ORIGINAL ARTICLE

Favourable outcomes of 177Lu-octreotate peptide receptorchemoradionuclide therapy in patients with FDG-avidneuroendocrine tumours

Raghava Kashyap & Michael S. Hofman &

Michael Michael & Grace Kong & Timothy Akhurst &Peter Eu & Diana Zannino & Rodney J. Hicks

Received: 16 July 2014 /Accepted: 25 August 2014# Springer-Verlag Berlin Heidelberg 2014

AbstractPurpose Increased glycolytic activity on FDG PET/CT de-f i n e s a subg roup o f pa t i e n t s w i t h me t a s t a t i cgastroenteropancreatic neuroendocrine tumour (NET) with apoor prognosis. A limited range of systemic treatment optionsexist for more aggressive NET. The role of peptide receptorchemoradionuclide therapy (PRCRT) in such patients is, how-ever, unclear. This retrospective study assessed the outcomesof patients with FDG-avid NET treated with PRCRT.Methods Clinical, biochemical and imaging response wasassessed after completion of induction treatment of PRCRTwith 5-fluorouracil in 52 patients selected for treatment on thebasis of somatostatin-receptor imaging without spatially dis-cordant FDG-avid disease. Of the cohort, 67 % had receivedprior chemotherapy. Overall survival (OS) and progression-free survival (PFS) were also analysed.Results PRCRT was well tolerated with negligible grade 3/4toxicities. After a median follow-up period of 36 months, the

medianOSwas not achievedwith amedian PFS of 48months.At 3 months after completion of PRCRT 2 % of patientsshowed a complete anatomical response, 28 % a partial re-sponse, 68 % stable disease, and only 2 % progression. OnFDG PET/CT, 27 % achieved a complete metabolic responseduring the follow-up period. A biochemical response (>25 %fall in chromogranin-A levels) was seen in 45 %.Conclusion PRCRT is an effective treatment in patients withFDG-avid NET, even in patients who have failed conventionaltherapies. Given apparently higher response rates than withalternative therapeutic options and low toxicity, further re-search is needed to establish whether PRCRT should be usedas a first-line treatment modality in this patient population.

Keywords Peptide receptor radionuclide therapy (PRRT) .

Lutetium-177DOTA-TATE . 5-Fluorouracil . 18F-FDG .

Neuroendocrine tumour

Dr. R. Kashyap and A/Prof M.S. Hofman are joint first authors.

Electronic supplementary material The online version of this article(doi:10.1007/s00259-014-2906-4) contains supplementary material,which is available to authorized users.

R. Kashyap :M. S. Hofman :G. Kong : T. Akhurst : P. Eu :R. J. Hicks (*)Centre for Cancer Imaging,Peter MacCallum Cancer Center,St Andrews Place, 3002 Melbourne, Australiae-mail: rod.hicks@petermac.org

R. Kashyape-mail: kashyapk@americanoncology.com

M. S. Hofmane-mail: michael.hofman@petermac.org

M. MichaelDepartment of Medicine, University of Medicine, Melbourne,Australia

M. S. Hofman :G. Kong : T. Akhurst : P. Eu :R. J. HicksNeuroendocrine Tumour Service, Peter MacCallum Cancer Centre,Melbourne, Australia

M. MichaelDivision of Cancer Medicine, Peter MacCallum Cancer Centre,Melbourne, Australia

D. ZanninoBiostatistics and Clinical Trials, Peter MacCallum Cancer Centre,Melbourne, Australia

M. Michael : R. J. HicksThe Sir Peter MacCallum Department of Oncology, University ofMelbourne, Melbourne, Australia

Eur J Nucl Med Mol ImagingDOI 10.1007/s00259-014-2906-4

Introduction

Neuroendocrine tumours (NET) demonstrate a variable natu-ral history. Although generally indolent, they can demonstrateaggressive behaviour with rapid growth and early death if notmodified by effective therapeutic intervention. In 2010, in anattempt to provide prognostic information and guide treat-ment, the European Neuroendocrine Tumour Society(ENETS) and World Health Organization (WHO) establisheda new pathological grading system that stratifies patientsbased on Ki-67 staining [1]. Grade 1 NET (Ki-67 ≤2 %) arewell differentiated and generally have a long natural historybut a poor response to chemotherapy. Conversely, grade 3neuroendocrine carcinoma (NEC, Ki-67 >20%) progressivelylose features of differentiation and demonstrate rapid progres-sion despite having higher response rates, albeit transiently, tochemotherapy [2]. Grade 2 NET (Ki-67 3 – 20 %) has anintermediate and often unpredictable natural history betweenthese extremes. Consequently, optimal treatment options forthis group of patients remain unclear. Several chemotherapyregimens have been evaluated and recommendations havebeen formulated based on the site of origin and tumour grade[3, 4]. However, responses in a given individual are unpre-dictable and the adverse effects of combined chemotherapycan be significant.

