Evaluation of ¹¹¹In-DTPA-Folate as a Receptor-Targeted Diagnostic Agent for Ovarian Cancer: Initial Clinical Results

Study Protocol.

The study protocol was approved by the institutional review board at each participating institution. Each subject gave written, informed consent before inclusion in the study. Eligible subjects were women aged 18 y or older with one of the following: an ovarian mass, demonstrated by CT or ultrasonography, that was suspected to represent ovarian cancer and for which pathologic evaluation after laparotomy, laparoscopy, or percutaneous biopsy was planned; suspected recurrent ovarian or endometrial cancer, demonstrated by CT or ultrasonography, and for which pathologic evaluation after laparotomy, laparoscopy, or percutaneous biopsy was planned; or pathologically proven ovarian or endometrial cancer (primary, metastatic, or recurrent), with at least one site of measurable disease demonstrated by CT or MRI. Pregnant and lactating patients were excluded, as were those with a history of congestive heart failure, unstable angina, or recent myocardial infarction (within 6 mo); septicemia, severe infection, or acute hepatitis; inadequate bone marrow, renal, or hepatic function; known or suspected kidney disease; or a Zubrod performance score greater than 2. In addition, patients were ineligible if they had taken nonsteroidal anti-inflammatory medication within the previous 5 d, could not stop taking folic acid supplements 2 d before imaging, or had any preexisting medical condition that precluded the intravenous administration of fluids required by the protocol.

Patients were studied in 8 groups that received varying amounts of DTPA-folate, ¹¹¹In, and preinjected folic acid (Table 1). Preliminary estimates of radiation absorbed dose were made from mouse biodistribution data and the results of a prior study of 3 healthy women who received a 3.7-MBq dose of ¹¹¹In-DTPA-folate (30). The initial group of 3 patients in the present study received a 37-MBq dose of ¹¹¹In-DTPA-folate and underwent limited imaging, confirming that an administered dose of 185 MBq would be acceptable for use in the remaining 7 groups. The radiation absorbed dose estimates calculated from the human biodistribution data of the present study will be reported later.

¹¹¹In-DTPA-folate (Fig. 1) was administered by bolus intravenous injection through a peripheral venous catheter. Before administration of the radiopharmaceutical, a venous blood sample was drawn for determination of serum folate. An infusion of 1 L of 0.9% saline solution began 30 min before radiopharmaceutical administration and continued for 5 h. Patients who were pretreated with free folic acid were given the injection intravenously 1–2 min before injection of the radiopharmaceutical. Patients were monitored over the entire course of the imaging study for signs of adverse reactions.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 1.

Structural formula of ¹¹¹In-DTPA-folate.

Imaging Procedure.

Anterior and posterior whole-body planar scintigrams were obtained at 30 min, 4 h, and (for the first 19 patients) 24 h after injection with a dual-detector gamma camera equipped with medium-energy parallel-hole collimators. Whole-body scanning was performed at a maximum speed of 8 cm/min with an optimal total imaging time of 30 min. SPECT images of the pelvis were obtained at the 4- and 24-h imaging times as well. SPECT images were collected using a dual- or triple-head camera with 120 total projections over 360°. The scanning duration was 30 s per step. Images were acquired in a 128 × 128 matrix. SPECT images were reconstructed by filtered backprojection; a Butterworth filter (order, 5; cutoff, 0.035 cycles per centimeter) or similar Wiener filter was used.

The images were interpreted, independently, by 2 nuclear medicine physicians and were evaluated for abnormal uptake of ¹¹¹In-DTPA-folate. In the first reading, the physicians were unaware of clinical history, including CT, sonographic, or MR images. In the second reading, designed to mirror the typical clinical setting, both physicians were given access to clinical information and prior imaging studies that would normally be available during presurgical scan interpretation. Images were scored as showing definitely increased uptake (++), faint uptake (+), or no abnormal uptake (0), with the positive interpretation for malignancy being based on focal increased uptake of radiotracer at sites not corresponding to obvious normal anatomic structures.

After the imaging was completed, surgery was performed by a gynecologic oncologist, and resected tissue was examined by a pathologist to confirm the presence or absence of neoplastic tissue. The pathology report was used as the reference standard for comparison with the scintigraphic imaging analysis.

Imaging Parameters.

Planar images were obtained at 30 min primarily for dosimetry evaluation; these images were dominated by renal and urinary bladder activity and were not further evaluated for detection of lesions. Images obtained at 4 and 24 h provided comparable diagnostic information; accordingly, the 24-h imaging time was eliminated after the first 19 patients to simplify the phase II component of the protocol. As predicted from preclinical biodistribution studies in mice (26), uptake of ¹¹¹In-DTPA-folate was evident in the kidneys of all patients. The partial retention of ¹¹¹In-DTPA-folate by the kidneys is presumed to be due to the presence of folate receptors in the proximal tubules (24,26,27,31,32). In some patients, increased ¹¹¹In uptake was also observed in the liver and spleen.

