- Email: [email protected]
Prognostic Value of Positron Emission Tomography Using F-18-Fluorodeoxyglucose in Patients with Cervical Cancer Undergoing Radiotherapy
Gynecologic Oncology 84, 289 –295 (2002) doi:10.1006/gyno.2001.6504, available online at http://www.idealibrary.com on
Prognostic Value of Positron Emission Tomography Using F-18-Fluorodeoxyglucose in Patients with Cervical Cancer Undergoing Radiotherapy Yuji Nakamoto, M.D.,* Avraham Eisbruch, M.D.,† Eric D. Achtyes,† Yoshifumi Sugawara, M.D.,‡ Kevin R. Reynolds, M.D.,§ Carolyn M. Johnston, M.D.,§ and Richard L. Wahl, M.D.* ,1 *Division of Nuclear Medicine, Johns Hopkins University, Baltimore, Maryland 21287– 0817; †Department of Radiation Oncology, and §Department of Obstetrics and Gynecology, University of Michigan Medical Center, Ann Arbor, Michigan 48109; and ‡Department of Radiology, Ehime University, Matsuyama 790-8577, Japan Received July 20, 2001
Objective. The purpose of this study was to determine whether positron emission tomography (PET) using F-18-fluorodeoxyglucose (FDG) before and after radiotherapy would predict whether local control of cervical cancer had been achieved. Methods. FDG-PET scans were performed prior to therapy and at a mean of 4.6 months after radiation in 20 patients (pts) with histologically proven uterine cervical cancer who were undergoing a “curative” course of radiation therapy. FDG uptake was interpreted visually by two readers using a 5-point grading system (0 ⴝ normal, 1 ⴝ probably normal, 2 ⴝ equivocal, 3 ⴝ probably abnormal, and 4 ⴝ definitely abnormal). The standardized uptake values corrected by lean body mass (SUL) were calculated for suspicious areas. The percentage of residual activity (%RA) for the posttherapy SUL was also evaluated as a percentage of the pretherapy SUL. Results. At baseline before irradiation, 17 of 20 (85.0%) primary tumors were detected. Following irradiation, no or low (grade 0 –2) uptake was observed in 9 pts, and none of these had local recurrence. Among the remaining 11 pts with grade 3 or 4 uptake, the correct diagnosis was made for 5 pts with active tumor; SULs (mean ⴞ SD ⴝ 4.17 ⴞ 2.52) and %RAs (57.9 ⴞ 16.8). Six patients without active tumor showed relatively low SULs (2.67 ⴞ 0.69) and %RAs (43.0 ⴞ 18.3). No significant differences were observed between the recurrent and nonrecurrent groups for these parameters. Overall, sensitivity, specificity, and accuracy were 100, 60, and 70%, respectively. Conclusion. These preliminary data indicate that FDG-PET is a sensitive tool for detecting active cervical cancer after radiation, however, the method, without anatomic correlation had suboptimal specificity. © 2002 Elsevier Science Key Words: FDG; PET; recurrence; uterine cervical cancer.
1 To whom correspondence and reprint requests should be addressed at Division of Nuclear Medicine, 601 N Caroline Street, Room 3223A, Baltimore, MD 21287-0817. Fax: 410-614-3896. E-mail: [email protected].
INTRODUCTION Failure to achieve local control, local recurrence, and locoregional or distant metastasis are often major problems for patients who have undergone treatment with curative intent for malignant diseases. Although surgery, radiation therapy, and chemotherapy have long-established roles in the management of uterine cervical cancer, recurrence cannot be avoided in many patients [1]. It is difficult to treat recurrent cancer, resulting in poor prospects of survival [2, 3]. However, more accurate diagnosis during the early posttreatment period is important for the quality of life and care of these patients. Recent progress in diagnostic imaging, especially magnetic resonance imaging (MR) and helical computed tomography (CT) with contrast enhancement, has made possible the characterization of gynecological malignancies. Typical manifestations of recurrent cervical cancer such as pelvic masses and lymphadenopathy are well recognized. However, since some measurable tumors are difficult to differentiate from scar tissue, the distinction between pelvic recurrence and postoperative or radiation-induced changes presents a significant diagnostic challenge [4, 5]. Positron emission tomography (PET) using F-18-fluoro-2deoxy-D-glucose (FDG) has been shown to be a useful tool for imaging recurrent cancers, including colorectal cancers [6 – 8]. As others and we have previously reported, PET may be used not only in evaluating primary cervical tumors [9], but also in staging of uterine cervical cancer [10]. Until now, however, the clinical application of FDG-PET for recurrent uterine cervical cancer has been limited [11, 12], and its benefit in contributing to clinical decision-making remains unknown. In the present study, we assessed whether quantification of metabolic activity or tumor response using FDG-PET could predict persistent or local tumor recurrence and overall survival
289
0090-8258/02 $35.00 © 2002 Elsevier Science All rights reserved.
290
NAKAMOTO ET AL.
TABLE 1 Patients’ Profiles
Patient No.
Age (years)
Histology
Stage (FIGO)
Size (cm)
Total dose (Gy)
Chemotherapy
Time interval between RT and postscan (month)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
55 46 34 49 77 75 35 52 65 82 38 72 41 47 37 48 82 45 26 36
SqCCa SqCCa SqCCa SqCCa SqCCa SqCCa SqCCa SqCCa AdenoCa SqCCa AdenoSqCa SqCCa SqCCa AdenoCa SqCCa SqCCa SqCCa SqCCa SqCCa (rec) SqCCa
IIB IIB IIB IIIB IIB IIIB IIB IIIB IIB IB IIB IIIB IB IIB IVA IB IB IIA IB IIIB
4 8 7 8 2 4 8 4 6 3 4 n.a. 3 8 8 6 n.a. 8 7 10
45.0 40.0 50.0 45.8 45.0 40.0 45.0 40.0 51.0 20.0 45.0 50.0 40.0 45.0 45.0 40.0 30.6 46.0 73.8 45.5
(⫹) (⫺) (⫹) (⫹) (⫺) (⫹) (⫹) (⫺) (⫹) (⫺) (⫹) (⫹) (⫹) (⫹) (⫹) (⫺) (⫹) (⫹) (⫹) (⫹)
4 3 6 4 5 4 5 4 4 5 7 5 4 6 5 3 4 5 4 5
Status (F/U period, month) Alive (61) Alive (24) Alive (20) Dead, local rec (7) Alive (18) Alive (17) Dead, lung mets (7) Dead, lung ca (9) Alive (6) Alive (37) Alive (29) Alive (12) Dead, lung mets (24) Dead, local rec (17) Dead, para-aortic LN (3) Alive (19) Dead, local rec (22) Alive (10) Dead, persistent disease (6) Dead, persistent disease (4)
Note. FIGO, International Federation of Gynecology and Obstetrics; RT, radiation therapy; F/U, follow-up. SqCCa, squamous cell carcinoma; AdenoCa, adenocarcinoma; AdenoSqCa, adenosquamous carcinoma. local rec, local recurrence; lung mets, lung metastases; para-aortic LN, para-aortic lymph node. n.a., not available.
in patients who underwent a curative course of radiation therapy (RT) for treatment of uterine cervical cancer. PATIENTS AND METHODS Patients Twenty patients (age range 26 – 82, mean 52.1 years old) with histologically proven cervical cancer were enrolled in this study. The patients’ profiles are summarized in Table 1. Among them, 10 patients, who had been previously evaluated [9], were reevaluated before and after radiation therapy. There were 5 patients with International Federation of Gynecology and Obstetrics (FIGO) stage I B, 1 with stage II A, 8 with stage II B, 5 with stage III B, and 1 with IV A disease. The histopathologic types were squamous cell carcinoma in 17 patients, adenosquamous cell carcinoma in 1 patient, and adenocarcinoma in 2 patients. Of the 20 patients, 19 had newly diagnosed cervical cancer, and 1 had recurrent cervical cancer after surgical treatment. In addition to a baseline scan prior to radiation treatment, all patients underwent a second PET study 3 to 7 months following therapy (mean ⫽ 4.6 months). Written informed consent was obtained from all patients for the study, which was approved by the institutional review board and conducted under the guidelines for a physician-sponsored investigational drug application.