Advances in molecular imaging, including PET with so-matostatin receptor (SSTR) analogues such as 68Ga-DOTA-octreotate (GaTate) and 18FDG have provided new insightsinto NET biology. By performing both scans sequentially, it ispossible to identify heterogeneity of phenotype between andwithin patients. While GaTate avidity is a feature of well-differentiated disease, FDG avidity tends to be associated withmore aggressive, de-differentiated disease [5]. Grade 1 NETtend to be GaTate-avid but negative on FDG PET, whereasgrade 3 NEC generally show the opposite imaging phenotype.Grade 2 NET may demonstrate uptake of both tracers.Irrespective of pathological grade, the distribution of thesetracers may not be spatially concordant, with some lesionshaving either GaTate or FDG avidity, but not both. Thishighlights the limitations of relying on histopathological gradefrom a single biopsy site to predict disease behaviour. Despitethe prognostic utility of pathological grading, FDG PET pos-itivity has been consistently shown to be independently asso-ciated with a poor prognosis [6–9]. A recent prospective studydemonstrated strong prognostic value with FDG PET positiv-ity conferring a hazard ratio of 10.3 for risk of death, exceed-ing the value of traditional markers such as Ki-67,chromogranin-A (CgA) and anatomical stage [10].

SSTR expression on the surface of NET enables the use ofsomatostatin analogues labelled with particle-emitting radio-nuclides for targeted peptide receptor radionuclide therapy(PRRT). The principal agents used are 90Y-DOTA-octreotateand 177Lu-DOTA-octreotate (LuTate). Some groups,

including ours, have combined radiosensitizing chemotherapywith peptide receptor chemoradionuclide therapy (PRCRT)which demonstrates a toxicity profile similar to that ofPRRT alone [11, 12]. The median progression-free survival(PFS) achieved with LuTate PRRT in a large series from theErasmus group was 40 months [13]. This compares veryfavourably with the results of recent phase III studies ofeverolimus and sunitinib, which both demonstrated PFS ofapproximately 11 months in patients randomized to the treat-ment arm [14, 15]. However, as a feature of well-differentiateddisease, the requirement for SSTR expression in selectingpatients for PRRT in the Erasmus study may itself haveintroduced a bias for longer survival when compared to ther-apeutic trials in which this was not an eligibility criterion.Further, the Dutch series did not utilize FDG PET/CT as partof the diagnostic work-up and, although not recorded, theremay have been an excess of grade 1 NET. In contradistinctionto this, we have recently reported our initial experience withLuTate PRCRT, which included 68 patients treated on com-passionate grounds for progressive disease within 12 months(58 patients) or uncontrolled symptoms (10 patients) on con-ventional therapy [16]. As a probable consequence of oureligibility criteria, at least 44 % of our study population hadgrade 2 NETand at least 35 % had FDG-avid disease (>liver).Nevertheless, after a median follow-up of 60 months, a medi-an overall survival (OS) was not reached.

These data are somewhat at odds with the poor prognosisreported for FDG-avid disease in earlier series. We thereforeevaluated the survival, efficacy and toxicity of treatment withLuTate PRCRT in an expanded but overlapping subgroup ofpatients undergoing PRCRT despite a positive FDG PET/CTscan prior to treatment.

Materials and methods

Patients

We conducted a single-centre analysis of consecutive patientswith FDG-avid NETwho were treated with LuTate PRCRT incombination with radiosensitizing 5-fluorouracil (5-FU). Ourinstitutional protocol is to perform FDG PET/CTas part of thebaseline evaluation of all NET patients with a Ki-67 >5 %, or,irrespective of histopathological grade, progression of diseaseover a period of less than 6 months or with lesions on ana-tomical imaging that lack SSTR expression [5]. Malignantdisease was considered FDG-avid if the intensity of uptakewas equal to or greater than that of the liver. Eligibility criteriafor performing PRCRT included inoperable locally advancedor unresectable metastatic gastroenteropancreatic NET withconcordant high SSTR expression as defined by a Krenningscore of 3 or more (uptake >liver) [17] on 68Ga-DOTA-TATEPET/CTor 111In-octreotide (Octreoscan;Mallinckrodt, Petten,

Eur J Nucl Med Mol Imaging

The Netherlands) SPECT/CT at known sites of disease.Patients also had to qualify for compassionate use of thistherapy by having either uncontrolled symptoms on maximal-ly tolerated somatostatin analogue therapy, or progressivedisease. Progressive disease was defined as: (1) new sites ofdisease on SSTR imaging, (2) progression on CT byResponseEvaluation Criteria in Solid Tumors (RECIST) 1.0, or (3) asignificant increase in CgA levels within the previous12 months. At the time of this study, treatment with chemo-therapy was mandated in all patients with highly FDG-aviddisease prior to being considered for PRCRT, unless consid-ered unsuitable because of medical comorbidity.

Exclusion criteria for PRCRT included: (1) spatially dis-cordant FDG-avid disease showing low or absent SSTR ana-logue avidity, (2) uncorrected carcinoid heart disease withsevere valvular dysfunction, (3) glomerular filtration rate<30 ml/min, (4) Eastern Cooperative Oncology Group(ECOG) score of 4, (5) expected survival <3 months, (6)platelet count rate <50,000/L, or (7) serum albumin level<25 g/L indicating advanced hepatic biosynthetic dysfunc-tion, particularly if accompanied by ascites and peripheraloedema. Data were gathered from a prospective NET databasesupplemented by review of data from our hospital electronicmedical records system. The study was approved by theinstitutional ethics board. All patients provided written in-formed consent to receive PRCRT.

