ONCH-1930; No. of Pages 16
ARTICLE IN PRESS
Critical Reviews in Oncology/Hematology xxx (2015) xxx–xxx
Treating metastatic sarcomas locally: A paradoxe, a rationale, an evidence? Timothée Olivier a,b , Daniel Pop c , Amina Chouiter Djebaili a , Alexander Tuan Falk d , Antoine Iannessi e , Esma Saada f , Willy Nettekoven a , Jean-Yves Blay g , Patrick Baque h , Didier Cupissol b , Antoine Thyss f , Juliette Thariat d,∗ a
Hôpital neuchâtelois, Département d’Oncologie, Chasseral 20, 2300 La Chaux-de-Fonds, Switzerland Institut régional du Cancer de Montpellier, Parc Euromédecine, 208 rue des Apothicaires, 34298 Montpellier Cedex 5, France c Centre Hospitalo-Universitaire Pasteur, Service de Chirurgie Thoracique, 30 Avenue de la Voie Romaine, 06000 Nice, France Centre Antoine Lacassagne, Département d’oncologie, Département de radiothérapie, 227, Avenue de la Lanterne, 06200 Nice, France e Centre Antoine Lacassagne, Département de radiologie, 33, Avenue Valombrose, 06189 Nice, France f Centre Antoine Lacassagne, Département d’oncologie médicale, 33, Avenue Valombrose, 06189 Nice, France g Département d’Oncologie, Centre Léon Bérard, 28 Rue Laennec, 69008 Lyon, France h Centre Hospitalo-Universitaire Saint Roch, Service de Chirurgie, 06000 Nice, France b
d
Accepted 6 January 2015
Contents 1. 2. 3. 4. 5.
Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . History of local treatments in sarcomas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Strategy adapted on metastatic burden . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Design: review criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Surgery of lung metastases from sarcomas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.1. Outcomes and prognostic factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2. Surgery and development of mini-invasive procedures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6. Radiofrequency ablation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7. Stereotactic body radiation therapy (SBRT) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8. Systemic and local treatments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9. Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conflict of interest . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Reviewer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
Abstract Purpose: The mainstay of first line treatment in metastatic sarcomas is chemotherapy with response rates of ≈25% but the optimal management of further events is debated. We assessed the benefit of local metastatic treatment in metastatic sarcomas. Results: Local control of local treatment strategies (≈85%) is excellent but highly institution-dependent and subject to selection biases. Formal evidence of an improvement of survival with local ablative treatments has been limited to retrospective studies. On the other hand, some chemotherapy trials are inconclusive because about 20% of patients receive local metastatic ablation as it is considered unethical to omit
∗
Corresponding author at: Department of Radiation Oncology, 33 Av Valombrose, 06189 Nice, France. Tel.: +33 492031083; fax: +33 492031096. E-mail address: [email protected] (J. Thariat).
http://dx.doi.org/10.1016/j.critrevonc.2015.01.004 1040-8428/© 2015 Elsevier Ireland Ltd. All rights reserved.
Please cite this article in press as: Olivier T, et al. Treating metastatic sarcomas locally: A paradoxe, a rationale, an evidence? Crit Rev Oncol/Hematol (2015), http://dx.doi.org/10.1016/j.critrevonc.2015.01.004
ONCH-1930; No. of Pages 16
2
ARTICLE IN PRESS T. Olivier et al. / Critical Reviews in Oncology/Hematology xxx (2015) xxx–xxx
local treatment in these patients. Further, technology has made surgery, stereotactic irradiation and radiofrequency ablation highly effective on local control with limited morbidity. Conclusion: The benefit on survival of metastatic ablation deserves prospective studies integrating quality of life, cost effectiveness and patient-reported outcomes assessment. © 2015 Elsevier Ireland Ltd. All rights reserved.
Keywords: Sarcoma; Metastases; Chemotherapy; Ablative treatment; Metastasectomy; Stereotactic radiotherapy
1. Introduction
2. History of local treatments in sarcomas
Sarcomas represent a rare (1% of cancers) and heterogeneous group of about 100 subtypes of soft-tissue and bone tumors of mesenchymal origin [1,2]. When surgically removed at an early stage, sarcomas can be cured [1]. However, up to 50% of patients develop metastases during the course of their disease [3]. Median survival measured from the diagnosis of metastases has improved from 12 to over 18 months in the last decades [4]. Such improvement is multifactorial, including advances in imaging, lines of systemic treatment, better supportive care and metastatic ablation. Chemotherapy remains the standard of care of metastatic sarcomas but hardly cures any patient and rather prolongs survival by a few weeks or months. Performed alone, it results in 5-year overall survival rates of less than 10% with exceptional long survivors [5,6]. Further, most series include both oligometastatic (defined as 1–3 or 6 synchronous metastases) and polymetastatic patients. These might overestimate the benefit of chemotherapy in polymetastatic patients. The management of metastatic sarcomas raises several other controversial issues. It is uncertain whether local metastatic ablation truly improves outcomes or rather represents a selected population with slowly progressive limited tumor bulk of spontaneously better prognosis regardless of treatment. Additionally, recent evidence suggests that numerous metastatic patients indeed receive metastatic resection in addition to systemic treatment [7]. Local metastatic treatments are associated with chemotherapy delivered sequentially or perioperatively, or may be used as exclusive therapy. Lung metastasectomy has been performed in many cancers for more than 20 years [8]. In sarcomas, lung metastases are the predominating site of metastases and the exclusive site in over 50% of patients at first metastatic event [9–11]. Lung metastases are rarely symptomatic and often discovered during follow-up Computed Tomography scan (CT). Extra-pulmonary and atypical metastatic sites become more likely with prolonged disease course [12,13]. While surgery has the largest ground of evidence, other local ablative techniques like radiofrequency ablation (RFA), and focused high-dose whole-body stereotactic radiation therapy (SBRT) also show promising local control rates and excellent tolerance for pulmonary and extrapulmonary metastases. In the present review, we addressed the current level of evidence for local ablative treatments in metastatic sarcomas.
Pulmonary metastasectomy was first reported by Weinlecher in 1882 and by Divis in 1927 as a planned procedure [8]. Following publication of a first case series, Thomford described the principles of lung metastasectomy in 1965 [8]. Both cytotoxic chemotherapy and metastasectomy emerged in the 1970s as major advances in sarcomas (and in colorectal cancers). Moreover, sarcomas occur in young patients and are often relatively chemoresistant. Both reasons provided a strong rationale for aggressive local management of metastases from sarcomas. Among 184 osteosarcoma patients treated at the Memorial Sloan Kettering Cancer Center by limb amputation, 75% presented lung metastases within 18 months of their primary and there were no 5year survivors [14]. In contrast, a pilot study by Martini et al., in 1971, reported a 3-year survival of 40% in 22 patients after lung metastasectomy [15]. Eligibility criteria for surgery were controlled primary tumor, planned complete resection, adequate cardio-pulmonary function and absence of progressive extra-thoracic metastases. Since then, lung metastasectomy has been performed on a case-by-case basis, with yet limited overall evidence. In the same period, doxorubicin and ifosfamid-based chemotherapy also considerably changed the management of metastatic sarcomas. However, first-line chemotherapy yields limited overall response rates in the order of 10–25% and aims at improving three-month progression-free survival (PFS), suggesting that sustained response is rarely achievable [5]. Second-line (and over) systemic therapies are even less supported despite being routinely performed in practice [5]. Moreover, such sarcoma subtypes as alveolar and clear cell sarcomas have been very poorly sensitive to chemotherapy (although this may change with new histology-driven targeted therapies) (Fig. 1).
3. Strategy adapted on metastatic burden The justification of treating metastases aggressively includes [16]: (1) Anecdotal experience suggests that selected patients benefit from metastatic ablation, irrespective of systemic therapy. (2) Consolidative local therapy yields good outcomes in partial responders to systemic therapy.
Please cite this article in press as: Olivier T, et al. Treating metastatic sarcomas locally: A paradoxe, a rationale, an evidence? Crit Rev Oncol/Hematol (2015), http://dx.doi.org/10.1016/j.critrevonc.2015.01.004
ONCH-1930; No. of Pages 16
ARTICLE IN PRESS T. Olivier et al. / Critical Reviews in Oncology/Hematology xxx (2015) xxx–xxx
3
Fig. 1. Timeline: major breakthroughs and change of paradigm in the management of metastases from sarcomas.
