ORIGINAL ARTICLE

Treatment outcomes of patients with cervical cancerwith complete metabolic responsesafter definitive chemoradiotherapy

Cem Onal & Mehmet Reyhan & Ozan C. Guler &

Ali Fuat Yapar

Received: 29 September 2013 /Accepted: 29 January 2014# Springer-Verlag Berlin Heidelberg 2014

AbstractPurpose We sought to evaluate failure patterns and prognosticfactors predictive of recurrences and survival in cervical can-cer patients who are treatedwith definitive chemoradiotherapy(ChRT), who have a subsequent complete metabolic response(CMR) with 18F-fluorodeoxyglucose positron-emission to-mography (FDG-PET) after treatment.Methods The records of 152 cervical cancer patients whowere treated with definitive chemoradiotherapy were evaluat-ed. All patients underwent pre-treatment positron emissiontomography (PET-CT), and post-treatment PET-CT was per-formedwithin amedian of 3.9months (range, 3.0–9.8months)after the completion of ChRT. The prognoses of partialresponse/progressive disease (PR/PD) cases (30 patients,18 %) and CMR cases (122 patients, %82) were evaluated.Univariate and multivariate analysis effecting the treatmentoutcome was performed in CMR cases.Results The median follow-ups for all patients and survivingpatients were 28.7 (range, 3.3–78.7 months) and 33.2 months(range, 6.23–78.7 months), respectively. Four-year overallsurvival (OS) rate was significantly better in patients withCMR compared to patients with PR/PD (66.9 % vs. 12.4 %,p<0.001, respectively). Patients with PR/PD had higher max-imum standardized uptake value (SUVmax) of primary cervical

tumor (26.4±10.1 vs. 15.9±6.3; p<0.001) and larger tumor(6.4 cm±2.3 cm vs. 5.0 cm±1.4 cm; p<0.001) compared topatients with CMR. Of the 122 patients with post-treatmentCMRs, 25 (21 %) developed local, locoregional, or distantfailure. In univariate analysis, tumor size ≥ 5 cm, ‘InternationalFederation of Obstetricians and Gynecologists’ (FIGO)stage≥IIB, and pelvic and/or para-aortic lymph node metasta-sis were predictive of both overall survival (OS) and disease-free survival (DFS), while histology was predictive of onlyOS. In multivariate analysis, tumor size, stage and lymph nodemetastasis were predictive of OS and DFS.Conclusion Although CMR is associated with better out-comes, relapses remain problematic, especially in patientswith bulky tumors (≥ 5 cm), extensive stage (≥ IIB) or pelvicand/or para-aortic lymph node metastasis. These findingscould support the need for more aggressive treatment oradjuvant chemotherapy regimens.

Keywords Cervical cancer . Gynecological cancer .

Positron-emission tomography . Radiotherapy .

Chemotherapy

Introduction

18F-fluorodeoxyglucose positron-emission tomography(FDG-PET) integrated with computed tomography (CT) in-corporates metabolic tumor function with anatomic localiza-tion. PET-CT has become an increasingly important compo-nent of staging tumor localization and assessing treatmentresponse for many malignancies. Furthermore, some studiesconcluded that the efficacy of FDG uptake measured as themaximum standardized uptake value (SUVmax) was predictiveof outcome [1–4].

It has also been demonstrated that metabolic responsedetermined by FDG-PET performed within 8–16 weeks after

C. Onal :O. C. GulerDepartment of Radiation Oncology, Baskent University Faculty ofMedicine, Adana, Turkey

M. Reyhan :A. F. YaparDepartment of Nuclear Medicine, Baskent University Faculty ofMedicine, Ankara, Turkey

C. Onal (*)Adana Research and Treatment Centre, Department of RadiationOncology, Baskent University Faculty of Medicine, Adana 01120,Turkeye-mail: hcemonal@hotmail.com

Eur J Nucl Med Mol ImagingDOI 10.1007/s00259-014-2719-5

the completion of therapy can be used to predict outcome[2, 5]. Schwarz et al. [6] reported 3-year cause-specific sur-vival rates of 100 % in patients with complete metabolicresponses (CMRs), 51 % for patients with partial responses(PRs), and 17 % for patients with progressive disease (PD). Inthis study, the majority (85 %) of patients who relapsed haddistant failures. In another study, Beriwal et al. [7] reported 3-year actuarial disease-free survival (DFS) and overall survival(OS) rates of 79 and 88 %, respectively, in patients with CMRafter a median follow-up of 15 months. Furthermore, wepreviously reported that the primary tumor pretreatmentSUVmax for patients with CMR was significantly lower thanthat of patients with PR or PD (15.6±5.7 vs 28.0±9.9, respec-tively; p<0.001) [3].

