Respiratory gated and prolonged acquisition 18F-FDG PET improve preoperative assessment of colorectal liver metastases

Abstract

Background

Detection of small liver metastases from colorectal cancer by 18F-FDG PET/CT is hampered by high physiologic uptake in the liver parenchyma and respiratory movements during image acquisition.

Purpose

To investigate whether two tailored 18F-FDG PET liver acquisitions (prolonged liver acquisition time [PL-PET] and repeated breath-hold respiratory gated liver acquisition [RGL-PET]) would improve detection of colorectal liver metastases, when added to a standard whole body PET (WB-PET).

Material and Methods

Twenty consecutive patients referred to our hospital for surgical treatment of colorectal liver metastases diagnosed with contrast-enhanced CT underwent preoperative 18F-FDG PET/CT tailored for detection of liver metastases. Concordance between preoperative imaging results and true findings (histology and/or follow-up imaging) as well as changes in clinical management, based on 18F-FDG PET/CT findings, were documented. Background noise, defined as the standard deviation measured in a reference region within the normal liver parenchyma, was compared between the three 18F-FDG PET/CT protocols.

Results

WB-PET, PL-PET, and RGL-PET showed suspicious liver lesions in 18 out of 20 patients. Compared to WB-PET alone, the combination of PL-PET and RGL-PET showed additional lesions in the liver in seven out of the 18 patients. The combination of all three PET acquisitions changed clinical management in four patients. Two patients with negative PET results were later found to have benign liver lesions.

Conclusion

The addition of tailored liver-specific 18F-FDG PET/CT protocols (PL-PET and RGL-PET) to a WB-PET, improved the detection of intrahepatic colorectal metastases, compared to WB-PET alone. Such add-ons can change clinical patient management of potentially resectable colorectal liver metastases.

Keywords

18F-FDG PET colorectal cancer liver metastasis respiratory gating prolonged acquisition

Introduction

The liver is the most common metastatic site of colorectal cancer (1). The overall survival of patients with metastatic colorectal cancer without treatment is about eight months (2). If the liver is amenable to surgical resection evaluated from contrast-enhanced computed tomography (CT) or magnetic resonance imaging (MRI), a 5-year overall survival of up to 55% can be expected (3 –5).

Accurate diagnostic imaging in the preoperative assessment of patients with potentially resectable colorectal liver metastases is essential in order to identify candidates for curative liver surgery (6,7). Current guidelines (October 2013) from the National Comprehensive Cancer Network (NCCN) recommend a staging 2-deoxy-2-(18F)fluoro-D-Glucose positron emission tomography (18F-FDG PET) examination for patients with potentially surgically curable metastatic colorectal cancer (8). Even though most hepatic colorectal metastases show increased 18F-FDG-uptake, the modality’s ability to detect small sub-centimeter intrahepatic lesions is limited by several factors: limited spatial resolution causing intensity diffusion (9,10) hampers the correct rendering of small lesions, physiologic 18F-FDG-uptake in the liver parenchyma, respiratory movements during acquisition causing smearing of focal 18F-FDG uptakes (11), and possible erroneous attenuation correction due to respiratory phase mismatch between the PET and CT data (10 –12).

The aim of the present study was to investigate whether the addition of two tailored 18F-FDG PET liver protocols (prolonged liver acquisition time [PL-PET] and repeated breath-hold respiratory gated liver acquisition [RGL-PET]) to a standard whole body 18F-FDG PET protocol would improve detection of 18F-FDG-uptake in colorectal liver metastases.

Material and Methods

Study population

Twenty patients, five women and 15 men, with a median age of 68 years (range, 55–80 years) and a median body mass index (BMI) of 26.5 kg/m² (range, 21.5–30.8 kg/m²), referred to our hospital for determination of eligibility for surgery, were consecutively included in the study over a 5-month period (May to September 2011). Inclusion criteria were: colorectal liver metastasis on CT (performed at the referral hospital), prior resection of the primary tumor, and no surgical treatment received the last 3 months before PET imaging. Written informed consent according to the institutional and national regulations was obtained from all patients and the institutional ethics board approved the project.