(3) Oligometastases, as described by Hellman and Weichselbaum [17,18], represent a subgroup of patients of intermediate disease stage between primary and widely metastatic disease. Eradication of oligometastases might lead to prolonged remission or cure. Identification of patients truly who benefiting is a challenge. In the 1980s, metastatic ability was found heterogeneous amongst mouse KHT sarcoma cells [19]. Lussier et al. recently demonstrated that there may be a micro-RNA signature for either oligoor polymetastatic progression [20,21]. These works, among others, are preliminary evidence that the oligometastatic state may have a specific biologic background and may be predicted with subsequent therapeutic implications. (4) Based on the Norton-Simon hypothesis, debulking may make residual tumor more sensitive to chemotherapy (based on Gompertzian kinetics) [22]. In the metastatic cascade model, tumor doubling time increases with subsequent metastases [23]. Interactions between primary and metastases can lead to metastatic growth or regression (such as after treatment of primary renal cancer) [24]. Local metastatic ablation aims primarily at eradicating macroscopic disease through targeted intervention. Along with chemotherapy [25], it may also stimulate antitumor immunity [26]. Such non-targeted effects (including the abscopal effect) might also be involved in the antitumoural effects of local metastatic ablation. Practically, local
metastatic treatment may alleviate symptoms, prolong PFS and/or cure selected patients depending on metastatic burden, i.e. number, volume and metastatic growth kinetics. Technical improvements (minimally-invasive surgery, RFA, SBRT, etc.) have rendered ablative treatments feasible for metastases situated nearly anywhere in the body with limited morbidity. The physical, psychological, social or economic impact of local therapy of metastases should thus be balanced against the expected benefits of systemic treatment only. An illustration of different case scenarios is presented in Fig. 2.
4. Design: review criteria A literature search, according to PRISMA statement, was performed on MEDLINE (http://pubmed.gov). Articles were searched using key words (radiofrequency, ablation, stereotactic, radiosurgery, pulmonary resection, metastasectomy, lung resection/metastasis, metastases, metastatic/sarcoma, sarcomas, osteosarcoma, osteosarcomas) and evaluated on title and abstract. Full text articles meeting criteria were collected to assess for definite inclusion. Inclusion in Table 1 and prognostic factors analysis of surgical series (numerous) was restricted to series with more than 20 patients published from 1994 to 1st January 2014. For SBRT and RFA (scarce) series, these limitations were not applied. A PRISMA flowchart is described in Fig. 3.
Please cite this article in press as: Olivier T, et al. Treating metastatic sarcomas locally: A paradoxe, a rationale, an evidence? Crit Rev Oncol/Hematol (2015), http://dx.doi.org/10.1016/j.critrevonc.2015.01.004
ARTICLE IN PRESS
ONCH-1930; No. of Pages 16
4
T. Olivier et al. / Critical Reviews in Oncology/Hematology xxx (2015) xxx–xxx
Table 1 Series on lung surgical resection for sarcoma metastases. First Author, Year
Patients with lung meta (treated with PMa )
Inclusion criteria for metastasectomy when available (inclusion period)
Survival data
Comments/Complications (when available) (number of patients)
Robinson 1994 [29]
189(44) STSb
Not homogeneous for all patients (1970–1990)
5 year-OSc after PM = 52%
Choong 1995 [111]
274(214) STS
(1976–1991)
Kawai 1995 [120]
(23) STS
(1970–1992)
Temeck 1995 [46]
(152) OSTd and STS (pediatric)
No extrapulmonary metastases (1975–1993)
5 year-OS after PM = 40% 10 year-OS = 20% Median survival = 28 months 5 year actuarial OS = 32% Median survival = 2.2 year
Complications = 13.6% including: One post operative death pulmonary embolism(1) Not available
Van Geel 1996 [36]
(255) STS
(Not available–1993)
Billingsley 1999 [38]
719(213) STS
(1982–1997)
Antunes 1999 [121]
198(31) OST
Suri 2005 [52]
(103) MFH
Inclusion: no local recurrence, no other metastases, all meta operable with low operative risk (1989–1997) (1976–2000)
Suemitsu 2005 [49]
(43) Bone and STS
Pfannschmidt 2006 [122]
(50) STS
Pfannschmidt 2006 [123]
(21) OST
Suzuki 2006 [124]
(105) OST and STS
Harting 2006 [53]
137(99) OST
Inclusion: primary controllable, good general condition, no extrathoracic lesions, anatomically and functionally resectable (1981–1999) Inclusion: primary controlled, resectable, general and functional risks tolerable, no extrathoracic lesions (1996–2002) Inclusion: primary controlled, resectable, general and functional risks tolerable, no extrathoracic lesions (1997–2001) Inclusion: operable, resectable, control primary, no extrathoracic lesions (1990–2002) Exclusion: unresectable disease, death before surgery, refused surgery, apparent resolution with chemo, extensive local recurrence before surgery (1980–2000)
3 year-OS = 54% 5 year OS = 38% All patients: Median survival = 15 months Actuarial 3 year S = 25% PM complete resection (n = 161): Median OS = 33 months Actuarial 3 year-S = 46% 3 year-S = 61%
Median FUe = 18 months 5 year-OS 21%
Median survival = 30 months 5 year-OS = 20.7%
Not available
Complications = 6.5% (after 1st procedure): Prolonged air leak(3), pneumonia(2), superficial wound infection(2) respiratory insufficiency(1), pneumonitis(1) atrial tear(1) One post operative death Not available
Ventilation for 48 h(1)
Operative mortality = 1% Morbidity = 11%: prolonged air leak(7), pneumonia, empyema, atelectasis, and sepsis. No post operative death
Median survival = 39.5 months
No operative mortality
Median survival = 35.8 months
No operative mortality 2 patients requiring re-thoracotomy for hemorrhage
5 year-S = 43.6%
Not available
PMa : Median survival = 33.6 months No PM = 10.1 months 3 year-OS = 46.2% 5 year-OS = 29.0%
11.8% intra or post-operative complications: pneumonia, transfusions, drain more than 7 days, reoperation for hemorrhage, mechanic ventilation >7 days
Please cite this article in press as: Olivier T, et al. Treating metastatic sarcomas locally: A paradoxe, a rationale, an evidence? Crit Rev Oncol/Hematol (2015), http://dx.doi.org/10.1016/j.critrevonc.2015.01.004
ARTICLE IN PRESS
ONCH-1930; No. of Pages 16
T. Olivier et al. / Critical Reviews in Oncology/Hematology xxx (2015) xxx–xxx
5
Table 1 (Continued ) First Author, Year
Patients with lung meta (treated with PMa )
Inclusion criteria for metastasectomy when available (inclusion period)
Survival data
Comments/Complications (when available) (number of patients)
Rehders 2007 [125]
121(61) STS
Median survival after PMa = 33 months
Complications: Prolonged pleural drain = 9, wound infection = 3, hypoventilation = 3, No perioperative mortality
Liebl 2007 [48] Chen 2008 [51]
(42) STS (23) STS
3 year-OS = 31% 5 year-OS = 31%
Not available No deaths directly from surgery
Chen 2009 [71]
(23) OST
Inclusion: less than 6 metastases, no contraindication i.e.:insufficient control of the primary tumor, unresectable lung disease, extensive involvement of the mediastinum or the chest wall, and unresectable metastatic disease outside the lung (1991–2002) (1990–2005) Inclusion: resectable, operable, limited to lung, primary controlled or potentially controlled (1989–2007) Inclusion: resectable, operable, limited to lung, primary controlled (1990–2007)
No deaths directly from surgery
Chen 2009 [40]
(25) OST
Inclusion: resectable, operable, limited to lung, primary controlled (1989–2007)
Franco 2009 [126]
(22) STS
Inclusion: primary under control, resectable, operable, -no other metastases (1996–2006)
Franco 2009 [127]
(52) OST
Inclusion: primary under control, resectable, operable, -no other metastases (1996–2006)
Median FU = 13 months 2 year-OS (after PMa ) = 52% 5 year-OS = 43% 10 year-OS = 29% Median FU = 16 months 2 year-OS (after PM) = 42.9% 5 year-OS = 19.0 Median FU = 14months Median survival = 19 months 5 year-OS = 23.1% 3 year-OS = 47.1% Median FU = 28months Median survival = 27 months 3 year-OS = 43% 5 year-OS = 31%
Smith 2009 [28]
(94) STS
Inclusion: all patients receiving PM (1976–2000)
Buddingh 2009 [50]
88(56) OST
Blackmon 2009 [128]
(234) Bone and STS
Exclusion: insufficient data, unresectable primary (1990–2008) (1998–2006)
Aljubran 2009 [129]
85(47)OST
(1986–2003)
Sardenberg 2010 [41]
(77) STS
Inclusion: primary control, exclusively pulmonary metastases, resectable, operable, no better treatment (1990–2006)
Briccoli 2010 [70]
(323) OST
Inclusion: primary tumor controlled, no pleural or pericardial effusion, no other metastases, with adequate residual pulmonary function following resection (1985–2005)
Median survival = 16 months Actual 5 year-DFS = 5% Actual 5 year-OS = 15% No difference in OS between synchronous and metachronous Median survival = 36.2 months 3 year-OS = 38% in the group surgery 5 year-OS = 34.7%
5 year-OS = 37%
No deaths directly from surgery