Treatment protocol

177Lu was obtained from IDB-Holland (Petten/Baarle Nassau,Netherlands). Each patient received 6 – 10 GBq of LuTateadapted to tumour burden as part of a serial treatment regimencomprising an induction course consisting of three to fivecycles spaced 6 – 8 weeks apart followed by maintenancedosages as needed 12 – 18 months after induction.Radiosensitizing 5-FU was administered as a continuous in-fusion at a dose of 200 mg/m2/day from the second cycle ofPRRT starting 4 days prior to radionuclide administration andcontinuing for up to 3 weeks. During the period of this study,5-FU was omitted from the first cycle to minimize flare as aresult of hormone secretion. An infusion of renoprotectiveamino acids consisting of 25 g lysine and 25 g arginine in1 L of normal saline was administered intravenously over 4 hbeginning 30 min before LuTate administration.

Efficacy and safety assessment

Response was assessed 3 months after completion of induc-tion PRCRT.

Biochemical response The change in CgA levels from base-line to the end of induction was assessed.

Anatomical imaging response RECIST 1.1 [18] was usedmodified by the use ofmolecular imaging to select appropriateindex lesions. In particular, primarily necrotic or cystic lesionsas indicated by extensive central photopaenia on both GaTateand FDG PET/CT imaging were excluded. Lesions weremeasured on diagnostic CT images. If lesions were poorlydemarcated on contrast-enhanced CT images, or thesewere not available, the lesion margins were measuredon the low-dose noncontrast CT image. On follow-up,lesions that had enlarged as a result of necrosis orhaemorrhage as identified by loss of uptake on bothFDG PET/CT and SSTR images were not consideredto reflect progressive disease and evaluation of RECISTresponse was limited to target lesions not demonstratingthese features on functional imaging.

Molecular imaging response For both SSTR and FDG PET/CT, a complete response was defined as resolution of uptake atall sites, a partial response as a decrease in avidity of lesions ora decrease in the number of lesions, progressive disease as anincrease in avidity of lesions or the appearance of new malig-nant lesions, and stable disease as a response not conformingto these groups.

Safety assessment All adverse effects occurring from thetime of PRCRT administration to the end of induction,defined as 3 months after the last dose of LuTate, wererecorded. Toxicity to the treatment was defined in termsof the Common Terminology Criteria for AdverseEvents version 4.0.

Statistical analysis

PFS and OS were evaluated using Kaplan-Meier curvesmeasured from the start of treatment induction. For PFS,the patient was determined to have progressive disease ifthere was evidence of development of new lesions on atleast one of the imaging modalities used in follow-up.The log-rank test was used to compare PFS and OSbetween baseline subgroups (CgA level <100 U/L vs.≥100 U/mL, ECOG performance status, Ki-67 grade,and FDG uptake grade) and where appropriate the log-rank test for trend is reported. A subgroup analysis basedon FDG avidity was done with a SUV cut-off of 9.0, asthis was demonstrated in a previous study to be a prog-nostic marker for OS [10]. The QLQ-C30 and QLQ-GINET21 quality of life (QoL) questionnaires were com-pleted at baseline and follow-up. Changes in QoL scoresfrom baseline to follow-up were analysed using theWilcoxon signed-ranks test. Analyses were carried outusing SAS version 9.2 (SAS Institute, Cary, NC). Allstatistical inferences were based on two-sided p values of≤0.05, with no adjustment for multiple comparisons.

Eur J Nucl Med Mol Imaging

Results

Patients

Between October 2005 and March 2012 a total of 52 patients(27 men , 25 women) were eligible for the study. The baselinedemographic and clinical characteristics of the patients arepresented in Table 1. Their mean age was 56 years (range

23 – 75 years). The majority of the patients (34, 65 %) hadnonfunctional NET. The Ki-67 histopathological grading wasavailable in 36 patients, with the majority (28, 78 %) havingENETS grade II tumours. As a quaternary referral centre notall patients had their initial biopsy at our facility, but archivalmaterial was sought and reviewed if available.

On FDG PET/CT, the median SUVmax of the cohort was7.2 (range 2.9 – 47), with 25 % of the cohort having intensemetabolic abnormality, as defined by a SUVmax greater than10. Prior chemotherapy had been given in 35 patients. Thisincluded carboplatin and etoposide in 27, while the otherregimens used either alone or in addition were streptozotocinwith 5-FU in 8 patients, cyclophosphamide, doxorubicin andvincristine (CAV) in three patients, and single-agent temozo-lomide in one patient. In some patients more than one chemo-therapy regimen had been tried. Prior somatostatin analoguetherapy was given in 24 patients. The median follow-upperiod from the beginning of PRCRT was 36 months (range9 – 72 months). The median induction administered activitywas 33 GBq (range 9 – 45 GBq) over a median of four cycles(range one to six).

Efficacy

Survival analysis

At the median follow-up of 36 months, 35 patients (67 %)were alive and amedian OSwas not achieved (Fig. 1).MedianPFS was 48 months. At 24 months of follow-up 74 % patientswere alive. Univariate analysis was done to determine ifbaseline CgA levels, Ki-67 grade or FDG uptake grade dem-onstrated any effect on OS, but none was statistically signif-icant. The median PFS in patients with SUVmax less than 9was 48 months while that in patients with SUVmax ≥9 it was26 months, but this difference was not statistically significant(log-rank p=0.2).

Biochemical response

A total of 41 patients had an abnormal CgA level atbaseline, defined as more than twice the upper normallimit of 17.2 U/L. In 16 of these 41 patients (31 %) theCgA level decreased by more than 50 % from the pre-treatment level, in 17 (14 %) the level decreased by25 – 50 %, in 14 (35 %) the level was stable, and in 4(10 %) the level increased (Fig. 2a).