Radiation Therapy Patients were treated between August 1994 and August 1999. Therapy consisted of whole pelvic irradiation using the four-field technique (median, 45 Gy, range 20 –70 Gy) followed with one or two intracavitary implants utilizing FletcherSuit applicators (16 patients) or an interstitial implant (3 patients). One patient received external RT alone (74 Gy). The total dose prescribed to patients receiving intracavitary implants was in the range of 77–90.1 Gy (median, 85 Gy), and the patients receiving interstitial radiation received a total of 75 Gy minimal tumor dose. Ten patients received cisplatin-containing chemotherapy concurrent with radiation, 5 received the radiosensitizer bromodeoxyuridine according to a previously published protocol [13], and 5 patients received radiation alone. After therapy, patients had been followed with history and physical examination, including bimanual pelvic examination and cervical cytologic smears, every 2–3 months before the follow-up PET scan was done. PET Scanning FDG was produced by a standard nucleophilic fluorination method as previously described [14]. FDG-PET scans were performed with a Model 921 Exact (47 scanning planes, 15-cm longitudinal field of view) scanner (CTI, Knoxville, TN, dis-
291
FOLLOW-UP OF CERVICAL CANCER WITH FDG-PET
tributed by Siemens Medical System, Iselin, NJ). The reconstructed x–y resolution with the Hanning filter cutoff value of 0.3 was approximately 1.2 cm full width at half maximum. All patients fasted for at least 4 h before PET scanning. Transmission scan (at least 10 min) was obtained using a 68Ge rod source for the purpose of attenuation correction, followed by intravenous tracer injection (approximately 370 MBq FDG). Sequential dynamic scans at the level of suspected tumors were obtained through 60 min. These dynamic scans were generally six 10-s scans, three 20-s scans, two 90-s scans, one 5-min scan and five 10-min scans. Then, some patients were instructed to void, and the postvoid emission and transmission scans, each of 10-min duration, were obtained over the pelvis to minimize bladder activity. Generally, two contiguous levels of scanning were imaged from the level of the symphysis pubis to the level of the middle abdomen. In general, CT was not available at the time of PET and was not used to guide patient positioning for PET scans. Software used to perform the postinjection emission and transmission scans was provided by Siemens/CTI. CT Scanning In 17 of 20 cases, CT scans were obtained with 10-mm-thick contiguous axial sections with a GE-9800 scanner (Hi-lite Advantage; GE Medical Systems, Milwaukee, WI). Intravenous contrast material (Omnipaque-300; Winthrop Pharmaceuticals, New York, NY) and oral contrast material (1.5% Hypaque solution; Winthrop Pharmaceuticals) were used. Three cases had CT scans performed in outside institutions. Image Analysis All PET and CT scans were interpreted separately. PET images were analyzed on an interactive computer display by two observers who were blinded to other imaging results and clinical data. Any obvious foci of increased FDG uptake were evaluated using transaxial, sagittal, and coronal displays and compared between prevoid and postvoid scans, if available. Thus, FDG uptake in the last (50 – 60 min) frames of dynamic scans were assessed visually, and the degree of abnormality of FDG accumulation was classified into five grades by consensus interpretations: 0 ⫽ normal, 1 ⫽ probably normal, 2 ⫽ equivocal, 3 ⫽ probably abnormal, and 4 ⫽ definitely abnormal for both sets of images. Cases with grade 3 or 4 uptake were considered positive for disease. For a semiquantitative index of FDG uptake in tumors, standardized uptake value (SUV), which is the decay-corrected tissue activity divided by the injected dose per patient body weight, was calculated for areas suspicious for recurrence. Then, those values were corrected by predicted lean body mass, as described previously [15], and SUV-lean (SUL) was used for quantitative analyses. All SULs were calculated from regions of interest (ROIs) that were placed by means of automated algorithm on the maximal area (16 pixels in size) of FDG uptake within a larger ROI covering
the tumors. No correction was applied for partial volume effects. In addition, the percentage of residual activity (%RA) for the posttherapy SUL was also evaluated as a percentage of the pretherapy SUL. Diagnostic accuracy was evaluated by comparing the PET results with final diagnoses, determined by histopathological examination (n ⫽ 4) or clinical follow-up (n ⫽ 16), including radiological findings. The follow-up period was 6 to 61 months (mean, 23.0 months) for living patients and 3 to 24 months (mean, 11.1 months) for dead patients. RESULTS PET findings were shown in Table 2. Among all 20 patients (pts) primary cervical tumors were detected visually by PET (score 3 or 4) prior to RT in 17 cases (85.0%), and SULs ranged from 4.17 to 14.23, with a mean of 7.77 ⫾ 3.11. After radiation therapy with or without chemotherapy for radiosensitization, 8 of 20 patients were diagnosed as positive recurrent or disease on the basis of physical examinations, cytology, and radiological findings, e.g., local recurrence (n ⫽ 5), paraaortic nodes (n ⫽ 1), or distant metastases only (n ⫽ 2). However, one paraaortic node and two distant metastatic lesions were out of scan of PET images, so five local lesions were further evaluated. In the posttherapeutic scans, uptakes of grades 0 to 2 were observed in 9 pts. The range of posttherapy SULs of the remaining 11 pts was 1.83 to 8.50 with a mean of 3.35 ⫾ 1.84. Of these 11 pts, we made correct visual diagnoses for 5 pts with recurrence, with %RA values ranging from 32.5 to 74.1 (mean ⫽ 57.9 ⫾ 16.8). Recurrence in these patients was confirmed by biopsy for 4 pts and by increasing radiological tumor size for 1 pt. However, 6 pts with elevated posttreatment SULs also had high %RAs ranging from 18.2 to 72.7 (mean ⫽ 43.0 ⫾ 18.3). PET was able to provide additional clinical information for a representative case shown in Fig. 1. The relationship between visual scoring and the current status of the patients is shown in Table 3. A statistically significant difference was observed between the recurrent and nonrecurrent cases (P ⬍ 0.05) with regard to their visual score, but there was no difference between the two groups based on the current status of the patient. In semiquantitative analysis, there was no significant difference between the positive and negative local recurrent groups with regard to some quantitative parameters, such as SUL, lesions-to-gluteus muscle ratio, and %RA. The percentage of recurrent positive patients based on the visual scoring is shown in Fig. 2. Based on visual analysis, overall sensitivity, specificity, and accuracy were 100, 60, and 70%, respectively (Table 4). Receiver operating characteristic (ROC) analysis was done using the binormal curve-fitting routine of the CORROC2 program developed at the University of Chicago by Metz et al. [16]. There were no statistically significant differences among areas under the ROC curves using these parameters (Fig. 3).
292
NAKAMOTO ET AL.
TABLE 2 PET Results Primary tumor Before therapy
Patient No.