All primary tumors had been resected at external hospitals. The primary tumors were located in the rectum (n = 6), in the left colon (n = 14), and in the right colon (n = 1). One patient had first been diagnosed with cancer in the left colon and later developed a second primary cancer in the rectum (patient 18). Thirteen patients had received adjuvant systemic chemotherapy. The time between completion of adjuvant systemic chemotherapy and 18F-FDG PET/CT was for 12 out of 13 patients at least 6 weeks. The remaining patient received systemic chemotherapy only one day prior to 18F-FDG PET/CT (patient 15). An overview of patient and tumor characteristics is shown in Table 1.

Table 1.

Summary of patient and tumor characteristics in 20 patients.

Patient	Gender	Age (years)	BMI	Location primary	Prior chemotherapy
1	M	72	23.6	Left	Y
2	M	80	24.4	Left	Y
3	M	66	n/a	Left	Y
4	F	56	n/a	Rectum	N
5	M	61	26.9	Left	Y
6	M	70	30.8	Left	Y
7	M	55	25.4	Left	Y
8	M	64	26.7	Rectum	N
9	M	67	n/a	Rectum	Y
10	F	72	27.5	Left	Y
11	F	73	30.5	Left	Y
12	M	67	22.0	Left	N
13	M	66	28.4	Left	N
14	M	62	26.5	Rectum	N
15	F	74	21.5	Left	Y
16	M	66	26.5	Left	Y
17	M	73	29.1	Left	Y
18	M	68	n/a	left/rectum	Y
19	F	79	24.8	Right	N
20	M	77	n/a	Rectum	N

BMI, body mass index.

Preparation for 18F-FDG PET/CT

All patients underwent 18F-FDG PET/CT using a hybrid PET/CT system (Siemens Biograph 64, Siemens Medical Systems, Erlangen, Germany) equipped with a Somatom Sensation 64 CT scanner. After fasting for at least 6 h and confirmation that the blood glucose level was less than 10 mmol/L, 5 MBq/kg 18F-FDG was injected intravenously. Imaging started after the patient had rested for approximately 60 min.

The imaging sequence for individual patients was low-dose CT for WB PET, low-dose CT for PL-PET, WB-PET, PL-PET, CT for RGL-PET, and finally RGL-PET. The total time used for PL-PET and RGL-PET was 30 min, including the acquisition of low-dose CT.

Low-dose CT

The acquisition of free-breathing, low-dose, contrast-enhanced CT corresponding to the WB-PET and the PL-PET images started 40 and 90 s, respectively, after intravenous injection of contrast medium (80 mL iodixanol, 350 mg/mL at a rate of 2.5 mL/s [Visipaque®, GE Healthcare, Oslo, Norway]). The CT parameters for the WB-examination were 120 kV, 50 mAs (automatic exposure control (AEC)), 0.5 s tube rotation time, pitch of 1.35, collimation of 24 x 1.2 mm, 28.8 mm per rotation, reconstruction slice thickness of 3 mm, and a reconstruction increment of 3 mm. The CT parameters corresponding to the PL examination were 120 kV, 80 mAs (AEC), 0.5 s tube rotation time, pitch of 0.9, collimation of 64x 0.6 mm, 19.2 mm per rotation, reconstruction slice thickness of 3 mm, and a reconstruction increment of 2 mm.

The CT images corresponding to RGL-PET were acquired without administration of contrast medium. The CT images were acquired while the patient held his/her breath in inspiration with these parameters: 120 kV, 50 mAs (AEC) and axial slices of 3 mm.

Standard whole body PET (WB-PET)

WB-PET was acquired from the skull base to the upper thigh with an acquisition time of 3 min per bed position and 30% overlap between bed positions. The data were reconstructed with a matrix size of 168 x 168 pixels (pixel size 4.06 mm) using the OSEM algorithm with four iterations, eight subsets (4i/8 s), and Gaussian post-reconstruction filter with full-width at half maximum (FWHM) of 3.5 mm. This reconstruction yielded an effective point spread function with FWHM of 5.8 and 5.1 mm in the axial and longitudinal planes, respectively.

Prolonged liver acquisition time PET (PL-PET)

PL-PET was performed immediately after WB-PET with the patient still lying on the examination table. The PL-PET acquisition covered the liver with two bed positions each of 8 min duration with all other PET acquisition parameters being identical to WB PET.

