Role of 18F FDG PET/CT in evaluation of post-operative ovarian carcinoma

Abstract

Background

We evaluated the sensitivity, specificity and accuracy of fluorine-18-fluorodeoxyglucose with positron emission tomography/computed tomography (18F-FDG-PET/CT) in a group of patients with suspicion of ovarian cancer recurrence. It is considered a diagnostic dilemma, particularly in the 2 years following first-line therapy. CA125 serum levels computed tomography (CT) and other modalities are used during routine follow-up. These traditional modalities could provide a significant number of false-negative or equivocal results even in the presence of elevated CA125 levels. So the performance of 18F-FDG-PET/CT is essential for the optimal diagnosis of recurrence and treatment planning.

Results

Studying PET/CT behaviour in the detection of ovarian cancer recurrence, 18F-FDG-PET/CT had an accuracy of 98% with sensitivity and specificity of 98% and 100%, respectively. 18F-FDG-PET/CT had a PPV of 100% and NPV of 83%. While studying the CA125 level (> 35 U/ml) to detect the ovarian cancer recurrence during patient follow-up, the CA125 level had an accuracy of 50% with a sensitivity ratio and specificity ratio of 47% and 80%, respectively. CA-125 level had a PPV of 95% and NPV of 14%. In comparison between conventional CT and PET/CT studies, the PET/CT diagnosed local tumor recurrence in 16 patients (32%), while CT scan diagnosed local tumor recurrence in only 3 patients (6%), and PET/CT detected peritoneal recurrence in 34 patients (68%). CT scan found peritoneal deposits in 11 patients (22%), also the PET/CT showed suspicious abdominal LNS in 22 patients (44%) while, CT scan showed suspicious abdominal LNS in 4 patients (8%), and PET/CT showed suspicious pelvic LNS in 16 patients (32%). CT scan showed suspicious pelvic LNS in 7 patients (14%). PET/CT detected distant organ metastases in 18 patients (36%). CT scan detected distant organ metastasis in only 8 patients (16%). Comparison between CT and PET/CT in 32 follow-up cases for the detection of local tumor recurrence, peritoneal deposits, suspicious abdominal/pelvic LNs and distant organ metastasis. There was a statistically significant difference between CT and PET/CT the end results (p < 0.0001, p = 0.0047, p = 0.001, p = 0.03).

Conclusions

18F-FDG-PET/CT is more sensitive in detection and localization of ovarian cancer recurrence and more superior than the other imaging modalities.

Background

Significant progress has been made in the treating ovarian carcinoma. However, within 5 years of diagnosis, most of the patients with advanced OC experience disease recurrence and they received 2nd line therapy and several lines of treatment [1, 2].

Surgery, chemotherapy, and local radiation are the widely identified treatment options for OC. For deciding the proper treatment plane and specifying patients for a certain treatment option, early detection of recurrent OC is essential. Accompanied by contrast-enhanced computed tomography (CECT), magnetic resonance imaging (MRI), or positron emission tomography–magnetic resonance imaging (PET–MRI) with CA125 measurement, is advisable to be conducted for more precise radiological evaluation [3].

It was suggested that positron emission tomography (PET) with fluorine-18-fluorodeoxyglucose (FDG) can overcome the previously mentioned limitations. It provides combined benefits of the functional and anatomical images as it localizes the anatomical areas of increased FDG uptake and accurately detects or excludes cancer recurrence [4].

The efficiency of combined PET/CT employing FDG in post-treatment surveillance of OC has been reported [6]. However, limited data on the diagnostic accuracy of 18F-FDG-PET/CT in suspected recurrent epithelial OC with non-disseminated lesions are available, which may indicate for secondary cytoreduction [7].

The purpose of this work was to assess the diagnostic performance of 18F-FDG PET/CT for detecting OC recurrence in patients with high levels of serum CA-125 and patients who are clinically suspected to have OC recurrence.

Methods

We retrospectively reviewed the clinical and radiological records of 50 female patients with suspected recurrent ovarian cancer (OC) and who underwent PET/CT scan in our institute (Siemens biographTruePoint 64 PET/CT) from October 2018 to October 2020 have been reviewed in which PET-CT examination was followed by diagnostic contrast-enhanced CT examination and permits the acquisition of co-registered CT and PET images in one session.