Post operative complications: 72 h ventilation mechanical(1), persistent air leak(1) No mortality Major morbidity 4 patients: mechanical ventilation more than 24 hrs due to acute respiratory distress syndrome, pneumonia(2), persistent air leak-2) Overall 30 day mortality = 3.7% Actual FU2 ≥5 years Not available
Comparison between groups with different metastatic presentation and treatment No difference for age Complications (9.1%): fever, atelectasis, prolonged air leakage >7days, infection, prolonged hospital stay >30 days No 30 days-mortality Mixed synchronous and metachronous No data complications
Please cite this article in press as: Olivier T, et al. Treating metastatic sarcomas locally: A paradoxe, a rationale, an evidence? Crit Rev Oncol/Hematol (2015), http://dx.doi.org/10.1016/j.critrevonc.2015.01.004
ARTICLE IN PRESS
ONCH-1930; No. of Pages 16
6
T. Olivier et al. / Critical Reviews in Oncology/Hematology xxx (2015) xxx–xxx
Table 1 (Continued ) First Author, Year
Patients with lung meta (treated with PMa )
Inclusion criteria for metastasectomy when available (inclusion period)
Survival data
Comments/Complications (when available) (number of patients)
Predina 2011 [30] Kim 2011 [42]
(48) STS (89) STS and OST
(1995–2007) Inclusion: complete local control, exhibited isolated pulmonary metastases (2002–2008)
5 year-OS = 52% After pulmonary resection: Median OS = 10.9months 5 year-OS = 50% 5 year-DFS = 18.9%
Salah 2013 [130]
71(30) OST and STS
Inclusion: primary site has to be controlled, and metastases are amenable to surgical resection without residual disease (2001–2012)
From PM: Median survival = 39.6 months
Dear 2012 [44]
(114) Bone and STS
(1978–2008)
Median OS = 37months 3 year OS = 53% 5 year-OS = 43%
Mizuno 2012 [43]
(52) OST and STS
From PM Median survival = 33.3months 5 year-OS = 50.9%
Toussi 2013 [45]
(34) STS
Stanelle 2013 [131]
41 (31)Synovial sarcoma pediatric
Inclusion: primary and other distant site controlled, resectable But PM indicated on background, respiratory function, length of the DFIf after treatment of primary site, number and locations (1996–2011) Inclusion: no active uncontrollable extrapulmonary disease, high probability of complete resection, sufficient pulmonary reserve, operable (1996–2007) (1971–2011)
No perioperative death Reintubation(3), tracheotomy(1), 2 wound complications(2) The more aggressive tumor first received chemotherapy, making a selection bias Post-operative mortality = 1.7%. Major morbidity (10% patients): stroke(1), peripheral arterial ischemia(1), severe sepsis(2) No perioperative death Morbidity = 6%: sputum retention and pneumonia, persistent air leak, neuropathic pain, empyema, wound breakdown, swollen flap, accidental pulmonary artery division Post operative complications 4 patients: pneumonia, prolonged air leakage, wound infection, arrhythmia
a b c d e f
Median survival = 42 months
No perioperative mortality Morbidity = 14% (4 prolonged air leak, 1 wound infection)
5 year OS = 24% post PM (0% in the no PM group)
Not available
PM = pulmonary metastasectomy. STS = soft tissue sarcoma. OS = overall survival. OST = osteosarcoma. FU = follow up. DFI = disease free interval.
5. Surgery of lung metastases from sarcomas 5.1. Outcomes and prognostic factors The largest ground of evidence for metastatic ablation in sarcomas comes from retrospective and one matched-control group series of lung metastasectomy [27]. Table 1 shows 5year survival rates of 15% [28] to 52% [29,30]. Retrospective studies are intrinsically biased but show some of the prognostic factors used in decision-making algorithms to select patients to undergo metastatic ablation. Indirect comparisons suggest that pulmonary metastasectomy yields better overall (and disease-free) survival than a systemic-only attitude, with median survivals of 51 and 30 months, respectively
[27]. A meta-analysis of 18 publications (5 bone, 6 soft tissue, 4 various) reported overall 5-year survival rates of 25% for metastatic bone and 15% for metastatic soft tissue sarcomas, with corresponding 5-year rates of 34% and 25% for patients undergoing a first metastasectomy [31]. Data from the Registry of Lung Metastases also showed slightly worse median survival (after complete resection) for soft tissue compared to bone sarcomas (and colorectal cancer) [32]. Disease-free interval (DFI), the time between diagnosis of primary and of first metastasis, is another prognostic factor for survival. Time to identification of lung metastases was found to be an important prognostic factor in osteosarcoma patients [33]. Kane et al. showed no benefit of metastasectomy in patients with synchronous metastases and primary
Please cite this article in press as: Olivier T, et al. Treating metastatic sarcomas locally: A paradoxe, a rationale, an evidence? Crit Rev Oncol/Hematol (2015), http://dx.doi.org/10.1016/j.critrevonc.2015.01.004
ONCH-1930; No. of Pages 16
ARTICLE IN PRESS T. Olivier et al. / Critical Reviews in Oncology/Hematology xxx (2015) xxx–xxx
7
Fig. 2. Case scenarios: “benefit/no benefit” is assumed based on the ability to significantly improve survival by at least 3 months (with a possibility to perform another ablation). AWOD: alive without disease, DOMD: dead of metastatic disease, RT: radiotherapy, DOD: dead of disease, MFH: malignat fibrous histiocytoma, PR: partial response.
[34]. Conversely, metastasectomy appears beneficial in soft tissue sarcomas when the DFI exceeds ≈6/7 months, this “cut-off” varying from 6 months to over 2 years among studies [35–37]. Another major prognostic factor is completeness of resection [28,36,38–46]. Gadd et al. reported an 83% complete resection rate in 78 patients, and 19, 10 and 8-months survival after complete, incomplete or no resection (p = 0.005) [47]. Of note, incomplete resection is less likely achievable in larger or poorly accessible tumors. Additionally, distinction between relapse at the site of metastasectomy or a distant site is not systematically reported. Other unfavorable prognostic factors for survival include high grade [36,38,48], age (≥40 [36] or 50 years [38], or 60 ◦ C, causing necrosis and destruction around a conductive electrode inserted under image-guidance. Nakamura et al. reported 3-year survival rates of 59% vs. 0% in metastatic sarcoma patients undergoing complete and incomplete RFA, respectively (Table 2) [73]. While the decision for metastasectomy or RFA may depend on age, survival rates were similar between patients ≥65 years or younger [74]. Minor side-effects of RFA include transient pain and discomfort. Death has been reported in up to 0.5% of cases in one series [75]. Pneumothorax occurs in up to 65% and requires chest tube insertion in up to 59% (Table 2) [76]. A single-institution retrospective study of 66 patients with liver metastases from
Please cite this article in press as: Olivier T, et al. Treating metastatic sarcomas locally: A paradoxe, a rationale, an evidence? Crit Rev Oncol/Hematol (2015), http://dx.doi.org/10.1016/j.critrevonc.2015.01.004
ARTICLE IN PRESS
ONCH-1930; No. of Pages 16
T. Olivier et al. / Critical Reviews in Oncology/Hematology xxx (2015) xxx–xxx
9
Table 2 Series on RFAa for sarcoma metastases. First Author Year of Publication
(Number of patients treated), organ treated
Local control
Survival data
Berber 2005 [132]
(18) Liver
Median survival = 25 months
Ding et Chua 2009 [133]
(4) Lung
Tumors recurred locally for 17% of the le-sions over a mean follow-up period of 24 months Not available
Nakamura 2009 [73]
(20) Lung
Local tumor progression 10%
Jones 2010 [134]
(25) Lung, liver
Not available
3 year survival = 29% median survival = 12.9 months 2 year survival from RFAa = 77%(GISTb )/40%(no GISTb )
Palussière 2011 [76]
(29) Lung
5 recurrences on RFA sites were noted during follow-up for a total of 47 treated metastases (10.6%)
3 year survival = 65.2%
Yamanaka 2013 [135]
(21) Liver from GISTb
95.2% local control at median follow-up of 30.6 months
Koelblinger 2013 [136]
(22) Lung
Primary local control rate = 95%
Overall survival = 85.7% (95% confidence interval, 33.6–97.8%) Mean survival = 51 months 2 year overall survival = 94% 3 year overall survival = 85%
a b
Median Disease Free Interval following RFAa = 19 months
Comments/complications
3 patients: no complication 1 patient: small pneumothorax Pneumothorax (65%) and chest tube (38%) Atrial fibrillation/infection RFAa cavity/sepsis(liver)/small pneumothorax(lung) Pneumothorax (68.7%) chest tube (59%), one sublobar alveolar hemorrhage, one chronic pneumothorax Median duration stay at hospita = 3days 30-day mortality = 0%
6.6% grade 3 CTC complications
RFA = radiofrequency ablation. GIST = gastro intestinal stromal tumor.
sarcomas treated by RFA and/or surgery found that treatment by RFA (alone or combined) was associated with shorter disease-free survival in comparison with surgical treatment [77]. RFA offers an alternative to surgery in inoperable patients but is most effective for lesions ≤35 mm.