Repeated breath-hold respiratory gated liver acquisition PET (RGL-PET)

The RGL PET protocol covered one bed centred over the liver hilum. A PET image volume equivalent to 5 min static inspiration breath-hold was obtained, using an in-house developed electronic circuit and the PET/CT channel for physiological triggers: The patient was instructed to breathe according to information provided by a display repeatedly counting down from 7 to 0 and at the same time alternating between green (breathe) and red number (breath-hold) in consecutive cycles. During the 10-min acquisition of list mode data, physiological triggers were inserted for every 7-s breath-hold period. The standard Biograph software was used to create the raw data (sinograms) and to reconstruct the image volume. The patient was instructed to reach the same depth of inspiration in all the breath-holds, also for the CT to be used for attenuation correction. The image format and the image volumes were created with the same reconstruction algorithms and filter as the ones used in the WB-PET.

Interpretation of PET examinations

The images were read and quantitatively analyzed by two experienced nuclear medicine physicians. Metastases were identified as areas having a focal uptake above the level of surrounding tissue. The 18F-FDG uptake was defined as the maximum standardized uptake value (SUV_max, the highest activity concentration per injected dose per lean body mass after correction for decay) in a volume of interest (VOI).

The background liver uptake and image noise, defined as the standard deviation (SD), in units of SUV were obtained in a region of the right liver lobe with homogenous uptake. Comparable SUV_mean values were assumed (13). In patients with more than one metastatic lesion, the 18F-FDG uptake in the smallest lesion detected in all PET series was selected for comparison.

The nuclear medicine physicians read the three PET acquisitions in a single session and had access to the diagnostic CT findings. Based on the findings on diagnostic CT acquired before inclusion to the study and PET (WB-PET, PL-PET, and RGL-PET), a multidisciplinary team decided to go for surgical or non-surgical treatment. Histology of surgical specimen and/or follow-up imaging was used to validate all PET imaging findings.

Results

The median number of liver metastases detected at WB-PET, PL-PET, and RGL-PET were 1.5, 2.5, and 2, respectively. The mean SUV_max for all lesions detected at WB-PET, PL-PET, and RGL-PET was 5.8 g/mL (range, 2.3–15 g/mL), 6.2 g/mL (range, 2.9–15.7 g/mL), and 6.6 g/mL (range, 2.9–15.7 g/mL), respectively. The mean SUV_mean physiologic background liver uptake was 1.8 for WB-PET and 1.7 for both PL-PET and RGL-PET. Compared to WB-PET, the mean value of SD measured in a uniform region of the liver without signs of disease was reduced by 31% for the PL-PET and 41% for the RGL-PET.

Combination of all three PET protocols showed liver lesions in 18 out of 20 patients. However, three patients failed to perform the respiratory gated study due to lack of compliance to the breathing protocol (patients 1 and 10) or presence of red–green color blindness, which led to inability to follow the colored number display (patient 15). One patient (patient 14) with no suspicious liver lesions upon PET imaging underwent liver resection of a solitaire lesion in liver segment 6 where histopathology revealed a hemangioma. For this particular patient, the interpretation of the diagnostic CT was false positive (patient 14). The other patient with no suspicious liver lesion on PET had a 5 mm suspicious structure in liver segment 6 on diagnostic CT, but did not show any sign of malignancy upon follow-up (patient 4).

Compared to WB-PET, the combined PL-PET and RGL-PET detected a higher number of intrahepatic lesions in seven out of 18 patients. All of these lesions were confirmed to be metastases, five by histopathology and two by follow-up imaging. For two out of the seven patients, PL-PET and RGL-PET performed equally (patients 2 and 5), for two patients the RGL-PET detected the most lesions (patients 13 and 17) and for three patients the PL-PET detected the most lesions (patients 10, 15 and 18). However, in two out of the three patients were PL-PET detected the most lesions, RGL-PET was not performed. The smallest lesion was only detected by RGL-PET (patient 13).

Extrahepatic malignancy was suspected in four patients after performing WB-PET. Three of these had suspected recurrence at the site of the primary tumor (patients 4, 9, and 20) and one had suspicious lesions in both lungs and in an adrenal gland (patient 10). An overview of liver metastasis characteristics by WB-PET, PL-PET, and RGL-PET imaging as well as extrahepatic findings with WB-PET is shown in Table 2.