Visual score
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
4 4 4 4 1 0 4 4 4 0 4 4 4 4 4 4 4 4 4 4
After therapy
SUL
L/M ratio
Postvoid scan
Visual score
PET results
SUL
4.17 4.55 5.32 5.15 n.a. n.a. 10.85 11.84 14.23 n.a. 10.16 6.08 4.77 6.58 10.36 6.34 5.23 8.64 6.41 11.47
4.48 7.11 7.09 7.15 n.a. n.a. 13.23 7.95 12.82 n.a. 15.63 6.54 4.92 8.55 11.26 8.81 5.51 10.16 8.90 9.25
Yes Yes Yes Yes Yes No No Yes Yes Yes Yes Yes Yes Yes No No Yes Yes Yes Yes
3 4 3 3 2 0 1 1 3 1 1 0 2 3 0 4 3 3 4 4
FP FP FP TP TN TN TN TN FP TN TN TN FP TP TN TN TP FP TP TP
2.17 3.31 1.83 3.69 n.a. n.a. n.a. n.a. 2.59 n.a. n.a. n.a. n.a. 2.14 n.a. 2.45 2.73 3.65 3.77 8.50
Nodal mets
L/M ratio
Postvoid scan
2.17 2.88 2.23 3.08 n.a. n.a. n.a. n.a. 3.65 n.a. n.a. n.a. n.a. 2.85 n.a. 3.55 2.97 4.10 5.39 9.66
Yes Yes Yes No Yes No No Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes No Yes Yes
%RA
Before therapy, visual score, SUL
After therapy, visual score, SUL
52.0 72.7 34.4 71.7 n.a. n.a. n.a. n.a. 18.2 n.a. n.a. n.a. n.a. 32.5 n.a. 38.6 52.2 42.2 58.8 74.1
0, 0, 0, 3, 1, 0, 0, 0, 3, 0, 0, 0, 3, 0, 3, 1, 2, 4, 0, 2,
0, 0, 0, 1, 2, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 4, 0, 0,
n.a. n.a. n.a. 2.65 n.a. n.a. n.a. n.a. 6.63; 3, 3.73 n.a. n.a. n.a. 2.12 n.a. 3.59; 4, 3.65 n.a. n.a.; 2, n.a. 8.12; 3, 1.98 n.a. n.a.
n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a.; 0, n.a. n.a. n.a. n.a. n.a. n.a. n.a.; 0, n.a. n.a. n.a.; 0, n.a. 6.85; 0, n.a. n.a. n.a.
Note. L/M ratio, lesion-to-muscle ratio; %RA, percentage of residual activity; n.a., not available. TP, true positive; TN, true negative; FP, false positive; FN, false negative.
DISCUSSION Our study expands upon the small, but promising literature regarding the role of FDG-PET in uterine cervical cancer. Although the number of patients was limited, our data indicate that all recurrent cancers were detected by PET, consistent with the preliminary report by Umesaki et al., who showed that all recurrent foci had relatively high SUV compared to normal cervix [11, 12]. Specificity, however, was lower than has been reported under some other conditions, such as posttreatment lung cancer [17]. In addition, semiquantitative analysis could not differentiate between recurrent and nonrecurrent patients and was not useful in predicting prognostic status, at least in this small patient cohort. There are nine cases where PET and CT studies were done within approximately 3 weeks of each other (mean 13.1 days). Both PET and CT showed true negative results in five of these nonrecurrent cases. However, in four of the recurrent cases, CT yielded two positive, one negative, and one inconclusive result, whereas PET revealed positive findings for all cases. Thus, additional information by PET was obtained in only two patients in our study, but the clinical contribution of this modality
would likely increase with broader application of the technique during the follow-up of recurrent cervical cancer. With regard to the relatively low specificity observed among our patient population, it is well known that FDG accumulates in inflammatory foci, not just malignant tumors, which can lead to false-positive findings [18]. Furthermore, it has been shown that once irradiation is initiated, FDG uptake can rise in the treated and dying tumors [19]. In this study, posttreatment scans were performed 4 to 5 months after radiation therapy was completed, but it is possible that inflammatory processes were continuing in these localized areas in the cases with falsepositive results. A significant difference in the time interval between radiation therapy and posttreatment scan was not observed (mean 4.6 months) between recurrent and nonrecurrent groups. In other cases, physiological uptake in the ureters or bladder and alimentary tract may have led to false-positive findings. In 32 of the 40 scans, we obtained postvoid images, which made it easier to evaluate pelvic lesions, especially when images were reconstructed using a filtered backprojection technique. However, physiological accumulation observed in the urinary system seemed not to contribute to the falsepositive results, since five of six false-positive diagnoses were
293
FOLLOW-UP OF CERVICAL CANCER WITH FDG-PET
FIG. 2. Percentages of recurrent positive patients based on the visual scoring grade. As would be expected, higher grading scores correlated with higher percentages of pts with recurrences.
FIG. 1. A 47-year-old female, recurrent cervical cancer (patient No. 14). PET (a, left) shows intense uptake behind the bladder (arrow, SUL ⫽ 6.58), corresponding to the primary cervical cancer. At 6 months after irradiation, focal uptake was observed (a, right: arrow, SUL ⫽ 2.14), indicating local recurrence. However, definite diagnosis of local recurrence could not be obtained by CT performed following the PET scan (b). The recurrence was confirmed by biopsy.
made in postvoid images. Physiological uptake in the gut remains a difficult problem, which, unlike that of urinary system, cannot be eliminated simply by voiding. In case of continuous linear accumulation, it may be differentiated from pathological uptake, but in other cases the cause of the increased uptake in this region remains enigmatic. Further study is needed to avoid both false-positive and false-negative results, which occur when lesions lie in close proximity to organs in the abdomen or pelvis, such as the gut, which can display a high level of physiological uptake.
With regard to the posttherapeutic evaluation of head and neck cancers with PET, Kitagawa et al. reported that a pretreatment PET scan was useful in predicting the response to treatment and that posttreatment scans were helpful in detecting residual viable cells [20]. Rege et al. described how pretreatment PET findings might have prognostic implications for patient outcome following radiation therapy [21]. Similarly, Anzai et al. reported PET to be significantly more accurate than MRI in detecting the presence or absence of recurrent squamous cell cancers [22]. Their separation was made on a qualitative basis only. However, in our study of uterine cervical cancer, semiquantitative analysis using SUL or %RA showed no clear correlation to the prognostic outcome of the patients. Mean values of SUL or lesion-to-muscle ratio after therapy and %RA in recurrent cases tended to be higher than those in negative cases, but there were no statistically significant differences between the two groups. It is not yet possible to differentiate between recurrent and nonrecurrent patients simply by referring to these values. TABLE 4 Diagnostic Accuracy of PET Recurrence or persistent disease
TABLE 3 Visual Scores in Posttreatment Scan Visual score
0
1
2
3
4
Negative local recurrence (n ⫽ 15) Positive local recurrence (n ⫽ 5) Alive (n ⫽ 11) Dead (n ⫽ 9)
3 0 2 1
4 0 2 2
2 0 1 1
4 3 4 3
2 2 2 2
Note. The statistically significant difference was found between the nonrecurrent and recurrent cases (P ⫽ 0.03), but there was no significant difference between alive and dead cases (Mann–Whitney U test).
PET positive PET negative Total Sensitivity Specificity Positive predictive value Negative predictive value Accuracy
Positive
Negative
5 0 5
6 9 15 100.0% 60.0% 45.5% 100.0% 70.0%
294
NAKAMOTO ET AL.
REFERENCES 1. Estape R, Angioli R. Surgical management of advanced and recurrent cervical cancer. Semin Surg Oncol 1999;16:236 – 41. 2. Krebs HB, Helmkamp BF, Sevin BU, Poliakoff SR, Nadji M, Averette HE. Recurrent cancer of the cervix following radical hysterectomy and pelvic node dissection. Obstet Gynecol 1982;59:422–7. 3. Ito H, Shigematsu N, Kawada T, Kubo A, Isobe K, Hara R, Yasuda S, Aruga T, Ogata H. Radiotherapy for centrally recurrent cervical cancer of the vaginal stump following hysterectomy. Gynecol Oncol 1997;67:154 – 61. 4. Yamashita Y, Harada M, Torashima M, Takahashi M, Miyazaki K, Tanaka N, Okamura H. Dynamic MR imaging of recurrent postoperative cervical cancer. J Magn Reson Imaging 1996;6:167–17.