7. Stereotactic body radiation therapy (SBRT) SBRT, also known as Stereotactic Ablative Radiotherapy (SABR), can be delivered by dedicated photon-based CyberKnife® SBRT or multivalent linear accelerators. It is the combination of highly conformal image-guided radiation with very steep dose gradients, which allow to deliver very high doses and to treat metastatic masses anywhere in the body using non-coplanar fields in a few fractions. Radiosurgery (mostly by GammaKnife) is a variant name for single fraction SBRT. SBRT allows higher doses while better preserving normal tissues from high and median doses than conventional radiation therapy [78]. Such high doses given in a few fractions (hypofractionation) potentially help overcome the so-called radioresistance of sarcomas (illustrated by low ␣/ [79,80] ≤1 Gy for liposarcomas [79] and osteosarcomas [80], respectively, versus 10 in head and neck carcinomas). Definitive conventional radiation therapy can alleviate symptoms but response rates are poor. The
ablative (eradication of macroscopic disease) radiation use was originally reported in 1999 for lung metastases from osteosarcoma [81]. Since then, SBRT has been used for intracranial and extracranial metastases. Corbin et al. recently highlighted the role of SBRT in treating extracranial metastases in a curative intent (not restricted to sarcomas) [82–84]. With a median overall survival after diagnosis of (exceptional) brain metastases from sarcomas of 2.7 months [85], surgery [86,87] and SBRT (16–18 Gy [88–91]) seem to offer similar results. The corresponding local control rates and median survival with SBRT range from 88% to 91.3%, and 9.8–16 months, respectively [88,92]. SBRT also yields 3-year local control rates of ≥80% in extracranial metastases with excellent tolerance (Table 3) although morbidity has been reported in sites like vertebrae treated with very high-dose fractions [93]. Such excellent control rates cannot be achieved with palliative doses of 30 Gy in 10 fractions. Contrary to conventional radiation therapy, SBRT delivers ablative doses [94] of ≈50 Gy in 10 fractions of 5 Gy over 2 weeks for metastatic sarcomas [95,96]. For example, CyberKnife-SBRT in spinal sarcomatous metastases can deliver an optimal tumoricidal dose that is superior to the tolerance dose of the spinal dose [97]. Whether total dose should be customized on sarcoma subtype has been questioned based on the observation that uterine sarcoma metastases yielded a 71% complete response rate that was higher than in other histological subtypes [94].
Please cite this article in press as: Olivier T, et al. Treating metastatic sarcomas locally: A paradoxe, a rationale, an evidence? Crit Rev Oncol/Hematol (2015), http://dx.doi.org/10.1016/j.critrevonc.2015.01.004
ARTICLE IN PRESS
ONCH-1930; No. of Pages 16
10
T. Olivier et al. / Critical Reviews in Oncology/Hematology xxx (2015) xxx–xxx
Table 3 Series on SBRT for sarcoma metastases (radiosurgery for brain metastases excluded). First Author Year of Publication
(Number of patients treated), organ treated
Local control
Survival data
Schedule/complications/comments
Rock 2002 [137] Levine 2009 [97]
(1) Lumbar spine of Ewing sarcoma (10) Spine
60–70%, when available
Excellent clinical control at 1 year Mean survival of 11.1 months (range: 1–21 months)
Novalis Shaped Beam Radiosurgery Case report CyberKnife, median treatment: 30 Gy at the 80% isodose in 3 fraction No major side effects or complications in this group of patients. Preferred dose and fractionation: 50 Gy in 5 Gy fractions over 2 weeks 3 patients: Grade 2 toxicity 1 Grade 3 toxicity (non-malignant pleural effusion)
Dhakal 2012 [96]
(15) Lung
7/74 local failures Estimated 3 year local control rate of 82%
Stragliotto 2012 [94]
(46) Lung, liver and other
Chang 2012 [138]
(27) Spine (included primary)
Mehta 2013 [139]
(16) Lung STS and bone sarcoma
136 tumors treated, as best response: 29% complete response (CR) 20% partial response (PR) 39% stable disease (SD) 12% progression disease (PD) Local control = CR + PR + SD = 88% Local control= 96.7% (29/30) lesions at 6 months, = 76.9% (10/13) at 2 years At 43 months, local control was 94%.
Folkert 2014 [140]
(88) Spine
Local contral = 87.9% at 12 months
The current strategy is rather to deliver a sufficient dose to the tumor regardless of histology, but to reduce the dose per fraction for large tumors and tumors close to normal tissues at risk of generating significant toxicities over a given dose. Hadrontherapy (proton or carbon therapy) is an interesting alternative to sophisticated photon-based SBRT in unresectable primary sarcomas [98,99]. Preliminary unpublished data suggest that carbon therapy may be particularly interesting in sarcomatous metastases owing to better radiobiological efficacy. However, the management of moving targets like lung metastases require adequate image-guidance systems, which are yet less developed than in photon-based SBRT. Based on similar outcomes and good tolerance, prospective surgery/SBRT randomized studies would be of interest. Non-surgical metastatic ablation, based on control rates, excellent tolerance and low morbidity, probably offers a valuable alternative to surgery that is not restricted to inoperable or marginally operable patients. It is valuable in metastatic
Median (range) survival: 2.1 years (0.8–11.5) for SBRT 0.6 years (0.1–7.8) for no SBRT (p value = 0.002). Median overall survival time = 26.3 months 3-year overall survival = 33.6% 5-year overall survival was 20%. 31% long-term survivors = survived longer than 36 months after first SBRT. Overall median survival = 29 months
Overall survival at 4 years was 72%
Overal survival at 1 year = 60.6%, median S = 16.9 months
4–20 Gy per fraction in 1–5 fractions with total doses of 10–48 Gy All patients but one: metastatic disease and medically inoperable. Very heterogeneous situations 1 perforation of the colon 1 contracture of the hip. No lethal side effect 16–45 Gy in 1–3 fractions, median single session equivalent dose = 21.8 Gy 54 Gy (36–54) in 3–4 fractions (mainly 54 Gy in 3 fractions) 0 grade 2–4 radiation pneumonitis, 0 radiation esophagitis 28.5 Gy (median) in 3–6 fractions Or 24 Gy (median) in single fraction 1% acute grade 3 toxicity, 4.5% chronic grade 3 toxicity, and no grade >3 toxicity
sites less prone to be operated on like bone [100] and rare metastatic sites. Liver, bone, brain, subcutaneous, soft tissue metastases and others locations are exceptional but increasingly likely with prolonged disease course. Metastatic liver ablation shows promising results in selected patients with surgery (median survival 24–54 months [101–104]) or RFA (Table 2). Preliminary data in numerous cancers suggest that SBRT can be performed safely and with good local control rates in liver metastases. Conversely, for intra-abdominal metastases, surgery is yet the preferred option, although with limited data [105,106].
8. Systemic and local treatments First-line standard treatment of metastatic sarcoma is chemotherapy but later lines are little evidence-supported [5]. Subgroup analyses suggest that chemotherapy is not equally
Please cite this article in press as: Olivier T, et al. Treating metastatic sarcomas locally: A paradoxe, a rationale, an evidence? Crit Rev Oncol/Hematol (2015), http://dx.doi.org/10.1016/j.critrevonc.2015.01.004
ONCH-1930; No. of Pages 16
ARTICLE IN PRESS T. Olivier et al. / Critical Reviews in Oncology/Hematology xxx (2015) xxx–xxx
important for all histological subtypes (with a spectrum going from Ewing sarcomas (chemosensitive) to clear cell sarcomas (chemoresistant)) [5,107,108]. In partial responders to systemic treatment, metastatic ablation might help consolidate local control at the treated metastatic sites [7]. On the other hand, chemotherapy given before surgery for patients with short DFI and large tumor bulk, may help assess tumor kinetics, tumor response [109] and select for further treatments [110–112]. Old data before the era of ifosfamid did not support the use of perioperative chemotherapy [113] and the EORTC-STBSG 62933 trial randomizing between chemotherapy followed by metastasectomy and metastasectomy alone failed because of slow accrual. Canter et al. compared 85 metastatic extremity soft-tissue sarcoma patients undergoing pulmonary resection alone with 53 patients undergoing surgery and perioperative chemotherapy [39]. DFI were 14 and 8 months in the resection alone and chemotherapy groups, respectively. After adjustment on propensity scores (statistical method accounting for biases), perioperative chemotherapy had minimal effect on survival, if any, in patients undergoing pulmonary resection. The large Rizzoli Institute series of metastatic bone sarcomas showed similar results [70].