Although patients with CMRs had better outcomes, thepatterns and factors predicting recurrence and survival inpatients with CMRs after definitive radiotherapy (RT) hasnot yet been well studied. Given our recent study demonstrat-ing the prognostic importance SUVmax of cervical tumor [3],with a longer median follow-up compared to previous reports[7], we aimed to assess failure patterns and prognostic factorspredicting local or distant recurrences and survival in patientswith cervical cancer who were treated with definitive chemo-radiotherapy (ChRT), with a subsequent CMR after the com-pletion of treatment.

Methods and materials

Patients

We reviewed the records of 152 patients with biopsy-provencervical cancer who were treated with definitive ChRTwith acurative intent between November 2006 and August 2012atour institution, and who underwent follow-up FDG-PET atleast 3 months following the completion of therapy. This studywas approved by Baskent University Institutional ReviewBoard (Project no: KA 13/288) and supported by BaskentUniversity Research Fund.

All patients underwent routine clinical staging, includingmedical history reviews, physical and gynecological exami-nations, blood chemistry tests, and complete blood counts.The patients were staged according to the ‘InternationalFederation of Obstetricians and Gynecologists’ (FIGO) stag-ing system. All patients had also undergone PET-CT for initialdiagnosis, staging, and RT planning. Eighty-two patients(54 %) had undergone magnetic resonance imaging [8] toevaluate local disease. The exclusion criteria were a previousdiagnosis of another malignant disease and a follow-up dura-tion of less than 6 months.

Patients without distant metastasis were treated with acombination of three-dimensional conformal external beamRT with concurrent weekly 40 mg/m2 cisplatin and high-

dose rate brachytherapy as previously described [9]. Intotal, 50.4 Gy of external RT (1.8 Gy per fraction, daily,Monday through Friday) were delivered using 18-MV pho-tons. Three-dimensional brachytherapy planning was per-formed, using 7 Gy per fraction prescribed to the targetminimum and given in four fractions. Brachytherapy wasdelivered after completion of external RT, and given twiceper week in 2 weeks. A CT scan with 2.5-mm slicethickness through the pelvis was performed for each highdose rate brachytherapy in each patient, with the CT com-patible applicator in place. Delineation of the target volumewas performed based on CT information at the time of thebrachytherapy and was supported by clinical and radio-graphic findings, as recommended by 'Image-guidedBrachytherapy Working Group' [10].

PET/CT Technique

The patients were imaged using a dedicated PET/CT system(Discovery-STE 8; General Electric Medical System,Milwaukee, WI), as previously described [11]. The patientsfasted for at least 6 hours before intravenous administrationof 370 to 555 MBq (10–15 mCi) FDG. Preinjection bloodglucose levels were measured to ensure that they were below150 mg/dL. During the distribution phase, the patients laysupine in a quiet room. Combined image acquisition began60 minutes after FDG injection. The patients were scannedon a flat-panel carbon fiber composite table insert. First, anunenhanced CT scan (5-mm slice thickness) from the base ofthe skull to the inferior border of the pelvis was acquiredusing a standardized protocol (140 kV and 80 mA). Thesubsequent PET scan was acquired in the three-dimensionalmode from the base of the skull to the inferior border of thepelvis (six to seven bed positions, 3 minutes per bed posi-tion) without repositioning the patient on the table.Computed tomographic and PET images were acquired withthe patient breathing shallowly. Attenuation was correctedusing the CT images. Areas of FDG uptake were categorizedas malignant based on location, intensity, shape, size, andvisual correlation with CT images to differentiate physiologicuptake from pathologic uptake. In particular, abnormal met-abolic uptake was defined as moderately increased if SUVmax

ranged from 2.5 to 3.5 and markedly increased if SUVmax

was higher than 3.5 [12].

