Radiotherapy and Oncology 102 (2012) 335342

Contents lists available at SciVerse ScienceDirect

Radiotherapy and Oncology

journal homepage: www.thegreenjournal.com

Systematic review

Radiographic changes after lung stereotactic ablative radiotherapy (SABR) Can we distinguish recurrence from brosis? A systematic review of the literature
Kitty Huang a,b, Max Dahele c, Suresh Senan c, Matthias Guckenberger d, George B. Rodrigues a,b,e, Aaron Ward b, R. Gabriel Boldt a, David A. Palma a,b,
a

Department of Radiation Oncology, London Health Sciences Centre; b Department of Oncology, University of Western Ontario, London, Canada; c Department of Radiation Oncology, VU University Medical Center, Amsterdam, The Netherlands; d Department of Radiation Oncology, University Hospital Wuerzburg, Wuerzburg, Germany; e Department of Epidemiology and Biostatistics, University of Western Ontario, London, Canada

a r t i c l e

i n f o

a b s t r a c t
Background: Changes in lung density on computed tomography (CT) are common after stereotactic ablative radiotherapy (SABR) and can confound the early detection of recurrence. We performed a systematic review to describe post-SABR ndings on computed tomography (CT) and positron-emission tomography (PET), identify imaging characteristics that predict recurrence and propose a follow-up imaging algorithm. Methods: A systematic review was conducted of studies providing detailed radiologic descriptions of anatomic and metabolic lung changes after SABR. Our search returned 824 studies; 26 met our inclusion criteria. Data are presented according to PRISMA guidelines. Results: Acute changes post-SABR predominantly appear as consolidation or ground glass opacities. Late changes often demonstrate a modied conventional pattern of brosis, evolving beyond 2 years after treatment. Several CT features, including an enlarging opacity, correlate with recurrence. Although PET SUVmax may rise immediately post-SABR, an SUVmax P 5 carries a high predictive value of recurrence. Conclusions: CT density changes are common post-SABR. The available evidence suggests that recurrent disease should be suspected if high-risk CT changes are seen with SUVmax P 5 on PET. Further studies are needed to validate the predictive values of such metrics, and for advanced analysis of CT changes to allow early detection of potentially curable local recurrence. 2012 Elsevier Ireland Ltd. All rights reserved. Radiotherapy and Oncology 102 (2012) 335342

Article history: Received 4 October 2011 Received in revised form 15 December 2011 Accepted 23 December 2011 Available online 1 February 2012 Keywords: Stereotactic radiation Lung cancer Computed tomography Positron emission tomography Local recurrence

Stereotactic ablative radiotherapy (SABR), also known as stereotactic body radiotherapy (SBRT), is now a standard treatment option for patients with stage I non-small cell lung cancer (NSCLC) who are unt for surgery or who are medically operable but refuse surgery [13]. Image-guided SABR is characterized by large fraction sizes coupled with accurate and precise delivery [4]. Although data from randomized trials comparing SABR with surgery are not available, non-randomized single-institution and populationbased studies have reported short/medium-term clinical outcomes comparable to those of primary surgery, with reported local control rates of over 90% [2,5,6]. Comparable overall survival (OS) rates in operable patients is also supported by a Markov model-based decision analysis [7]. Radiation-induced CT lung changes after SABR differ from those observed after conventionally fractionated radiotherapy. Lung changes characteristic to conventional RT consist of CT changes with straight borders corresponding to conventional treatment

Corresponding author. Address: London Regional Cancer Program, 790 Commissioners Rd. E., London, Ontario, Canada. E-mail addresses: david.palma@lhsc.on.ca, david.palma@uwo.ca (D.A. Palma).
0167-8140/$ - see front matter 2012 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.radonc.2011.12.018

eld edges [8]. SABR employs complex beam arrangements to conform high-dose regions to the tumor and create steep dose gradients around the target volume, with a relatively large volume of lung receiving low/intermediate doses (e.g. 540 Gy). In some instances, CT changes after SABR can develop as mass-like patterns that mimic the appearance of recurrent disease [9,10]. Although CT changes after SABR are common, and in-eld recurrences are uncommon [11], distinguishing between the two entities is of paramount importance in patients who may be candidates for salvage therapy, including surgical resection [12]. Current follow-up relies mainly on CT imaging, with positron emission tomography (PET) or biopsy when recurrence is strongly suspected. The drawback of over-investigation is unnecessary imaging and the risks of invasive procedures [13,14], while under-investigation of a growing recurrence risks missing the window of opportunity for salvage therapy. However the optimal imaging paradigm after SABR is yet to be dened. As tter patients with longer life expectancies are now undergoing treatment with SABR, understanding long-term imaging changes will take on increased importance, yet a standardized approach to follow-up and interpretation of imaging after SABR is