Imaging response

Follow-up somatostatin imaging was available in all but onepatient who had a rapid deterioration in hepatic function andsuccumbed prior to imaging follow-up (Table 2). Anatomicalresponse according to RECIST was evaluable in 40 patients

Table 1 Baseline patient demographics and clinical characteristics

Characteristic Number(%)

Total patients 52 (100)

Sex

Male 27 (52)

Female 25 (48)

ENETS grade (Ki-67) (n=36)

1 (<2 %) 4

2 (3 – 20 %) 28

3 (≥20 %) 4

FDG avidity prior to PRCRT

Equal to Liver 4 (8)

More than Liver 23 (44)

Intense uptake (SUVmax >10) 25 (48)

Pretreatment CgA levels (ng/ml)) (n=41)

35 – 500a 16

501 – 1,000 10

>1,000 15

Previous therapy

Surgery 23

Chemotherapy 35

Somatostatin analogue (octreotide or lanreotide) 24

Radiotherapy (to skeletal lesion) 1

Radiofrequency ablation followed by intraarterial therapy 1

Diagnosis

Carcinoid syndrome 9

Small intestine 4

Colonic 1

Rectal 2

Unknown primary 2

Functional pancreatic metastatic NET

Insulinoma 9

Gastrinoma 3

Glucagonoma 1

VIPoma 2

Nonfunctional metastatic NET 34

Unknown primary 12

Pancreatic 21

Colonic 1

a 35 ng/ml: more than twice the upper limit of normal

Eur J Nucl Med Mol Imaging

with a partial response in 11 patients (28 %), stable disease in27 (68%), a complete response in one and progressive diseasein one (Fig. 2b). In nine patients, target lesions were notnonevaluable on CT precluding RECIST measurements. In a

further three patients, baseline or restaging CT data weremissing precluding assessment.

Follow-up FDG PET imaging at 3 months was available in31 patients, and the best response on follow-up was also

Fig. 1 Kaplan-Meier curves for (a) OS and (b) PFS (with 95 % confidence bands)

Fig. 2 Changes in (a) CgA and(b) RECIST 1.1 anatomicalresponse 3 months aftercompletion of PRCRT

Eur J Nucl Med Mol Imaging

recorded. At 3 months, 5 patients (16 %) had a completemetabolic response (CMR), 17 (55 %) a partial metabolicresponse (PMR), 5 (16 %) stable disease, and 4 (13 %4)progressive metabolic disease. Three patients with a PMRachieved a CMR on later follow-up PET, bringing the totalnumber of patients achieving a CMR to eight (27 %). The PFSwas significantly different in patients belonging to differentcategories of best FDG PET response, but the OS was notsignificantly different. In patients with a CMR median PFSwas not achieved, in those with a PMR it was 26 months, inthose with stable disease it was 17 months and in those withprogressive disease it was 5 months, and these differenceswere statistically significant (p<0.001).

Quality of life

The ECOG scores improved in 41 patients (79 %). Overall, 24patients had clinically significant NET-related symptoms.There was a significant improvement in the physical function-ing of the patients on the QLQ-C30 scale and in endocrine-related symptoms, social function and body image on theQLQ-GINET 21 scale (all p<0.05).

Safety

Acute transient exacerbation of endocrine symptoms such asflushing during administration of therapy was noted in sixpatients. A transient cutaneous reaction at the site of peripheralamino acid infusion was seen in four patients. PRCRT wasvery well tolerated with grade 3 or 4 toxicities in only fourpatients during or following induction PRCRT (see supple-mentary Table 1). One patient had grade 2 mucositis, probablydue to 5-FU infusion. The most common adverse effects werehaematological toxicity with leucopenia in 42 % (but thisprimarily reflected lymphopenia rather than neutropenia; nograde 3/4), thrombocytopenia in 38 % (grade 3/4 in 6 %) andanaemia in 38 % (no grade 3/4). The sole patient with grade 4thrombocytopenia had disseminated intravascular coagulationand died following three cycles of PRCRT. Two patients hadgrade 3 thrombocytopenia which persisted until death from

progressive disease 20 and 34 months after induction PRRT.Both these patients had received carboplatin and etoposidechemotherapy prior to PRCRT. In all, six patients had hepa-totoxicity (grade 1 in three, grade 2 in two, grade 3 in one).The one patient with grade 3 hepatotoxicity died due toprogressive hepatic metastases.

Discussion

The role of PRRT in the management of NET is now wellestablished but continues to evolve [11, 13, 19–26]. Whetherthe addition of a radiosensitizing agent such as capecitabineincreases the efficacy of PRRT is being evaluated in a ran-domized trial, but small studies have demonstrated minimaltoxicity with the addition of 5-FU or capecitabine to LuTatePRRT [27]. In a phase II study, Claringbold et al. [11] foundthat combination therapy with capecitabine and temozolomideresulted in tumour control and stabilization of disease in 94 %of 33 patients evaluated with a median follow-up of16 months. Kesavan et al. [28] further demonstrated thathaematological toxicity was not significantly increased withthis combination. The use of concomitant chemotherapy isparticularly attractive in patients with more aggressive dis-ease. In this study, we used metabolic abnormality on FDGPET/CT to define such a population on the basis of prospec-tive data demonstrating that such patients have significantlyworse OS and PFS [10]. FDG avidity is commensurate withtumour grade, with tumours of poorer differentiation (ENETSgrade 2 and 3) tending to show higher FDG avidity [6–8, 29].