Clinical follow-up

The follow-up FDG-PET and CT scans (PET-CT) were per-formed at least 3 months after the completion of RT to eval-uate the response to therapy. CMR was defined as FDG-PETuncovering no evidence of local or distant disease; PR wasdefined as the persistence of FDG uptake exceeding the back-ground level in the liver at the site of initial disease, whereas

Eur J Nucl Med Mol Imaging

PD was defined as a new site of FDG uptake on PET scan [6].One hundred and twenty-two patients (80 %) had CMRs onfollow-up FDG-PET, whereas 23 (13 %) and seven patients(5 %) had PRs and PD, respectively.

Clinical follow-up of the patients was performed every3 months until 2 years after the completion of therapy, every6 months up to 5 years, and then annually thereafter. At eachvisit, a detailed gynecological examination was made by thereferred physician. Biopsy was not performed prior to6 months after the completion of ChRT. Failure was definedas biopsy-proven recurrence or the documentation of diseaseprogression on serial imaging studies [2]. Six of nine patientswith locoregional failure were proven with biopsy, and two of16 patients with distant failure were diagnosed withmetastasectomy. Failure patterns were determined by follow-up imaging studies and divided into four groups: none, isolat-ed local failure (central pelvis), locoregional failure (pelviclymph node recurrence with/without local failure), distantfailure (which includes para-aortic and supraclavicular lymphnodes), or combined local plus distant failure.

Statistical analysis

The time to event was calculated as the time from the date ofdiagnosis to the date of the first finding in a clinical or imagingexamination that suggested disease recurrence. DFS and OSrates were calculated using the Kaplan–Meier method. Thechi-square test or Student’s t-test was utilized for univariateanalysis. The effect of age, stage, tumor size, histologic typeand lymph node metastasis on survival was investigated usingthe log rank test. Multivariate analysis was performed usingthe Cox proportional hazards model including all factors thatwere significant on univariate analysis. A p value ≤ 0.05 wasconsidered statistically significant.

Results

Patient characteristics

All patients had pre-treatment and post-treatment PET-CTimages. Post-treatment PET-CT images were taken within amedian of 3.9 months (range, 3.0–9.8 months) after the com-pletion of ChRT.

The characteristics of patients with CMR and PR/PD afterChRT are included in Table 1. As demonstrated in Fig. 1, the4-year OS and DFS rates were significantly lower in patientswith PR/PD compared to patients with CMR (12.4 % vs.66.9 %; p<0.001 and 0 % vs. 64.3 %; p<0.001). More thanhalf of the patients had FIGO stage IIB disease, and mostpatients had squamous cell carcinoma. The incidence of pelvicor para-aortic lymph node metastasis is higher in patients withPR/PD (63 %) compared to patients with CMR (51 %)

(Table 1). The 4-year OS and DFS rates were significantlyworse in patients with advanced stage (FIGO stage≥IIB),larger tumors (≥ 5 cm) and lymph node metastasis (Table 2).The mean SUVmax value of the primary cervical tumor wassignificantly lower in patients with CMR compared to PR/PDpatients (15.9±6.3 vs. 26.4±10.1; p<0.001). The mean tumorsize in CMR patients was 5.0 cm±1.4 cm, and it was 6.4 cm±2.3 cm in patients with PR/PD (p<0.001)

All patients were treated with concurrent ChRT: 108 pa-tients (89 %) completed at least four cycles of ChT, tenpatients (8 %) completed three cycles, and four patients(3 %) received two cycles of ChT during RT. All patientsreceived full dose RT, as planned.

Treatment outcome

The median follow-ups of all patients and surviving patientswere 28.7 (range, 3.3–78.7 months) and 33.2 months (range,6.23–78.7 months), respectively. Three-year OS and DFSrates were 78.4 and 74.6 %, respectively. Of the 122 patientswith post-treatment CMR, 25 (21 %) developed local,locoregional, or distant failure. Of these, two patients (2 %)had local or locoregional failure, fifteen patients (12 %) de-veloped distant metastasis and four patients (3 %) had bothlocoregional and distant failure. At the time of the last follow-up, 94 patients (77 %) were alive (six patients (5 %) withdisease), and 28 patients (23 %) had died. Of the 28 patientswho died, nine patients (7 %) died from other causes.

