Download - Low dose non-enhanced CT versus standard dose contrast-enhanced CT in combined PET/CT protocols for staging and therapy planning in non-small cell lung cancer

Original article

Low dose non-enhanced CT versus standard dosecontrast-enhanced CT in combined PET/CT protocols for stagingand therapy planning in non-small cell lung cancerAnna C. Pfannenberg1, Philip Aschoff1, Klaus Brechtel1, Mark Müller2, Roland Bares2, Frank Paulsen3,Jutta Scheiderbauer3, Godehard Friedel4, Claus D. Claussen1, Susanne M. Eschmann2

1 Department of Diagnostic Radiology, Eberhard-Karls-University, Tübingen, Germany2 Department of Nuclear Medicine, Eberhard-Karls-University, Tübingen, Germany3 Clinic for Radiation Oncology, Eberhard-Karls-University, Tübingen, Germany4 Schillerhöhe Centre of Thoracic Medicine, Gerlingen, Germany

Received: 10 February 2006 / Accepted: 3 April 2006 / Published online: 1 August 2006© Springer-Verlag 2006

Abstract. Purpose: To evaluate low dose non-enhancedCT and standard dose contrast-enhanced CT in combinedPET/CT protocols for staging and therapy planning ofnon-small cell lung cancer (NSCLC).Methods: Retrospective analysis was performed of 50consecutive patients with proven NSCLC who had beenreferred for primary staging (n=41) or restaging (n=9). Allpatients underwent a multi-phase PET/CT consisting of alow dose non-enhanced attenuation scan and an arterialand portal-venous contrast-enhanced CT scan followed bywhole-body PET. Fused datasets of non-enhanced andcontrast-enhanced PET/CT were compared per patient byusing the TNM staging system, and per lesion regardinglocalisation, characterisation and delineation of tumourlesions. The staging results were validated either byhistopathology or by clinical–radiological follow-up for≥6 months.Results: In 47/50 patients, the results of T staging did notdiffer between the two PET/CT protocols. Three patientscould only be correctly classified as having T4 tumoursafter contrast application. Regarding N staging, bothprotocols yielded the same results. In M staging, therewas only one patient with an improvement of the results asa result of contrast application. The lesion-based analysis of92 sites showed no difference in the accuracy of lesionlocalisation and only one revision of lesion characterisationby contrast-enhanced PET/CT. The assessment of tumourdelineation was altered by contrast application in 58/92sites (p<0.0001). In 10/50 patients, contrast-enhancedPET/CT detected additional clinically important findings.Conclusion: In patients with advanced NSCLC, contrast-enhanced CT as part of the PET/CT protocol more

accurately assessed the TNM stage in 8% of patientscompared with non-contrast PET/CT. However, forplanning of 3D conformal radiotherapy and non-conven-tional surgery, contrast-enhanced PET/CT protocols areindispensable owing to their superiority in preciselydefining the tumour extent.

Keywords: Dual-modality imaging – PET/CT –Scanning protocol – Contrast-enhanced CT –Non-enhanced CT

Eur J Nucl Med Mol Imaging (2007) 34:36–44DOI 10.1007/s00259-006-0186-3

Introduction

Accurate staging is crucial for both the treatment and theprognosis of patients with non-small cell lung cancer(NSCLC) [1]. The goal is to identify those patients whowould benefit from surgical resection and those who wouldnot, in order to optimise individual treatment. Surgery haslimited success among patients with locally advanceddisease. However, combined modality regimens utilisingsurgery, radiotherapy and chemotherapy with newer agentsin the neoadjuvant and adjuvant setting are currently underinvestigation in order to improve response rate, even inlocally advanced tumour stages. Clinical staging ofNSCLC includes delineation of the primary tumour(T staging), detection of lymph node involvement (Nstaging) and the discovery of distant metastatic disease (Mstaging). When used alone, the majority of the stagingmethods are not sufficiently accurate. Although both lesionsize and local invasion by the primary tumour are bestdetermined with computed tomography (CT) [2, 3], severalstudies have demonstrated superior sensitivity and speci-

Anna C. Pfannenberg ())Department of Diagnostic Radiology,Eberhard-Karls-University,Tübingen, Germanye-mail: [email protected].: +49-7071-2982087, Fax: +49-7071-295845