Table 2.

Comparison of findings after conventional radiological imaging, WB-PET, PL-PET, and RGL-PET imaging.

Patient	Diagnostic CT
Patient	Size of smallest lesion (mm)	WB-PET liver metastases (n)	PL-PET liver metastases (n)	RGL-PET liver metastases (n)	Additional detected lesions (PL-/RGL-PET vs. WB-PET)
1	7	4	4	NP	0
2	14	0	1	1	1 (PL/RGL)
3	9	4	4	4	0
4	5	0	0	0	0
5	7	2	3	3	1 (PL/RGL)
6	9	1	1	1	0
7	11	1	1	1	0
8	10	1	1	1	0
9	13	6	6	6	0
10	6	2	4	NP	2 (PL)
11	35	1	1	1	0
12	8	3	3	3	0
13	4	3	5	6	2 (PL), 3 (RGL)
14	6	0	0	0
15	9	0	10	NP	10 (PL)
16	13	2	2	2	0
17	5	1	1	2	0 (PL), 1 (RGL)
18	6	2	3	2	1 (PL), 0 (RGL)
19	17	1	1	1	0
20	25	1	1	1	0

FP, false positive; NP, not performed; PL-PET, prolonged liver acquisition; RGL-PET, respiratory gated PET; WB-PET, standard whole body 18F-FDG PET.

Combined PL-PET, RGL-PET, and WB-PET changed the clinical management in four patients (Table 3). One of these patients was found to have more progressive intrahepatic disease than anticipated from diagnostic CT and WB-PET, addition of PL-PET and RGL-PET led to more extensive liver resection than initially planned (patient 13) (Fig. 1). The remaining three patients were spared futile surgery; two of these were found to have extrahepatic disease (patients 10 and 20) and one underwent combined PL-PET, RGL-PET, and WB-PET without confirmation of intrahepatic malignancy (patient 4, as described above).

Fig. 1.

Maximum intensity projection (MIP) of standard whole body 18F-FDG PET (WB-PET) and two liver-specific scans consistent of prolonged acquisition time and repeated breath-hold respiratory gated liver acquisition (PL-PET and RGL-PET) in a patient with synchronous colorectal liver metastases (patient 13). The arrows point out focal 18F-FDG uptakes in the liver. (a) Standard whole body 18F-FDG PET (WB-PET), (b) prolonged liver acquisition time 18F-FDG PET (PL-PET), and (c) respiratory gated liver acquisition 18F-FDG PET (RGL-PET).

Table 3.

Overview of surgical procedures, validation of PET findings, and corresponding patient management.

Patient	Surgery	Procedure	Confirmation of PET-detected liver metastases	Changed management
1	Y	Open right hemihepatectomy	Y (histology)	N
2	Y	Laparoscopic atypical resection	N (necrotic tissue)	N
3	N	NA	Y (follow-up)	N
4	N	NA	Y (follow-up)	Y
5	Y	Repetitive laparoscopic atypical resection and RFA	Y (histology)	N
6	Y	Laparoscopic atypical resection	Y (histology)	N
7	Y	Laparoscopic atypical resection	Y (histology)	N
8	Y	Laparoscopic atypical resection	Y (histology)	N
9	N	NA	Y (follow-up)	N
10	N	NA	Y (follow-up)	Y
11	Y	Open atypical resection and adrenalectomy	Y (histology)	N
12	Y	Explorative laparotomy (peritoneal carcinomatosis)	Y (histology)	N
13	Y	Open right hemihepatectomy	Y (histology)	Y
14	Y	Laparoscopic atypical resection	Y (histology, hemangioma)	N
15	Y	Open atypical resections	Y (histology)	N
16	Y	Explorative laparotomy (peritoneal carcinomatosis)	Y (histology)	N
17	Y	Open right hemihepatectomy and local atypical resection	Y (histology)	N
18	Y	Left hemihepatectomy, local atypical resections and RFA	Y (histology)	N
19	Y	Laparoscopic atypical resection	Y (histology)	N
20	N	NA	Y (follow-up)	Y

RFA, radiofrequency ablation.