Our study included a patient cohort with a number ofadverse prognostic factors in addition to FDG avidity, includ-ing prior progression in the preceding 12 months, failure ofconventional therapy including aggressive chemotherapy reg-imens in the majority of patients, and a high proliferationindex. Despite expectations of a poor prognosis in our cohort,during the median follow-up period of 36 months, the medianOS was not reached and the median PFS was 48 months. Thisis significantly higher than reported for other therapeuticmodalities including everolimus [15, 30], sunitinib [14], 5-FU/streptozocin [31], capecitabine/temozolomide [32] or so-matostatin analogues alone or in combination with 5-FU[33–35]. In these studies, PFS ranged from 5.3 to 22.6months,which appears significantly inferior to that following PRCRT.

On the basis of these data, we postulate that tumours withFDG avidity, reflecting higher rates of proliferation, are moresensitive to the effect of radiation and provided that they retainsufficient SSTR to deliver adequate radiation doses, may beuniquely sensitive to PRCRT. As tumours become morededifferentiated and lose SSTR expression, targeting withPRRT becomes constrained despite potentially enhanced ra-diosensitivity. Thus, we believe that there may be a “sweet

Table 2 Imaging responses at 3 months after completion of PRCRT

RECIST 1.1(n=40)

FDG PET(n=31)

SSTR imaging(s=51)

Complete response 1 5 1

Partial response 11 17 32

Stable disease 27 5 1

Progressive disease 1 4 17a

a A proportion of these patients had a mixed response with response indominant preexisting lesions but progression of existing small-volumedisease defining progression

Eur J Nucl Med Mol Imaging

spot” for PRCRT response in NET biology corresponding totumours in mid-to-high grade 2 and low grade 3. The additionof a radiosensitizing agent to the regimen may have contrib-uted to the benefit in the setting of more rapidly proliferatingdisease. Some patients were continued on cold octreotidetherapy between the PRRT cycles as well as after PRCRT.This could have partially contributed to the disease control,although almost all of these patients had already been on, andprogressed through, this therapy previously.

In our study, RECIST anatomical response could only beassessed in 40 patients (77 %). This is not uncommon in NETsince many lesions seen on SSTR imaging are poorly seen onconventional imaging, and some lesions can only be transient-ly visualized in certain phases of contrast-enhanced studies,rendering accurate, reproducible measurement difficult. Wesought to overcome this limitation by modifying our evalua-tion of RECIST response by referencing CT findings to thoseon functional imaging. Even using this approach, a significantproportion of lesions and patients could not be assessed.Nevertheless, 30 % of evaluable patients showed a completeor partial response while disease stability was achieved in68 % of patients 3 months after completing their inductioncourse of PRCRT.

On FDG PET/CT follow-up, 27 % of patients achieved aCMR despite most patients having residual disease evident onSSTR imaging (see Figs. 3 and 4). This suggests that PRCRT

was able to eradicate the poorly differentiated disease withremaining disease being of a more well-differentiated pheno-type. Our favourable OS and PFS results support this hypoth-esis. The survival plots show a trend for patients with higherbaseline CgA level, higher Ki-67 grade and higher FDGuptake to show an OS period that was shorter relative to othergroups (Supplementary Fig. 1), which is in agreement with theresults of previous studies [10, 36, 37]. Only four patients hadprimary refractory disease with progression on metabolicimaging. Of these patients, one was treated for symptomaticreasons after failing chemotherapy but, in retrospect, did notmeet our eligibility criteria given the presence of discordantFDG-avid disease at baseline. Two patients progressed withFDG-avid SSTR-negative disease, and one patient progressedwith FDG-avid disease that, however, remained SSTR-posi-tive. This patient developed cord compression requiring ur-gent external beam radiotherapy but succumbed due to pro-gressive disease.

Two patients had a complete response on all imagingmodalities. One of these patients, who was initially assessedas inoperable after carboplatin and etoposide chemotherapy,underwent successful R0 surgical resection following PRCRTand has previously been reported [38]. Despite achievinginitial control, disease progression still occurred in some ofthe patients. The most common pattern of relapse was multi-focal bone disease. Four patients developed spatially

Fig. 3 A 46-year-old man with metastatic rectal neuroendocrine carci-noma (ENETS grade 3, Ki-67 40 %). There was progressive diseasedespite six cycles of carboplatin/etoposide chemotherapy and long-actingoctreotide hormonal therapy, with worsening liver function tests, increas-ing pain, weight loss and deteriorating performance status. 68Ga-DOTA-TATE PET/CT (a, c) demonstrates high SSTR expression in widespread

diffuse hepatic metastases. The patient was treated with four cycles ofLuTate PRCRT, with a near complete imaging response (b, d) andmarked clinical improvement including resolution of pain. The patientreturned to work, and at the time of this report remained alive and well38 months after induction PRCRT, despite two recurrences both treatedwith maintenance PRCRT

Eur J Nucl Med Mol Imaging

discordant FDG-positive SSTR-negative disease renderingthem ineligible for further PRCRT. Some patients showed amixed response on SSTR imaging with the dominant baselinelesions showing favourable response but progression or de-velopment of new disease foci usually less than 1 cm in sizeon the 3-month scan. We postulate that these sites did notreceive an adequate radiation dose from PRCRT due to smalltumour volume and consequent inefficient cross-fire effects.We consider these patients to be still suitable for furtherPRCRT.

