Abstract Gastrointestinal stromal sarcomas (GISTs) are the most common mesenchymal tumours originating in the digestive tract. These tumours have become a model of multidisciplinary work in oncology: the participation of several specialities (oncologists, pathologists, surgeons, molecular biologists, radiologists and others) has allowed advances in the understanding of this tumour and the consolidation of a targeted therapy, imatinib, as the first molecular treatment that is efficacious in solid tumours. Following the introduction of this drug, median survival of patients with advanced stage GIST has gone from 18 to more than 60 months. Therapy planning of GIST must be considered within a multidisciplinary context, and it is ad-visable that it takes place in reference centres for the care of sarcomas and GIST participating in clinical trials.

Keywords GIST · Imatinib · Sunitinib · GEIS · CD117 · c-kit · DOG1

Introduction

Gastrointestinal stromal sarcomas (GISTs) are the most common mesenchymal tumours originating in the digestive tract. They have a characteristic morphology, are gener-ally positive for CD117 (c-kit) and are primarily caused by activating mutations in the KIT or PDGFRA [1]. On rare occasions, they occur in extravisceral locations such as the omentum, mesentery, pelvis and retroperitoneum.

GISTs have become a model of multidisciplinary work in oncology: the participation of several specialities (on-

cologists, pathologists, surgeons, molecular biologists, radiologists,…) has allowed advances in the understanding of this tumour and the consolidation of a targeted therapy, imatinib, as the first molecular treatment that is efficacious in solid tumours. Following the introduction of this drug, median survival of patients with advanced stage GIST has gone from 18 to more than 60 months. Additionally, sunitinib is another targeted drug registered as second-line treatment for metastatic GIST.

Risk stratification

There are several classifications to calculate the risk of relapse. The most frequently used are: the Armed Forces Institute of Pathology or Miettinen criteria (AFIP) [2] (Ta-ble 1), the modified NIH [3] (Joensuu criteria) and Gold’s nomogram [4]. Other, less frequently used, systems are the National Institute of Health (NIH) [5] (or Fletcher) criteria or the AJCC [6].

Oncologists should analyse risk based on mitotic rate, tumour size and location (essentially intestinal or gastric). Additionally, capsule rupture of the tumour is a known fac-tor for very poor prognosis of GIST relapse [7].

A genotype study is considered mandatory to be able to recommend adjuvant therapy or for advanced disease. The genotype has proven its role as a predictive factor: there are mutations resistant to treatment with imatinib, such as D842V. Furthermore, the genotype has a prognostic value for localised disease [8], and it is expected that in the fu-ture it will be included in risk classification systems.

Diagnosis

Radiology

Radiological diagnosis of GIST is similar to that of other digestive tract tumours. In barium studies, GISTs ap-pear as submucosal lesions [9] and in ultrasound studies

A. Poveda () · F. Rivera · J. MartínServicio de Oncología MédicaFundación Instituto Valenciano de OncologíaC/ Profesor Beltrán Báguena, 19ES-46009 Valencia, Spaine-mail: apoveda@fivo.org

Clin Transl Oncol (2012) 14:536-540DOI 10.1007/s12094-012-0837-8

C l I N I C A l G u I D E S I N O N C O l O G y

SEOM guidelines for gastrointestinal stromal sarcomas (GIST)

Andrés Poveda · Fernando Rivera · Javier MartínSpanish Group for Sarcoma Research (GEIS)

Received: 4 May 2012 / Accepted: 2 June 2012

Clin Transl Oncol (2012) 14:536-540 537

as hypoechogenic masses that, when large, can displace neighbouring structures and show cystic, necrotic or haem-orrhagic areas.

A computerized tomography (CT) scan and magnetic resonance imaging (MRI) are the first choice to study lo-cation and extension [10]. A CT with contrast and image acquisitions of the arterial and portal phases allows identi-fication of hypervascular hepatic lesions that would other-wise go unnoticed and become evident when they become hypodense with treatment. The latter could wrongly sug-gest progression due to the development of new lesions. On the contrary, a CT scan without endovenous contrast allows detection of a haemorrhage or intratumoral calcification.

In a CT, tumours appear as well circumscribed exolu-minal masses that, after the contrast, show heterogeneous enhancement, especially large tumours, which may have ne-crotic-haemorrhagic areas or degenerative components [11].

An MRI is useful for the local study of tumours located in the pelvic area [12] as well as for the study of the mes-enteric and peritoneal extension. With PET, small GISTs have a homogeneously increased uptake, while in large le-sions (>4 cm) uptake may be heterogeneous [13].

The most frequent locations are stomach (60%); small intestine: jejunum-ileum (30%), duodenum (5%); rectum (2–3%); and colon (1–2%). They are far less frequent in the oesophagus (<1%). Some cases present disseminated tumours without a known primary tumour and a few cases

originate in the omentum, mesentery and retroperitoneum [9, 14]. Metastases are usually intra-abdominal, in the peri-toneum and the liver; lymph node or distance metastases are very rare and in the lung they are exceptional.

Microscopy

The microscopic aspect is heterogeneous with three dis-tinct main types: spindle cell (77%), epithelioid cell (8%) and mixed (15%). Given its prognostic importance, the procedure should be standardised counting mitosis in the most active areas, in a total of 50 HPF (corresponding to 10 mm2).

Immunohistochemistry

Immunohistochemistry shows that over 95% of GISTs ex-press CD117 (c-kit). Between 70 and 90% express CD34, 20 and 30% actin, 8 and 10% protein S-100 and 2 and 4% desmin. An IHC study generally helps confirm the GIST diagnosis. DOG1 may be optionally included in the initial panel and is highly recommendable in negative c-kit cases [15], as it is expressed in over 35% of these.