Table 1 Patient characteristics

Characteristics CMR PR/PDNo. of patients (%) No. of patients (%)

Median age (year) 58 (30–84) 57 (21–86)

FIGO stage

IB2 13 (11) 3 (10)

IIA 6 (5) 0 (0)

IIB 72 (59) 15 (50)

IIIA 12 (10) 2 (7)

IIIB 19 (16) 8 (26)

IVA 0 (0) 2 (7)

Tumor size

< 5 cm 60 (49) 7 (23)

≥ 5 cm 62 (51) 22 (77)

Histology

Squamous cell carcinoma 115 (94) 27 (90)

Adenocarcinoma 7 (6) 3 (10)

Lymph node metastasis

None 60 (49) 11 (37)

Pelvic lymph nodemetastasis 51 (42) 15 (50)

Para-aortic lymph nodemetastasis

11 (9) 4 (13)

Eur J Nucl Med Mol Imaging

The mean and median times to local recurrence were 16.6and 11.4 months, respectively (range, 2.9–26.8 months). Allof these patients had recurrences in the cervix and/orparametria. Three patients with local recurrence were re-irradiated, two who were alive and one who died with disease7 months after re-irradiation. Two patients were salvaged withtotal pelvic exenteration. One patient was disease-free after 9months of follow-up after the surgery, and one patient was lostto follow-up 3 months after surgery. Lastly, one patient re-fused salvage treatment, and this patient was lost to follow-up3 months after recurrence. Of the 19 patients with distantmetastasis, six patients (32 %) had lung metastasis, fourpatients (21 %) had liver metastasis, four patients (21 %)had peritoneal carcinomatosis, two patients (11 %) hadsupraclavicular lymph node metastasis, and three patients(15 %) had metastasis at multiple sites. In total, 12 of 19patients received systemic chemotherapy after the detectionof metastasis, and seven patients refused treatment.

Prognostic factors

On univariate analysis, tumor size ≥ 5 cm, FIGO stage≥IIB,and pelvic and/or para-aortic lymph node metastasis werepredictive of both OS and DFS, while histology was predic-tive of OS only (Tables 3 and 4). On multivariate analysis,tumor size, stage and lymph node metastasis were predictiveof increased risk of local recurrence and distant metastasis(Tables 3 and 4).

Discussion

The present study investigated the treatment outcomes andprognostic factors predicting local or locoregional recurrence(six patients, 5 %) and distant metastasis (19 patients, 16%) inpatients with cervical cancer who were treated with definitiveChRT and who exhibited a CMR during PET-CT performedduring follow-up. Tumor size, stage and lymph node metas-tasis were identified as important prognostic factors predictingOS, DFS, local recurrence, and distant metastasis in multivar-iate analysis. In this study, we also found that nearly 3/4 of allrecurrences are distant metastasis, which suggests a need foreffective systemic treatment.

The treatment of choice in patients with locoregional cer-vical cancer is definitive ChRT, which is associated with a 5-year local failure rate of 20–30% [8, 13, 14]. Local failure canbe defined by either clinical examination or serial imagingstudies. However, clinical examination in some cases under-assesses the local disease response, and it does not account fornodal response or any new onset metastatic disease [15, 16]. Ithas been demonstrated that metabolic response demonstratedby post-therapy PET-CT has utility in predicting long-termoutcome [2, 7, 17]. The metabolic CMR rate on PET-CT after

CMR

PR/PD

CMR

PR/PD

a

b

Fig. 1 (a) Overall survival and (b) disease-free survival graphs of pa-tients with complete metabolic response (CMR) and partial response (PR)or progressive disease (PD)

Table 2 The 4-year overall survival (OS) and disease-free survival (DFS)according to prognostic factors

Variable 4-y OS (%) p 4-y DFS (%) p

Age

≤ 40 years 67 0.96 33 0.42

> 40 years 70 65

Stage

< IIB 100 0.007 88 0.04

≥ IIB 63 58

Tumor size

< 5 cm 77 0.009 69 0.003

≥ 5 cm 61 59

Histology

SCC 43 0.35 48 0.42

Adenocarcinoma 73 65

L.n. metastasis

Absent 88 0.001 77 0.004

Present 53 53

Eur J Nucl Med Mol Imaging

definitive ChRT in cervical cancer patients is 70–80 % [2,5–7]. Grigsby et al. [17] reported a 2-year PFS rate of 86% forpatients with no abnormal FDG uptake at any site on post-treatment PET and 40 % for those with persistent abnormaluptake. In a prospective study of 92 patients imaged withFDG-PET, Schwarz et al. [2] reported 3-year PFS rates of78, 33, and 0 % for patients with CMRs, PRs, and PD,respectively. Similarly Beriwal et al. [7] reported 3-year DFSrates of 78, 31, and 0 % for patients with CMRs, PRs, and PD,respectively. In our series, the 2-year and 3-year DFS rateswere 84 and 75 %, respectively, which is consistent withpublished series.