This studywas limited by its retrospective nature in a singleinstitution setting. Although our protocol mandates restagingat 3 months following completion of PRCRT, this was notperformed in each patient resulting in incomplete imagingresponse data at this time point. Nevertheless, no patients werelost to clinical follow-up and we observed remarkable anddurable responses in patients with a poor prognosis and per-formance status who failed multiple lines of prior therapy, and

in this context, we contend that this constitutes a high level ofevidence of efficacy [18, 39, 40]. Moreover, there is a cleardirect biological mechanism of action, further strengthened bya personalized approach with SSTR and FDG imaging thatpreselects individual patients that are likely to respond totherapy. A significant limitation was the inability to quantifythe additional value of radiosensitizing chemotherapy in thisstudy.

Conclusion

PRCRT is a highly effective treatment option in patients withFDG-avid NET, even in patients who have failed conventionaltherapies. Although most studies of PRRT and PRCRT haveincluded patients with indolent NET, our study shows thatpatients with higher tumour proliferation are also sensitive toradionuclide therapy. Given apparently higher response ratescompared to historical controls and low toxicity, further

Fig. 4 A 62-year-old woman with pancreatic gastrinoma invading theduodenum with multiple hepatic metastases (Ki-67 20 %). There wasprogressive clinical and radiological disease despite carboplatin/etoposide, streptozocin/5-FU chemotherapy and a trial of everolimus,which was not tolerated. Following this, SSTR SPECT/CT and FDGPET/CT imaging demonstrated concordant high uptake in the primarysite (red arrows) and hepatic metastases (awhole-body planar image after177Lu-DOTA-TATE therapy, b FDG PET/CT maximum intensity projec-tion, e axial SPECT/CT image of primary lesion after therapy). At thestart of PRCRT the patient was symptomatic with bleeding from the

duodenum (haemoglobin 79 g/L). Following four cycles of LuTate he-patic metastases had disappeared on all imaging, with a major reductionin size and a PMR on FDG PET/CT in the primary mass (c). Gastrin andCgA decreased from 21,000 to 1,400 pmol/L and 217 to 139 U/L,respectively. Following two further cycles of PRCRT there was a CMRon FDG PET/CT at all sites (not shown) and a further reduction in theprimary mass (d SSTR planar image and f SPECT/CT image followingcycle 6 of PRCRT). At the time of this report at 32 months follow-up, thepatient was asymptomatic with a sustained imaging response

Eur J Nucl Med Mol Imaging

research is needed to establish whether PRCRT could be usedas a first-line treatment modality and to define the optimalconcurrent chemotherapy regimens. In particular, comparisonwith emerging treatments advocated for patients with grade 2NET, such as capecitabine and temozolomide, everolimus orsunitinib, which were not available at the time of this study,are required.

Conflicts of interest None.

Funding Professor Rodney Hicks was supported by a translationalresearch grant of the Victorian Cancer Agency. Dr. K. RaghavaKashyap was supported by Endeavour Awards, a venture ofAustraining International in partnership with the Department ofEducation, Employment and Workplace Relations, Government ofAustralia.

References

1. Klimstra DS, Modlin IR, Coppola D, Lloyd RV, Suster S. Thepathologic classification of neuroendocrine tumors: a review of no-menclature, grading, and staging systems. Pancreas. 2010;39:707–12. doi:10.1097/MPA.0b013e3181ec124e.

2. Sorbye H,Welin S, Langer SW, Vestermark LW, Holt N, Osterlund P,et al. Predictive and prognostic factors for treatment and survival in305 patients with advanced gastrointestinal neuroendocrine carcino-ma (WHO G3): the NORDIC NEC study. Ann Oncol. 2013;24:152–60. doi:10.1093/annonc/mds276.

3. Oberg KE. The management of neuroendocrine tumours: current andfuture medical therapy options. Clin Oncol (R Coll Radiol). 2012;24:282–93. doi:10.1016/j.clon.2011.08.006.

4. Pavel M, Baudin E, Couvelard A, Krenning E, Oberg K, SteinmullerT, et al. ENETS Consensus Guidelines for the management of pa-tients with liver and other distant metastases from neuroendocrineneoplasms of foregut, midgut, hindgut, and unknown primary.Neuroendocrinology. 2012;95:157–76. doi:10.1159/000335597.

5. HofmanMS, Hicks RJ. Changing paradigms with molecular imagingof neuroendocrine tumors. Discov Med. 2012;14:71–81.

6. Adams S, BaumR, Rink T, Schumm-Drager PM,Usadel KH, Hor G.Limited value of fluorine-18 fluorodeoxyglucose positron emissiontomography for the imaging of neuroendocrine tumours. Eur J NuclMed. 1998;25:79–83.

7. Kayani I, Bomanji JB, Groves A, Conway G, Gacinovic S, Win T,et al. Functional imaging of neuroendocrine tumors with combinedPET/CT using 68Ga-DOTATATE (DOTA-DPhe1, Tyr3-octreotate)and 18F-FDG. Cancer. 2008;112:2447–55. doi:10.1002/cncr.23469.

8. Pasquali C, Rubello D, Sperti C, Gasparoni P, Liessi G, ChierichettiF, et al. Neuroendocrine tumor imaging: can 18F-fluorodeoxyglucosepositron emission tomography detect tumors with poor prognosis andaggressive behavior? World J Surg. 1998;22:588–92.