336

Detection of recurrence after lung SABR

lacking. The goals of this study were to systematically review the literature describing CT and PET ndings following lung SABR, to identify imaging characteristics that might predict for local recurrence, and to propose imaging guidelines based on the available evidence for early detection of local recurrence after SABR. Methods A systematic review of the literature was performed according to PRISMA guidelines [15]. We searched for English-language papers published between 1995 and June 2011 that met the following criteria: (1) Studies reporting on primary lung tumors, or tumors metastatic to lung, treated with SABR. (2) Studies providing a detailed radiologic description of benign or malignant lung changes on follow-up CT or PET imaging. The following exclusion criteria were used: (1) Studies only reporting toxicity or tumor response but not detailed radiologic changes. (2) Case studies. (3) Review articles. Search strategy Studies were identied by searching the MEDLINE (n = 439) and EMBASE (n = 332) databases. These searches were completed in June 2011. Search strategies are described in online Appendix e1 and selection process is illustrated in Fig. 1. After elimination of duplicate papers, 439 pertinent papers were identied based on their titles and abstracts. Of these, 43 were selected for detailed review, and 22 ultimately met our inclusion criteria and were selected for this review. 8729 conference abstracts from 2005 to 2011 were searched by textword: stereotactic, yielding 385 studies

for additional review. Of these, 10 were selected for detailed review, one abstract was selected and its corresponding unpublished manuscript was obtained from the author as a personal communication [16]. Finally, a review of reference lists of selected papers yielded two additional abstracts of interest and both were selected for the study. One abstract was obtained through expert suggestion, and was included in this review. A total of 23 articles and three abstracts were reviewed, summarized in online Appendix e2 and e3. Results Acute and late CT lung changes post-SABR The pattern of changes in lung parenchyma on CT post-SABR has been described in several studies (Table 1 and Appendix e4) [9,1722] and can generally be categorized as acute (within 6 months, corresponding to pneumonitis) or late (after 6 months, corresponding to brosis) [18,23]. Several papers have classied acute changes into one of ve general patterns: diffuse consolidation, patchy consolidation, diffuse ground-glass opacities (GGO), patchy GGO, and no change (Table 1) [9,1719]. Representative images showing each of these changes are well illustrated in reference [9]. These classications are based on descriptions of changes after conventional radiotherapy [24,25]. Based on weighted averages across studies, the overall incidences of increased CT density in the acute setting was 62%, the majority of which is consolidation (45% of patients), with GGO less common. Dahele et al. analyzed the temporal pattern of CT changes and found that the actuarial median time to acute CT change was 17 weeks [9]. Late changes include modied conventional (consolidation, volume loss, and bronchiectasis similar to but less extensive than conventional radiation brosis), scar-like brosis (linear opacity in the region of the original tumor), mass-like brosis and no change, again illustrated in the study by Dahele et al. [9]. The terms

Fig. 1. Electronic search strategy.

K. Huang et al. / Radiotherapy and Oncology 102 (2012) 335342 Table 1 Incidence and patterns of CT lung changes with SABR. Study Benign acute CT changes (%) Diffuse consolidation Patchy consolidation Diffuse GGO Patchy GGO No evidence of increasing density 46 Benign late CT changes (%) Modied conventional Scarlike brosis Masslike brosis No evidence of increasing density 11 Recurrence features

337

Dahele et al. [9]

Acute n = 67 lesions Late n = 68 lesions n = 52 lesions

16

24

71

11

Recurrences excluded from study

Kimura et al. [17]

38

15

12

33

62

22

17

Palma et al. [18]

3DCRT n = 50 patients RA n = 25 patients 6 months n = 33 patients 12 months n = 35 patients Acute n = 227 Late n = 155

14

22

16

44

Four based on CT enlargement, evolved from scar-like (n = 2) and mass-like (n = 2) brosis Excluded from study

32 27

8 33

4 12

16 6

40 21 46 14 20 20 Three based on CT enlargement and increased SUV, evolved from diffuse GGO (n = 1) patchy consolidation and GGO (n = 2)

Trovo et al. [19]

Weighted Averages

24

21

38

62

15

14

Abbreviations: computed tomography (CT), stereotactic ablative radiotherapy (SABR), ground glass opacity (GGO), three-dimensional conformal radiotherapy (3DCRT), RapidArc (RA). Table 2 Summary of malignant CT lung changes with SABR. Study n recurrence/ total 5/27 Distinguishing features of recurrence

diffuse versus patchy have also been objectively dened according to size criteria and severity [18]. Across studies, the mean incidence of all CT lung changes was 91% of cases. The predominant CT density patterns were modied conventional (62%) and less commonly scar-like (15%) and mass-like brosis (14%). The appearance of mass-like brosis appears to peak at 1 to 2 years post-SABR [9]. A study by Matsuo et al. analyzed only mass-like consolidation and noted its appearance in 68% of cases, at a median of 5 months post-SABR [26]. Other classication schemes for CT patterns have also been reported (Appendix e4) [2022]. The incidence of early and late radiation-induced lung changes ranged from 5479% to 80100%, respectively. Although early ndings of severe radiation pneumonitis (RP) are predictive of late changes [9] they are not a pre-requisite [10]. Acute and late CT changes have also been correlated with dose and volume as summarized in Appendix e5 [20,2730]. The graphical RP is generally correlated with doses above 20 Gray [29].