Discussion

Our study shows that the addition of respiratory gated (RGL-PET) and prolonged acquisition 18F-FDG PET (PL-PET) can improve image quality and thereby increase the detection rate of intrahepatic colorectal metastases, compared to standard WB-PET alone. In some patients, the 30 min extra scanner time necessary for the RGL-PET and PL-PET may change clinical management of potentially resectable colorectal liver metastases by either preventing futile laparotomies or by leading to a change in surgical strategy. Moreover, the study confirms the importance of performing a 18F-FDG PET before surgical resection of colorectal liver metastases to rule out extrahepatic malignancy and to confirm the CT findings of liver lesions suspicious of malignancy (6,7,14).

In accordance with our findings, Fin et al. (15) recently showed an improved detection of intrahepatic colorectal metastases by the use of a CT-based gated PET method. The value of respiratory gating in PET has also been shown in a study of tumors in the liver, pancreas and bile duct by Nagamachi et al. (16). In this study deep inspiration breath-hold for both PET and CT changed the location of the tumor from incorrect to correct organ in 30% of the cases, and resulted in a SUV_max increases of 15 to 59%.

The image quality of the two tailored PET protocols was superior to the standard WB-PET as reflected by reduced image noise and detection rate for both protocols. Improved image quality of the PL-PET is anticipated to be a result of increased image statistics, whereas the effect of displacement of anatomical structures reduces image noise in RGL-PET. In the present study, both repeated breath-hold PET and breath-hold CT acquisitions over the organ of interest during inspiration, were performed to obtain phase-matched PET- and CT images (17). This is crucial as mismatch between CT and PET images may, due to incorrect attenuation correction, result in over- or underestimation of the uptake, lack of exact co-registration with CT, as well as overestimation of tumor size (18). Due to the relatively long acquisition time of PET there will always be a mismatch between CT and PET, the degree being directly linked to the breathing pattern.

It could be argued that the increased detection rate of PL-PET and RGL-PET is related to the fact that these series were acquired after WB-PET as studies have shown that malignant lesions often continue to accumulate 18F-FDG from an early phase to a delayed phase, whereas 18F-FDG-uptake tends to decrease in benign lesions (19,20). However, in our study the delay between WB-PET and the tailored PET-protocols was only 12 min for PL-PET and 28 min for RGL-PET.

One patient had systemic chemotherapy 1 day prior to PET examination. This patient had focal 18F-FDG uptake in all metastatic lesions detected at PL-PET (RGL-PET was not performed), but they were not visible at the WB-PET (patient 15). Systemic chemotherapy may reduce the sensitivity of PET due to decreased cellular metabolic activity following chemotherapy (21,22).

A limitation of this study is the small patient cohort, not allowing for advanced statistics. However, the side-by-side comparison of different PET protocols used in this study illustrates the potential for more tailored PET protocols. The use of iodine contrast medium may affect the attenuation correction by increasing the SUV in focal regions that accumulate the iodine. The use of iodine contrast media for the low-dose CT was necessary to locate the PET findings within the different lobes of the liver, and to obtain better visualization of lymph nodes in the liver hilum, information being essential for the surgeons. Since the CT study was obtained with a long delay between injection and data acquisition, no focal region was present, and a possible error in the attenuation correction caused by the low concentration of iodine would presumably only lead to slight parallel shifts of SUV values within larger regions.

The two additional protocols (RGL-PET and PL-PET) seem to be complementary and lesions undetected by RGL-PET were detected by PL-PET and vice versa. The challenge with RGL-PET is the need for a cooperative patient who acts strictly according to instructions. Non-compliance can be a reason to why PL-PET displayed a tumor that RGL-PET did not.

In conclusion, this study shows that the addition of two tailored 18F-FDG PET liver-specific protocols to standard whole body PET improves the detection rate of intrahepatic colorectal metastases.

Footnotes

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

References

Fortner

Silva

Golbey

. Multivariate analysis of a personal series of 247 consecutive patients with liver metastases from colorectal cancer. I. Treatment by hepatic resection. Ann Surg 1984; 199: 306–316.