9. van Essen M, Sundin A, Krenning EP, Kwekkeboom DJ.Neuroendocrine tumours: the role of imaging for diagnosis andtherapy. Nat Rev Endocrinol. 2014;10:102–14. doi:10.1038/nrendo.

10. Binderup T, Knigge U, Loft A, Federspiel B, Kjaer A. 18F-fluorodeoxyglucose positron emission tomography predicts survivalof patients with neuroendocrine tumors. Clin Cancer Res. 2010;16:978–85. doi:10.1158/1078-0432.CCR-09-1759.

11. Claringbold PG, Brayshaw PA, Price RA, Turner JH. Phase II studyof radiopeptide 177Lu-octreotate and capecitabine therapy of

progressive disseminated neuroendocrine tumours. Eur J Nucl MedMol Imaging. 2011;38:302–11. doi:10.1007/s00259-010-1631-x.

12. Hubble D, Kong G, Michael M, Johnson V, Ramdave S, Hicks RJ.177Lu-octreotate, alone or with radiosensitising chemotherapy, issafe in neuroendocrine tumour patients previously treated withhigh-activity 111In-octreotide. Eur J Nucl Med Mol Imaging.2010;37:1869–75. doi:10.1007/s00259-010-1483-4.

13. Kwekkeboom DJ, de Herder WW, Kam BL, van Eijck CH, vanEssen M, Kooij PP, et al. Treatment with the radiolabeled somato-statin analog [177Lu-DOTA 0, Tyr3]octreotate: toxicity, efficacy, andsurvival. J Clin Oncol. 2008;26:2124–30. doi:10.1200/JCO.2007.15.2553.

14. Raymond E, Dahan L, Raoul JL, Bang YJ, Borbath I, Lombard-Bohas C, et al. Sunitinib malate for the treatment of pancreaticneuroendocrine tumors. N Engl J Med. 2011;364:501–13. doi:10.1056/NEJMoa1003825.

15. Yao JC, Shah MH, Ito T, Bohas CL, Wolin EM, Van Cutsem E, et al.Everolimus for advanced pancreatic neuroendocrine tumors. N EnglJ Med. 2011;364:514–23. doi:10.1056/NEJMoa1009290.

16. Kong G, Thompson M, Collins M, Herschtal A, Hofman MS,Johnston V. Assessment of predictors of response and long-termsurvival of patients with neuroendocrine tumour treated with peptidereceptor chemoradionuclide therapy (PRCRT). Eur J Nucl Med MolImaging. 2014. doi:10.1007/s00259-014-2788-5.

17. Krenning EP, Valkema R, Kooij PP, Breeman WA, Bakker WH, deHerder WW. Scintigraphy and radionuclide therapy with [indium-111-labelled-diethyl triamine penta-acetic acid-D-Phe1]-octreotide.Ital J Gastroenterol Hepatol. 1999;31 Suppl 2:S219–223.

18. Eisenhauer EA, Therasse P, Bogaerts J, Schwartz LH, Sargent D,Ford R, et al. New response evaluation criteria in solid tumours:revised RECIST guideline (version 1.1). Eur J Cancer. 2009;45:228–47.

19. Krenning EP, Kooij PP, Pauwels S, Breeman WA, Postema PT, DeHerder WW, et al. Somatostatin receptor: scintigraphy and radionu-clide therapy. Digestion. 1996;57 Suppl 1:57–61.

20. Kong G, Johnston V, Ramdave S, Lau E, Rischin D, Hicks RJ. High-administered activity In-111 octreotide therapy with concomitantradiosensitizing 5FU chemotherapy for treatment of neuroendocrinetumors: preliminary experience. Cancer Biother Radiopharm.2009;24:527–33. doi:10.1089/cbr.2009.0644.

21. Claringbold PG, Price RA, Turner JH. Phase I-II study ofradiopeptide 177Lu-octreotate in combination with capecitabineand temozolomide in advanced low-grade neuroendocrine tumors.Cancer Biother Radiopharm. 2012;27:561–9. doi:10.1089/cbr.2012.1276.

22. Cwikla JB, Sankowski A, Seklecka N, Buscombe JR, Nasierowska-Guttmejer A, Jeziorski KG, et al. Efficacy of radionuclide treatmentDOTATATE Y-90 in patients with progressive metastaticgastroenteropancreatic neuroendocrine carcinomas (GEP-NETs): aphase II study. Ann Oncol. 2010;21:787–94. doi:10.1093/annonc/mdp372.

23. Horsch D, Ezziddin S, HaugA,Gratz KF, Dunkelmann S, Krause BJ,et al. Peptide receptor radionuclide therapy for neuroendocrine tu-mors in Germany: first results of a multi-institutional cancer registry.Recent Results Cancer Res. 2013;194:457–65. doi:10.1007/978-3-642-27994-2_25.

24. Romer A, Seiler D,Marincek N, Brunner P, Koller MT, NgQK, et al.Somatostatin-based radiopeptide therapy with [177Lu-DOTA]-TOC versus [90Y-DOTA]-TOC in neuroendocrine tumours.Eur J Nucl Med Mol Imaging. 2014;41:214–22. doi:10.1007/s00259-013-2559-8.

25. Villard L, Romer A,Marincek N, Brunner P, Koller MT, Schindler C,et al. Cohort study of somatostatin-based radiopeptide therapy with[(90)Y-DOTA]-TOC versus [(90)Y-DOTA]-TOC plus [(177)Lu-DOTA]-TOC in neuroendocrine cancers. J Clin Oncol. 2012;30:1100–6. doi:10.1200/JCO.2011.37.2151.