9. Discussion With advances in techniques of metastatic ablation and limited, transient response rates with first-line systemic treatments, local ablative treatments are increasingly used. Changes in definitions of “limited metastatic burden”, “pauciprogressive” and “oligometastatic disease” appear to parallel changes in technological improvements. As experienced with intensity modulated radiation therapy or proton therapy in other cancers, showing the benefit of innovative ablative techniques for metastases from sarcomas is a challenge. The results of recent chemotherapy trials underpowered by the unplanned use of metastatic ablation techniques (mainly surgery or radiotherapy) in >20% of the patients suggest that omission of metastatic ablation is considered unethical at least as consolidation following incomplete response to chemotherapy [7]. Metastatic ablation shows a potential to eradicate macroscopic tumor (>80% local control rates with surgery and SBRT) and prolong survival in selected cases, i.e. to realize a switch from purely palliative intent to more curative intent. However, demonstration of a benefit from metastatic ablation is a road with hurdles due to our inability to resolve the methodological issues (intrinsic biases of retrospective studies) and overcome the reluctance/conviction involved in the institution-dependent patient selection processes for aggressive local metastatic treatments. Two scenari can be identified: upfront chemotherapy (+/adjuvant) in patients who have relatively short DFI (
ARTICLE IN PRESS
Critical Reviews in Oncology/Hematology xxx (2015) xxx–xxx
Treating metastatic sarcomas locally: A paradoxe, a rationale, an evidence? Timothée Olivier a,b , Daniel Pop c , Amina Chouiter Djebaili a , Alexander Tuan Falk d , Antoine Iannessi e , Esma Saada f , Willy Nettekoven a , Jean-Yves Blay g , Patrick Baque h , Didier Cupissol b , Antoine Thyss f , Juliette Thariat d,∗ a
Hôpital neuchâtelois, Département d’Oncologie, Chasseral 20, 2300 La Chaux-de-Fonds, Switzerland Institut régional du Cancer de Montpellier, Parc Euromédecine, 208 rue des Apothicaires, 34298 Montpellier Cedex 5, France c Centre Hospitalo-Universitaire Pasteur, Service de Chirurgie Thoracique, 30 Avenue de la Voie Romaine, 06000 Nice, France Centre Antoine Lacassagne, Département d’oncologie, Département de radiothérapie, 227, Avenue de la Lanterne, 06200 Nice, France e Centre Antoine Lacassagne, Département de radiologie, 33, Avenue Valombrose, 06189 Nice, France f Centre Antoine Lacassagne, Département d’oncologie médicale, 33, Avenue Valombrose, 06189 Nice, France g Département d’Oncologie, Centre Léon Bérard, 28 Rue Laennec, 69008 Lyon, France h Centre Hospitalo-Universitaire Saint Roch, Service de Chirurgie, 06000 Nice, France b
d
Accepted 6 January 2015
Contents 1. 2. 3. 4. 5.
Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . History of local treatments in sarcomas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Strategy adapted on metastatic burden . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Design: review criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Surgery of lung metastases from sarcomas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.1. Outcomes and prognostic factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2. Surgery and development of mini-invasive procedures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6. Radiofrequency ablation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7. Stereotactic body radiation therapy (SBRT) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8. Systemic and local treatments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9. Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conflict of interest . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Reviewer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
Abstract Purpose: The mainstay of first line treatment in metastatic sarcomas is chemotherapy with response rates of ≈25% but the optimal management of further events is debated. We assessed the benefit of local metastatic treatment in metastatic sarcomas. Results: Local control of local treatment strategies (≈85%) is excellent but highly institution-dependent and subject to selection biases. Formal evidence of an improvement of survival with local ablative treatments has been limited to retrospective studies. On the other hand, some chemotherapy trials are inconclusive because about 20% of patients receive local metastatic ablation as it is considered unethical to omit
∗
Corresponding author at: Department of Radiation Oncology, 33 Av Valombrose, 06189 Nice, France. Tel.: +33 492031083; fax: +33 492031096. E-mail address: [email protected] (J. Thariat).
http://dx.doi.org/10.1016/j.critrevonc.2015.01.004 1040-8428/© 2015 Elsevier Ireland Ltd. All rights reserved.
Please cite this article in press as: Olivier T, et al. Treating metastatic sarcomas locally: A paradoxe, a rationale, an evidence? Crit Rev Oncol/Hematol (2015), http://dx.doi.org/10.1016/j.critrevonc.2015.01.004
ONCH-1930; No. of Pages 16
2
ARTICLE IN PRESS T. Olivier et al. / Critical Reviews in Oncology/Hematology xxx (2015) xxx–xxx
local treatment in these patients. Further, technology has made surgery, stereotactic irradiation and radiofrequency ablation highly effective on local control with limited morbidity. Conclusion: The benefit on survival of metastatic ablation deserves prospective studies integrating quality of life, cost effectiveness and patient-reported outcomes assessment. © 2015 Elsevier Ireland Ltd. All rights reserved.
Keywords: Sarcoma; Metastases; Chemotherapy; Ablative treatment; Metastasectomy; Stereotactic radiotherapy
1. Introduction
2. History of local treatments in sarcomas
Sarcomas represent a rare (1% of cancers) and heterogeneous group of about 100 subtypes of soft-tissue and bone tumors of mesenchymal origin [1,2]. When surgically removed at an early stage, sarcomas can be cured [1]. However, up to 50% of patients develop metastases during the course of their disease [3]. Median survival measured from the diagnosis of metastases has improved from 12 to over 18 months in the last decades [4]. Such improvement is multifactorial, including advances in imaging, lines of systemic treatment, better supportive care and metastatic ablation. Chemotherapy remains the standard of care of metastatic sarcomas but hardly cures any patient and rather prolongs survival by a few weeks or months. Performed alone, it results in 5-year overall survival rates of less than 10% with exceptional long survivors [5,6]. Further, most series include both oligometastatic (defined as 1–3 or 6 synchronous metastases) and polymetastatic patients. These might overestimate the benefit of chemotherapy in polymetastatic patients. The management of metastatic sarcomas raises several other controversial issues. It is uncertain whether local metastatic ablation truly improves outcomes or rather represents a selected population with slowly progressive limited tumor bulk of spontaneously better prognosis regardless of treatment. Additionally, recent evidence suggests that numerous metastatic patients indeed receive metastatic resection in addition to systemic treatment [7]. Local metastatic treatments are associated with chemotherapy delivered sequentially or perioperatively, or may be used as exclusive therapy. Lung metastasectomy has been performed in many cancers for more than 20 years [8]. In sarcomas, lung metastases are the predominating site of metastases and the exclusive site in over 50% of patients at first metastatic event [9–11]. Lung metastases are rarely symptomatic and often discovered during follow-up Computed Tomography scan (CT). Extra-pulmonary and atypical metastatic sites become more likely with prolonged disease course [12,13]. While surgery has the largest ground of evidence, other local ablative techniques like radiofrequency ablation (RFA), and focused high-dose whole-body stereotactic radiation therapy (SBRT) also show promising local control rates and excellent tolerance for pulmonary and extrapulmonary metastases. In the present review, we addressed the current level of evidence for local ablative treatments in metastatic sarcomas.
Pulmonary metastasectomy was first reported by Weinlecher in 1882 and by Divis in 1927 as a planned procedure [8]. Following publication of a first case series, Thomford described the principles of lung metastasectomy in 1965 [8]. Both cytotoxic chemotherapy and metastasectomy emerged in the 1970s as major advances in sarcomas (and in colorectal cancers). Moreover, sarcomas occur in young patients and are often relatively chemoresistant. Both reasons provided a strong rationale for aggressive local management of metastases from sarcomas. Among 184 osteosarcoma patients treated at the Memorial Sloan Kettering Cancer Center by limb amputation, 75% presented lung metastases within 18 months of their primary and there were no 5year survivors [14]. In contrast, a pilot study by Martini et al., in 1971, reported a 3-year survival of 40% in 22 patients after lung metastasectomy [15]. Eligibility criteria for surgery were controlled primary tumor, planned complete resection, adequate cardio-pulmonary function and absence of progressive extra-thoracic metastases. Since then, lung metastasectomy has been performed on a case-by-case basis, with yet limited overall evidence. In the same period, doxorubicin and ifosfamid-based chemotherapy also considerably changed the management of metastatic sarcomas. However, first-line chemotherapy yields limited overall response rates in the order of 10–25% and aims at improving three-month progression-free survival (PFS), suggesting that sustained response is rarely achievable [5]. Second-line (and over) systemic therapies are even less supported despite being routinely performed in practice [5]. Moreover, such sarcoma subtypes as alveolar and clear cell sarcomas have been very poorly sensitive to chemotherapy (although this may change with new histology-driven targeted therapies) (Fig. 1).
3. Strategy adapted on metastatic burden The justification of treating metastases aggressively includes [16]: (1) Anecdotal experience suggests that selected patients benefit from metastatic ablation, irrespective of systemic therapy. (2) Consolidative local therapy yields good outcomes in partial responders to systemic therapy.
Please cite this article in press as: Olivier T, et al. Treating metastatic sarcomas locally: A paradoxe, a rationale, an evidence? Crit Rev Oncol/Hematol (2015), http://dx.doi.org/10.1016/j.critrevonc.2015.01.004
ONCH-1930; No. of Pages 16
ARTICLE IN PRESS T. Olivier et al. / Critical Reviews in Oncology/Hematology xxx (2015) xxx–xxx
3
Fig. 1. Timeline: major breakthroughs and change of paradigm in the management of metastases from sarcomas.