FIG. 3. ROC curves for FDG-PET using semiquantitative parameters, SUL at posttreatment scan (bold line), lesion-to-muscle (L/M) ratio (normal line), and %RA (thin line). There were no statistically significant differences between values of area under the ROC curves (A z). A z (SUL) 0.80, A z (L/M ratio) 0.71, and A z (%RA) 0.74.
Uterine cervical cancer is a locally invasive and highly metastatic tumor. Distant metastases to the lung and other areas occur often, and unexpected metastases to atypical locations are not uncommon [23]. In our cases, three patients died from lung metastases (n ⫽ 2) or persistent aortic nodal metastasis (n ⫽ 1) despite good local control of the primary tumor. Recent advances in PET camera technology have enabled us to evaluate the whole body and could be used to assess metastatic spread in patients with cervical cancer. Finally, tumor markers such as squamous cell carcinoma (SCC) antigen are widely used for evaluating patients’ clinical conditions during the follow-up period [24 –26]. A blood test is a noninvasive technique that is easily repeatable. With SCC, reelevation of the tumor antigen was shown to correlate closely with early relapse. However, the rising of serum tumor markers provides no information about the location or extent of the disease and is not useful in all cases. It is also sometimes difficult to identify recurrent lesions with standard imaging modalities, such as CT and MR. As the high sensitivity of PET was shown in our study, Umesaki et al. demonstrated a case report where PET detected a recurrent tumor even when the tumor marker remained normal [11]. Although PET would not completely replace the monitoring of tumors using these modalities, this noninvasive technique could come to have a greater role for screening patients during follow-up, due to its high sensitivity. In summary, FDG-PET is a highly sensitive tool for predicting recurrent or persistent uterine cervical cancer. Despite its current relatively low specificity, it may be helpful for screening of follow-up patients undergoing a curative course of radiotherapy. Further investigations are needed to reduce falsepositive results.
5. Kinkel K, Ariche M, Tardivon AA, Spatz A, Castaigne D, Lhomme C, Vanel D. Differentiation between recurrent tumor and benign conditions after treatment of gynecologic pelvic carcinoma: value of dynamic contrast-enhanced subtraction MR imaging. Radiology 1997; 204:55– 63. 6. Strauss LG, Clorius JH, Schlag P, Lehner B, Kimmig B, Engenhart R, Marin-Grez M, Helus F, Oberdorfer F, Schmidlin P. Recurrence of colorectal tumors: PET evaluation. Radiology 1989;170:329 –32. 7. Delbeke D, Vitola JV, Sandler MP, Arildsen RC, Powers TA, Wright JK Jr, Chapman WC, Pinson CW. Staging recurrent metastatic colorectal carcinoma with PET. J Nucl Med 1997;38:1196 –201. 8. Flamen P, Stroobants S, Cutsem EV, Dupont P, Bormans G, Vadder ND, Penninckx F, Hoe LV, Mortelmans L. Additional value of whole-body positron emission tomography with fluorine-18-2-fluoro-2-deoxy-D-glucose in recurrent colorectal cancer. J Clin Oncol 1999;17:894 –901. 9. Sugawara Y, Eisbruch A, Kosuda S, Recker BE, Kison PV, Wahl RL. Evaluation of FDG PET in patients with cervical cancer. J Nucl Med 1999;40:1125–31. 10. Rose PG, Adler LP, Rodriguez M, Faulhaber PF, Abdul-Karim FW, Miraldi F. Positron emission tomography for evaluating para-aortic nodal metastasis in locally advanced cervical cancer before surgical staging: a surgicopathologic study. J Clin Oncol 1999;17:41–5. 11. Umesaki N, Tanaka T, Miyama M, Tokuyama O, Kawamura N, Ogita S, Kawabe J, Okamura T, Koyama K, Ochi H. Early diagnosis and evaluation of therapy in postoperative recurrent cervical cancers by positron emission tomography. Oncol Rep 2000;7:53– 6. 12. Umesaki N, Tanaka T, Miyama M, Kawabe J, Okamura T, Koyama K, Ochi H, Ogita S. The role of 18F-fluoro-2-deoxy-D-glucose positron emission tomography ( 18F-FDG-PET) in the diagnosis of recurrence and lymph node metastasis of cervical cancer. Oncol Rep 2000;7:1261– 4. 13. Eisbruch A, Robertson JM, Johnston CM, Tworek J, Reynolds KR, Roberts JA, Lawrence TS. Radiation alternating with bromodeoxyuridine for advanced uterine cervix cancer: a phase I and drug incorporation study. J Clin Oncol 1999;17:31– 40. 14. Toorongian SA, Mulholland GK, Jewett DM, Bachelor MA, Kilbourn MR. Routine production of 2-deoxy-2-[ 18F]fluoro-D-glucose by direct nucleophilic exchange on a quatenary 4-aminopyridinium resin. Int J Rad Appl Instrum B 1990;17:273–9. 15. Zasadny KR, Wahl RL. Standardized uptake values of normal tissues at PET with 2-[fluorine-18]-fluoro-2-deoxy-D-glucose: variations with body weight and a method for correction. Radiology 1993;189:847–50. 16. Metz CE, Wang PL, Kronman HB. A new approach for testing the significance of differences between ROC curves measured from correlated data. In: Deconinck F, editor. Information processing in medical imaging. The Hague, The Netherlands: Martius Nijhoff, 1984:432– 45.
FOLLOW-UP OF CERVICAL CANCER WITH FDG-PET 17. Patz EF Jr, Lowe VJ, Hoffman JM, Paine SS, Harris LK, Goodman PC. Persistent or recurrent bronchogenic carcinoma: detection with PET and 2-[F-18]-2-deoxy-D-glucose. Radiology 1994;191:379 – 82. 18. Strauss LG. Fluorine-18 deoxyglucose and false-positive results: a major problem in the diagnostics of oncological patients. Eur J Nucl Med 1996;23:1409 –15. 19. Higashi K, Clavo AC, Wahl RL. In vitro assessment of 2-fluoro-2-deoxyD-glucose, L-methionine and thymidine as agents to monitor the early response of a human adenocarcinoma cell line to radiotherapy. J Nucl Med 1993;34:773–9. 20. Kitagawa Y, Sadato N, Azuma H, Ogasawara T, Yoshida M, Ishii Y, Yonekura Y. FDG PET to evaluate combined intra-arterial chemotherapy and radiotherapy of head and neck neoplasms. J Nucl Med 1999;40: 1132–7. 21. Rege S, Safa AA, Chaiken L, Hoh C, Juillard G, Withers HR. Positron emission tomography: an independent indicator of radiocurability in head and neck carcinomas. Am J Clin Oncol 2000;23:164 –9.
295
22. Anzai Y, Carroll WR, Quint DJ, Bradford CR, Minoshima S, Wolf GT, Wahl RL. Recurrence of head and neck cancer after surgery or irradiation: prospective comparison of 2-deoxy-2-[F-18]fluoro-D-glucose PET and MR imaging diagnoses. Radiology 1996;200:135– 41. 23. Fulcher AS, O’Sullivan SG, Segreti EM, Kavanagh BD. Recurrent cervical carcinoma: typical and atypical manifestations. Radiographics 1999; 19:S103– 6. 24. Abe A, Nakano T, Morita S, Oka K. Clinical evaluation of serum and immunohistochemical expression of SCC and CA19-9 in radiation therapy for cervical cancer. Anticancer Res 1999;19:829 –36. 25. Juang CM, Wang PH, Yen MS, Lai CR, Ng HT, Yuan CC. Application of tumor markers CEA, TPA, and SCC-Ag in patients with low-risk FIGO stage IB and IIA squamous cell carcinoma of the uterine cervix. Gynecol Oncol 2000;76:103– 6. 26. Suzuki Y, Nakano T, Ohno T, Abe A, Morita S, Tsujii H. Serum CYFRA 21-1 in cervical cancer patients treated with radiation therapy. J Cancer Res Clin Oncol 2000;126:332– 6.