In animal studies, tumor-to-background contrast was optimized by coinjection of a low dose of free folic acid with the ¹¹¹In-DTPA-folate (26). In this initial clinical study, images were evaluated qualitatively, with patients who had received a low-dose preinjection of free folate being compared with patients who had not. Although the interpatient variability could not be controlled, preinjection of free folate did appear to reduce tissue background activity; however, the free folate preinjection did not appear to appreciably improve ¹¹¹In-DTPA-folate tumor-to-background contrast. Definitive assessment of the effects of folate loading would require repeated studies of individual patients, which unfortunately was not practical within the limits of the present study.

Newly Diagnosed Pelvic Masses.

The scintigraphic results for 26 patients presenting with a new pelvic mass are summarized in Tables 3 and 4. Seven of the newly detected masses were malignant tumors (6 ovarian and 1 endometrial). The mass in 1 patient was classified as a borderline malignancy, and the remaining 18 were benign pelvic masses. Representative scintigrams from patients with malignant and benign lesions are shown in Figure 2.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 2.

Anterior and posterior planar scintigrams obtained 4 h after administration of ¹¹¹In-DTPA-folate. The 2 leftmost images, of patient 15, demonstrate increased tracer uptake (arrow) corresponding to solid tumor component in rim of poorly differentiated adenocarcinoma of left ovary. The 2 rightmost images, of patient 9, show region of focally decreased activity (arrow) corresponding to simple serous benign cyst. (Foci of increased activity between legs of each patient are reference sources used for dosimetry calculations.)

Both readers scored all 7 malignant tumors as having increased uptake of ¹¹¹In-DTPA-folate (sensitivity, 100%). The masked and unmasked interpretations of these studies were identical (Table 4). The 1 borderline-malignant tumor (patient 6) was scored as showing faint uptake by reader A on the masked and unmasked readings. Reader B scored the borderline-malignant tumor as showing no uptake on the masked reading and did not complete the unmasked reading.

Of the 18 benign masses, 11 were scored by both readers, on both masked and unmasked readings, as showing no uptake (Table 4). Two additional masses (patients 27 and 29) were scored as showing no uptake by one or both readers. One benign mass (patient 12) was scored by both readers as having definitely increased uptake of ¹¹¹In-DTPA-folate on SPECT images, on both masked and unmasked analyses. This mass was classified pathologically as adenomatoid tumor/hilar Leydig cell hyperplasia. Another benign mass (patient 23) was scored consistently by both readers as showing faint uptake. Variable scoring was recorded on the remaining 3 benign masses. For patient 17, with a mature cystic teratoma, both readers scored the images as showing definitely increased uptake on the masked readings but as showing no uptake when the reading was unmasked. Only 2 benign masses were scored differently by the readers (patients 22 and 34; Table 3).

The specificities for the masked analysis of the ¹¹¹In-DTPA-folate images were 78% and 75% for readers A and B, respectively (Table 4). For the unmasked readings, their corresponding specificities were 82% and 81%, respectively.

The agreement between the nuclear medicine physicians on the readings of newly diagnosed masses was excellent (92%), with all but 4 sets of images being scored identically, including both masked and unmasked readings (43/47 readings; 5 incomplete image sets excluded). The overall accuracies of ¹¹¹In-DTPA-folate imaging for distinguishing malignant and benign lesions in patients with newly diagnosed ovarian masses, excluding the borderline malignancy, were 83% (40/48) and 89% (41/46) for masked and unmasked readings, respectively.

Recurrent Malignancies.

Seven patients with suspected recurrent ovarian or endometrial cancers were studied (Table 5), and subsequent pathologic examination showed all 7 to have recurrent malignant disease (5 ovarian and 2 endometrial). In the masked readings, only 2 of the malignancies were scored as showing increased uptake. The remaining 5 image sets were scored as showing no or faint uptake. When the readers were later unmasked, they scored 4 of the malignancies as showing increased uptake, 2 as showing slight uptake, and only 1 as showing no uptake. The average sensitivity values of both readers for recurrent ovarian and endometrial tumors were 38% and 85% for masked and unmasked analyses, respectively (Table 6). The agreement between the nuclear medicine physicians on the readings of recurrent ovarian and endometrial masses was quite high (92%), with all but 1 set of images having been scored identically, including both masked and unmasked readings (11/12 readings; 2 incomplete image sets excluded).

One patient with recurrent ovarian cancer (patient 33) had a 4 × 6 cm mass seen on CT and shown by pathology to represent poorly differentiated adenocarcinoma. This patient’s ¹¹¹In-DTPA-folate images were interpreted as negative. Of note, however, this patient was under treatment and, uniquely, was in the midst of a third cycle of chemotherapy at the time of the imaging study.

Another patient (patient 31) presented with a recurrent 2 × 3 cm mass adjacent to the spleen. This mass was not included in the field of view of the first SPECT study. The later SPECT study included the upper abdomen and was interpreted as showing definitely increased uptake on the unmasked readings.