As cervical tumors are radiosensitive, the tumor massregresses in size during the course of irradiation.Most patientshad complete tumor regression, and in some cases, persistenttumors of the cervix, which are predictive of poor outcome,can be detected by pelvic examination. However, evaluationof the lymph node response or distant metastasis could not beperformed via physical examination. Additionally, CT andMRI are not specific for evaluating the tumor response inlymph nodes, because lymph nodes that are unchanged in sizedo not reliably indicate whether tumor is present or absent. Forthis reason, previous reports documented relatively high rates

of local or locoregional failure rates compared to those ofdistant metastasis [13, 18]. However, when tumor responsewas evaluated with FDG-PET, the rate of distant relapse wasfound to be higher than those of previous series [6, 7].Although CMR after definitive ChRT in patients with cervicalcancer is predictive of better outcome, recurrences will devel-op in 8–23 % of these patients [5, 6, 16, 18]. Park et al. [18]reported a failure rate of 23 % in 349 patients with stage IIBcervical carcinoma treated with ChRT. The incidences ofdistant metastases alone and pelvic with distant metastaseswere 24 and 9%, respectively. In a study by Schwarz et al. [6],the relapse rate was 23 % (40 patients) in 173 patients withCMRs detected on post-treatment PET. The mean time tofailure was 27±20 months. Isolated local recurrence wasobserved in six patients, and distant failure was observed inforty patients. In another study of 112 patients having CMRson post-treatment FDG-PET, 11 (9.8 %) had a recurrence, six(5.4 %) had isolated local recurrences, and five (4.5 %) haddistant recurrences. The median time to recurrence was9 months. In our study, we observed failure in 25 patients;of them, two patients (2 %) had local or locoregional failure,15 patients (12 %) and four patients (3 %) had bothlocoregional failure and distant metastasis.

Table 3 Univariate and multi-variate analysis of prognosticfactors for overall survival in pa-tients with complete metabolicresponse

Variables Risk factors HR (95 % CI) p

Univariate analysis

Age (years) 1.04 (0.99 – 1.08) 0.11

Histology SCC vs. adenocarcinoma 12.33 (1.43 – 63.7) 0.02

Tumor size < 5 cm vs. ≥ 5 cm 3.11 (1.08 – 8.90) 0.04

Stage < IIB vs. ≥ IIB 2.19 (1.31 – 3.66) 0.003

Lymph node metastasis Present vs. absent 4.12 (1.32 – 12.84) 0.01

Multivariate analysis

Histology SCC vs. adenocarcinoma 2.56 (0.74 – 8.78) 0.14

Tumor size < 5 cm vs. ≥ 5 cm 2.79 (1.22 – 6.40) 0.02

Stage < IIB vs. ≥ IIB 2.03 (1.37 – 3.00) 0.001

Lymph node metastasis Present vs. absent 4.82 (1.92 – 12.09) 0.01

Table 4 Univariate and multi-variate analysis of prognosticfactors for disease-free survival inpatients with complete metabolicresponse

Variables Risk factors HR (95 % CI) p

Univariate analysis

Age (years) 0.98 (0.94 – 1.08) 0.32

Histology SCC vs. adenocarcinoma 3.12 (0.45 – 21.56) 0.25

Tumor size < 5 cm vs. ≥ 5 cm 3.63 (1.28 – 10.28) 0.02

Stage < IIB vs. ≥ IIB 1.71 (1.45 – 2.40) 0.008

Lymph node metastasis Present vs. absent 2.51 (1.75 – 5.98) 0.02

Multivariate analysis

Tumor size < 5 cm vs. ≥ 5 cm 3.17 (1.46 – 6.88) 0.003

Stage < IIB vs. ≥ IIB 1.58 (1.11 – 2.25) 0.01

Lymph node metastasis Present vs. absent 2.56 (1.19 – 5.50) 0.02

Eur J Nucl Med Mol Imaging

Several clinical risk factors that are predictive of treat-ment outcome have been identified, including age, tumortype, lymph node metastasis, FIGO stage, bulky tumorsize, and parametrial infiltration [14, 19]. In this study,tumor size, stage, and lymph node metastasis were signif-icant factors for DFS and OS in univariate analysis. Inmultivariate analysis, tumor size ≥5 cm, stage≥IIB and thepresence of lymph node metastasis were significant factors(Tables 3 and 4).