(3) Oligometastases, as described by Hellman and Weichselbaum [17,18], represent a subgroup of patients of intermediate disease stage between primary and widely metastatic disease. Eradication of oligometastases might lead to prolonged remission or cure. Identification of patients truly who benefiting is a challenge. In the 1980s, metastatic ability was found heterogeneous amongst mouse KHT sarcoma cells [19]. Lussier et al. recently demonstrated that there may be a micro-RNA signature for either oligoor polymetastatic progression [20,21]. These works, among others, are preliminary evidence that the oligometastatic state may have a specific biologic background and may be predicted with subsequent therapeutic implications. (4) Based on the Norton-Simon hypothesis, debulking may make residual tumor more sensitive to chemotherapy (based on Gompertzian kinetics) [22]. In the metastatic cascade model, tumor doubling time increases with subsequent metastases [23]. Interactions between primary and metastases can lead to metastatic growth or regression (such as after treatment of primary renal cancer) [24]. Local metastatic ablation aims primarily at eradicating macroscopic disease through targeted intervention. Along with chemotherapy [25], it may also stimulate antitumor immunity [26]. Such non-targeted effects (including the abscopal effect) might also be involved in the antitumoural effects of local metastatic ablation. Practically, local
metastatic treatment may alleviate symptoms, prolong PFS and/or cure selected patients depending on metastatic burden, i.e. number, volume and metastatic growth kinetics. Technical improvements (minimally-invasive surgery, RFA, SBRT, etc.) have rendered ablative treatments feasible for metastases situated nearly anywhere in the body with limited morbidity. The physical, psychological, social or economic impact of local therapy of metastases should thus be balanced against the expected benefits of systemic treatment only. An illustration of different case scenarios is presented in Fig. 2.
4. Design: review criteria A literature search, according to PRISMA statement, was performed on MEDLINE (http://pubmed.gov). Articles were searched using key words (radiofrequency, ablation, stereotactic, radiosurgery, pulmonary resection, metastasectomy, lung resection/metastasis, metastases, metastatic/sarcoma, sarcomas, osteosarcoma, osteosarcomas) and evaluated on title and abstract. Full text articles meeting criteria were collected to assess for definite inclusion. Inclusion in Table 1 and prognostic factors analysis of surgical series (numerous) was restricted to series with more than 20 patients published from 1994 to 1st January 2014. For SBRT and RFA (scarce) series, these limitations were not applied. A PRISMA flowchart is described in Fig. 3.
Please cite this article in press as: Olivier T, et al. Treating metastatic sarcomas locally: A paradoxe, a rationale, an evidence? Crit Rev Oncol/Hematol (2015), http://dx.doi.org/10.1016/j.critrevonc.2015.01.004
ARTICLE IN PRESS
ONCH-1930; No. of Pages 16
4
T. Olivier et al. / Critical Reviews in Oncology/Hematology xxx (2015) xxx–xxx
Table 1 Series on lung surgical resection for sarcoma metastases. First Author, Year
Patients with lung meta (treated with PMa )
Inclusion criteria for metastasectomy when available (inclusion period)
Survival data
Comments/Complications (when available) (number of patients)
Robinson 1994 [29]
189(44) STSb
Not homogeneous for all patients (1970–1990)
5 year-OSc after PM = 52%
Choong 1995 [111]
274(214) STS
(1976–1991)
Kawai 1995 [120]
(23) STS
(1970–1992)
Temeck 1995 [46]
(152) OSTd and STS (pediatric)
No extrapulmonary metastases (1975–1993)
5 year-OS after PM = 40% 10 year-OS = 20% Median survival = 28 months 5 year actuarial OS = 32% Median survival = 2.2 year
Complications = 13.6% including: One post operative death pulmonary embolism(1) Not available
Van Geel 1996 [36]
(255) STS
(Not available–1993)
Billingsley 1999 [38]
719(213) STS
(1982–1997)
Antunes 1999 [121]
198(31) OST
Suri 2005 [52]
(103) MFH
Inclusion: no local recurrence, no other metastases, all meta operable with low operative risk (1989–1997) (1976–2000)
Suemitsu 2005 [49]
(43) Bone and STS
Pfannschmidt 2006 [122]
(50) STS
Pfannschmidt 2006 [123]
(21) OST
Suzuki 2006 [124]
(105) OST and STS
Harting 2006 [53]
137(99) OST
Inclusion: primary controllable, good general condition, no extrathoracic lesions, anatomically and functionally resectable (1981–1999) Inclusion: primary controlled, resectable, general and functional risks tolerable, no extrathoracic lesions (1996–2002) Inclusion: primary controlled, resectable, general and functional risks tolerable, no extrathoracic lesions (1997–2001) Inclusion: operable, resectable, control primary, no extrathoracic lesions (1990–2002) Exclusion: unresectable disease, death before surgery, refused surgery, apparent resolution with chemo, extensive local recurrence before surgery (1980–2000)
3 year-OS = 54% 5 year OS = 38% All patients: Median survival = 15 months Actuarial 3 year S = 25% PM complete resection (n = 161): Median OS = 33 months Actuarial 3 year-S = 46% 3 year-S = 61%
Median FUe = 18 months 5 year-OS 21%
Median survival = 30 months 5 year-OS = 20.7%
Not available
Complications = 6.5% (after 1st procedure): Prolonged air leak(3), pneumonia(2), superficial wound infection(2) respiratory insufficiency(1), pneumonitis(1) atrial tear(1) One post operative death Not available
Ventilation for 48 h(1)
Operative mortality = 1% Morbidity = 11%: prolonged air leak(7), pneumonia, empyema, atelectasis, and sepsis. No post operative death
Median survival = 39.5 months
No operative mortality
Median survival = 35.8 months
No operative mortality 2 patients requiring re-thoracotomy for hemorrhage
5 year-S = 43.6%
Not available
PMa : Median survival = 33.6 months No PM = 10.1 months 3 year-OS = 46.2% 5 year-OS = 29.0%
11.8% intra or post-operative complications: pneumonia, transfusions, drain more than 7 days, reoperation for hemorrhage, mechanic ventilation >7 days
Please cite this article in press as: Olivier T, et al. Treating metastatic sarcomas locally: A paradoxe, a rationale, an evidence? Crit Rev Oncol/Hematol (2015), http://dx.doi.org/10.1016/j.critrevonc.2015.01.004
ARTICLE IN PRESS
ONCH-1930; No. of Pages 16
T. Olivier et al. / Critical Reviews in Oncology/Hematology xxx (2015) xxx–xxx
5
Table 1 (Continued ) First Author, Year
Patients with lung meta (treated with PMa )
Inclusion criteria for metastasectomy when available (inclusion period)
Survival data
Comments/Complications (when available) (number of patients)
Rehders 2007 [125]
121(61) STS
Median survival after PMa = 33 months
Complications: Prolonged pleural drain = 9, wound infection = 3, hypoventilation = 3, No perioperative mortality
Liebl 2007 [48] Chen 2008 [51]
(42) STS (23) STS
3 year-OS = 31% 5 year-OS = 31%
Not available No deaths directly from surgery
Chen 2009 [71]
(23) OST
Inclusion: less than 6 metastases, no contraindication i.e.:insufficient control of the primary tumor, unresectable lung disease, extensive involvement of the mediastinum or the chest wall, and unresectable metastatic disease outside the lung (1991–2002) (1990–2005) Inclusion: resectable, operable, limited to lung, primary controlled or potentially controlled (1989–2007) Inclusion: resectable, operable, limited to lung, primary controlled (1990–2007)
No deaths directly from surgery
Chen 2009 [40]
(25) OST
Inclusion: resectable, operable, limited to lung, primary controlled (1989–2007)
Franco 2009 [126]
(22) STS
Inclusion: primary under control, resectable, operable, -no other metastases (1996–2006)
Franco 2009 [127]
(52) OST
Inclusion: primary under control, resectable, operable, -no other metastases (1996–2006)
Median FU = 13 months 2 year-OS (after PMa ) = 52% 5 year-OS = 43% 10 year-OS = 29% Median FU = 16 months 2 year-OS (after PM) = 42.9% 5 year-OS = 19.0 Median FU = 14months Median survival = 19 months 5 year-OS = 23.1% 3 year-OS = 47.1% Median FU = 28months Median survival = 27 months 3 year-OS = 43% 5 year-OS = 31%
Smith 2009 [28]
(94) STS
Inclusion: all patients receiving PM (1976–2000)
Buddingh 2009 [50]
88(56) OST
Blackmon 2009 [128]
(234) Bone and STS
Exclusion: insufficient data, unresectable primary (1990–2008) (1998–2006)
Aljubran 2009 [129]
85(47)OST
(1986–2003)
Sardenberg 2010 [41]
(77) STS
Inclusion: primary control, exclusively pulmonary metastases, resectable, operable, no better treatment (1990–2006)
Briccoli 2010 [70]
(323) OST
Inclusion: primary tumor controlled, no pleural or pericardial effusion, no other metastases, with adequate residual pulmonary function following resection (1985–2005)
Median survival = 16 months Actual 5 year-DFS = 5% Actual 5 year-OS = 15% No difference in OS between synchronous and metachronous Median survival = 36.2 months 3 year-OS = 38% in the group surgery 5 year-OS = 34.7%
5 year-OS = 37%
No deaths directly from surgery
Post operative complications: 72 h ventilation mechanical(1), persistent air leak(1) No mortality Major morbidity 4 patients: mechanical ventilation more than 24 hrs due to acute respiratory distress syndrome, pneumonia(2), persistent air leak-2) Overall 30 day mortality = 3.7% Actual FU2 ≥5 years Not available
Comparison between groups with different metastatic presentation and treatment No difference for age Complications (9.1%): fever, atelectasis, prolonged air leakage >7days, infection, prolonged hospital stay >30 days No 30 days-mortality Mixed synchronous and metachronous No data complications
Please cite this article in press as: Olivier T, et al. Treating metastatic sarcomas locally: A paradoxe, a rationale, an evidence? Crit Rev Oncol/Hematol (2015), http://dx.doi.org/10.1016/j.critrevonc.2015.01.004
ARTICLE IN PRESS
ONCH-1930; No. of Pages 16
6
T. Olivier et al. / Critical Reviews in Oncology/Hematology xxx (2015) xxx–xxx
Table 1 (Continued ) First Author, Year
Patients with lung meta (treated with PMa )
Inclusion criteria for metastasectomy when available (inclusion period)
Survival data
Comments/Complications (when available) (number of patients)
Predina 2011 [30] Kim 2011 [42]
(48) STS (89) STS and OST
(1995–2007) Inclusion: complete local control, exhibited isolated pulmonary metastases (2002–2008)
5 year-OS = 52% After pulmonary resection: Median OS = 10.9months 5 year-OS = 50% 5 year-DFS = 18.9%
Salah 2013 [130]
71(30) OST and STS
Inclusion: primary site has to be controlled, and metastases are amenable to surgical resection without residual disease (2001–2012)
From PM: Median survival = 39.6 months
Dear 2012 [44]
(114) Bone and STS
(1978–2008)
Median OS = 37months 3 year OS = 53% 5 year-OS = 43%
Mizuno 2012 [43]
(52) OST and STS
From PM Median survival = 33.3months 5 year-OS = 50.9%
Toussi 2013 [45]
(34) STS
Stanelle 2013 [131]
41 (31)Synovial sarcoma pediatric
Inclusion: primary and other distant site controlled, resectable But PM indicated on background, respiratory function, length of the DFIf after treatment of primary site, number and locations (1996–2011) Inclusion: no active uncontrollable extrapulmonary disease, high probability of complete resection, sufficient pulmonary reserve, operable (1996–2007) (1971–2011)
No perioperative death Reintubation(3), tracheotomy(1), 2 wound complications(2) The more aggressive tumor first received chemotherapy, making a selection bias Post-operative mortality = 1.7%. Major morbidity (10% patients): stroke(1), peripheral arterial ischemia(1), severe sepsis(2) No perioperative death Morbidity = 6%: sputum retention and pneumonia, persistent air leak, neuropathic pain, empyema, wound breakdown, swollen flap, accidental pulmonary artery division Post operative complications 4 patients: pneumonia, prolonged air leakage, wound infection, arrhythmia
a b c d e f
Median survival = 42 months
No perioperative mortality Morbidity = 14% (4 prolonged air leak, 1 wound infection)
5 year OS = 24% post PM (0% in the no PM group)
Not available
PM = pulmonary metastasectomy. STS = soft tissue sarcoma. OS = overall survival. OST = osteosarcoma. FU = follow up. DFI = disease free interval.
