The Diagnostic Performance of 18F-FDG PET/CT in Recurrent Pancreatic Cancer: A Systematic Review and Meta-analysis.

Purpose: The CT scan is the best common screening test for pancreatic cancer recurrence after surgery. The goal of our meta-analysis was to assess the diagnostic accuracy of 18F-FDG PET/CT for pancreatic cancer recurrence.
Methods: We examined PubMed and Embase for suitable papers between 2009 and 2022. The researchers considered studies that looked at the diagnostic usefulness of 18F-FDG PET/CT in identifying local and/or distant disease recurrence throughout the follow-up following pancreatic cancer resection. The Quality Assessment of Diagnostic Performance Studies-2 (QUADAS-2) method was used to evaluate the quality of each study. For each of the publications included, two researchers extracted data independently. The extracted data included general data (authors, year of publication), literature characteristics (country, type of literature, and design of study), characteristics of the patient (patients' number, mean or median age, and treatment regimen), and technical aspects (scanner, injection activity, and image analysis).
Results: The analysis includes 7 trials with a total of 263 patients. The sensitivity and specificity of 18F-FDG PET/CT in detecting recurrent pancreatic cancer following definitive treatment were 0.89 (95 percent CI: 0.83-0.93) and 0.88 (95 percent CI: 0.72-0.96), respectively, according to the pooled estimates. PET/CT performed well in the diagnosis of recurrent pancreatic cancer, with an AUC of 0.94. (0.91-0.95). Conclusions: 18F-FDG PET-CT was found to be a reliable detection method in recurrent pancreatic tumor.

1. Introduction

Adjuvant therapy following surgical excision of pancreatic cancer offers the highest chance for long-term survival as a gastrointestinal tumor with a poor prognosis [1]. Nevertheless, the 5-year continued existence time of pancreatic cancer patients after treatment aimed at cure remains dismal [2-4].

Studies have found that up to eighty percent of individuals who undertook pancreatic cancer resection will feel the occurrence of local or distant disease reappearance, so early detection can help take appropriate treatment measures for patients [5]. However, recommendations for pancreatic cancer patients' postsurgical surveillance strategies are controversial, and various national and international guidelines give their answers [6-9]. The international group strongly suggested resection in fit patients with main duct IPMNs larger than 10 mm. Surveillance is considered an appropriate option for branch-duct IPMNs in patients who are older or unfit or for cysts lacking high-risk stigmata.

Although there is no defined follow-up protocol for pancreatic cancer resection, clinical evaluation is frequently included, CA19-9 measurement, and imaging. Tumor recurrence may be suspected if symptoms associated with recurrence or a sudden increase in serum tumor markers (CA19-9) [10] occur during follow-up. And further relevant information regarding the extent and location of tumor recurrence is provided by imaging.

The CT scan is the best common screening test for pancreatic cancer recurrence after surgery [11]. However, CT-postoperative imaging evaluation of pancreatic cancer poses a significant challenge due to the extensive postoperative changes after surgical treatment, including postoperative fibrosis and lymph node enlargement, which may be mistaken for tumor recurrence [12, 13]. Although MR becomes less sensitive to local tumor recurrence than FDG-PET, it is more sensitive to liver metastasis [14]. Therefore, it is expected that MR will only be used as a complementary test for liver metastasis after pancreatic cancer surgery. Currently, 18F-FDG PET/CT scans are commonly used to identify cancer recurrence in a variety of settings, and excellent and effective results have been received [15-17]. PET/CT is more sensitive and specific than CT in monitoring pancreatic cancer recurrence after surgery [18].

The goal of this systematic review and meta-analysis is to offer a complete overview of the literature regarding the diagnostic performance of 18F-FDG PET/CT in detecting recurrent disease following pancreatic cancer resection.

2. Material and Methods

2.1. Search Strategy

We examined all existing literature in both the PubMed and EMBASE databases as of April 8, 2022, using an algorithm that combined the following phrases: (1)“positron emission tomography”(Mesh)/PET; (2) regeneration/recurrence/recurrent/relapse∗/Recrudescence/recidive; and (3) “pancreas tumor”/pancreas neoplasia/pancreas neoplasm/pancreas tumor/pancreatic neoplasm/pancreatic neoplasms/pancreatic tumor/pancreatic tumor/Cancer of Pancreas/Pancreas Cancers/Pancreatic Cancer/Pancreatic Cancers/Cancer of the Pancreas. Other keywords identified during the search are incorporated into the search strategy. In addition, a reference list of identifying publications is manually searched for potentially relevant researches.