Simmonds

. Palliative chemotherapy for advanced colorectal cancer: systematic review and meta-analysis. Colorectal Cancer Collaborative Group. BMJ 2000; 321: 531–535.

Nanji

Cleary

Ryan

. Up-front hepatic resection for metastatic colorectal cancer results in favorable long-term survival. Ann Surg Oncol 2013; 20: 295–304.

Wei

Greig

Grant

. Survival after hepatic resection for colorectal metastases: a 10-year experience. Ann Surg Oncol 2006; 13: 668–676.

Wiering

Krabbe

Jager

. The impact of fluor-18-deoxyglucose-positron emission tomography in the management of colorectal liver metastases. Cancer 2005; 104: 2658–2670.

Ruers

Wiering

van dS

. Improved selection of patients for hepatic surgery of colorectal liver metastases with (18)F-FDG PET: a randomized study. J Nucl Med 2009; 50: 1036–1041.

Strasberg

Dehdashti

. Role of FDG-PET staging in selecting the optimum patient for hepatic resection of metastatic colorectal cancer. J Surg Oncol 2010; 102: 955–959.

Benson

III Bekaii-Saab

Chan

. Metastatic colon cancer, version 3.2013: featured updates to the NCCN Guidelines. J Natl Compr Canc Netw 2013; 11: 141–152.

Skretting

. ‘Intensity diffusion' is a better description than ‘partial volume effect'. Eur J Nucl Med Mol Imaging 2009; 36: 536–537.

10.

Vogel

van Dalen

Wiering

. Evaluation of image registration in PET/CT of the liver and recommendations for optimized imaging. J Nucl Med 2007; 48: 910–919.

11.

Skretting

Revheim

Knudtsen

. An implementation of time-efficient respiratory-gated PET acquisition by repeated breath-holds. Acta Radiol 2013; 54: 672–675.

12.

Hamill

Bosmans

Dekker

. Respiratory-gated CT as a tool for the simulation of breathing artifacts in PET and PET/CT. Med Phys 2008; 35: 576–585.

13.

Wahl

Jacene

Kasamon

. From RECIST to PERCIST: Evolving Considerations for PET response criteria in solid tumors. J Nucl Med 2009; 50(Suppl. 1): 122S–150S.

14.

Fernandez

Drebin

Linehan

. Five-year survival after resection of hepatic metastases from colorectal cancer in patients screened by positron emission tomography with F-18 fluorodeoxyglucose (FDG-PET). Ann Surg 2004; 240: 438–447.

15.

Fin

Daouk

Bailly

. Improved imaging of intrahepatic colorectal metastases with 18F-fluorodeoxyglucose respiratory-gated positron emission tomography. Nucl Med Commun 2012; 33: 656–662.

16.

Nagamachi

Wakamatsu

Kiyohara

. Usefulness of a deep-inspiration breath-hold 18F-FDG PET/CT technique in diagnosing liver, bile duct, and pancreas tumors. Nucl Med Commun 2009; 30: 326–332.

17.

Knudtsen

Rodal

Brustugun

. Dynamic respiratory gated (18)FDG-PET of lung tumors - a feasibility study. Acta Oncol 2011; 50: 889–896.

18.

Daouk

Fin

Bailly

. Respiratory-gated positron emission tomography and breath-hold computed tomography coupling to reduce the influence of respiratory motion: methodology and feasibility. Acta Radiol 2009; 50: 144–155.

19.

Chan

Ramsay

Szeto

. Dual-time-point (18)F-FDG-PET/CT imaging in the assessment of suspected malignancy. J Med Imaging Radiat Oncol 2011; 55: 379–390.

20.

Miyake

Nakamoto

Togashi

. Dual-time-point (18)F-FDG PET/CT in patients with colorectal cancer: clinical value of early delayed scanning. Ann Nucl Med 2012; 26: 492–500.

21.

Glazer

Beaty

Abdalla

. Effectiveness of positron emission tomography for predicting chemotherapy response in colorectal cancer liver metastases. Arch Surg 2010; 145: 340–345.

22.

Angliviel

Benoist

Penna

. Impact of chemotherapy on the accuracy of computed tomography scan for the evaluation of colorectal liver metastases. Ann Surg Oncol 2009; 16: 1247–1253.