Patients with a history of managed OC (with suspected recurrence) at any age were included in the study provided the following: (1) previously treated from OC by combined surgery and chemotherapy or chemotherapy alone with complete radiographic responses to treatment, (2) presented with elevated CA125 or clinical signs and symptoms suspicious for recurrence. However, patients who were diagnosed with OC and didn’t previously receive treatment or patients with other concurrent malignancy were excluded from the study.

Patient preparation

• Obtaining an informed written consent by the patient before PET/CT examination.

• Fasting for 6 h with proper hydration will be advised for at least 4 h before contrast injection.

• Pre-procedural assessment of serum creatinine.

• Measuring patient body weight was for the calculation of the number of contrast media and 18F-FDG administration, measures for contrast allergy will be available.

• The dministration of 1 intravenous cannula in the antecubital vein.

Precautions of PET/CT study

The patients were instructed to avoid any stress activity before the examination and the following injection of the radioisotope to avoid physiologic muscle uptake of FDG and the patient asked to void before scanning. Our strategy for decreasing brown fat providing a controlled temperature (warm) environment for patients before 18F- FDG injection.

Dosage administration

10–20 mCi (370 MBq; approximate dose to patient 3-5 MBq/Kg). 18F-FDG administered for each patient 45–90 min before the examination. This period was referred to as the uptake phase and the necessary amount of time for the FDG to be adequately bio-distributed and transported into the patient’s cells. Patients asked to rest in a quiet room, devoid of distractions, and they also asked to keep their movements, including talking, at an absolute minimum. This minimises physiologic uptake of FDG into skeletal muscle, which can confound the interpretation of the scan. Patients should be comfortable and relaxed.

Patient position

The patients were introduced to the PET-CT machine lying in the supine position with head fixation and arms up.

Examination time

We conducted low dose non-enhanced CT scan first for attenuation correction, then a whole body PET study followed by diagnostic enhanced whole-body CT scan. The whole study takes approximately 25–35 min.

CT Technique

• For a typical whole body PET/CT study (neck, chest, abdomen, and pelvis), scanning began at the level of the skull base and extended caudally to the level of the upper thighs. The total length of CT coverage was equal to the integral number of bed positions scanned during the acquisition of PET data. The study was performed with the patient breathing quietly.

• Typical scanning parameters for low dose attenuation correction CT would be KV 120, MA 100, collimator width of 5.0 mm, pitch of 1, gantry rotation time of 0.75 s, and field of view of 50 cm. The helical data was retrospectively reconstructed at 1 mm interval.

• Typical scanning parameters for high dose diagnostic CT would be KV 120, MA 300, collimator width of 5.0 mm, pitch of 0.8, gantry rotation time of 0.75 s, and field of view of 50 cm. The helical data was retrospectively reconstructed at 1 mm interval. The patients wre injected about 125 ml of non-ionic iodinated contrast material using dual syringe Medrad (stellant) automated injector with an injection rate about 4 ml/s.

• The acquisition of post-contrast CT images of the whole body (from the skull vault to the mid-femur level) was done first followed by the acquisition of PET image. CECT images were interpreted first separately, then be fused PET/CT images.

Image interpretation Away to areas of physiologically high FDG uptake, any focus that showed increased FDG uptake compared to the surroundings was considered positive for OC recurrence on PET/CT. Maximum standardized uptake values (SUVmax) lesions with increased FDG uptake were measured on PET/CT fusion images using a region of interest (ROI). For OC recurrence, CECT images were evaluated in axial, coronal, and sagittal planes in terms of local pelvic mass, involved lymph nodes, intra-abdominal implants, and distant organ metastasis.

Standardized uptake value lesions found on PET/CT were considered to be true positive if proven to be malignant by histopathology or by clinical and radiological follow-up for 3 to 6 months when the lesions were not pathologically examined. When PET/CT detected no abnormality, the patients were considered as true negatives if there is no detected disease by radiological and clinical follow up for 3 to 6 months. The patients who were detected with recurrence on clinical and radiological follow-up after they had an unremarkable PET/CT study was considered true negatives.

The SUV is a used as a semiquantitative assessment of the radiotracer uptake from a static (single point in time) PET image. The SUV of a given tissue is calculated as following.

