findings or create ambiguous findings
leading to unnecessary further studies. The review by Forman et al.
[8] found that diagnostic CT
scans for prostate staging led to 21 additional studies in 273 patients with
prostate cancer. Diagnostic interpretation of planning scans in this study led
to 10 additional studies in 162 patients. The most additional studies were
performed in breast cancer patients, an occurrence that appeared to result
from the lack of IV contrast medium in studies of the liver and the
radiologist's lack of familiarity with the appearance of the postsurgical
axilla and breast. More intensive staging of breast and prostate patients at
presentation might decrease the likelihood of unexpected findings on
radiotherapy treatment planning scans, but this procedure would not be
consistent with current oncologic guidelines. Also, in current multimodality
protocols, radiotherapy planning CT scans may be obtained long after initial
staging, with the attendant possibility of new findings.
Significant benign or cancer-related findings were uncommon in this study,
but radiation centers with a different distribution of disease sites, stages,
or background patient comorbidities could encounter different results.
Although such findings may be rare, they are not nonexistent. One study
suggested that detailed review by the radiation oncologist could screen
planning CT scans for significant findings
[11]. Without a prospective
comparison, it is unknown whether radiation oncology review can substitute for
diagnostic expertise. In our breast cancer patients, we had a 53% incidence of
incidental findings overall as compared with 11% in the previously cited study
[11]. Alternatively, hard
copies of CT scans could be selectively submitted for diagnostic review for
patients for whom no pretreatment scan had been obtained or who had received
interval therapy or had a long interval since the last scan. However, a
funding mechanism for professional interpretation might need to be provided.
Our center currently provides Medicare-equivalent reimbursement to the
diagnostic group when other insurance coverage is not forthcoming
(approximately 15-20% of these patients). Patients should perhaps be made
aware through appropriate consent forms that treatment planning scans are not
of diagnostic quality and will not be routinely reviewed by diagnostic
radiologists. In addition to relevant benign findings in two patients and new
metastatic disease in one, diagnostic expertise may have been useful in an
additional 19 patients in whom IV contrast administration was needed for tumor
delineation or for whom formal staging reevaluation was sought by the
referring physician. It is unclear whether radiation oncologists currently
have optimal knowledge or experience in terms of contrast medium reaction
prophylaxis and treatment and contrast medium administration protocols. Such
knowledge could, perhaps, be obtained by a coordination of policies with local
diagnostic radiologists or by dissemination of appropriate professional
guidelines. The indications for and use of contrast media in conformal therapy
planning scans may increase.
In summary, diagnostic interpretation of radiotherapy planning scans in 162
patients treated curatively showed important benign disease in two patients
and resulted in a change of planned cancer treatment in one patient
(<1%).
References
-
American Joint Committee on Cancer. AJCC cancer staging
manual, 5th ed. Philadelphia: Lippencott-Raven,
1997
-
National Comprehensive Cancer Network. NCCN proceedings:
oncology practice guidelines, vols. 1-7.
Rutledge, PA: National Comprehensive Cancer Network,
2000
-
Photon Treatment Planning Collaborative Working Group. Evaluation
of high-energy photon external beam treatment planning: project summary.
Int J Radiat Oncol Biol Phys
1991;21:3
-8[Medline]
-
Sandler HM, McLaughlin PW, Ten Haken AK, et al. Three dimensional
conformal radiotherapy for the treatment of prostate cancer: low risk of
chronic rectal morbidity observed in a large series of patients.
Int J Radiat Oncol Biol Phys
1995;33:797
-801[Medline]
-
Hanks GE, Lee WR, Hanlon AL, et al. Conformal technique dose
escalation for prostate cancer: biochemical evidence of improved cancer
control with higher doses in patients with pre-treatment prostate specific
antigen > 10 mg/ml. Int J Radiat Oncol Biol Phys
1996;35:862
-868
-
Zelefsky J, Leibel SA, Kutcher GJ, et al. Dose escalation with
three-dimensional conformal radiation therapy affects the outcome in prostate
cancer. Int J Radiat Oncol Biol Phys
1998;41:491
-500[Medline]
-
Owen JB, Coia LR, Hanks GE. The structure of radiation oncology in
the United States in 1994. Int J Radiat Oncol Biol
Phys 1997;39:179
-185[Medline]
-
Forman HP, Heikin JP, Brink JA, Glazer HS, Fox LA, McClennan BL. CT
screening for comorbid disease in patients with prostatic carcinoma: is it
cost-effective? AJR
1994;162:1125
-1128[Abstract/Free Full Text]
-
Burcombe RJ, Ostler PJ, Ayoub AW, et al. The role of staging CT
scans in the treatment of prostate cancer: a retrospective audit.
Clin Oncol
2000;12:32
-35
-
Miller JS, Puckett ML, Johnstone PAS. Frequency of coexistent
disease at CT in patients with prostate carcinoma selected for definitive
radiation therapy: is limited treatment planning CT adequate?
Radiology
2000;215:41
-44[Abstract/Free Full Text]
-
Mehta VK, Goffinet DR. Unsuspected abnormalities noted on CT
treatment planning scans obtained for breast and chest wall irradiation.
Int J Radiat Oncol Biol Phys
2001;49:723
-725[Medline]
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
Top
Abstract
Introduction