Kato et al. [31]

 Bulging margin 80% (versus 5%)  Linear margin disappearance 100% (versus 9%)  Air bronchogram disappearance 100% (versus 19%)  Ipsilateral pleural effusion 100% (versus 23%)  Opacity increased in size after 12 months  Lymph node enlargement 60% (versus 18%)  Inhomogeneous enhancement 100% (versus 67%)  Difference in size after 12 months was the only statistically signicant predictor of recurrence (p < 0.01) Initial regrowth with recurrence n = 3 (2 pathology proven)  At 5.9, 9.2, and 12.0 months; shape: solid n=3 Initial regrowth without recurrence n = 14 (4 pathology proven)  Median time 21.4 months (8.635.6); shape: solid n = 9

Matsuo et al. [26] Takeda et al. [10]

3/27 masslike lesions 3/50

Post-SABR CT ndings suggestive of recurrence Three studies directly analyzed the radiological appearance of recurrent disease in the post-SABR setting [10,26,31]. A summary of post-SABR imaging characteristics of recurrent disease is presented in Table 2. The presence of an enlarging mass as a sign of recurrence is supported by all three studies. According to Kato et al., the most reliable indicator of local recurrence is increased size of the opacity after 12 months and additional CT ndings suggestive of recurrence include a bulging margin, disappearance of air bronchograms and appearance of pleural effusion [31]. Of these, disappearance of air bronchograms and emergence of pleural effu-

Abbreviations: computed tomography (CT), stereotactic ablative radiotherapy (SABR).

sion tended to appear early (within 12 months) and has been suggested as an aid to the early detection of recurrence. Matsuo et al. similarly found a correlation between an increase in size after 12 months and recurrence (p < 0.01). They found that recurrence was not predicted by the initial timing of the abnormality, conformity of the abnormality to dose distribution or the presence of ec-

338

Detection of recurrence after lung SABR

tatic (distended/dilated) bronchi, some of which had been previously found to be suggestive of recurrence after conventional RT [26]. An enlarging mass may be sensitive, but not specic for recurrence as Takeda et al. report that the diagnosis of only a minority of patients with an enlarging mass was ultimately recurrence, based upon biopsy and further imaging studies [10].

Expected metabolic ndings on post-SABR PET The role of 2-uoro-2-deoxy-D-glucose (FDG)-PET in staging and evaluation of treatment response after conventional radiotherapy treatment is well understood [32]. An increase in standardized uptake value (SUV) is normally a sign of increased metabolic activity of tumor cells, which is a marker for tumor recurrence. However, in the post SABR setting, FDG avidity can also be related to normal acute radiation inammatory response of lung tumors/ parenchyma [33].

Several studies report normal tumor response values after SABR either as stable SUV or as a decline to background SUV [3340]. Transient rises can occur within the rst 6 months after SABR, but usually not over an SUVmax of 5, consistent with acute postradiation normal tumor/tissue response (Table 3) [33,36,41,42]. Henderson et al. noted a trend in the expected SUVs post-SABR; tumors with low pre-SABR SUVs were correlated with transient rises at 2 weeks post-SABR, whereas high pre-SABR SUVs were correlated with a post-SABR SUV decline at 2 weeks [36].

Post-SABR metabolic ndings of recurrence FDG-PET is commonly considered when local recurrence is suspected on CT imaging [36]. However, the use of PET in the evaluation of treatment response and detection of tumor recurrence after SABR is controversial. Some authors have suggested that metabolic changes on PET may provide an earlier indicator of tumor recur-

Table 3 Selected studies reporting on metabolic ndings after SABR on FDG-PET. Study Bollineni et al. [44] Whole cohort n = 132  Median SUVmax 3.0 Recurrence/treatment failure features n=6  SUVmax 5.0 at 3 months as cut-off: 2 year local control for high and low SUVmax groups were 78.8% versus 97.1% (p = 0.001)  SUVmax > 5.0: better predictor for local control than lesion size (p = 0.025)  SUVmax at 36 months > 5.0 predictive of recurrence: Positive predictive value = 83% Negative predictive value = 75%  High SUVmax at 9 months signicantly predictive of failure  Initial SUVmax decrease of 0.4% then local progression (not dened)  Progressive disease at 7 month  Initial complete anatomic and metabolic response at 3 month  SUV was stable or increased  3 recurrences of 5 patients with stable or increase in SUVmax at 912 months  1 recurrence of 4 patients with SUVmax > 2.5 at 3 months  SUV failed to decline at 12 weeks  Failure to decline <3.0 may indicate resistance to therapy