Region of interest activity (KBq/ml) × body weight (kg)/injected dose (MBq)

We use visual assesment and SUV in assessing suspicious lesions or for the follow-up of FDG-avid masses. Usually, malignant tumours have an SUV greater than 2.5–3.0, whereas normal tissues such as the liver, lung, and marrow have SUVs ranging from 0.5 to 2.5. Differences in the SUV were described, such as the glucose-corrected SUV and SUV normalised by surface area or lean body mass. It is helpful to assess the tumor SUV before therapy to assess tumor grade and the treatment response following the therapy. The time interval between injection of the radiotracer and the PET study should be standardised because SUV variability with time has been well documented [5].

Statistical methods SPSS (Statistical Package for the Social Sciences) version 25 was used to code and input the data. Standard deviation, mean, minimum, median, and maximum have been used to represent quantitative data, while numbers and percentages have been used to summarise categorical data. Analyses were conducted in accordance with Galen [8] who described standard diagnostic indexes, including specificity, sensitivity, the negative predictive value (NPV), and the positive predictive value (PPV). We used Chi-square (χ2) test to compare categorical data while for comparing numerical data, an unpaired t-test was applied. When the frequency was expected to be < 5, an exact test was done. Statistically, significant results were considered with a p value < 0.05.

Results

The included 50 females of post-operative follow-up with pathologically proven ovarian carcinoma who were candidates for PET/CT imaging after resection of intra-pelvic malignancies and suspected for recurrence due to elevated tumor markers Their ages ranged between 18 and 77 years, with a mean age of 54.00 ± 13.78 years. We found that female patients aged > 50 years are more susceptible to OC. Metastatic adenocarcinomas represent the most common pathological subtype of OC. (Table 1).

Table 1 Distribution of pathological subtype of all studied cases

The CA125 ranged from 6.4 to 256 u/ml with a mean ± SD of 57.36 ± 55.15 u/ml. Among 50 patients, CA125 Level was normal (< 35) in 28 patients (56%), while it was elevated (> 35) in 22 patients (44%) at the time of the PET/CT scan. (Table 2).

Table 2 Distribution of CA 125 level of all studied cases

Studying CA125 level (> 35 U/ml) in the detection of OC recurrence during patient follow-up, CA125 level had an accuracy of 50% with sensitivity and specificity of 47% and 80%, respectively. CA-125 level had a PPV of 95% and a NPV of 14%. (Table 3) (Fig. 1).

Table 3 Diagnostic performance of serum CA 125 level in OC recurrence

Fig. 1

ROC curve of CA 125 level

In Table 4, we compared CT and PET/CT in terms of the detection of local tumor recurrence, peritoneal deposits, suspicious abdominal or pelvic LNs and distant organ metastasis. There was a statistically significant difference between CT and PET/CT with regard the detection of suspicious lymph nodes (p = 0.0001*). However, there was no statistically significant difference between CT and PET/CT as regards local tumor recurrence, peritoneal deposits, suspicious abdominal lymph nodes and distant organ metastasis.

Table 4 Comparison between CT and PET/CT in detection of local tumor recurrence, peritoneal deposits, regional LNS and distant organ metastasis

Studying PET/CT behaviour in the detection of ovarian cancer recurrence, PET/CT had an accuracy of 98% with sensitivity and specificity of 98% and 100%, respectively. PETCT had a PPV of 100% and a NPV of 83%. (Table 5) (Fig. 2).

Table 5 Diagnostic performance of combined PET/CT in ovarian cancer recurrence

Fig. 2

ROC curve of PET/CT of local recurrence

Thirty-two cases continued to follow-up twice; the first follow-up was after receiving treatment by 6 months, and the second follow-up was after one year. In Table 6, we compared CT and PET/CT in the follow-up cases in terms of the detection of local tumor recurrence, peritoneal deposits, suspicious abdominal or pelvic LNs, and distant organ metastasis. There was a statistically significant difference between CT and PET/CT with regard the detection of peritoneal deposits, suspicious abdominal lymph nodes, suspicious pelvic lymph nodes and the end results (p < 0.0001, p = 0.0047, p = 0.001, p = 0.03). However, there was no statistically significant difference between CT and PET/CT as regards local tumor recurrence and distant organ metastasis. (Table 6).