European Journal of Nuclear Medicine and Molecular Imaging Vol. 34, No. 1, January 2007

ficity of 18F-fluorodeoxyglucose (FDG) positron emissiontomography (PET) in determining lymph node involve-ment as well as distant spread of disease [4–11]. Thelimited anatomical resolution of PET can be overcome byco-registration of functional PET images with morpholog-ical CT data with an integrated PET/CT system. It has beenshown that the use of dual-modality PET/CT significantlyimproves the accuracy of staging of NSCLC in comparisonto PET or CT alone, and positively affects therapeuticmanagement [12–14]. In the context of maximising theefficiency of PET/CT so that it becomes a real “one-stopshop” examination in patients with NSCLC, a centralquestion is how much CT is needed in order to addressspecific issues in staging NSCLC [15–17]. Today, differentapproaches are adopted for PET/CT scanning, with the CTprotocol positioned between two possible extremes: (a) CTis used as a fast transmission source for attenuationcorrection and approximate anatomical mapping and isperformed with a low radiation dose (“low dose CT”), and(b) CT is used for both attenuation correction anddiagnostic purposes, being performed with a standardradiation dose and intravenous and oral contrast (“diag-nostic CT”). The multi-phase PET/CT protocol used in ourinstitution offers the possibility of comparing state-of-the-art contrast-enhanced CT with the non-contrast low dosescan for attenuation correction in the same patient at thesame time, both integrated in the multi-phase study [18].The objective of our study was to evaluate the need fordiagnostic CT protocols in staging and therapy planning ofpatients with NSCLC by comparing non-enhanced andcontrast-enhanced PET/CT protocols.

Materials and methods

Patient population

Data of 50 consecutive patients with proven NSCLC who underwenta multi-phase contrast-enhanced PET/CT at our institution betweenOctober 2004 and May 2005 were analysed retrospectively. Patientcharacteristics are shown in Table 1. Most patients were referred forprimary tumour staging (n=41) before surgery (n=4), neoadjuvantradiochemotherapy (n=24), definitive radiotherapy (n=8) or pallia-tive chemotherapy (n=5). Nine patients were referred for restagingduring follow-up owing to suspected local recurrence and/or distantmetastases. All patients considered for surgery or neoadjuvantradiochemotherapy had undergone conventional clinical staging atan external hospital (Schillerhöhe Hospital, Gerlingen) before thePET/CT examination. The conventional staging was based onmedical history, physical findings, blood tests, bronchoscopy andendosonography, mediastinoscopy with lymph node sampling, bonescan and CTof the chest. The combination of these tests and invasiveprocedures provided the basis for the initial staging with the TNM(tumour-node-metastasis) and AJCC (American Joint Committee onCancer) staging systems [19]. In most cases, this initial staging CTwas performed on an out-patient basis at different institutions with atime interval often too long to provide the basis for accurate therapyplanning. Therefore, in all patients without recently performed state-of-the-art diagnostic CT but designated to receive tumour resection,neoadjuvant therapy or definitive radiotherapy, the PET/CT protocolincluded a contrast-enhanced multi-phase CT which met the

requirements of our thoracic surgeons and radio-oncologists. Allpatients gave their written informed consent after having beeninformed about the PET/CT examination in detail.

Patient preparation

Patients fasted overnight (for at least 6 h) prior to the intravenousadministration of 18F-FDG. The injected dose of 18F-FDG variedbetween 350 and 450 MBq depending on the patient’s weight. Inpatients with BMI <25, 350–400 MBq was administered, while thosewith a BMI >25 received 400–450 MBq. The blood glucose wasmeasured before injection of the tracer to ensure blood glucose levels<11.1 mmol/l. During the uptake phase of 60 min, the patients wereinstructed to rest comfortably. All patients were asked to drink1,000 ml mannitol 2% as a negative oral contrast agent prior toscanning in order to distend the bowel.

PET/CT imaging protocol

In all patients, PET/CTwas performed using the Hi-Rez Biograph 16(Siemens Medical Solutions, Knoxville, TN, USA), consisting of ahigh-resolution 3D LSO PET scanner and a state-of-the-art 16-rowmulti-slice CT. The PET scanner has an axial field of view (FOV) of15.5 cm and a slice thickness of 4.25 mm. Emission data wereacquired for six to eight bed positions, typically from the base of theskull to the upper thigh. The PET acquisition time was usually 3 minper FOV, though patients with a BMI >25 were examined for 4 minper FOV. CT was operated with a peak voltage of 120 kV, a tubecurrent of 30 mAs (low dose) or 120–160 mAs (standard dose), arotation time of 0.5 s, collimation of 0.75 mm (thorax) and 1.5 mm(abdomen), a table feed of 12/24 mm, and a reconstructed slicethickness/increment of 5/5 mm (axial) and 3/2 mm (coronal),respectively. Patients were positioned on the scanning table with theirarms raised in order to reduce beam-hardening artefacts. The multi-phase CT protocol included a low dose (30 mAs) non-enhanced scan