5. Surgery of lung metastases from sarcomas 5.1. Outcomes and prognostic factors The largest ground of evidence for metastatic ablation in sarcomas comes from retrospective and one matched-control group series of lung metastasectomy [27]. Table 1 shows 5year survival rates of 15% [28] to 52% [29,30]. Retrospective studies are intrinsically biased but show some of the prognostic factors used in decision-making algorithms to select patients to undergo metastatic ablation. Indirect comparisons suggest that pulmonary metastasectomy yields better overall (and disease-free) survival than a systemic-only attitude, with median survivals of 51 and 30 months, respectively
[27]. A meta-analysis of 18 publications (5 bone, 6 soft tissue, 4 various) reported overall 5-year survival rates of 25% for metastatic bone and 15% for metastatic soft tissue sarcomas, with corresponding 5-year rates of 34% and 25% for patients undergoing a first metastasectomy [31]. Data from the Registry of Lung Metastases also showed slightly worse median survival (after complete resection) for soft tissue compared to bone sarcomas (and colorectal cancer) [32]. Disease-free interval (DFI), the time between diagnosis of primary and of first metastasis, is another prognostic factor for survival. Time to identification of lung metastases was found to be an important prognostic factor in osteosarcoma patients [33]. Kane et al. showed no benefit of metastasectomy in patients with synchronous metastases and primary
Please cite this article in press as: Olivier T, et al. Treating metastatic sarcomas locally: A paradoxe, a rationale, an evidence? Crit Rev Oncol/Hematol (2015), http://dx.doi.org/10.1016/j.critrevonc.2015.01.004
ONCH-1930; No. of Pages 16
ARTICLE IN PRESS T. Olivier et al. / Critical Reviews in Oncology/Hematology xxx (2015) xxx–xxx
7
Fig. 2. Case scenarios: “benefit/no benefit” is assumed based on the ability to significantly improve survival by at least 3 months (with a possibility to perform another ablation). AWOD: alive without disease, DOMD: dead of metastatic disease, RT: radiotherapy, DOD: dead of disease, MFH: malignat fibrous histiocytoma, PR: partial response.
[34]. Conversely, metastasectomy appears beneficial in soft tissue sarcomas when the DFI exceeds ≈6/7 months, this “cut-off” varying from 6 months to over 2 years among studies [35–37]. Another major prognostic factor is completeness of resection [28,36,38–46]. Gadd et al. reported an 83% complete resection rate in 78 patients, and 19, 10 and 8-months survival after complete, incomplete or no resection (p = 0.005) [47]. Of note, incomplete resection is less likely achievable in larger or poorly accessible tumors. Additionally, distinction between relapse at the site of metastasectomy or a distant site is not systematically reported. Other unfavorable prognostic factors for survival include high grade [36,38,48], age (≥40 [36] or 50 years [38], or 60 ◦ C, causing necrosis and destruction around a conductive electrode inserted under image-guidance. Nakamura et al. reported 3-year survival rates of 59% vs. 0% in metastatic sarcoma patients undergoing complete and incomplete RFA, respectively (Table 2) [73]. While the decision for metastasectomy or RFA may depend on age, survival rates were similar between patients ≥65 years or younger [74]. Minor side-effects of RFA include transient pain and discomfort. Death has been reported in up to 0.5% of cases in one series [75]. Pneumothorax occurs in up to 65% and requires chest tube insertion in up to 59% (Table 2) [76]. A single-institution retrospective study of 66 patients with liver metastases from
Please cite this article in press as: Olivier T, et al. Treating metastatic sarcomas locally: A paradoxe, a rationale, an evidence? Crit Rev Oncol/Hematol (2015), http://dx.doi.org/10.1016/j.critrevonc.2015.01.004
ARTICLE IN PRESS
ONCH-1930; No. of Pages 16
T. Olivier et al. / Critical Reviews in Oncology/Hematology xxx (2015) xxx–xxx
9
Table 2 Series on RFAa for sarcoma metastases. First Author Year of Publication
(Number of patients treated), organ treated
Local control
Survival data
Berber 2005 [132]
(18) Liver
Median survival = 25 months
Ding et Chua 2009 [133]
(4) Lung
Tumors recurred locally for 17% of the le-sions over a mean follow-up period of 24 months Not available
Nakamura 2009 [73]
(20) Lung
Local tumor progression 10%
Jones 2010 [134]
(25) Lung, liver
Not available
3 year survival = 29% median survival = 12.9 months 2 year survival from RFAa = 77%(GISTb )/40%(no GISTb )
Palussière 2011 [76]
(29) Lung
5 recurrences on RFA sites were noted during follow-up for a total of 47 treated metastases (10.6%)
3 year survival = 65.2%
Yamanaka 2013 [135]
(21) Liver from GISTb
95.2% local control at median follow-up of 30.6 months
Koelblinger 2013 [136]
(22) Lung
Primary local control rate = 95%
Overall survival = 85.7% (95% confidence interval, 33.6–97.8%) Mean survival = 51 months 2 year overall survival = 94% 3 year overall survival = 85%
a b
Median Disease Free Interval following RFAa = 19 months
Comments/complications
3 patients: no complication 1 patient: small pneumothorax Pneumothorax (65%) and chest tube (38%) Atrial fibrillation/infection RFAa cavity/sepsis(liver)/small pneumothorax(lung) Pneumothorax (68.7%) chest tube (59%), one sublobar alveolar hemorrhage, one chronic pneumothorax Median duration stay at hospita = 3days 30-day mortality = 0%
6.6% grade 3 CTC complications
RFA = radiofrequency ablation. GIST = gastro intestinal stromal tumor.
sarcomas treated by RFA and/or surgery found that treatment by RFA (alone or combined) was associated with shorter disease-free survival in comparison with surgical treatment [77]. RFA offers an alternative to surgery in inoperable patients but is most effective for lesions ≤35 mm.