Table 6 Comparison between CT and PET/CT between 1st and 2nd follow-up cases regarding local recurrence and distant metastasis:

Discussion

The most widely used biomarker in the prediction and management of OC recurrence is the CA125 [9]. However, lack of specificity is the major problem facing the use of serum CA125 in diagnosing OC recurrence. CA125 may be elevated in other malignant or benign diseases such as pancreatic, breast, and endometrial cancers [10].

In literature, the reported CA125 value denoting OC recurrence has 62–94% sensitivity and 91–100% specificity [11, 12]. In this study, elevated CA 125 level (> 35 U/ml) in detecting OC recurrence during patient follow up, CA125 level had an accuracy of 50% with sensitivity and specificity of 47% and 80%, respectively. CA125 level had a PPV of 95% and a NPV of 14%, indicating a moderate diagnostic performance of CA125 level in detecting recurrent ovarian cancer.

In accordance with our results, Hopkins et al. did not find any prognostic benefit for the use of CA125 level alone for detecting of recurrent OC. They recommended the use of CA125 level only for OC surveillance [13].

Also, Gronlund et al. documented that CA125 concentration is not a sufficient independent prognostic factor for recurrence in different cut-off [14].

In contrast, in the Gu et al. meta-analysis, the CA125 recorded the highest specificity (93%) in diagnosing OC recurrence [15].

Yang et al. found that the sensitivity and specificity of serum CA125 for diagnosing epithelial ovarian carcinoma recurrence were 67.39% and 86.79%, respectively [16].

The undetected or inconclusive in cross sectional imaging, 18F-FDG PET/CT plays a crucial role in the confirmation or exclusion of suspected ovarian cancer recurrence. So, when suspicion of radically treatable disease recurrence is confirmed, this study provides additional information in the detection of non-suspected disease, also it altering the initially proposed therapeutic management protocol [17].

In this study regarding 18F-FDG-PET/CT behaviour in the detection of OC recurrence during patient follow-up, 18F-FDG-PET/CT had an accuracy of 98% with sensitivity and specificity of 98% and 100%, respectively. 18F-FDG-PET/CT had PPV of 100% and a NPV of 83%, indicating an excellent diagnostic performance of 18F-FDG-PET/CT level in the detection of ovarian cancer recurrence.

Batra et al. reported that the sensitivity, specificity, PPV, NPV, and accuracy of PET-CT to detect OC recurrence were 90%, 66.7%, 83.7%, 77.7%, and 81.9%, respectively. These measures are lesion-based compared to histopathology, not patient-based, which may explain the low accuracy of PET/CT in this study [18].

Additionally, Tawakol et al. concluded that 18F-FDG PET/CT is superior to CECT concerning the diagnosis of OC recurrence. 18F-FDG PET/CT and CECT had a sensitivity of 92 versus 59%, a specificity of 96 versus 84%, PPV of 90 versus 59%, NPV of 97 versus 84%, and accuracy of 95 versus 76%, respectively [19].

Furthermore, Limei et al. noted that the specificity, sensitivity, positive and negative ratios, and the area under the curve of PET/CT scan that detect OC recurrence was 91.0%, 89.7%, 6.140%, 0.123%, and 0.9497%, respectively [20].

Moreover, Hebel et al. retrospectively studied 48 cases with suspicious recurrent OC who were referred for 18F-FDG-PET/CT. The recorded 18F-FDG-PET/CT sensitivity and PPV of 97% and specificity and NPV of 90% as in one case, 18F-FDG-PET/CT scan showed false-positive result and in another case showed false-negative result. Most patients changed the management modalities after 18F-FDG-PET/CT. The survival rate was significantly higher in 18F-FDG-PET/CT negative than in positive cases (p = 0.04) [21].

Previous studies examined the effect of post-treatment (from 1 to 109 months after treatment) PET/CT on the prognosis of the disease [22, 24]. Avril et al. found that consecutive 18F-FDG PET/CT has more accuracy than clinical and histopathologic parameters, including changes in CA125 for evaluating the prolonged response to different therapeutic options [23]. Also, Chu et al. reported that the initial post-treatment PET/CT within 3 to 9 months has more sensitivity than CA125 in detecting OC recurrence [22].

This was in the same agreement as our results that showed that PET/CT is the most valuable modality for the long-term follow-up for patients with high or normal CA125, and in patients with negative or inconclusive CT imaging results in terms of local recurrence, peritoneal deposits, suspicious lymph nodes and distant organ metastases.