Table 1. Patient characteristics

Characteristic No. of patients

Total number 50SexMale 40Female 10Age (years)Range 39–81Mean 61IndicationPrimary staging 41Restaging 9Tumour localisationCentrally 33Peripherally 17HistologySquamous cell carcinoma 18Adenocarcinoma 15Large cell carcinoma 8Undifferentiated carcinoma 9

37


first for attenuation correction owing to the fact that the currentlyavailable scanner software does not allow for the use of partial-bodyspiral CT for attenuation correction. The low dose scan is followedby a bolus-triggered arterial phase thorax/liver scan, a portal-venousabdomen/pelvis scan and the whole-body PET from the base of theskull to the upper thigh. The intravenous contrast volume of 120 mlwas administered with a flow of 2 ml/s. To prevent contrast-inducedartefacts, we optimised the injection protocol with a 40 ml salinechaser. During preliminary studies, we tested different scanning andbreathing protocols to optimise contrast-enhanced CT studies [18].According to the results of our tests, patients were asked to stopbreathing in normal expiration during the contrast-enhanced CTscans, as well as during the low dose attenuation scan so as to obtainoptimal co-registration.

Image reconstruction

Diagnostic CT scans were reconstructed with 3 mm slice thickness inthe coronal and 5 mm in the axial plane (reconstruction increment 2and 5 mm, respectively). In addition, several special post-processingreconstructions and window level settings (lung and bone window,MIP reconstruction) were applied. The non-enhanced CT data wereused for attenuation correction of PET emission images. PET imageswere reconstructed by using an iterative algorithm (ordered-subsetexpectation maximisation: 2 iterations, 8 subsets). The reconstructedPET, CT and fused images were displayed by commercially availablesoftware (e-soft/VSIM, Siemens Medical Solutions) in axial, coronaland sagittal planes using a matrix of 128×128 pixels for the PET and512×512 pixels for the CT.

Data interpretation

Tumour staging with PET/CT was based on the AJCC/TNM systemfor the classification of lung cancer [19]. Image analysis wasperformed in several steps. In the first step, the fused contrast-enhanced PET/CT images (CE PET/CT) were evaluated by twoexperienced nuclear medicine specialists (S.M.E., M.M.) and twoexperienced CT radiologists (A.C.P., P.A.) in consensus. The readerswere blinded to the results of other staging examinations. In a secondinterpretation session (3 months later), the fused datasets of low dosenon-enhanced PET/CT (LD PET/CT) were analysed by thecombined reader team and compared with the results of fused CEPET/CT to determine the additional value of contrast application perpatient and per lesion. On PET images, any focal tracer uptakeexceeding normal regional tracer accumulation at qualitative analysisand an average standardised uptake value (SUVavg) of more than 2.5for mediastinal lesions (derived from region of interests applying the50% isocontour technique) were assessed as a malignant lesion.Lymph node involvement in CT relating to the relevant lymph nodelevels was based on nodal size with a short-axis diameter greater than10 mm or presence of necrosis defined as pathological [20]. Weanalysed lymph node levels instead of single lymph nodes accordingto the clinical needs and practice as well as to prevent bias inevaluation by over-representation of single patients with multiplelymph node metastases. In the case of equivocal findings of bothexaminations, the classification was based on the PET resultsregardless of the size of lymph nodes. The following criteria wereapplied for the per patient analysis: (a) change of TNM classificationby CE PET/CT in comparison to the classification based on LDPET/CT (0=no change, 1=upstaging, 2=downstaging, 3=staging onlypossible with contrast application), (b) detection of NSCLC-relatedand non-NSCLC-related additional findings, triggering therapy orfurther work-up. The per lesion evaluation of CE PET/CT and LD

PET/CT images was based on assessment of accuracy of each methodin terms of certainty of lesion localisation (0=uncertain localisation,1=definite localisation), lesion characterisation (0=not suspicious,1=suspicious for malignancy) and certainty of lesion delineation(0=uncertain lesion delineation, 1=definite lesion delineation).Lesion delineation means the precise definition of tumour extentincluding assessment of invasion into the adjacent tissue and appliesto the primary tumour as well as to involved lymph nodes and distantmetastases.

Standard of reference

The standard of reference for confirming regional lymph nodeinvolvement in all patients before neoadjuvant therapy was given byhistological results of mediastinoscopy with lymph node sampling.All mediastinoscopies were performed before the PET/CT at anexternal hospital (Schillerhöhe Centre of Thoracic Medicine,Gerlingen). The time interval between the two examinations was3–4 weeks. On PET/CT images, no postoperative tissue changes inthe mediastinum have been detected. The evaluation of primarytumours was accomplished by the combination of invasive (bron-choscopy, endosonography) and non-invasive clinical stagingprocedures (initial chest CT) and the follow-up. In patients whounderwent surgery, the T stage and N stage were verified byhistopathological results. The validation of distant metastases wasperformed by complementary imaging methods and/or clinical–radiological follow-up for at least 6 months.