Chang et al. [16, and unpublished manuscript]

n = 128

n=8

Coon et al. [34] Dahele et al. [35]

n = 28 n = 31

 Median SUVmax 1.7  Median SUVmax: 1.7 (up to 4.8)  Median reductions SUVmax: 3.6 or 64%  Mean SUVmax was 2.6 (max 7.0)  13 of 15 with drop in SUVmax  Drop in the SUVmax by 3 months correlated with local control  SUV declined to < 3.0 at 12 weeks in all 28 patients without recurrence  Median SUVmax 6.04 at 2 weeks, 2.80 at 6 months, 3.58 at 12 months  6 of 13 patients had SUVmax > 3.5 at 12 months recurrence free  Median SUV = 4.2. In subset with delayed PET, hypermetabolic activity (SUV 2.55.87) may exist for years (22 26 months) following SABR without recurrence  SUVmax post-SABR signicantly higher at < 6 months (mean SUVmax 4.9) than 612 months (2.6), 1224 months (3.0), or > 24 months (2.3)  Rapid reduction in mean SUVmax to 3.4 at 6 weeks, 3.0 at 16 weeks, and 3.7 at 52 weeks  Beyond 6 weeks, inammatory changes are evident as low residual activity that persists over 1 year  3 patients without recurrence had SUVmax < 5.0  Mean SUVmax of 2.3(up to 5.7) as early as 3 month, all peaked prior to 1824 months and returned to background levels with mean SUVmax 2.0(up to 2.8) at 1824 months  Transient elevations seen as radiation pneumonitis on CT

n=2 n=1

Feigenberg et al. [39] Abstract

n = 18

n=3

Fuss et al. [40] Abstract Henderson et al. [36] Pilot study Hoopes et al. [33]

n = 30

n=2

n = 14

n=0

n = 28 (subset n = 4 with delayed PET) n = 23

n=0

Matsuo et al. [37]

Excluded

Mohammed et al. [38]

n = 39

n=0

Takeda et al. [10] Vahdat et al. [41]

n=4 n = 20

n=1 n=1

 SUVmax 5.0  SUVmax of 8.4 at 24 months  Corresponding benign CT changes

Abbreviations: stereotactic ablative radiotherapy (SABR), 18F-uorodeoxyglucose positron emission tomography (FDG-PET), computed tomography (CT), standardized uptake values (SUV).

K. Huang et al. / Radiotherapy and Oncology 102 (2012) 335342

339

rence than tumor size on CT, even arguing for routine PET followup imaging and prompting a pilot study on routine PET imaging post-SABR [36], which remains investigational at present. PET may incorrectly identify patients at risk for recurrence, subjecting patients to the unnecessary risk of biopsy [14]. The limited availability of PET imaging in many centers further restricts its use in clinical practice. Due to the low incidence of local recurrences, studies examining their metabolic response signature are limited. A selected summary of FDG-PET values associated with both tumor response and local failure is provided in Table 3. The total number of local recurrences in these 12 PET reports is very small (n = 24), follow-up is heterogenous, and the PET imaging strategies vary. Acknowledging this, the following preliminary observations are made: (i) SUVmax associated with recurrence may fail to decline or it increases over time, sometimes after an initial fall [34,35,39,40], (ii) if a PET scan is done on CT suspicion of recurrence then an SUVmax of 5 or more has generally been reported as indicative of recurrence, and if the scan is performed 36 months after SABR in an attempt to predict response, then an SUVmax of 5 or more has also been associated with a higher risk of recurrence [10,16,41,43], (iii) transient uctuations in SUVmax have been reported, for example, recurrences may exhibit a transient initial SUV decrease [34,35], perhaps as the effects of acute radiation pneumonitis resolve before increased metabolic activity of tumor cells is detected. A transient decrease and subsequent increase in SUV has also been associated with apparent infection rather than recurrence [36]. Only one study examined the positive predictive values (PPV) and negative predictive values (NPV) of a PET SUVmax cutoff for discriminating recurrence from post-SABR changes. Data from Chang et al. use a cutoff of SUVmax = 5 on FDG PET approximately 36 months post-SABR for predicting recurrence, nding a PPV of 83% and a NPV of 75% after a median follow-up of 31 months. In addition, SUVmax at 9 months was also correlated with development of recurrence ([16] and associated unpublished manuscript). A separate study, presented only in abstract form, used an SUVmax cut-off of 5.0 at 3 months post-SABR and found improved local control with the SUVmax < 5 group (97.1% if SUVmax 6 5, vs. 78.8% if SUVmax > 5, p = 0.001) at 2 years, with a trend toward improved OS [43]. The performance of FDG-PET in these circumstances may change as the follow-up duration increases. The limited data available suggest that 39months post-SABR acute radiation changes may be resolving, allowing the metabolic activity of recurrent tumors to be more reliably detected, and that an SUVmax threshold of 5 at 36 months post-SABR may serve as a useful cut-off value for identifying lesions at high risk of subsequent local failure. This may not apply equally to lesions that were initially PET negative or weakly FDG-avid nor should PET ndings be interpreted in isolation. Recommendations for follow-up imaging Based on the above ndings, a proposed algorithm for imaging follow-up of SABR patients who are candidates for salvage treatment is shown in Fig. 2. The primary imaging modality is CT thorax with contrast. In the absence of data, a frequency of 36 months is suggested, at least for the rst year, thereafter the frequency might be tailored to whether the patient is medically operable (e.g. 612 m, depending on time frame), or not (e.g. annually), and the suspicion of recurrence (increased frequency if recurrence is a possibility). FDG-PET is indicated when recurrence is suspected on CT imaging as demonstrated by an enlarging mass with 20% increase in sum of diameters (SOD) and absolute increase in SOD of at least 5 mm from baseline measurements post-SABR, based on RECIST 1.1 criteria [44], with or without the presence of other high-risk features on CT. These include sequentially enlarging mass-like le-