2.2. Inclusion Criteria

Studies that match all of the criteria below will be considered for inclusion: (a) FDG's diagnostic performance for recurrence of pancreatic cancer following therapy has been studied in several studies; (b) English literature; (c) a number of patients ≥10; and (d) histological pathology or follow-up imaging as exclusion criteria.

2.3. Exclusion Criteria

Exclusion criteria were (a) case reports, reviews, conferences, no abstracts, meta-analyses, letters, and comments; (b) data could not be extracted.

Two researchers appraised the relevance of publications depending on the titles and abstracts of the papers returned via screening, using the inclusion and exclusion criteria mentioned above, and disputes were resolved through discussion and consensus.

2.4. Quality Assessment

Two researchers independently evaluated the quality of the included studies by using the Quality Assessment of Diagnostic Accuracy Studies (QUADAS-2) instrument. The following domains were used to evaluate each study: the selection of the patient, the index test, the reference standard, and the flow and timing. Based on the risk of bias, the applicability of these domains was assessed as “high,” “bad,” or “unclear.” The consensus was used to settle disagreements among the researchers.

2.5. Data Extraction

For each of the publications included, two researchers extracted data independently. The extracted data included general data (authors and year of publication), literature characteristics (country, type of literature, and design of study), characteristics of the patient (patients' number, mean or median age, and treatment regimen), technical aspects (scanner, injection activity, and image analysis), and results for the total number of patients, true positive cases, false-positive cases, true negative cases, and false-negative cases which are all counted. Calculations were generated based on sensitivity, specificity, PPV, and NPV results if these values were not provided.

2.6. Statistical Analysis

PET/CT detection rates were calculated using random-effects analysis, as well as estimates of their sensitivities, specificities, and 95 percent confidence intervals (CIs). After the summary receiver-operating characteristic (SROC) curve was created, the area under the curve (AUC) was computed.

3. Results

3.1. Study Selection and Literature Search

A systematic search of the PubMed and Embase databases found 479 and 1126 items, respectively. After removing duplicates, 1302 articles remained, and the titles and abstracts were reviewed in detail. There were 49 publications found to be potentially relevant to this investigation. After reading the remaining studies' full-text chapters, 42 articles were excluded because they included the following: data not available (n = 7), full text not relevant to the present study (n = 29), and only PET but not CT was studied (n = 6). Eventually, 7 investigations met our inclusion and exclusion criteria, encompassing 263 individuals with pancreatic cancer who had 18F-FDG PET/CT after completing previous treatment [19-25] (Figure 1).

Figure 1

The study selection flow diagram.

3.2. Description & Evaluation of the Study

Table 1 lists the literature criteria and patient factors of the included studies, while Table 2 lists the parameters and 18F-FDG PET/CT reference standards used. Seven investigations looked examined the diagnostic accuracy of 18F-FDG PET-CT in detecting recurrent pancreatic cancer, and the diagnostic value of 18F-FDG PET-CT was revealed. Only one of the seven studies was planned to answer this research topic prospectively [3]. Positive PET/CT findings were studied on a per-patient basis in six trials, whereas the results were assessed on a per-lesion basis in one research [4]. In 7 articles, all received surgical resection, of which 4 the treatment modality received (including surgical resection and adjuvant therapy) was described [5, 6].

Table 1

Study and patient characteristics of the included studies.

Author	Year	Study characteristics			Patient characteristics
		Country	Study design	Analysis	No. of patients	Mean age ± SD	Previous treatment
Bjerring et al.	2020	Denmark	Pro	PB	39	NA	Surg
El-Kholy et al.	2019	Egypt	Retro	PB	34	58.3 ± 10.3	Surg, Surg+Cx, Surg+Cx+RTx
Rayamajhi et al.	2017	America	Retro	PB	39	64.5(55-78)	Surg, Surg+Cx, Surg+Cx+RTx
Jung et al.	2016	South Korea	Retro	PB	110	62(35-84)	Surg, Surg+Cx, Surg+Cx+RTx
Peti et al.	2014	America	Retro	LB	97	64	Surg
Asagi et al.	2013	Japan	Retro	PB	17	NA	Surg
Kitajima et al.	2009	Japan	Retro	PB	45	58(45-81)	Surg, Surg+Cx, Surg+Cx+RTx

PB: patient-based; LB: lesion-based; Pro: prospective; Retro: retrospective; Surg: surgery; Cx: chemotherapy; RTx: radiotherap.