Eur J Nucl Med Mol Imaging

26. Imhof A, Brunner P, Marincek N, Briel M, Schindler C, Rasch H,et al. Response, survival, and long-term toxicity after therapywith theradiolabeled somatostatin analogue [90Y-DOTA]-TOC in metasta-sized neuroendocrine cancers. J Clin Oncol. 2011;29:2416–23. doi:10.1200/JCO.2010.33.7873.

27. van Essen M, Krenning EP, Kam BL, de Herder WW, van Aken MO,KwekkeboomDJ. Report on short-term side effects of treatments with177Lu-octreotate in combination with capecitabine in seven patientswith gastroenteropancreatic neuroendocrine tumours. Eur J Nucl MedMol Imaging. 2008;35:743–8. doi:10.1007/s00259-007-0688-7.

28. Kesavan M, Claringbold PG, Turner JH. Hematological toxicity ofcombined Lu-octreotate radiopeptide chemotherapy ofgastroenteropancreatic neuroendocrine tumors in long-term follow-up. Neuroendocrinology. 2014;99:108–17. doi:10.1159/000362558.

29. Binderup T, Knigge U, Loft A, Mortensen J, Pfeifer A, Federspiel B,et al. Functional imaging of neuroendocrine tumors: a head-to-headcomparison of somatostatin receptor scintigraphy, 123I-MIBG scin-tigraphy, and 18F-FDG PET. J Nucl Med. 2010;51:704–12. doi:10.2967/jnumed.109.069765.

30. Yao JC, Phan AT, Chang DZ, Wolff RA, Hess K, Gupta S, et al.Efficacy of RAD001 (everolimus) and octreotide LAR in advancedlow- to intermediate-grade neuroendocrine tumors: results of a phaseII study. J Clin Oncol. 2008;26:4311–8. doi:10.1200/JCO.2008.16.7858.

31. Sun W, Lipsitz S, Catalano P, Mailliard JA, Haller DG; EasternCooperative Oncology Group. Phase II/III study of doxorubicin withfluorouracil compared with streptozocin with fluorouracil ordacarbazine in the treatment of advanced carcinoid tumors: EasternCooperative Oncology Group Study E1281. J Clin Oncol. 2005;23:4897–904.

32. Strosberg JR, Fine RL, Choi J, Nasir A, Coppola D, Chen DT, et al.First-line chemotherapy with capecitabine and temozolomide in pa-tients with metastatic pancreatic endocrine carcinomas. Cancer.2011;117:268–75. doi:10.1002/cncr.25425.

33. Rinke A, Muller HH, Schade-Brittinger C, Klose KJ, Barth P, WiedM, et al. Placebo-controlled, double-blind, prospective, randomized

study on the effect of octreotide LAR in the control of tumor growthin patients with metastatic neuroendocrine midgut tumors: a reportfrom the PROMID Study Group. J Clin Oncol. 2009;27:4656–63.doi:10.1200/JCO.2009.22.8510.

34. Brizzi MP, Berruti A, Ferrero A, Milanesi E, Volante M, CastiglioneF. Continuous 5-fluorouracil infusion plus long acting octreotide inadvanced well-differentiated neuroendocrine carcinomas. A phase IItrial of the Piemonte oncology network. BMC Cancer. 2009;9:388.doi:10.1186/1471-2407-9-388.

35. Delpassand ES, Samarghandi A, Mourtada JS, Zamanian S, EspenanGD, Sharif R, et al. Long-term survival, toxicity profile, and role of F-18 FDG PET/CT scan in patients with progressive neuroendocrinetumors following peptide receptor radionuclide therapy with highactivity In-111 pentetreotide. Theranostics. 2012;2:472–80. doi:10.7150/thno.3739.

36. Yao JC, Pavel M, Phan AT, Kulke MH, Hoosen S, St Peter J,et al. Chromogranin A and neuron-specific enolase as prog-nostic markers in patients with advanced pNET treated witheverolimus. J Clin Endocrinol Metab. 2011;96:3741–9. doi:10.1210/jc.2011-0666.

37. Pape UF, Berndt U, Muller-Nordhorn J, Bohmig M, Roll S, Koch M,e t a l . P rognos t i c f a c to r s o f l ong - t e rm ou t come ingastroenteropancreatic neuroendocrine tumours. Endocrinol RelatCancer. 2008;15:1083–97. doi:10.1677/ERC-08-0017.

38. Barber TW, Hofman MS, Thomson BN, Hicks RJ. The potential forinduction peptide receptor chemoradionuclide therapy to render in-operable pancreatic and duodenal neuroendocrine tumours resect-able. Eur J Surg Oncol. 2012;38:64–71. doi:10.1016/j.ejso.2011.08.129.

39. Kipritidis J, Siva S, Hofman MS, Callahan J, Hicks RJ, Keall PJ.Validating and improving CT ventilation imaging by correlating withventilation 4D-PET/CT using 68Ga-labeled nanoparticles. MedPhys. 2014;41:011910. doi:10.1118/1.4856055.

40. Stewart DJ, Kurzrock R. Fool’s gold, lost treasures, and the random-ized clinical trial. BMC Cancer. 2013;13:193. doi:10.1186/1471-2407-13-193.

Eur J Nucl Med Mol Imaging