Fig. 2. Proposed algorithm for follow-up of patients post-SABR who are candidates for salvage treatment. Enlargement of CT density: as described by RECIST 1.1 by an increase in the sum of the diameters (SOD) of the target lesion by at least 20% from baseline (i.e. post-SABR nadir size) and absolute increase in SOD of at least 5 mm [44]. High-risk features: sequential enlargement on repeat CT, opacity enlargement after 12 months, bulging margin, disappearance of air bronchograms, linear margin disappearance, ipsilateral pleural effusion or lymph node enlargement. Abbreviations: Stereotactic ablative radiotherapy (SABR), computed tomography (CT), 18Fuorodeoxyglucose positron emission tomography (FDG-PET), standardized uptake values (SUV).

sion, opacity enlargement after 12 months, bulging margin, disappearance of air bronchograms, linear margin disappearance, ipsilateral pleural effusion, or lymph node enlargement. When FDG-PET imaging is obtained, it should be performed preferably on the same machine or on a standardized machine, and residual increases in SUV even 1 year after SABR should not automatically be dened as a recurrence. The use of routine PET/ CT at a predened time point is still subject to investigation and, although useful in the setting of prospective clinical trials, cannot be recommended for routine follow-up at this time without current evidence to support it, particularly given very low rates of local recurrence, the limited availability of FDG-PET in many centers, and the high costs involved. Biopsy or resection should be considered on a case by case basis, taking into consideration the safety, availability of salvage options and the clinical suspicion for recurrence. The limited evidence suggests that when SUVmax equals 5 or greater, recurrence should be highly suspected, and a biopsy or resection should be considered if safe. Some patients may have

340

Detection of recurrence after lung SABR

recurrences that are not FDG avid, this is a particular concern if the original primary had a low SUV, and therefore in patients with SUVmax < 5 who have suspicious CT ndings, biopsy should still be considered. Operable patients can be considered for surgical salvage options whereas non-surgical salvage options can be considered in medically inoperable patients, preferably in prospective study protocols and with appropriate counseling for the expected increase in risk of severe toxicity. Outcomes from repeated use of SABR for in-eld salvage have not been reported, however SABR has been used in patients who have previously received fractionated radiotherapy to the chest, with higher risks of toxicity [45].

Discussion CT density changes are common post-SABR, yet high-level evidence to guide discrimination between recurrence and brosis is lacking. This is particularly important in patients who are candidates for salvage therapy in whom the inability to distinguish these scenarios is especially concerning. Benign CT lung changes follow common acute and late patterns, appearing predominantly as consolidation within the rst 6 months after treatment, with a modied conventional pattern of brosis emerging 6 months to years after treatment. An enlarging CT opacity after SABR is the most frequently reported feature of recurrence. Although PET SUVmax may transiently rise immediately post-SABR and persist for over 12 months without recurrence, SUVmax P 5 may serve as a useful cutoff for recurrence. Advances in the eld of SABR and CT/PET imaging in lung cancer continues to emerge [46,47]. The issue of discriminating between benign brosis and recurrence is expected to take on increasing clinical importance in the coming years. Radiographic changes post-SABR will be encountered more frequently as the use of SABR for the treatment of early stage NSCLC increases, as the use of screening CT increases the detection of early stage disease, and as tter patients with longer life expectancies are treated. The early detection of recurrence is especially important in the medically operable patient population or patients for whom curative surgical salvage treatments are still available. This study is the rst, to our knowledge, to systematically review the literature on the discriminative value of imaging and provide structured recommendations based on the available evidence. The fact that these appear to reect contemporary clinical practice is encouraging, however they also point to deciencies in the current evidence base. Our algorithm improves upon one presented previously by RTOG 0618, by providing more high-risk features on CT and including specic SUVmax cutoffs, again based on a systematic review of the literature [48]. A standardized denition of recurrence becomes especially important when comparing results across studies. Several Radiation Therapy Oncology Group (RTOG) clinical trials (0813, 0915, 0236, 0618) dene local recurrence as an increase in tumor size as noted on CT imaging, along with either increased SUV on FDGPET to original baseline levels, or biopsy proof of disease [48]. Recognizing the inability to obtain biopsy proof in all patients, the available evidence currently supports a denition of a recurrence as meeting either of two criteria: (1) increase in tumor size on CT imaging with SUVmax P5, or (2) pathology-proven disease. Where clinically appropriate, radiological ndings should be conrmed pathologically with biopsy and/or resection. The presence of any malignant cells on biopsy should indicate recurrence, regardless of their viability, since assessment of viability after SABR has not been validated. Although historical studies have reported lung changes after conventionally fractionated radiotherapy or chemoradiotherapy