Table 2

Technical aspects of included studies.

Author	Year	Scanner modality(PET/CT)	Ligand dose	Image analysis	18F-FDG-PET/CT
					Total	TP	FP	FN	TN
Bjerring et al.	2020	GE Medical Systems, Milwaukee, Wisconsin USA	4.0 MBQ/kg	Quantitative	39	17	9	5	8
El-Kholy et al.	2019	GE, PET/CT Discovery	NA	Quantitative	34	21	1	3	9
Rayamajhi et al.	2017	Discovery ST, STE, or RX, GE Healthcare	185-370 MBq/kg	Quantitative	39	30	0	3	6
Jung et al.	2016	Philips Gemini Dual (Best, The Netherlands)Siemens Biograph TruePoint (Germany)	5.18 MBq/kg	Quantitative	110	71	4	13	22
Peti et al.	2014	GE Medical Systems, Waukesha, WI	NA	Qualitative	97	52	6	4	35
Asagi et al.	2013	Toshiba, Otawara, Japan	3.0 MBq/kg	Quantitative	17	11	0	0	6
Kitajima et al.	2009	Siemens AG,Erlangen, Germany	4.0 MBq/kg	Quantitative	45	22	1	2	20

Table 1 summarizes the studies and patient factors of the 7 articles that included 263 patients. Figure 2 shows the overall results of the risk of bias and applicability problems for each study. The official outcome is determined by the quality of the acceptable consensus inclusion studies.

Figure 2

The listed studies' risk of bias & applicability problems are summarized.

3.3. Diagnostic Performance of 18F-FDG PET/CT for Pancreatic Cancer

For recurrent pancreatic cancer, the pooled sensitivity and specificity of 18F-FDG PET/CT were 0.89 (95 percent CI: 0.83-0.93) and 0.88 (95 percent CI: 0.72-0.95), respectively, with mild heterogeneity (29.37 percent) and 0.88 (95 percent CI: 0.72-0.96), respectively, and moderate heterogeneity (75.20 percent) (Figure 3). Figure 4 illustrates the SROC curve for 18F-FDG PET/CT, which has an AUC of 0.94 (95 percent confidence interval: 0.91-0.95). Meta-regression analysis was used to investigate the source of heterogeneity, and meta-regression analysis was used to investigate the source of heterogeneity (Table 3). It showed that the study type (specificity P = 0.02) was the possible cause of heterogeneity. Figure 5 shows that no publication bias was found for the treatment of pancreatic cancer (P = 0.74).

Figure 3

The sensitivity and specificity of 18F-FDG PET/CT for recurrent pancreatic cancer is shown in a forest plot.

Figure 4

SROC curve for 18F-FDG PET/CT with an AUC value.

Table 3

Subgroup analysis of diagnostic performance of 18F-FDG PET/CT.

Covariate/subgroup	Studies, n	Sensitivity (95% CI)	P value	Specificity (95% CI)	P value
Number of patients included			0.05		0.85
>50	2	0.88 (0.80–0.96)		0.85 (0.65–1.00)
≤50	5	0.89 (0.83–0.96)		0.90 (0.76 - 1.00)
Ethnicity			0.15		0.23
Asian	3	0.90 (0.82–0.98)		0.94 (0.84–1.00)
The rest	4	0.89 (0.82–0.96)		0.81 (0.64–0.98)
Treatment			0.07		0.30
Surg, Surg+Cx, Surg+Cx+RTx	4	0.89 (0.82–0.95)		0.93 (0.84–1.00)
Surg	3	0.90 (0.82–0.98)		0.79 (0.59–0.99)
Study			0.91		0.02
Retro	6	0.89 (0.85–0.93)		0.89 (0.83–0.95)
Pro	1			0.47 (0.23–0.71)
Analysis		0.77 (0.60–0.95)	0.80		0.81
LB	1
PB	6			0.86 (0.58–1.00)
Image analysis		0.93 (0.85–1.00)	0.80		0.81
		0.88 (0.82–		0.90 (0.77–1.00)
		0.94)
Qualitative	1	0.93 (0.85–1.00)		0.86 (0.58–1.00)
Quantitative	6	0.88 (0.82–0.94)		0.93 (0.77–1.00)

PB: patient-based; LB: lesion-based; Pro: prospective; Retro: retrospective; Surg: surgery; Cx: chemotherapy; RTx: radiotherapy.