Statistical analysis

The McNemar test was used to determine the statistical significanceof differences in lesion-based analysis between LD PET/CT and CEPET/CT. A p value of less than 0.05 was considered to indicatestatistical significance.

Results

Patient-based analysis

TNM staging

The results of TNM staging according to the findings of LDPET/CT and CE PET/CT and compared with the referencestandard are listed in Table 2. Comparing the results of Tstaging between CE PET/CTand LD PET/CT, there was nodifference in 47/50 patients (94%). The remaining threepatients could only be correctly classified as having T4tumours by performing CE PET/CT (Fig. 1). In the 4/50patients referred for restaging after surgery, no local tumourrecurrence was found.

Regarding N staging, there was no difference betweenthe results of LD PET/CT and those obtained by CEPET/CT. The verification of N staging was established bylymph node sampling during mediastinoscopy in 31patients, by surgical histopathological results in fourpatients and by follow-up examinations in 15 patients.Comparing the N staging by PET/CT with the results ofmediastinoscopy, both LD PET/CT and CE PET/CT

38


correctly determined lymph node metastases in 25 patients(there was overstaging in three patients and understaging inthree), whereas mediastinoscopy missed N3 disease in one

supraclavicular and one very small contralateral lymphnode.

Evaluation of M staging revealed that, compared withLD PET/CT, CE PET/CT provided improved results inonly 1 of the 50 patients. This patient, with revised M stageupon contrast application, showed an adrenal lesion withmoderate FDG uptake (Fig. 2a–c) that was suspicious formalignancy on PET, but demonstrated the typical morpho-logical and dynamic pattern (lipid content and washout) foran adenoma on the diagnostic CT (Fig. 2d). Tworadiological follow-up examinations with lack of changeover a period of >1 year confirmed the presence of a benignlesion. The sites of distant metastases are listed in Table 3.Most metastases were found in the lungs and bones, withno liver metastases being detected. Four patients with smallpulmonary nodules (10 mm diameter or less) in lobes otherthan the lobe containing the primary carcinoma werefalsely classified as having M1 disease by both LDPET/CT and CE PET/CT, but downstaging by follow-updid not alter the treatment plan in these patients.

AJCC staging

The results of AJCC staging according to the findings ofLD PET/CT and CE PET/CT and compared with thereference standard are listed in Table 4. The majority of thepatients (82%) presented with advanced NSCLC and wereclassified as stage III and IV. Comparing the overall staging

Table 2. TNM staging of 50 patients with NSCLC according to lowdose (LD) PET/CT and contrast-enhanced (CE) PET/CT

TNM LD PET/CT CE PET/CT Referenceb

No tumoura 4 (8%) 4 (8%) 4 (8%)T1 1 (2%) 1 (2%) 3 (6%)T2 12 (24%) 12 (24%) 9 (18%)T3 4 (8%) 4 (8%) 4 (8%)T4 26 (52%) 29 (58%) 30 (60%)Unclear (T3/T4) 3 (6%) 0 (0%) 0 (0%)N0 19 (38%) 19 (38%) 19 (38%)N1 3 (6%) 3 (6%) 3 (6%)N2 11 (22%)c 11 (22%)c 11 (22%)N3 17 (34%)c 17 (34%)c 17 (34%)M0 35 (70%) 36 (72%) 40 (80%)M1 15 (30%) 14 (28%) 10 (20%)

aNo local recurrence tumour detectablebT stage: surgery (n=4), initial conventional staging (n=46);N stage: mediastinoscopy (n=31), follow-up (n=15), surgery (n=4);M stage: complementary imaging and/or follow-up (n=50)cOverstaging of three patients (N2/N3) and understaging of threepatients (N3/N2)

Fig. 1. A 75-year-old malepatient with proven NSCLC andstaging before neoadjuvanttherapy. LD 18F-FDG PET/CTprovides an inadequate assess-ment of tumour extension (a, b).Only on CE PET/CT (c, d) isit possible to precisely delineatetumour invasion of themediastinum and oesophagus(asterisk), as well as theencasement of the right pulmo-nary vein (arrow), consistentwith a T4 tumour stage. LA leftatrium

39


results between LD PET/CT and CE PET/CT, there was achange in TNM classification on the basis of CE PET/CTin 4/50 (8%) patients, affecting the T staging in threepatients and the M staging in one patient. The difference inT stage (T3/T4) did not result in an alteration of the AJCCstage (IIIb); the patient with a modified M stage (M1 toM0) was the only one with a change in the AJCC stage byCE PET/CT.