[8], which tend to be less conformal than that of SABR, to our knowledge no study has rigorously scored morphological changes for direct comparison across these treatment modalities. Limited conclusions can be drawn from two similar studies that measured hounseld unit (HU) density changes after either SABR [49] or chemoradiotherapy [50], whereby the HU changes observed after SABR appear larger at each dose level. A current study is underway using HU density changes pooled from several centers, including patients treated with SABR, radiotherapy alone, and chemoradiotherapy, however the results will not be available for several months. The conclusions of this study must be considered in the context of its strengths and limitations. Large studies examining this question are lacking, and studies vary in their patient selection, radiotherapy doses, follow-up regimens, and CT reporting criteria. Biopsy proof of recurrence was not obtained in the vast majority of patients, and denitions of recurrence varied between studies. SUV readings from FDG-PET scans can vary between institutions using different protocols and can uctuate due to differences in blood glucose levels and PET agents used [35,42]. The individual studies reported here were mainly retrospective, and independent validation of many predictive factors is lacking. There are limitations inherent in our follow-up imaging recommendations, as they are based on data detailing only 35 total local/ in-eld recurrences as seen on CT or PET. The most reliable indicator for recurrence is an enlarging mass, and additional high-risk features based on results of a single study were also included in our algorithm to help guide treatment and would require further validation. We acknowledge that our post-SABR SUVmax recommendations are based on SUVs gathered from both routine and non-routine PET/CT data, however we feel these results can be extrapolated to the setting of PET/CT at time of suspected recurrence. To validate and rene these recommendations, more studies are needed. Although the evidence is imperfect, the algorithm presented provides a framework that can be tested prospectively and modied in the future, and may assist clinicians who are currently following SABR patients, until further data becomes available. Future directions include the objective quantication of CT changes; studies examining CT density, CT-based volumetric assessment and other quantication methods are emerging [49]. Additional imaging studies with larger sample sizes and longer follow-up times are needed to detect the number of recurrences required to draw more robust conclusions. A notable challenge in comparing sequential CT chest images after SABR may be lung volume retraction due to lung brosis, however this may be amenable to deformable image registration, another potential area of future study [49,51]. Further characterization of expected SUV uctuations with time may also help to distinguish abnormal SUV rises that may be correlated with recurrence. Recent data suggests pre-treatment SUVmax P6 signicantly predicts local recurrence after SABR [47], and one may argue for tailored follow-up schedules for such higher risk patients, a concept that requires validation. Future studies should also present receiver-operator characteristic curves and standard metrics such as sensitivity and specicity. These data will allow for a renement of the preliminary management algorithm presented herein. In conclusion, as SABR becomes more widely used in clinical practice, the detection of local recurrence during follow-up becomes increasingly important. This systematic review suggests that post-SABR lung changes demonstrate common acute and late patterns on CT. Furthermore, although the available evidence is limited, recurrence after SABR generally correlates with sequential opacity enlargement on follow-up CT imaging and SUV of P5. More studies, including those rigorously analyzing CT and PET cutoffs for prediction of recurrence are urgently needed to facilitate