Figure 5

Funnel plot based on the data of PET/CT for the diagnosis of recurrent pancreatic cancer.

4. Discussion

Difficulty distinguishing between local tumor recurrence and postoperative fibrosis after pancreatic cancer surgery has been complex in imaging evaluation [11]. CT and tumor markers (CA19-9) are now commonly used by many clinicians to monitor postoperative follow-up of pancreatic cancer and have moderate diagnostic value, with 72.2 percent sensitivity and 66.6 percent specificity, correspondingly [21]. In adding to concerns about the efficacy of CT in local recurrences, there is also the potential to miss metastatic recurrences in other unscanned areas. MR only shows good diagnostic performance in hepatic recurrence of pancreatic cancer [14]. PET/CT has broadly been utilized in pancreatic cancer management as a new device to help diagnose the location of cancer through the metabolism of tumor cells [26-28]. Furthermore, the relevance of PET/CT in the recurrence of pancreatic cancer has long been researched and established [25, 29]. According to the literature we reviewed, PET/CT has prominent use to monitor distant recurrence, ambiguous CT findings, and in patients with normal CA19-9 levels [21-23]. The literature reported that the incidence of distant recurrence (60%) was more significant than local recurrence (17%) and simultaneous local, distant recurrence (23%) in the pattern of recurrent pancreatic cancer [30]. PET/CT, as a complementary test, may lead to an underestimation of its diagnostic performance in detecting pancreatic cancer recurrence. With an AUC value of 0.94, our study showed that PET/CT had a strong diagnostic performance in recurrent pancreatic cancer (0.91-0.95). The surveillance method for detecting recurring pancreatic cancer after surgery needs to be improved further.

The purpose of this study was to demonstrate how well PET/CT might detect recurring pancreatic cancer. PET/CT exhibited a sensitivity of 0.89 (95 percent CI 0.83-0.93) and specificity of 0.88 (95 percent CI 0.72-0.96) in detecting pancreatic cancer recurrence in our study. The combined sensitivity and specificity estimates for 18F-FDG PET/CT in detecting pancreatic cancer recurrence were 0.88 and 0.89 in a prior meta-analysis [31]. It demonstrates PET/relatively CT's good diagnostic effectiveness in pancreatic cancer recurrence. The previous meta-literature only included studies of PET/CT compared with CT, whereas our literature without this restriction. With the further development of treatment options and the need for early diagnosis of cancer recurrence, PET/CT in the clinic should not be a passive option in case of poor CT results.

The limitations of this study are that most of the data were collected from retrospective studies and only one piece of literature was prospective. Moreover, only seven studies were eligible, most of which had relatively small numbers of patients. These biases may overestimate FDG PET-CT specificity and sensitivity. In addition to the majority of the studies being retrospective comparisons of PET/CT and CT, one looked at the performance of intracavitary ultrasound versus PET/CT in detecting recurrence, and another looked at PET/CT in identifying benign and malignant recurrent pancreatic lesions at two-time points. As a result, the potential bias that can arise from utilizing different reference standards to validate disease recurrence is a source of worry.

When we pooled the overall specificity of PET/CT, there was moderate heterogeneity. We utilized meta-regression to figure out where the heterogeneity came from. Finally, it was identified that this work design might be the source of heterogeneity (P = 0.02); it could be explained by selection and information bias between the two study designs.

Finally, this meta-analysis shows that 18 F-FDG PET/CT for recurrent pancreatic cancer is a reliable diagnostic with good sensitivity and specificity. However, these potentially beneficial results need to be considered in the context of the additional financial and psychological burden that may be imposed on patients. More high-quality prospective studies are needed in the future to analyze whether PET/CT will serve as a more aggressive diagnostic measure for patients with postoperative pancreatic cancer.

5. Conclusions

18F-FDG PET-CT was found to be a reliable detection method in recurrent pancreatic tumor.