Lesion-based analysis

The lesion-based analysis covered a total of 92 sites,including 46 primary tumours, 31 involved lymph nodelevels and 15 distant metastases. The evaluation referred tolesion localisation, lesion characterisation and lesiondelineation (Fig. 3). The results demonstrated no differencein the accuracy of lesion localisation between the LD PET/CT and the CE PET/CT. Regarding lesion characterisation,CE PET/CT correctly changed classification of one distantmetastasis (adrenal lesion) from malignant to benign, asconfirmed by follow-up. In the remaining 91 suspicioussites, no differences in lesion classification between LDPET/CT and CE PET/CT could be found. In contrast,precise lesion delineation was uncertain through LD PET/

CT in 59/92 sites (64%), whereas CE PET/CT failed inonly 1 of 92 sites, relating to faintly visible bonemetastases. In 63% of tumour sites, CE PET/CT alteredlesion delineation in comparison to LD PET/CT. Thesechanges referred to 2/15 distant metastases, 27/31 lymphnode involvements and 29/46 primary tumours, 27 ofwhich were centrally situated. The differences in lesiondelineation between LD PET/CT and CE PET/CT werestatistically significant (p<0.0001).

Additional findings

In 10/50 (20%) patients, CE PET/CT detected clinicallyimportant additional findings that could not be diagnosedwith LD PET/CT; these additional findings influencedclinical management by triggering therapeutic interventionand/or further diagnostic work-up (Table 5, Fig. 4).

Discussion

Several studies have demonstrated that dual-modalityPET/CT offers the most efficient and accurate approach tothe staging of NSCLC. However, the optimal PET/CT

Fig. 2. a–c A 62-year-old malepatient with proven NSCLC ands/p resection of the left upperlobe who was referred forrestaging. The MIP reconstruc-tion (a), axial 18F-FDG PET(b) and the fused image (c)demonstrate a hypermetabolicfocus in the left adrenal gland(SUVavg 3.1) (red arrow)consistent with an adrenal me-tastasis of the NSCLC. d Multi-phase CE CT reveals thetypical morphological anddynamic pattern (lipid contentand washout) of a corticaladenoma (confirmed by 1-yearof follow-up)

40


scanning protocol is still a point of debate [12–14, 16, 17,21–23]. Summarising the literature, the CT part of thecombined PET/CT protocol in NSCLC staging is com-monly applied as a whole-body low dose scan withoutintravenous contrast [13, 21, 26] or as a contrast-enhancedsingle-phase CT in a few centres [15, 27, 28]. The questionis how much CT, enhanced or non-enhanced, is needed forthe specific requirements in staging and therapy planning ofNSCLC [15, 16]. The results of our study comparing LDPET/CTand CE PET/CT in patients with proven NSCLC of

stages III and IV indicate a limited influence of intravenouscontrast and standard dose application on the overall TNMstaging in patients with advanced NSCLC. TNM stagingwas changed by diagnostic CT in 8% of patients, but in onlyone patient was there a change in the clinical (AJCC) stageand a relevant impact on treatment by virtue of downstagingfrom M1 to M0. In the other three cases, the upstaging of

Fig. 2 (continued)

Table 3. Site of distant metastases in 50 patients with NSCLCaccording to LD PET/CT and CE PET/CTa

Site LD PET/CT CE PET/CT Referenceb

Lung 7 7 3Bone 6 6 6Distant lymph nodes 3 3 3Adrenal glands 4 3 3Abdominal wall 1 1 1

aPatients with distant metastases in more than one site arementioned several timesbComplementary imaging and/or clinical–radiological follow-upfor ≥6 months

Table 4. AJCC staging of 50 patients with NSCLC according to LDPET/CT and CE PET/CT

AJCC stage LD PET/CT CE PET/CT Reference

No tumoura 3 (6%) 4 (8%) 4 (8%)b

Ia 0 0 0Ib 4 (8%) 4 (8%) 4 (8%)c

IIa 0 0 0IIb 1 (2%) 1 (2%) 1 (2%)b

IIIa 5 (10%) 5 (10%) 5 (10%)d

IIIb 22 (44%) 22 (44%) 26 (52%)d

IV 15 (30%) 14 (28%) 10 (20%)b

aExclusion of local tumour recurrence and distant metastasesbFollow-up ≥6 months (T, N and M staging)cSurgery (T and N staging)dMediastinoscopy (N staging), follow-up (T staging)

41


the T stage by CE PET/CT did not affect patient manage-ment. As expected, the classification of lymph nodeinvolvement was not influenced by contrast application:both groups, LD PET/CT and CE PET/CT achieved thesame results. This is in accordance with the results ofAntoch et al. [12]. PET/CT results in N staging are mainlybased on the high accuracy of PET in the detection of lymphnode involvement, as has been demonstrated in severalstudies [1, 5–8, 10]. Concerning the M staging, the resultswere similar. Except in one patient, the application ofcontrast had no additional diagnostic impact on thedetection and characterisation of distant metastases. How-ever, there were no liver metastases in our patientpopulation, for which CE CT as part of the PET/CTwould have been of substantial benefit.