Prognostic Value of Positron Emission Tomography Using F-18-Fluorodeoxyglucose in Patients with Cervical Cancer Undergoing Radiotherapy Yuji Nakamoto, M.D.,* Avraham Eisbruch, M.D.,† Eric D. Achtyes,† Yoshifumi Sugawara, M.D.,‡ Kevin R. Reynolds, M.D.,§ Carolyn M. Johnston, M.D.,§ and Richard L. Wahl, M.D.* ,1 *Division of Nuclear Medicine, Johns Hopkins University, Baltimore, Maryland 21287– 0817; †Department of Radiation Oncology, and §Department of Obstetrics and Gynecology, University of Michigan Medical Center, Ann Arbor, Michigan 48109; and ‡Department of Radiology, Ehime University, Matsuyama 790-8577, Japan Received July 20, 2001
Objective. The purpose of this study was to determine whether positron emission tomography (PET) using F-18-fluorodeoxyglucose (FDG) before and after radiotherapy would predict whether local control of cervical cancer had been achieved. Methods. FDG-PET scans were performed prior to therapy and at a mean of 4.6 months after radiation in 20 patients (pts) with histologically proven uterine cervical cancer who were undergoing a “curative” course of radiation therapy. FDG uptake was interpreted visually by two readers using a 5-point grading system (0 ⴝ normal, 1 ⴝ probably normal, 2 ⴝ equivocal, 3 ⴝ probably abnormal, and 4 ⴝ definitely abnormal). The standardized uptake values corrected by lean body mass (SUL) were calculated for suspicious areas. The percentage of residual activity (%RA) for the posttherapy SUL was also evaluated as a percentage of the pretherapy SUL. Results. At baseline before irradiation, 17 of 20 (85.0%) primary tumors were detected. Following irradiation, no or low (grade 0 –2) uptake was observed in 9 pts, and none of these had local recurrence. Among the remaining 11 pts with grade 3 or 4 uptake, the correct diagnosis was made for 5 pts with active tumor; SULs (mean ⴞ SD ⴝ 4.17 ⴞ 2.52) and %RAs (57.9 ⴞ 16.8). Six patients without active tumor showed relatively low SULs (2.67 ⴞ 0.69) and %RAs (43.0 ⴞ 18.3). No significant differences were observed between the recurrent and nonrecurrent groups for these parameters. Overall, sensitivity, specificity, and accuracy were 100, 60, and 70%, respectively. Conclusion. These preliminary data indicate that FDG-PET is a sensitive tool for detecting active cervical cancer after radiation, however, the method, without anatomic correlation had suboptimal specificity. © 2002 Elsevier Science Key Words: FDG; PET; recurrence; uterine cervical cancer.
1 To whom correspondence and reprint requests should be addressed at Division of Nuclear Medicine, 601 N Caroline Street, Room 3223A, Baltimore, MD 21287-0817. Fax: 410-614-3896. E-mail: [email protected].
INTRODUCTION Failure to achieve local control, local recurrence, and locoregional or distant metastasis are often major problems for patients who have undergone treatment with curative intent for malignant diseases. Although surgery, radiation therapy, and chemotherapy have long-established roles in the management of uterine cervical cancer, recurrence cannot be avoided in many patients [1]. It is difficult to treat recurrent cancer, resulting in poor prospects of survival [2, 3]. However, more accurate diagnosis during the early posttreatment period is important for the quality of life and care of these patients. Recent progress in diagnostic imaging, especially magnetic resonance imaging (MR) and helical computed tomography (CT) with contrast enhancement, has made possible the characterization of gynecological malignancies. Typical manifestations of recurrent cervical cancer such as pelvic masses and lymphadenopathy are well recognized. However, since some measurable tumors are difficult to differentiate from scar tissue, the distinction between pelvic recurrence and postoperative or radiation-induced changes presents a significant diagnostic challenge [4, 5]. Positron emission tomography (PET) using F-18-fluoro-2deoxy-D-glucose (FDG) has been shown to be a useful tool for imaging recurrent cancers, including colorectal cancers [6 – 8]. As others and we have previously reported, PET may be used not only in evaluating primary cervical tumors [9], but also in staging of uterine cervical cancer [10]. Until now, however, the clinical application of FDG-PET for recurrent uterine cervical cancer has been limited [11, 12], and its benefit in contributing to clinical decision-making remains unknown. In the present study, we assessed whether quantification of metabolic activity or tumor response using FDG-PET could predict persistent or local tumor recurrence and overall survival
289
0090-8258/02 $35.00 © 2002 Elsevier Science All rights reserved.
290
NAKAMOTO ET AL.
TABLE 1 Patients’ Profiles
Patient No.
Age (years)
Histology
Stage (FIGO)
Size (cm)
Total dose (Gy)
Chemotherapy
Time interval between RT and postscan (month)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
55 46 34 49 77 75 35 52 65 82 38 72 41 47 37 48 82 45 26 36
SqCCa SqCCa SqCCa SqCCa SqCCa SqCCa SqCCa SqCCa AdenoCa SqCCa AdenoSqCa SqCCa SqCCa AdenoCa SqCCa SqCCa SqCCa SqCCa SqCCa (rec) SqCCa
IIB IIB IIB IIIB IIB IIIB IIB IIIB IIB IB IIB IIIB IB IIB IVA IB IB IIA IB IIIB
4 8 7 8 2 4 8 4 6 3 4 n.a. 3 8 8 6 n.a. 8 7 10
45.0 40.0 50.0 45.8 45.0 40.0 45.0 40.0 51.0 20.0 45.0 50.0 40.0 45.0 45.0 40.0 30.6 46.0 73.8 45.5
(⫹) (⫺) (⫹) (⫹) (⫺) (⫹) (⫹) (⫺) (⫹) (⫺) (⫹) (⫹) (⫹) (⫹) (⫹) (⫺) (⫹) (⫹) (⫹) (⫹)
4 3 6 4 5 4 5 4 4 5 7 5 4 6 5 3 4 5 4 5
Status (F/U period, month) Alive (61) Alive (24) Alive (20) Dead, local rec (7) Alive (18) Alive (17) Dead, lung mets (7) Dead, lung ca (9) Alive (6) Alive (37) Alive (29) Alive (12) Dead, lung mets (24) Dead, local rec (17) Dead, para-aortic LN (3) Alive (19) Dead, local rec (22) Alive (10) Dead, persistent disease (6) Dead, persistent disease (4)
Note. FIGO, International Federation of Gynecology and Obstetrics; RT, radiation therapy; F/U, follow-up. SqCCa, squamous cell carcinoma; AdenoCa, adenocarcinoma; AdenoSqCa, adenosquamous carcinoma. local rec, local recurrence; lung mets, lung metastases; para-aortic LN, para-aortic lymph node. n.a., not available.
in patients who underwent a curative course of radiation therapy (RT) for treatment of uterine cervical cancer. PATIENTS AND METHODS Patients Twenty patients (age range 26 – 82, mean 52.1 years old) with histologically proven cervical cancer were enrolled in this study. The patients’ profiles are summarized in Table 1. Among them, 10 patients, who had been previously evaluated [9], were reevaluated before and after radiation therapy. There were 5 patients with International Federation of Gynecology and Obstetrics (FIGO) stage I B, 1 with stage II A, 8 with stage II B, 5 with stage III B, and 1 with IV A disease. The histopathologic types were squamous cell carcinoma in 17 patients, adenosquamous cell carcinoma in 1 patient, and adenocarcinoma in 2 patients. Of the 20 patients, 19 had newly diagnosed cervical cancer, and 1 had recurrent cervical cancer after surgical treatment. In addition to a baseline scan prior to radiation treatment, all patients underwent a second PET study 3 to 7 months following therapy (mean ⫽ 4.6 months). Written informed consent was obtained from all patients for the study, which was approved by the institutional review board and conducted under the guidelines for a physician-sponsored investigational drug application.