K. Huang et al. / Radiotherapy and Oncology 102 (2012) 335342

341

early detection of potentially treatable recurrence while preventing unnecessary invasive interventions for benign disease. Conict of interest The VU University Medical Center has a research agreement with Varian Medical Systems, Inc. and Brainlab. Drs. Senan and Dahele have received travel support and/or honoraria from Varian Medical Systems, Inc. Acknowledgement Dr. Palma is the recipient of research funding from the Ontario Institute for Cancer Research. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.radonc.2011.12.018. References
[1] Onishi H, Shirato H, Nagata Y, et al. Stereotactic body radiotherapy (SBRT) for operable stage I non-small-cell lung cancer: can SBRT be comparable to surgery? Int J Radiat Oncol Biol Phys 2010;81:13528. [2] Lagerwaard FJ, Haasbeek CJ, Smit EF, et al. Outcomes of risk-adapted fractionated stereotactic radiotherapy for stage I non-small-cell lung cancer. Int J Radiat Oncol Biol Phys 2008;70:68592. [3] Lagerwaard FJ, Verstegen NE, Haasbeek CJ, et al. Stereotactic ablative radiotherapy (SABR) in potentially operable stage I non-small cell lung cancer patients. J Thorac Oncol 2011;6:S2789. [4] Dahele M, Senan S. The role of stereotactic ablative radiotherapy for earlystage and oligometastatic non-small cell lung cancer: evidence for changing paradigms. Cancer Res Treat 2011;43:7582. [5] McGarry RC, Papiez L, Williams M, et al. Stereotactic body radiation therapy of early-stage non-small-cell lung carcinoma: phase I study. Int J Radiat Oncol Biol Phys 2005;63:10105. [6] Onishi H, Shirato H, Nagata Y, et al. Hypofractionated stereotactic radiotherapy (HypoFXSRT) for stage I non-small cell lung cancer: updated results of 257 patients in a Japanese multi-institutional study. J Thorac Oncol 2007;2:S94S100. [7] Louie AV, Rodrigues G, Hannouf M, et al. Stereotactic body radiotherapy versus surgery for medically operable stage I non-small-cell lung cancer: a markov model-based decision analysis. Int J Radiat Oncol Biol Phys 2011;81:96473. [8] Libshitz HI, Shuman LS. Radiation-induced pulmonary change: CT ndings. J Comput Assist Tomogr 1984;8:159. [9] Dahele M, Palma D, Lagerwaard F, et al. Radiological changes after stereotactic radiotherapy for stage I lung cancer. J Thorac Oncol 2011;6:12218. [10] Takeda A, Kunieda E, Takeda T, et al. Possible misinterpretation of demarcated solid patterns of radiation brosis on CT scans as tumor recurrence in patients receiving hypofractionated stereotactic radiotherapy for lung cancer. Int J Radiat Oncol Biol Phys 2008;70:105765. [11] Chi A, Liao Z, Nguyen NP, et al. Systemic review of the patterns of failure following stereotactic body radiation therapy in early-stage non-small-cell lung cancer: clinical implications. Radiother Oncol 2010;94:111. [12] Neri S, Takahashi Y, Terashi T, et al. Surgical treatment of local recurrence after stereotactic body radiotherapy for primary and metastatic lung cancers. J Thorac Oncol 2010;5:20037. [13] Wiener RS, Schwartz LM, Woloshin S, Welch HG. Population-based risk for complications after transthoracic needle lung biopsy of a pulmonary nodule: an analysis of discharge records. Ann Intern Med 2011;155:13744. [14] Gupta S, Wallace MJ, Cardella JF, et al. Quality improvement guidelines for percutaneous needle biopsy. J Vasc Interv Radiol 2010;21:96975. [15] Moher D, Liberati A, Tetzlaff J, Altman DG. The PRISMA Group (2009). Preferred Reporting Items for Systematic Reviews and Meta-Analyses: The PRISMA Statement. BMJ 2009;339:b2535. doi:10.1136/bmj.b2535. [16] Chang JY, Zhang X, Komaki R, et al. PET/CT and outcome in lung cancer treated with stereotactic body radiation therapy. Chicago multidisciplinary symposium in thoracic oncology; 2010. p. 101 [abstract]. [17] Kimura T, Matsuura K, Murakami Y, et al. CT appearance of radiation injury of the lung and clinical symptoms after stereotactic body radiation therapy (SBRT) for lung cancers: are patients with pulmonary emphysema also candidates for SBRT for lung cancers? Int J Radiat Oncol Biol Phys 2006;66:48391. [18] Palma DA, Senan S, Haasbeek CJ, et al. Radiological and clinical pneumonitis after stereotactic lung radiotherapy: a matched analysis of three-dimensional conformal and volumetric-modulated arc therapy techniques. Int J Radiat Oncol Biol Phys 2011;80:50613.