In this study, we only analyzed patients with cervicalcancer with CMRs on post-treatment PET-CT. However, de-spite having good clinical outcomes, the relapse rates werehigher in high-risk patients. It is important for these patients tobe closely followed up, because local recurrence can be sal-vaged with surgery. In addition, there is need to evaluatewhether additional treatment may help to reduce this risk ofrecurrence. Univariate analysis identified tumor size and stageas predictive factors for a higher risk of local recurrence. Onmultivariate analysis, pre-treatment tumor size was associatedwith a higher risk of local failure (odds ratio [OR]=1.66; 95%confidence interval [CI]=1.15–2.40) and lymph node metas-tasis (OR=1.53; 95 % CI=1.06–2.19), which is consistentwith the literature [7]. This finding is most likely attributableto the fact that patients with larger tumor sizes have poorresponses to RT. In some series, tumor volume measured byMRI or PET-CT was an important variable for predictinglong-term outcome [15, 20]. For the majority of the patientsin our series, tumor size was recorded on the basis of the pre-treatment MRI or PET-CT findings, which may be morereliable than clinical examination. Additionally, we found thatlymph nodemetastasis is one of the most important prognosticfactors for predicting local recurrence, which may be correlat-ed with tumor aggressiveness. Pelvic or para-aortic lymphnode metastasis is one of the most important prognostic fac-tors for predicting survival in patients with cervical cancer [1,17]. In a study by Grigsby et al. [21], the presence of lymphnode FDG uptake correlated most significantly with DFS.Similarly, Xue et al. [1] found a significant correlation be-tween lymph node involvement on PET-CTand DFS.We alsofound that tumor size and lymph node metastasis were pre-dictive of distant recurrence in univariate analysis, and tumorsize and lymph node metastasis were predictive in multivari-ate analysis.

The optimal timing of PET-CT after the completion oftreatment is not well defined. Head and neck cancer datasuggest a high rate of false positives when post-therapyFDG PET is performed at 2 months [22]. In a prospectivestudy by Siva et al. [23], 103 patients underwent FDG-PET amedian of 4.9 months (range, 1–12 months) after the com-pletion of therapy. In another study by Beriwal et al. [7], post-treatment PET-CT was performed 10–16 weeks after thecompletion of RT. In our study, post-therapy PET-CT wasnot performed as early as 3 months, because of the formal

procedures of the social security systems. Therefore, wediminished the risk of false positivity due to early evaluation.The CMR rates were 70–75 %, which were consistent withour results. Further research using a well-designed case-controlled trial may aid in assessing the best timing of post-therapy FDG PET.

The present study possesses limitations. The retrospectivenature of our study is the largest limitation. Second, the post-therapy FDG response evaluations were performed at differenttimes varying from 3 to 10 months (median, 4 months).Typically, more than three-fourths of recurrences will occurwithin the first 2–3 years after the initial treatment, whichsuggests a role for increased surveillance during this timeframe [24, 25]. Thus, the guidelines recommend follow-upevaluation every 3–6 months for the first 2 years [25].Additionally, physical examination for cervical canceraccounted for the highest detection rate when compared withcytologic evaluation and imaging modalities [24, 26].Although the post-treatment PET-CT was delivered in variedtime period, only nine patients (7 %) had post-treatment PET-CT taken > 6 months after completion of treatment and all ofthese patients were evaluated with either CT or MRI.However, for evaluation of treatment response, all patientshad undergone a detailed gynecological examination every3 months for the first 2 years. Therefore, although we couldnot make a conclusion regarding the optimal timing of post-therapy PET-CT, at least a detailed gynecological examinationwas performed at each visit. Finally; the number of patientswas too small to make accurate therapy decisions. Despitethese limitations, our study is important in demonstrating theclinical outcomes in patients with CMR, in a study with longermedian follow-up compared to previous reports and in a groupwith more frequent high-risk features.

Conclusions

Our study confirms that patients with CMRs after definitiveChRT have excellent clinical outcomes. Although CMR ispredictive of better outcomes, relapses remain problematic.In particular, patients with bulky tumors, locally advancedstage and those with pelvic and/or para-aortic lymph nodemetastasis have higher risks for local or distant relapses. Theneed for additional local treatment with or without systemicChT in these patients is not yet known. Identification ofthese high-risk patients may permit the evaluation of theclinical benefit of additional treatments such as systemicChT. A multi-institutional clinical trial could confirm thesefindings and identify patients who are at increased risk ofrecurrence despite CMR. An ongoing GOG trial evaluatingthe role of adjuvant chemotherapy for locally advancedcervical cancer after definitive ChRT will clarify the needfor adjuvant treatment.