Radiation Therapy Patients were treated between August 1994 and August 1999. Therapy consisted of whole pelvic irradiation using the four-field technique (median, 45 Gy, range 20 –70 Gy) followed with one or two intracavitary implants utilizing FletcherSuit applicators (16 patients) or an interstitial implant (3 patients). One patient received external RT alone (74 Gy). The total dose prescribed to patients receiving intracavitary implants was in the range of 77–90.1 Gy (median, 85 Gy), and the patients receiving interstitial radiation received a total of 75 Gy minimal tumor dose. Ten patients received cisplatin-containing chemotherapy concurrent with radiation, 5 received the radiosensitizer bromodeoxyuridine according to a previously published protocol [13], and 5 patients received radiation alone. After therapy, patients had been followed with history and physical examination, including bimanual pelvic examination and cervical cytologic smears, every 2–3 months before the follow-up PET scan was done. PET Scanning FDG was produced by a standard nucleophilic fluorination method as previously described [14]. FDG-PET scans were performed with a Model 921 Exact (47 scanning planes, 15-cm longitudinal field of view) scanner (CTI, Knoxville, TN, dis-
291
FOLLOW-UP OF CERVICAL CANCER WITH FDG-PET
tributed by Siemens Medical System, Iselin, NJ). The reconstructed x–y resolution with the Hanning filter cutoff value of 0.3 was approximately 1.2 cm full width at half maximum. All patients fasted for at least 4 h before PET scanning. Transmission scan (at least 10 min) was obtained using a 68Ge rod source for the purpose of attenuation correction, followed by intravenous tracer injection (approximately 370 MBq FDG). Sequential dynamic scans at the level of suspected tumors were obtained through 60 min. These dynamic scans were generally six 10-s scans, three 20-s scans, two 90-s scans, one 5-min scan and five 10-min scans. Then, some patients were instructed to void, and the postvoid emission and transmission scans, each of 10-min duration, were obtained over the pelvis to minimize bladder activity. Generally, two contiguous levels of scanning were imaged from the level of the symphysis pubis to the level of the middle abdomen. In general, CT was not available at the time of PET and was not used to guide patient positioning for PET scans. Software used to perform the postinjection emission and transmission scans was provided by Siemens/CTI. CT Scanning In 17 of 20 cases, CT scans were obtained with 10-mm-thick contiguous axial sections with a GE-9800 scanner (Hi-lite Advantage; GE Medical Systems, Milwaukee, WI). Intravenous contrast material (Omnipaque-300; Winthrop Pharmaceuticals, New York, NY) and oral contrast material (1.5% Hypaque solution; Winthrop Pharmaceuticals) were used. Three cases had CT scans performed in outside institutions. Image Analysis All PET and CT scans were interpreted separately. PET images were analyzed on an interactive computer display by two observers who were blinded to other imaging results and clinical data. Any obvious foci of increased FDG uptake were evaluated using transaxial, sagittal, and coronal displays and compared between prevoid and postvoid scans, if available. Thus, FDG uptake in the last (50 – 60 min) frames of dynamic scans were assessed visually, and the degree of abnormality of FDG accumulation was classified into five grades by consensus interpretations: 0 ⫽ normal, 1 ⫽ probably normal, 2 ⫽ equivocal, 3 ⫽ probably abnormal, and 4 ⫽ definitely abnormal for both sets of images. Cases with grade 3 or 4 uptake were considered positive for disease. For a semiquantitative index of FDG uptake in tumors, standardized uptake value (SUV), which is the decay-corrected tissue activity divided by the injected dose per patient body weight, was calculated for areas suspicious for recurrence. Then, those values were corrected by predicted lean body mass, as described previously [15], and SUV-lean (SUL) was used for quantitative analyses. All SULs were calculated from regions of interest (ROIs) that were placed by means of automated algorithm on the maximal area (16 pixels in size) of FDG uptake within a larger ROI covering
the tumors. No correction was applied for partial volume effects. In addition, the percentage of residual activity (%RA) for the posttherapy SUL was also evaluated as a percentage of the pretherapy SUL. Diagnostic accuracy was evaluated by comparing the PET results with final diagnoses, determined by histopathological examination (n ⫽ 4) or clinical follow-up (n ⫽ 16), including radiological findings. The follow-up period was 6 to 61 months (mean, 23.0 months) for living patients and 3 to 24 months (mean, 11.1 months) for dead patients. RESULTS PET findings were shown in Table 2. Among all 20 patients (pts) primary cervical tumors were detected visually by PET (score 3 or 4) prior to RT in 17 cases (85.0%), and SULs ranged from 4.17 to 14.23, with a mean of 7.77 ⫾ 3.11. After radiation therapy with or without chemotherapy for radiosensitization, 8 of 20 patients were diagnosed as positive recurrent or disease on the basis of physical examinations, cytology, and radiological findings, e.g., local recurrence (n ⫽ 5), paraaortic nodes (n ⫽ 1), or distant metastases only (n ⫽ 2). However, one paraaortic node and two distant metastatic lesions were out of scan of PET images, so five local lesions were further evaluated. In the posttherapeutic scans, uptakes of grades 0 to 2 were observed in 9 pts. The range of posttherapy SULs of the remaining 11 pts was 1.83 to 8.50 with a mean of 3.35 ⫾ 1.84. Of these 11 pts, we made correct visual diagnoses for 5 pts with recurrence, with %RA values ranging from 32.5 to 74.1 (mean ⫽ 57.9 ⫾ 16.8). Recurrence in these patients was confirmed by biopsy for 4 pts and by increasing radiological tumor size for 1 pt. However, 6 pts with elevated posttreatment SULs also had high %RAs ranging from 18.2 to 72.7 (mean ⫽ 43.0 ⫾ 18.3). PET was able to provide additional clinical information for a representative case shown in Fig. 1. The relationship between visual scoring and the current status of the patients is shown in Table 3. A statistically significant difference was observed between the recurrent and nonrecurrent cases (P ⬍ 0.05) with regard to their visual score, but there was no difference between the two groups based on the current status of the patient. In semiquantitative analysis, there was no significant difference between the positive and negative local recurrent groups with regard to some quantitative parameters, such as SUL, lesions-to-gluteus muscle ratio, and %RA. The percentage of recurrent positive patients based on the visual scoring is shown in Fig. 2. Based on visual analysis, overall sensitivity, specificity, and accuracy were 100, 60, and 70%, respectively (Table 4). Receiver operating characteristic (ROC) analysis was done using the binormal curve-fitting routine of the CORROC2 program developed at the University of Chicago by Metz et al. [16]. There were no statistically significant differences among areas under the ROC curves using these parameters (Fig. 3).
292
NAKAMOTO ET AL.
TABLE 2 PET Results Primary tumor Before therapy
Patient No.