[19] Trovo M, Linda A, El Naqa I, et al. Early and late lung radiographic injury following stereotactic body radiation therapy (SBRT). Lung Cancer 2010;69:7785. [20] Aoki T, Nagata Y, Negoro Y, et al. Evaluation of lung injury after threedimensional conformal stereotactic radiation therapy for solitary lung tumors: CT appearance. Radiology 2004;230:1018. [21] Guckenberger M, Heilman K, Wulf J, et al. Pulmonary injury and tumor response after stereotactic body radiotherapy (SBRT): results of a serial followup CT study. Radiother Oncol 2007;85:43542. [22] Takeda T, Takeda A, Kunieda E, et al. Radiation injury after hypofractionated stereotactic radiotherapy for peripheral small lung tumors: serial changes on CT. AJR Am J Roentgenol 2004;182:11238. [23] Choi YW, Munden RF, Erasmus JJ, et al. Effects of radiation therapy on the lung: radiologic appearances and differential diagnosis. Radiographics 2004;24:98597. discussion 98. [24] Ikezoe J, Takashima S, Morimoto S, et al. CT appearance of acute radiationinduced injury in the lung. AJR Am J Roentgenol 1988;150:76570. [25] Koenig TR, Munden RF, Erasmus JJ, et al. Radiation injury of the lung after three-dimensional conformal radiation therapy. AJR Am J Roentgenol 2002;178:13838. [26] Matsuo Y, Nagata Y, Mizowaki T, et al. Evaluation of mass-like consolidation after stereotactic body radiation therapy for lung tumors. Int J Clin Oncol 2007;12:35662. [27] Hof H, Zgoda J, Nill S, et al. Time- and dose-dependency of radiographic normal tissue changes of the lung after stereotactic radiotherapy. Int J Radiat Oncol Biol Phys 2010;77:136974. [28] Kyas I, Hof H, Debus J, et al. Prediction of radiation-induced changes in the lung after stereotactic body radiation therapy of non-small-cell lung cancer. Int J Radiat Oncol Biol Phys 2007;67:76874. [29] Palma DA, van Sornsen de Koste J, Verbakel WF, et al. Lung density changes after stereotactic radiotherapy: a quantitative analysis in 50 patients. Int J Radiat Oncol Biol Phys 2011;81:9748. [30] Prendergast BM, Bonner JA, Popple RA, et al. Dosimetric analysis of imaging changes following pulmonary stereotactic body radiation therapy. J Med Imaging Radiat Oncol 2011;55:906. [31] Kato S, Nambu A, Onishi H, et al. Computed tomography appearances of local recurrence after stereotactic body radiation therapy for stage I non-small-cell lung carcinoma. Jpn J Radiol 2010;28:25965. [32] Choi NC, Fischman AJ, Niemierko A, et al. Doseresponse relationship between probability of pathologic tumor control and glucose metabolic rate measured with FDG PET after preoperative chemoradiotherapy in locally advanced nonsmall-cell lung cancer. Int J Radiat Oncol Biol Phys 2002;54:102435. [33] Hoopes DJ, Tann M, Fletcher JW, et al. FDG-PET and stereotactic body radiotherapy (SBRT) for stage I non-small-cell lung cancer. Lung Cancer 2007;56:22934. [34] Coon D, Gokhale AS, Burton SA, et al. Fractionated stereotactic body radiation therapy in the treatment of primary, recurrent, and metastatic lung tumors: the role of positron emission tomography/computed tomography-based treatment planning. Clin Lung Cancer 2008;9:21721. [35] Dahele M, Freeman M, Pearson S, et al. Early metabolic response evaluation after stereotactic radiotherapy for lung cancer: pilot experience with 18Fuorodeoxyglucose positron emission tomography-computed tomography. Clin Oncol (R Coll Radiol) 2011;23:35963. [36] Henderson MA, Hoopes DJ, Fletcher JW, et al. A pilot trial of serial 18Fuorodeoxyglucose positron emission tomography in patients with medically inoperable stage I non-small-cell lung cancer treated with hypofractionated stereotactic body radiotherapy. Int J Radiat Oncol Biol Phys 2010;76:78995. [37] Matsuo Y, Nakamoto Y, Nagata Y, et al. Characterization of FDG-PET images after stereotactic body radiation therapy for lung cancer. Radiother Oncol 2010;97:2004. [38] Mohammed N, Grills IS, Wong CY, et al. Radiographic and metabolic response rates following image-guided stereotactic radiotherapy for lung tumors. Radiother Oncol 2011;99:1822. [39] Feigenberg S, Yu J, Eade T. FDG PET response by 3 months following stereotactic body radiotherapy for non-small cell lung cancer may be an early surrogate of local failure. Int J Radiat Oncol Biol Phys 2007;69:S47980 [abstract]. [40] Fuss M, Salter B, Herman T, Thomas C. Metabolic PET imaging for stereotactic body radiotherapy (SBRT) planning and therapy response assessment of pulmonary malignancies. Int J Radiat Oncol Biol Phys 2005;63:S412 [abstract]. [41] Vahdat S, Oermann EK, Collins SP, et al. CyberKnife radiosurgery for inoperable stage IA non-small cell lung cancer: 18F-uorodeoxyglucose positron emission tomography/computed tomography serial tumor response assessment. J Hematol Oncol 2010;3:6. [42] Ishimori T, Saga T, Nagata Y, et al. 18F-FDG and 11C-methionine PET for evaluation of treatment response of lung cancer after stereotactic radiotherapy. Ann Nucl Med 2004;18:66974. [43] Bollineni VR, Wiegman EM, Pruim J, et al. 18F-FDG-PET 12 weeks after stereotactic body radiotherapy for stage I non-small-cell lung cancer predicts outcome. Ned Tijdschr Oncol 2011;8:218 [abstract]. [44] Eisenhauer EA, Therasse P, Bogaerts J, et al. New response evaluation criteria in solid tumours: revised RECIST guideline (version 1.1). Eur J Cancer 2009;45:22847. [45] Kelly P, Balter PA, Rebueno N, et al. Stereotactic body radiation therapy for patients with lung cancer previously treated with thoracic radiation. Int J Radiat Oncol Biol Phys 2010;78:138793.

342

Detection of recurrence after lung SABR [49] Palma DA, de Koste JR, Verbakel WF, Senan S. A new approach to quantifying lung damage after stereotactic body radiation therapy. Acta Oncol 2011;50:50917. [50] Phernambucq EC, Palma DA, Vincent A, et al. Time and dose-related changes in radiological lung density after concurrent chemoradiotherapy for lung cancer. Lung Cancer 2011;74:4516. [51] van Dam IE, de Koste JR, Hanna GG, et al. Improving target delineation on 4dimensional CT scans in stage I NSCLC using a deformable registration tool. Radiother Oncol 2010;96:6772.

[46] Ong CL, Palma D, Verbakel WF, et al. Treatment of large stage III lung tumors using stereotactic body radiotherapy (SBRT): planning considerations and early toxicity. Radiother Oncol 2010;97:4316. [47] Takeda A, Yokosuka N, Ohashi T, et al. The maximum standardized uptake value (SUVmax) on FDG-PET is a strong predictor of local recurrence for localized non-small-cell lung cancer after stereotactic body radiotherapy (SBRT). Radiother Oncol 2011;101:2917. [48] Radiation Therapy Oncology Group. Available from: <http://www.rtog.org> [accessed 12.10.11].