In contrast to the limited influence of diagnostic CT onthe overall TNM staging, our results demonstrate the clearsuperiority of CE PET/CT over LD PET/CT in theassessment of lesion delineation. This applies to thedistinct delineation of mediastinal and chest wall invasionby the primary tumour as well as to the exact demarcationof involved lymph nodes and distant metastases. Compre-hensive information regarding tumour delineation is aprecondition for both planning of 3D conformal radiother-apy and preparation of extended non-conventional resec-tions of locally advanced NSCLC [24, 25]. For planning ofsurgery, the essential questions to be answered by PET/CTare whether the tumour has spread to the hilar andmediastinal lymph nodes and whether the primary tumourhas deeply invaded the mediastinum, central vessels andchest wall. Radiotherapists need to know whether theprimary tumour and the nodal metastases will beencompassed within a suitable radiation field in such amanner that critical organ doses are not exceeded [25]. Thisentails precise delineation of all involved tumour sites.Only CE PET/CT including a state-of-the-art CT canadequately meet these requirements. This is in agreementwith the findings of Strobel et al., who reported a potentialbenefit of CE PET/CT in patients with central bronchialtumours [16]. In 63% of our patients, tumour delineationwas modified by CE PET/CT compared with LD PET/CTprotocols. As expected, this was especially true for patientswith centrally located NSCLCs presenting a complexanatomical situation. In addition, CE PET/CT identifiedclinically important NSCLC-related and non-NSCLC-related incidental findings in 20% of the patients, affectingthe clinical management to varying degrees. The properconsideration of incidental findings on the CT part of thecombined protocol was addressed by Schoeder et al., whounderlined the clinical importance of a comprehensiveinterpretation of both datasets of a PET/CT examination

Fig. 3. The lesion-based analy-sis of 92 suspicious sites, in-cluding 46 primary tumours, 31involved lymph node levels and15 distant metastases, showingno differences in lesion local-isation, one revision oflesion characterisation by CEPET/CT and statistically signif-icant differences between LD(non-enhanced) and CE PET/CTprotocols regarding lesiondelineation

Table 5. Patients with significant additional findings detected by CEPET/CT only

Additional findings No. of patients

NSCLC-related findingsLeft atrial involvement by primary tumour 1Compression of the SVC by bulky lymph nodes 1Precise delineation of bone/soft tissue metastasisof cervical spine

1

Non-NSCLC-related findingsHypervascularised liver mass suspicious for HCC 1Complicated renal cyst (indeterminate lesion) 1Pericardial cyst 1Solid pancreatic mass 1Large abdominal aortic aneurysm 1Thrombosis of portal and splenic vein 1Extensive uterine myomatosis 1

SVC superior vena cava, HCC hepatocellular carcinoma

42


[29]. Osman et al. found a 3% prevalence of significantincidental CT lesions on non-enhanced PET/CT studies[30]. The higher portion of significant findings in our studyis likely to have been due to the contrast-enhanced CTprotocol as well as the inclusion of a majority of patientspresenting with advanced NSCLC.

It is to be noted that the non-enhanced low-dose CT scan(30 mAs) as part of our multi-phase PET/CT protocol wasexclusively performed for attenuation correction (and notfor diagnostic purposes) owing to the fact that the currentlyavailable scanner software does not allow use of partial-body spiral CT for attenuation correction of the wholebody. In contrast, in the literature the term “low dose” CT isapplied to a wide dose range from 30 mAs up to 100 mAs,which affects CT image quality differently. The overallradiation dose of our multi-phase PET/CT protocol(including a low dose scan, two contrast CT phases andwhole-body PET) is ca. 28 mSv [18] and, therefore, higherthan that of a non-enhanced PET/CT or single-phase CEPET/CT protocol. However, regarding the radiation dose,two points have to be taken into consideration. First, in ourinstitution the PET/CT referral is understood as referral foran “all-in-one” examination that meets the different needsof referring physicians. Secondly, our study population hada mean age of 61 years, presented with advanced NSCLCand had in most cases been designated to receive radio-therapy; as a consequence, the radiation exposure fromPET/CT imaging is unlikely to have adversely affectedlong-term outcome.