However, this was in contrast with that of Kurosaki et al. who found that high serum CA125 level may be more beneficial than 18F-FDG-PET/CT imaging in the diagnosis of recurrent OC during the postoperative follow-up periods [24].

Limitations of the study

There was some limitation in our study as it was a retrospective study. It didn’t represent all the population (only included patients who referred to our inistitute). Some cases didn’t have other imaging modalities for comparison. Some cases didn’t follow up for enough period (3 months) and other didn’t come for annual follow-up. Serial measures of CA125 for some patients were not available, we used only the latest CA 125 measure before the PET/CT scan other patient.

Conclusions

18F-FDG-PET/CT is a convenient imaging modality in detecting recurrent OC, 18F-FDG-PET/CT is superior to cross-sectional imaging as it provides precise anatomical and functional information of suspected recurrence and better ability to detect intraperitoneal deposits and distant metastasis, especially in patients with post-operative fibrosis and unexplained elevated tumor marker levels. We recommended using PET/CT as a modality of choice in serial routine follow up every 6 to 12 months for the first 2 years after management.

Cases

Case 1

A-65-year female patient presented with elevated CA125 serum level in her routine follow-up investigations. She had history of managed OC with total abdominal hysterectomy and bilateral salpingioopherctomy followed by chemotherapy. Referred for PET/CT study which revealed multiple FDG avid para-aortic lymph node, peritoneal and omental nodules pathologically proved to be positive for serous ovarian cancer recurrence (Fig. 3).

Fig. 3

A–C Axial CT, PET and PET/CT showing FDG avid right para-aortic lymph node (blue arrow) measuring 1 cm with SUVmax 4.3, and anterior abdominal wall omental deposits (red arrows). D–I Axial CT, PET and PET/CT showing multiple nodular peritoneal/omental metabolically active deposits.These peritoneal nodules shows variable FDG uptake that is more obvious at peri-hepatic area measured up to 3 cm with SUV max = 7.2

Case 2

A-53- year post-menopausal female patient complaining of dull aching pelvic pain associated with elevated CA125 serum level. Serial PET/CT studies were done pre and post management. The 1st pre-mangement PET/CT study was done revealed FDG avid left complex adenexal neoplastic (ovarian) mass lesion showing coarse calcifications with peritoneal deposits having the same calcifications pattern (Fig. 4).

Fig. 4

1st pre management study A–C Axial CT, PET and PET/CT scan showing right complex neoplastic solid and cystic adnexal mass lesion (red arrow) measuring 5.5 × 3 cm with SUVmax 3.2, D–F Axial CT, PET and PET/CT scan showing multiple calcified metastatic ominto-peritoneal deposits at LV5 level (blue arrows) measuring up to 2 cm with SUV max = 4.5

She underwent debulking surgery followed by chemotherapy, the lesion pathologically proved to be ovarian papillary serous cystadenoma. A-6- month after surgery her CA125 serum level was normal but its 2nd PET/CT follow-up study showing minimal active omento-peritoneal infiltrative deposits with no active residual operative bed lesion detected (Fig. 5).

Fig. 5

The 2nd PET/CT follow-up study A–C Axial CT, PET and PET/CT scan showing small peritoneal deposit nodule residual at LV5 level measured up to 1.2 with SUVmax = 9.5. D–F Axial CT, PET and PET/CT scan showing residual minimal omento-peritoneal infiltrative thickening is seen beneath anerior abdominal SUVmax = 8.4 with successful total excision of the previously detected left complex neoplastic adnexal mass lesion

Figure 6 further 6 months follow-up 3rd PET/ CT study was done with also normal CA125 serum level revealed actively progressed omento-peritoneal infiltrative nodules as well as multiple newly developed FDG avid active omento-peritoneal soft tissue nodules.

Fig. 6

A–F Axial CT, PET and PET/CT scan (3rd PET/CT follow-up study) showing multiple newly developed FDG avid active omento-peritoneal soft tissue nodules and progressed previously noted peritoneal deposits at the right iliac region (paracolic) measuring 3.4 × 2.7 cm with SUVmax 10.7 (red arrows). G–I Axial CT, PET and PET/CT scan showing progressed omental nodules beneath anterior abdominal wall at level of Lv5 (blue arrow), measured 1.4 with SUV max = 11