Eur J Nucl Med Mol Imaging

Conflict of interest and funding The authors report no conflicts ofinterest. The authors alone are responsible for the content and writing ofthe paper.

References

1. Xue F, Lin LL, Dehdashti F, Miller TR, Siegel BA, Grigsby PW. F-18fluorodeoxyglucose uptake in primary cervical cancer as an indicatorof prognosis after radiation therapy. Gynecol Oncol. 2006;101:147–51. doi:10.1016/j.ygyno.2005.10.005.

2. Schwarz JK, Siegel BA, Dehdashti F, Grigsby PW. Association ofposttherapy positron emission tomography with tumor response andsurvival in cervical carcinoma. JAMA. 2007;298:2289–95. doi:10.1001/jama.298.19.2289.

3. Onal C, ReyhanM, Parlak C, Guler OC, Oymak E. Prognostic Valueof Pretreatment 18F-fluorodeoxyglucose Uptake in Patients WithCervical Cancer Treated With Definitive Chemoradiotherapy. Int JGynecol Cancer. 2013;23:1104–10. doi :10.1097/IGC.0b013e3182989483.

4. Chung HH, Nam BH, Kim JW, Kang KW, Park NH, Song YS, et al.Preoperative [18 F]FDG PET/CT maximum standardized uptakevalue predicts recurrence of uterine cervical cancer. Eur J Nucl MedMol Imaging. 2010;37:1467–73. doi:10.1007/s00259-010-1413-5.

5. Chung HH, Kim JW, KangKW, Park NH, Song YS, Chung JK, et al.Predictive role of post-treatment [18F]FDG PET/CT in patients withuterine cervical cancer. Eur J Radiol. 2012;81:e817–22. doi:10.1016/j.ejrad.2012.02.015.

6. Schwarz JK, Siegel BA, Dehdashti F, Grigsby PW. Metabolic re-sponse on post-therapy FDG-PET predicts patterns of failure afterradiotherapy for cervical cancer. Int J Radiat Oncol Biol Phys.2012;83:185–90. doi:10.1016/j.ijrobp.2011.05.053.

7. Beriwal S, Kannan N, Sukumvanich P, Richard SD, Kelley JL,Edwards RP, et al. Complete metabolic response after definitiveradiation therapy for cervical cancer: patterns and factors predictingfor recurrence. Gynecol Oncol. 2012;127:303–6. doi:10.1016/j.ygyno.2012.08.006.

8. Chen T, Jansen L, Gondos A, Emrich K, Holleczek B, Luttmann S,et al. Survival of cervical cancer patients in Germany in the early 21stcentury: a period analysis by age, histology, and stage. Acta Oncol.2012;51:915–21. doi:10.3109/0284186X.2012.708105.

9. Onal C, Arslan G, Topkan E, Pehlivan B, Yavuz M, Oymak E, et al.Comparison of conventional and CT-based planning for intracavitarybrachytherapy for cervical cancer: target volume coverage and organsat risk doses. J Exp Clin Cancer Res. 2009;28:95. doi:10.1186/1756-9966-28-95.

10. Nag S, Cardenes H, Chang S, Das IJ, Erickson B, Ibbott GS, et al.Proposed guidelines for image-based intracavitary brachytherapy forcervical carcinoma: report from Image-Guided BrachytherapyWorking Group. Int J Radiat Oncol Biol Phys. 2004;60:1160–72.doi:10.1016/j.ijrobp.2004.04.032.

11. Onal C, Oymak E, Findikcioglu A, Reyhan M. Isolated mediastinallymph node false positivity of [18F]-fluorodeoxyglucose-positronemission tomography/computed tomography in patients with cervicalcancer. Int J Gynecol Cancer. 2013;23:337–42. doi:10.1097/IGC.0b013e31827e00cc.

12. Signorelli M, Guerra L, Montanelli L, Crivellaro C, Buda A,Dell’Anna T, et al. Preoperative staging of cervical cancer: is 18-

FDG-PET/CT really effective in patients with early stage disease?Gynecol Oncol. 2011;123:236–40. doi:10.1016/j.ygyno.2011.07.096.