Based on the results of this study, we recommend astandard dose contrast-enhanced PET/CT in all patientswith proven NSCLC who are primarily under considera-tion for local therapy such as surgery, neoadjuvantradiochemotherapy or definitive radiotherapy. Low dosenon-enhanced PET/CT can be performed in patients

primarily under consideration for systemic therapy, patientswith recently performed state-of-the-art whole-body CT,patients referred for therapy monitoring and patients withrenal insufficiency or contrast allergy.

Limitations of the study

The results of our study may have been influenced by ourpatient population, with a majority presenting locallyadvanced, centrally situated NSCLC. The added value ofcontrast-enhanced CT as part of the PET/CT protocol couldbe minor in patients with localised disease of clinical stageI or II. Furthermore, the order in which the PET/CT datasets were evaluated, with the CE PET/CT first, may havebiased the study results. Even when considering the timeinterval of 3 months between the evaluation of the two datasets, this may have falsely improved the staging accuracyof the non-enhanced approach.

Conclusion

While contrast-enhanced CT as part of the PET/CTprotocol more accurately assessed the TNM stage in 8%of patients with advanced NSCLC compared with the non-contrast protocol, this change in the TNM stage had only aminor influence on patient management. However, con-trast-enhanced PET/CT protocols are indispensable inthese patients for planning 3D conformal radiotherapyand non-conventional extended surgical resections whichrequire precise definition of tumour extension.

Fig. 4. Detection of clinically important additional findings by CEPET/CT. A 66-year-old male patient with proven NSCLC of theright upper lobe and mediastinal lymph node involvement. The MIPreconstruction (a) and axial 18F-FDG PET of the thorax (b) andabdomen (e) demonstrate high tumour uptake, but no abdominallesion. The CE CT part of PET/CT shows significant compression of

the superior vena cava (d, black arrow) with venous congestion andcollateral flow via anterior thoracic veins (d, white arrow), as well asa hypervascularised liver mass suspicious for hepatocellular carci-noma (g, white arrow). Neither finding was reliably detectable onthe LD (non-enhanced) CT (c, f)

43


Acknowledgements.Wewould like to thank our technicians HenrietteHeners, Sylvia Storz, Benjamin Gerras, Anke Reckwell, Gabi Boutinand Tina Brutschy for their helpful assistance in the acquisition of thePET/CT data. We gratefully acknowledge the continuing support ofour colleagues of the Radiopharmacy Department.

References

1. Deslauriers J, Gregoire J. Clinical and surgical staging of non-small cell lung cancer. Chest 2000;117:96S–103S

2. Barker JM, Silvestri GA. Lung cancer staging. Curr Opin PulmMed 2002;8:287–293

3. Laroche C, Fairbairn I, Moss H, Pepke-Zaba J, Sharples L,Flower C, et al. Role of computed tomographic scanning of thethorax prior bronchoscopy in the investigation of suspectedlung cancer. Thorax 2000;55:359–363

4. Steinert HC, Hauser M, Allemann F, Engel H, Berthold T, vonSchulthess GK, et al. Non-small lung cancer: nodal stagingwith FDG PET versus CTwith correlative lymph node mappingand sampling. Radiology 1997;202:441–446

5. Pieterman RM, van Putten JWG, Meuzelaar JJ, Mooyaart EL,Vaalburg W, Koeter GH, et al. Preoperative staging of non-small-cell lung cancer with positron-emission tomography. NEngl J Med 2000;343:254–561

6. Sachs S, Bilfinger T. The impact of positron emission tomogra-phy on clinical decision making in a university-based multi-disciplinary lung cancer practice. Chest 2005;128:698–703

7. Lloyd C, Silvestri GA. Mediastinal staging of non-small-celllung cancer. Cancer control 2001;8:311–317

8. Gould MK, Kuschner WG, Rydzak CE, Maclean CC, DemasAN, Shigemitsu H, et al. Test performance of positron emissiontomography and computed tomography for mediastinal stagingin patients with non-small-cell lung cancer. Ann Intern Med2003;139:879–892

9. Gupta NC, Graeber GM, Bishop HA. Comparative efficacy ofpositron emission tomography with fluorodeoxyglucose inevaluation of small (<1 cm), intermediate (1 to 3 cm), and large(>3 cm) lymph node lesions. Chest 2000;117:773–778

10. Vansteenkiste JF, Stroobants SG, De Leyn PR, Dupont PJ,Bogaert J, Maes A, et al. Lymph node staging in non-small-celllung cancer with FDG-PET scan: a prospective study on 690lymph node stations from 68 patients. J Clin Oncol1998;16:2142–2149