7. Stereotactic body radiation therapy (SBRT) SBRT, also known as Stereotactic Ablative Radiotherapy (SABR), can be delivered by dedicated photon-based CyberKnife® SBRT or multivalent linear accelerators. It is the combination of highly conformal image-guided radiation with very steep dose gradients, which allow to deliver very high doses and to treat metastatic masses anywhere in the body using non-coplanar fields in a few fractions. Radiosurgery (mostly by GammaKnife) is a variant name for single fraction SBRT. SBRT allows higher doses while better preserving normal tissues from high and median doses than conventional radiation therapy [78]. Such high doses given in a few fractions (hypofractionation) potentially help overcome the so-called radioresistance of sarcomas (illustrated by low ␣/ [79,80] ≤1 Gy for liposarcomas [79] and osteosarcomas [80], respectively, versus 10 in head and neck carcinomas). Definitive conventional radiation therapy can alleviate symptoms but response rates are poor. The
ablative (eradication of macroscopic disease) radiation use was originally reported in 1999 for lung metastases from osteosarcoma [81]. Since then, SBRT has been used for intracranial and extracranial metastases. Corbin et al. recently highlighted the role of SBRT in treating extracranial metastases in a curative intent (not restricted to sarcomas) [82–84]. With a median overall survival after diagnosis of (exceptional) brain metastases from sarcomas of 2.7 months [85], surgery [86,87] and SBRT (16–18 Gy [88–91]) seem to offer similar results. The corresponding local control rates and median survival with SBRT range from 88% to 91.3%, and 9.8–16 months, respectively [88,92]. SBRT also yields 3-year local control rates of ≥80% in extracranial metastases with excellent tolerance (Table 3) although morbidity has been reported in sites like vertebrae treated with very high-dose fractions [93]. Such excellent control rates cannot be achieved with palliative doses of 30 Gy in 10 fractions. Contrary to conventional radiation therapy, SBRT delivers ablative doses [94] of ≈50 Gy in 10 fractions of 5 Gy over 2 weeks for metastatic sarcomas [95,96]. For example, CyberKnife-SBRT in spinal sarcomatous metastases can deliver an optimal tumoricidal dose that is superior to the tolerance dose of the spinal dose [97]. Whether total dose should be customized on sarcoma subtype has been questioned based on the observation that uterine sarcoma metastases yielded a 71% complete response rate that was higher than in other histological subtypes [94].
Please cite this article in press as: Olivier T, et al. Treating metastatic sarcomas locally: A paradoxe, a rationale, an evidence? Crit Rev Oncol/Hematol (2015), http://dx.doi.org/10.1016/j.critrevonc.2015.01.004
ARTICLE IN PRESS
ONCH-1930; No. of Pages 16
10
T. Olivier et al. / Critical Reviews in Oncology/Hematology xxx (2015) xxx–xxx
Table 3 Series on SBRT for sarcoma metastases (radiosurgery for brain metastases excluded). First Author Year of Publication
(Number of patients treated), organ treated
Local control
Survival data
Schedule/complications/comments
Rock 2002 [137] Levine 2009 [97]
(1) Lumbar spine of Ewing sarcoma (10) Spine
60–70%, when available
Excellent clinical control at 1 year Mean survival of 11.1 months (range: 1–21 months)
Novalis Shaped Beam Radiosurgery Case report CyberKnife, median treatment: 30 Gy at the 80% isodose in 3 fraction No major side effects or complications in this group of patients. Preferred dose and fractionation: 50 Gy in 5 Gy fractions over 2 weeks 3 patients: Grade 2 toxicity 1 Grade 3 toxicity (non-malignant pleural effusion)
Dhakal 2012 [96]
(15) Lung
7/74 local failures Estimated 3 year local control rate of 82%
Stragliotto 2012 [94]
(46) Lung, liver and other
Chang 2012 [138]
(27) Spine (included primary)
Mehta 2013 [139]
(16) Lung STS and bone sarcoma
136 tumors treated, as best response: 29% complete response (CR) 20% partial response (PR) 39% stable disease (SD) 12% progression disease (PD) Local control = CR + PR + SD = 88% Local control= 96.7% (29/30) lesions at 6 months, = 76.9% (10/13) at 2 years At 43 months, local control was 94%.
Folkert 2014 [140]
(88) Spine
Local contral = 87.9% at 12 months
The current strategy is rather to deliver a sufficient dose to the tumor regardless of histology, but to reduce the dose per fraction for large tumors and tumors close to normal tissues at risk of generating significant toxicities over a given dose. Hadrontherapy (proton or carbon therapy) is an interesting alternative to sophisticated photon-based SBRT in unresectable primary sarcomas [98,99]. Preliminary unpublished data suggest that carbon therapy may be particularly interesting in sarcomatous metastases owing to better radiobiological efficacy. However, the management of moving targets like lung metastases require adequate image-guidance systems, which are yet less developed than in photon-based SBRT. Based on similar outcomes and good tolerance, prospective surgery/SBRT randomized studies would be of interest. Non-surgical metastatic ablation, based on control rates, excellent tolerance and low morbidity, probably offers a valuable alternative to surgery that is not restricted to inoperable or marginally operable patients. It is valuable in metastatic
Median (range) survival: 2.1 years (0.8–11.5) for SBRT 0.6 years (0.1–7.8) for no SBRT (p value = 0.002). Median overall survival time = 26.3 months 3-year overall survival = 33.6% 5-year overall survival was 20%. 31% long-term survivors = survived longer than 36 months after first SBRT. Overall median survival = 29 months
Overall survival at 4 years was 72%
Overal survival at 1 year = 60.6%, median S = 16.9 months
4–20 Gy per fraction in 1–5 fractions with total doses of 10–48 Gy All patients but one: metastatic disease and medically inoperable. Very heterogeneous situations 1 perforation of the colon 1 contracture of the hip. No lethal side effect 16–45 Gy in 1–3 fractions, median single session equivalent dose = 21.8 Gy 54 Gy (36–54) in 3–4 fractions (mainly 54 Gy in 3 fractions) 0 grade 2–4 radiation pneumonitis, 0 radiation esophagitis 28.5 Gy (median) in 3–6 fractions Or 24 Gy (median) in single fraction 1% acute grade 3 toxicity, 4.5% chronic grade 3 toxicity, and no grade >3 toxicity
sites less prone to be operated on like bone [100] and rare metastatic sites. Liver, bone, brain, subcutaneous, soft tissue metastases and others locations are exceptional but increasingly likely with prolonged disease course. Metastatic liver ablation shows promising results in selected patients with surgery (median survival 24–54 months [101–104]) or RFA (Table 2). Preliminary data in numerous cancers suggest that SBRT can be performed safely and with good local control rates in liver metastases. Conversely, for intra-abdominal metastases, surgery is yet the preferred option, although with limited data [105,106].
8. Systemic and local treatments First-line standard treatment of metastatic sarcoma is chemotherapy but later lines are little evidence-supported [5]. Subgroup analyses suggest that chemotherapy is not equally
Please cite this article in press as: Olivier T, et al. Treating metastatic sarcomas locally: A paradoxe, a rationale, an evidence? Crit Rev Oncol/Hematol (2015), http://dx.doi.org/10.1016/j.critrevonc.2015.01.004
ONCH-1930; No. of Pages 16
ARTICLE IN PRESS T. Olivier et al. / Critical Reviews in Oncology/Hematology xxx (2015) xxx–xxx
important for all histological subtypes (with a spectrum going from Ewing sarcomas (chemosensitive) to clear cell sarcomas (chemoresistant)) [5,107,108]. In partial responders to systemic treatment, metastatic ablation might help consolidate local control at the treated metastatic sites [7]. On the other hand, chemotherapy given before surgery for patients with short DFI and large tumor bulk, may help assess tumor kinetics, tumor response [109] and select for further treatments [110–112]. Old data before the era of ifosfamid did not support the use of perioperative chemotherapy [113] and the EORTC-STBSG 62933 trial randomizing between chemotherapy followed by metastasectomy and metastasectomy alone failed because of slow accrual. Canter et al. compared 85 metastatic extremity soft-tissue sarcoma patients undergoing pulmonary resection alone with 53 patients undergoing surgery and perioperative chemotherapy [39]. DFI were 14 and 8 months in the resection alone and chemotherapy groups, respectively. After adjustment on propensity scores (statistical method accounting for biases), perioperative chemotherapy had minimal effect on survival, if any, in patients undergoing pulmonary resection. The large Rizzoli Institute series of metastatic bone sarcomas showed similar results [70].
9. Discussion With advances in techniques of metastatic ablation and limited, transient response rates with first-line systemic treatments, local ablative treatments are increasingly used. Changes in definitions of “limited metastatic burden”, “pauciprogressive” and “oligometastatic disease” appear to parallel changes in technological improvements. As experienced with intensity modulated radiation therapy or proton therapy in other cancers, showing the benefit of innovative ablative techniques for metastases from sarcomas is a challenge. The results of recent chemotherapy trials underpowered by the unplanned use of metastatic ablation techniques (mainly surgery or radiotherapy) in >20% of the patients suggest that omission of metastatic ablation is considered unethical at least as consolidation following incomplete response to chemotherapy [7]. Metastatic ablation shows a potential to eradicate macroscopic tumor (>80% local control rates with surgery and SBRT) and prolong survival in selected cases, i.e. to realize a switch from purely palliative intent to more curative intent. However, demonstration of a benefit from metastatic ablation is a road with hurdles due to our inability to resolve the methodological issues (intrinsic biases of retrospective studies) and overcome the reluctance/conviction involved in the institution-dependent patient selection processes for aggressive local metastatic treatments. Two scenari can be identified: upfront chemotherapy (+/adjuvant) in patients who have relatively short DFI (