13. Eifel PJ, Winter K, Morris M, Levenback C, Grigsby PW, Cooper J,et al. Pelvic irradiation with concurrent chemotherapy versus pelvicand para-aortic irradiation for high-risk cervical cancer: an update ofradiation therapy oncology group trial (RTOG) 90–01. J Clin Oncol.2004;22:872–80. doi:10.1200/JCO.2004.07.197.

14. Atahan IL, Onal C, Ozyar E, Yiliz F, Selek U, Kose F. Long-termoutcome and prognostic factors in patients with cervical carcinoma: aretrospective study. Int J Gynecol Cancer. 2007;17:833–42. doi:10.1111/j.1525-1438.2007.00895.x.

15. Narayan K, Fisher RJ, Bernshaw D, Shakher R, Hicks RJ. Patterns offailure and prognostic factor analyses in locally advanced cervicalcancer patients staged by positron emission tomography and treatedwith curative intent. Int J Gynecol Cancer. 2009;19:912–8. doi:10.1111/IGC.0b013e3181a58d3f.

16. Chung HH, Kim SK, Kim TH, Lee S, Kang KW, Kim JY, et al.Clinical impact of FDG-PET imaging in post-therapy surveillance ofuterine cervical cancer: from diagnosis to prognosis. Gynecol Oncol.2006;103:165–70. doi:10.1016/j.ygyno.2006.02.016.

17. Grigsby PW, Siegel BA, Dehdashti F, Mutch DG. Posttherapy sur-veillance monitoring of cervical cancer by FDG-PET. Int J RadiatOncol Biol Phys. 2003;55:907–13.

18. Park TK, Kwon JY, Kim SW, Kim SH, Kim SN, Kim GE. Patterns oftreatment failure following radiotherapy with combination chemo-therapy for patients with high-risk stage IIB cervical carcinoma. Int JClin Oncol. 2004;9:120–4. doi:10.1007/s10147-003-0378-9.

19. Horn LC, Fischer U, Raptis G, Bilek K, Hentschel B. Tumor size isof prognostic value in surgically treated FIGO stage II cervicalcancer. Gynecol Oncol. 2007;107:310–5. doi:10.1016/j.ygyno.2007.06.026.

20. Ma DJ, Zhu JM, Grigsby PW. Tumor volume discrepancies betweenFDG-PET and MRI for cervical cancer. Radiother Oncol. 2011;98:139–42. doi:10.1016/j.radonc.2010.10.004.

21. Grigsby PW, Siegel BA, Dehdashti F, Rader J, Zoberi I. Posttherapy[18F] fluorodeoxyglucose positron emission tomography in carcino-ma of the cervix: response and outcome. J Clin Oncol. 2004;22:2167–71. doi:10.1200/JCO.2004.09.035.

22. Gupta T, Master Z, Kannan S, Agarwal JP, Ghsoh-Laskar S,Rangarajan V, et al. Diagnostic performance of post-treatment FDGPET or FDG PET/CT imaging in head and neck cancer: a systematicreview and meta-analysis. Eur J Nucl Med Mol Imaging. 2011;38:2083–95. doi:10.1007/s00259-011-1893-y.

23. Siva S, Herschtal A, Thomas JM, Bernshaw DM, Gill S, HicksRJ, et al. Impact of post-therapy positron emission tomographyon prognostic stratification and surveillance after chemoradiother-apy for cervical cancer. Cancer. 2011;117:3981–8. doi:10.1002/cncr.25991.

24. Elit L, Fyles AW, Oliver TK, Devries-Aboud MC. Fung-Kee-FungM, members of the Gynecology Cancer Disease Site Group ofCancer Care Ontario’s Program in Evidence-Based C. Follow-upfor women after treatment for cervical cancer. Curr Oncol. 2010;17:65–9.

25. Greer BE, Koh WJ, Abu-Rustum NR, Apte SM, Campos SM, ChanJ, et al. Cervical cancer. J Natl Compr Cancer Netw JNCCN. 2010;8:1388–416.

26. Sartori E, Pasinetti B, Carrara L, Gambino A, Odicino F, Pecorelli S.Pattern of failure and value of follow-up procedures in endometrialand cervical cancer patients. Gynecol Oncol. 2007;107:S241–7. doi:10.1016/j.ygyno.2007.07.025.

Eur J Nucl Med Mol Imaging