11. van Tinteren H, Hoekstra OS, Smit EF, van den Bergh JH,Schreurs AJ, Stallaert RA, et al. Effectiveness of positronemission tomography in the preoperative assessment of patientswith suspected non-small-cell lung cancer: the PLUS multi-centre randomised trial. Lancet 2002;359:1388–1393

12. Antoch G, Stattaus J, Nemat AT, Marnitz S, Beyer T, Kuehl H,et al. Non-small cell lung cancer: dual-modality PET/CT inpreoperative staging. Radiology 2003;229:526–533

13. Lardinois D, Weder W, Hany TF, Kamel EM, Korom S, SeifertB, et al. Staging of non-small-cell lung cancer with integratedpositron emission tomography and computed tomography. NEngl J Med 2003;348:2500–2507

14. Goerres GW, von Schulthess GK, Steinert HC. Why most PETof lung and head-and-neck cancer will be PET/CT. J Nucl Med2004;45(suppl):66S–71S

15. Antoch G, Freudenberg LS, Beyer T, Bockisch A, Debatin J. Toenhance or not to enhance? 18F-FDG and CT contrast agents indual-modality 18F-FDG PET/CT. J Nucl Med 2004;45(suppl):56S–65S

16. Strobel K, Thuerl M, Hany TF. How much intravenous contrastis needed in FDG-PET/CT? Nuklearmedizin 2005;44:S32–S37

17. von Schulthess GK. Integrated modality imaging with PET-CTand SPECT-CT: CT issues. Eur Radiol Suppl 2005; 15 Suppl 4:D121–D126

18. Brechtel K, Klein M, Vogel M, Mueller M, Aschoff P, Beyer T,et al. Optimized contrast-enhanced CT protocols for diagnosticwhole-body 18F-FDG-PET/CT: technical aspects of single-phase versus multi-phase CT imaging. J Nucl Med2006;47:470–476

19. Mountain CF. Revisions in the international system for staginglung cancer. Chest 1997;111:1710–1717

20. Glazer GM, Gross BH, Quint LE, Francis IR, Bookstein FL,Orringer MB. Normal mediastinal lymph nodes: number andsize according to American Thoracic Society mapping. AJR1985;144:261–265

21. Schöder H, Erdi YE, Larson SM, Yeung HWD. PET/CT: a newimaging technology in nuclear medicine. Eur J Nucl Med MolImaging 2003;30:1419–1437

22. Beyer T, Antoch G, Müller S, Egelhof T, Freudenberg LS,Debatin J, et al. Acquisition protocol considerations forcombined PET/CT imaging. J Nucl Med 2004;45(suppl):25S–35S

23. Beyer T, Antoch G, Bockisch A, Stattaus J. Optimizedintravenous contrast administration for diagnostic whole-body18F-FDG PET/CT. J Nucl Med 2005;46:429–435

24. Detterbeck FC, Jones DR, Kernstine KH, Naunheim KS;American College of Physicians. Lung cancer. Special treat-ment issues. Chest 2003;123:244S–258S

25. Brianzoni E, Rossi G, Ancidei S, Berbellini A, Capoccetti F,Cidda C, et al. Radiotherapy planning: PET/CT scannerperformances in the definition of gross tumour volume andclinical target volume. Eur J Nucl Med Mol Imaging2005;32:1392–1399

26. Bar-Shalom R, Yefremov N, Guralnik L, Gaitini D, Frenkel A,Kuten A, et al. Clinical performance of PET/CT in evaluationof cancer: additional value for diagnostic imaging and patientmanagement. J Nucl Med 2003;44:1200–1209

27. Berthelsen AK, Holm S, Loft A, Klausen TL, Andersen F,Højgaard L. PET/CT with intravenous contrast can be used forPET attenuation correction in cancer patients. Eur J Nucl MedMol Imaging 2005;32:1167–7511

28. Kapoor V, McCook BM, Torok FS. An introduction to PET-CTimaging. RadioGraphics 2004;24:523–543

29. Schoeder H, Yeung HWD, Larson SM. CT in PET/CT:essentialfeatures of interpretation. J Nucl Med 2005;46:1249–1251

30. Osman MM, Cohade C, Fishman EK, Wahl RL. Clinicallysignificant incidental findings on the unenhanced CT portion ofPET/CT studies; frequency in 250 patients. J Nucl Med2005;46:1352–1355

44


Download - Low dose non-enhanced CT versus standard dose contrast-enhanced CT in combined PET/CT protocols for staging and therapy planning in non-small cell lung cancer

Top Related