Visual score
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
4 4 4 4 1 0 4 4 4 0 4 4 4 4 4 4 4 4 4 4
After therapy
SUL
L/M ratio
Postvoid scan
Visual score
PET results
SUL
4.17 4.55 5.32 5.15 n.a. n.a. 10.85 11.84 14.23 n.a. 10.16 6.08 4.77 6.58 10.36 6.34 5.23 8.64 6.41 11.47
4.48 7.11 7.09 7.15 n.a. n.a. 13.23 7.95 12.82 n.a. 15.63 6.54 4.92 8.55 11.26 8.81 5.51 10.16 8.90 9.25
Yes Yes Yes Yes Yes No No Yes Yes Yes Yes Yes Yes Yes No No Yes Yes Yes Yes
3 4 3 3 2 0 1 1 3 1 1 0 2 3 0 4 3 3 4 4
FP FP FP TP TN TN TN TN FP TN TN TN FP TP TN TN TP FP TP TP
2.17 3.31 1.83 3.69 n.a. n.a. n.a. n.a. 2.59 n.a. n.a. n.a. n.a. 2.14 n.a. 2.45 2.73 3.65 3.77 8.50
Nodal mets
L/M ratio
Postvoid scan
2.17 2.88 2.23 3.08 n.a. n.a. n.a. n.a. 3.65 n.a. n.a. n.a. n.a. 2.85 n.a. 3.55 2.97 4.10 5.39 9.66
Yes Yes Yes No Yes No No Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes No Yes Yes
%RA
Before therapy, visual score, SUL
After therapy, visual score, SUL
52.0 72.7 34.4 71.7 n.a. n.a. n.a. n.a. 18.2 n.a. n.a. n.a. n.a. 32.5 n.a. 38.6 52.2 42.2 58.8 74.1
0, 0, 0, 3, 1, 0, 0, 0, 3, 0, 0, 0, 3, 0, 3, 1, 2, 4, 0, 2,
0, 0, 0, 1, 2, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 4, 0, 0,
n.a. n.a. n.a. 2.65 n.a. n.a. n.a. n.a. 6.63; 3, 3.73 n.a. n.a. n.a. 2.12 n.a. 3.59; 4, 3.65 n.a. n.a.; 2, n.a. 8.12; 3, 1.98 n.a. n.a.
n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a.; 0, n.a. n.a. n.a. n.a. n.a. n.a. n.a.; 0, n.a. n.a. n.a.; 0, n.a. 6.85; 0, n.a. n.a. n.a.
Note. L/M ratio, lesion-to-muscle ratio; %RA, percentage of residual activity; n.a., not available. TP, true positive; TN, true negative; FP, false positive; FN, false negative.
DISCUSSION Our study expands upon the small, but promising literature regarding the role of FDG-PET in uterine cervical cancer. Although the number of patients was limited, our data indicate that all recurrent cancers were detected by PET, consistent with the preliminary report by Umesaki et al., who showed that all recurrent foci had relatively high SUV compared to normal cervix [11, 12]. Specificity, however, was lower than has been reported under some other conditions, such as posttreatment lung cancer [17]. In addition, semiquantitative analysis could not differentiate between recurrent and nonrecurrent patients and was not useful in predicting prognostic status, at least in this small patient cohort. There are nine cases where PET and CT studies were done within approximately 3 weeks of each other (mean 13.1 days). Both PET and CT showed true negative results in five of these nonrecurrent cases. However, in four of the recurrent cases, CT yielded two positive, one negative, and one inconclusive result, whereas PET revealed positive findings for all cases. Thus, additional information by PET was obtained in only two patients in our study, but the clinical contribution of this modality
would likely increase with broader application of the technique during the follow-up of recurrent cervical cancer. With regard to the relatively low specificity observed among our patient population, it is well known that FDG accumulates in inflammatory foci, not just malignant tumors, which can lead to false-positive findings [18]. Furthermore, it has been shown that once irradiation is initiated, FDG uptake can rise in the treated and dying tumors [19]. In this study, posttreatment scans were performed 4 to 5 months after radiation therapy was completed, but it is possible that inflammatory processes were continuing in these localized areas in the cases with falsepositive results. A significant difference in the time interval between radiation therapy and posttreatment scan was not observed (mean 4.6 months) between recurrent and nonrecurrent groups. In other cases, physiological uptake in the ureters or bladder and alimentary tract may have led to false-positive findings. In 32 of the 40 scans, we obtained postvoid images, which made it easier to evaluate pelvic lesions, especially when images were reconstructed using a filtered backprojection technique. However, physiological accumulation observed in the urinary system seemed not to contribute to the falsepositive results, since five of six false-positive diagnoses were
293
FOLLOW-UP OF CERVICAL CANCER WITH FDG-PET
FIG. 2. Percentages of recurrent positive patients based on the visual scoring grade. As would be expected, higher grading scores correlated with higher percentages of pts with recurrences.
FIG. 1. A 47-year-old female, recurrent cervical cancer (patient No. 14). PET (a, left) shows intense uptake behind the bladder (arrow, SUL ⫽ 6.58), corresponding to the primary cervical cancer. At 6 months after irradiation, focal uptake was observed (a, right: arrow, SUL ⫽ 2.14), indicating local recurrence. However, definite diagnosis of local recurrence could not be obtained by CT performed following the PET scan (b). The recurrence was confirmed by biopsy.
made in postvoid images. Physiological uptake in the gut remains a difficult problem, which, unlike that of urinary system, cannot be eliminated simply by voiding. In case of continuous linear accumulation, it may be differentiated from pathological uptake, but in other cases the cause of the increased uptake in this region remains enigmatic. Further study is needed to avoid both false-positive and false-negative results, which occur when lesions lie in close proximity to organs in the abdomen or pelvis, such as the gut, which can display a high level of physiological uptake.
With regard to the posttherapeutic evaluation of head and neck cancers with PET, Kitagawa et al. reported that a pretreatment PET scan was useful in predicting the response to treatment and that posttreatment scans were helpful in detecting residual viable cells [20]. Rege et al. described how pretreatment PET findings might have prognostic implications for patient outcome following radiation therapy [21]. Similarly, Anzai et al. reported PET to be significantly more accurate than MRI in detecting the presence or absence of recurrent squamous cell cancers [22]. Their separation was made on a qualitative basis only. However, in our study of uterine cervical cancer, semiquantitative analysis using SUL or %RA showed no clear correlation to the prognostic outcome of the patients. Mean values of SUL or lesion-to-muscle ratio after therapy and %RA in recurrent cases tended to be higher than those in negative cases, but there were no statistically significant differences between the two groups. It is not yet possible to differentiate between recurrent and nonrecurrent patients simply by referring to these values. TABLE 4 Diagnostic Accuracy of PET Recurrence or persistent disease
TABLE 3 Visual Scores in Posttreatment Scan Visual score
0
1
2
3
4
Negative local recurrence (n ⫽ 15) Positive local recurrence (n ⫽ 5) Alive (n ⫽ 11) Dead (n ⫽ 9)
3 0 2 1
4 0 2 2
2 0 1 1
4 3 4 3
2 2 2 2
Note. The statistically significant difference was found between the nonrecurrent and recurrent cases (P ⫽ 0.03), but there was no significant difference between alive and dead cases (Mann–Whitney U test).
PET positive PET negative Total Sensitivity Specificity Positive predictive value Negative predictive value Accuracy
Positive
Negative
5 0 5
6 9 15 100.0% 60.0% 45.5% 100.0% 70.0%
294
NAKAMOTO ET AL.
REFERENCES 1. Estape R, Angioli R. Surgical management of advanced and recurrent cervical cancer. Semin Surg Oncol 1999;16:236 – 41. 2. Krebs HB, Helmkamp BF, Sevin BU, Poliakoff SR, Nadji M, Averette HE. Recurrent cancer of the cervix following radical hysterectomy and pelvic node dissection. Obstet Gynecol 1982;59:422–7. 3. Ito H, Shigematsu N, Kawada T, Kubo A, Isobe K, Hara R, Yasuda S, Aruga T, Ogata H. Radiotherapy for centrally recurrent cervical cancer of the vaginal stump following hysterectomy. Gynecol Oncol 1997;67:154 – 61. 4. Yamashita Y, Harada M, Torashima M, Takahashi M, Miyazaki K, Tanaka N, Okamura H. Dynamic MR imaging of recurrent postoperative cervical cancer. J Magn Reson Imaging 1996;6:167–17.
FIG. 3. ROC curves for FDG-PET using semiquantitative parameters, SUL at posttreatment scan (bold line), lesion-to-muscle (L/M) ratio (normal line), and %RA (thin line). There were no statistically significant differences between values of area under the ROC curves (A z). A z (SUL) 0.80, A z (L/M ratio) 0.71, and A z (%RA) 0.74.
Uterine cervical cancer is a locally invasive and highly metastatic tumor. Distant metastases to the lung and other areas occur often, and unexpected metastases to atypical locations are not uncommon [23]. In our cases, three patients died from lung metastases (n ⫽ 2) or persistent aortic nodal metastasis (n ⫽ 1) despite good local control of the primary tumor. Recent advances in PET camera technology have enabled us to evaluate the whole body and could be used to assess metastatic spread in patients with cervical cancer. Finally, tumor markers such as squamous cell carcinoma (SCC) antigen are widely used for evaluating patients’ clinical conditions during the follow-up period [24 –26]. A blood test is a noninvasive technique that is easily repeatable. With SCC, reelevation of the tumor antigen was shown to correlate closely with early relapse. However, the rising of serum tumor markers provides no information about the location or extent of the disease and is not useful in all cases. It is also sometimes difficult to identify recurrent lesions with standard imaging modalities, such as CT and MR. As the high sensitivity of PET was shown in our study, Umesaki et al. demonstrated a case report where PET detected a recurrent tumor even when the tumor marker remained normal [11]. Although PET would not completely replace the monitoring of tumors using these modalities, this noninvasive technique could come to have a greater role for screening patients during follow-up, due to its high sensitivity. In summary, FDG-PET is a highly sensitive tool for predicting recurrent or persistent uterine cervical cancer. Despite its current relatively low specificity, it may be helpful for screening of follow-up patients undergoing a curative course of radiotherapy. Further investigations are needed to reduce falsepositive results.
5. Kinkel K, Ariche M, Tardivon AA, Spatz A, Castaigne D, Lhomme C, Vanel D. Differentiation between recurrent tumor and benign conditions after treatment of gynecologic pelvic carcinoma: value of dynamic contrast-enhanced subtraction MR imaging. Radiology 1997; 204:55– 63. 6. Strauss LG, Clorius JH, Schlag P, Lehner B, Kimmig B, Engenhart R, Marin-Grez M, Helus F, Oberdorfer F, Schmidlin P. Recurrence of colorectal tumors: PET evaluation. Radiology 1989;170:329 –32. 7. Delbeke D, Vitola JV, Sandler MP, Arildsen RC, Powers TA, Wright JK Jr, Chapman WC, Pinson CW. Staging recurrent metastatic colorectal carcinoma with PET. J Nucl Med 1997;38:1196 –201. 8. Flamen P, Stroobants S, Cutsem EV, Dupont P, Bormans G, Vadder ND, Penninckx F, Hoe LV, Mortelmans L. Additional value of whole-body positron emission tomography with fluorine-18-2-fluoro-2-deoxy-D-glucose in recurrent colorectal cancer. J Clin Oncol 1999;17:894 –901. 9. Sugawara Y, Eisbruch A, Kosuda S, Recker BE, Kison PV, Wahl RL. Evaluation of FDG PET in patients with cervical cancer. J Nucl Med 1999;40:1125–31. 10. Rose PG, Adler LP, Rodriguez M, Faulhaber PF, Abdul-Karim FW, Miraldi F. Positron emission tomography for evaluating para-aortic nodal metastasis in locally advanced cervical cancer before surgical staging: a surgicopathologic study. J Clin Oncol 1999;17:41–5. 11. Umesaki N, Tanaka T, Miyama M, Tokuyama O, Kawamura N, Ogita S, Kawabe J, Okamura T, Koyama K, Ochi H. Early diagnosis and evaluation of therapy in postoperative recurrent cervical cancers by positron emission tomography. Oncol Rep 2000;7:53– 6. 12. Umesaki N, Tanaka T, Miyama M, Kawabe J, Okamura T, Koyama K, Ochi H, Ogita S. The role of 18F-fluoro-2-deoxy-D-glucose positron emission tomography ( 18F-FDG-PET) in the diagnosis of recurrence and lymph node metastasis of cervical cancer. Oncol Rep 2000;7:1261– 4. 13. Eisbruch A, Robertson JM, Johnston CM, Tworek J, Reynolds KR, Roberts JA, Lawrence TS. Radiation alternating with bromodeoxyuridine for advanced uterine cervix cancer: a phase I and drug incorporation study. J Clin Oncol 1999;17:31– 40. 14. Toorongian SA, Mulholland GK, Jewett DM, Bachelor MA, Kilbourn MR. Routine production of 2-deoxy-2-[ 18F]fluoro-D-glucose by direct nucleophilic exchange on a quatenary 4-aminopyridinium resin. Int J Rad Appl Instrum B 1990;17:273–9. 15. Zasadny KR, Wahl RL. Standardized uptake values of normal tissues at PET with 2-[fluorine-18]-fluoro-2-deoxy-D-glucose: variations with body weight and a method for correction. Radiology 1993;189:847–50. 16. Metz CE, Wang PL, Kronman HB. A new approach for testing the significance of differences between ROC curves measured from correlated data. In: Deconinck F, editor. Information processing in medical imaging. The Hague, The Netherlands: Martius Nijhoff, 1984:432– 45.
FOLLOW-UP OF CERVICAL CANCER WITH FDG-PET 17. Patz EF Jr, Lowe VJ, Hoffman JM, Paine SS, Harris LK, Goodman PC. Persistent or recurrent bronchogenic carcinoma: detection with PET and 2-[F-18]-2-deoxy-D-glucose. Radiology 1994;191:379 – 82. 18. Strauss LG. Fluorine-18 deoxyglucose and false-positive results: a major problem in the diagnostics of oncological patients. Eur J Nucl Med 1996;23:1409 –15. 19. Higashi K, Clavo AC, Wahl RL. In vitro assessment of 2-fluoro-2-deoxyD-glucose, L-methionine and thymidine as agents to monitor the early response of a human adenocarcinoma cell line to radiotherapy. J Nucl Med 1993;34:773–9. 20. Kitagawa Y, Sadato N, Azuma H, Ogasawara T, Yoshida M, Ishii Y, Yonekura Y. FDG PET to evaluate combined intra-arterial chemotherapy and radiotherapy of head and neck neoplasms. J Nucl Med 1999;40: 1132–7. 21. Rege S, Safa AA, Chaiken L, Hoh C, Juillard G, Withers HR. Positron emission tomography: an independent indicator of radiocurability in head and neck carcinomas. Am J Clin Oncol 2000;23:164 –9.
295
22. Anzai Y, Carroll WR, Quint DJ, Bradford CR, Minoshima S, Wolf GT, Wahl RL. Recurrence of head and neck cancer after surgery or irradiation: prospective comparison of 2-deoxy-2-[F-18]fluoro-D-glucose PET and MR imaging diagnoses. Radiology 1996;200:135– 41. 23. Fulcher AS, O’Sullivan SG, Segreti EM, Kavanagh BD. Recurrent cervical carcinoma: typical and atypical manifestations. Radiographics 1999; 19:S103– 6. 24. Abe A, Nakano T, Morita S, Oka K. Clinical evaluation of serum and immunohistochemical expression of SCC and CA19-9 in radiation therapy for cervical cancer. Anticancer Res 1999;19:829 –36. 25. Juang CM, Wang PH, Yen MS, Lai CR, Ng HT, Yuan CC. Application of tumor markers CEA, TPA, and SCC-Ag in patients with low-risk FIGO stage IB and IIA squamous cell carcinoma of the uterine cervix. Gynecol Oncol 2000;76:103– 6. 26. Suzuki Y, Nakano T, Ohno T, Abe A, Morita S, Tsujii H. Serum CYFRA 21-1 in cervical cancer patients treated with radiation therapy. J Cancer Res Clin Oncol 2000;126:332– 6.