Differential diagnosis

The main differential diagnosis of spindle cell GISTs is with smooth muscle tumours (leiomyomas and leiomyo-sarcomas); schwannomas and malignant peripheral nerve sheath tumours; inflammatory myofibroblastic tumours; solitary fibrous tumours; sarcomatoid carcinoma; inflam-matory fibroid polyp; and desmoid fibromatosis. Fortunate-ly, morphological features and an adequate immunochem-istry panel facilitate diagnosis [10] (see Table 2).

Treatment

Therapy planning of GIST must be considered within a multidisciplinary context, and it is advisable that it takes

Table 1 Risk groups of GISTs according to Miettinen et al. [2]

Size Mitotic rate (50 HPF) location

Very low risk 2–5 cm ≤5 mitosis Gastriclow risk >5 and ≤10 cm ≤5 mitosis Gastric 2–5 cm ≤5 mitosis IntestinalIntermediate risk >10 cm ≤5 mitosis Gastric >5 and ≤10 cm ≤5 mitosis Intestinal 2–5 cm >5 mitosis GastricHigh risk 2–5 cm >5 mitosis Intestinal >10 cm ≤5 mitosis Intestinal >5 and ≤10 cm >5 mitosis Gastric >10 cm >5 mitosis Gastric >5 and ≤10cm >5 mitosis Intestinal >10 cm >5 mitosis Intestinal

Table 2 usefulness of immunohistochemistry in the differential diagnosis of GIST

CD117 CD34 S100 Actin Desmin AlK-1 Beta catenin Cytokeratin Synaptophysin/ chromogranin

GIST Positive Positive Smooth muscle T Positive Positive Peripheral nerve T Positive Inflammatory myofibroblastic T Positive Inflammatory fibroid polyp Positive Desmoid fibromatosis Positive Carcinoma Positive Endocrine T Positive

538 Clin Transl Oncol (2012) 14:536-540

place in reference centres for the care of sarcomas and GIST participating in clinical trials.

The therapeutic standard for localised primary GIST is surgical resection whenever this is technically possible, as there is no evidence of metastatic disease and morbimortal-ity is acceptable. The remaining patients, especially those with large GIST and those in specific locations such as the rectum, oesophagus and pancreas, where resection would entail a major functional disturbance, should be considered in a multidisciplinary context as potential candidates for neoadjuvant treatment. Optimal surgery should achieve type R0 disease-free resection margins, therefore at times it will be necessary to sacrifice organs next to the tumour and conduct “in block” surgical exeresis. Regional lymph-adenectomy is not indicated in GIST.

Exeresis of a GIST with laparoscopic surgery or as-sisted laparoscopy is currently well accepted, provided the same surgical radicalism (R0) criteria are met as for open surgery and the same care in the event of a possible tumour rupture. This surgery is specifically recommendable in cen-tres with advanced laparoscopic surgery programmes.

Patients with GISTs located in the rectum comprise less than 1% of total GISTs. In all cases, an extension study should be carried out with an abdominal CT. A pelvic MRI and an echo-endoscopy are mandatory to evaluate local extension and infiltration of the sphincter. Should there be anal sphincter infiltration, there is no evidence at all of any benefit of neoadjuvant treatment. If, after the extension study and evaluation by a surgeon with expertise in rectum surgery, it is considered that the patient is resectable, but not a candidate for sphincter conservative surgery, and the anal sphincter is unharmed, neoadjuvant treatment with imatinib may be assessed.

up to 50% of patients with complete tumour resec-tion may relapse according to some series. This has led to several studies to evaluate the role of imatinib as adjuvant treatment for the prevention of relapses. Evidence derived from two randomised studies, ACOSOG [16] and SSGX-VIII/AIO [17], shows the benefit of adjuvant imatinib for relapse-free survival. Furthermore, the SSGXVIII/AIO study has demonstrated survival advantages with 3-year administration of imatinib vs. 1 year in patients with a high risk of relapse (according to the NIH).

The consensus, supported by this study, is to administer adjuvant imatinib for 3 years to high-risk patients. Imatinib should not be recommended to low-risk patients. In patients with intermediate risk, there is insufficient data to recom-

mend adjuvant imatinib. The grey area of controversy lies in those patients with a non-gastric location and between 5 and 10 mitoses, as these were not included in a relapse high-risk group but they have a similar risk (Table 3).

Special situations

– Capsule rupture: It is generally accepted that these patients should be managed in the same way as those with disseminated disease since in the series pub-lished practically 100% progress at least to the peri-toneal level. In these patients it would be acceptable to administer imatinib for more than 3 years, just as patients with advanced disease would be treated.

– Specific genotypes: Administration of adjuvant ima-tinib is not recommended in patients with mutation D842V of the PDGRFα gene as it is known to be resistant to it. In carriers of exon 9 mutations of the KIT gene there is no agreement on the administration of doses higher than 400 mg of adjuvant imatinib a day. Patients who are carriers of the wild-type GIST should be offered administration of imatinib accord-ing to the above risk criteria.

– Patients with R1: It has not been proven whether low-risk patients with affected microscopic margins can benefit from adjuvant imatinib. Surgery should be reconsidered in those cases which have been R1 due to emergency surgery or if morbidity is acceptable.

Monitoring

Clinical and laboratory testing follow-up is recommended every 1–3 months during the adjuvant therapy period. Management of side effects, even if mild, may be relevant to ensure treatment compliance. We recommend radiologi-cal monitoring with CT or MRI every 3–6 months during the treatment phase. Once treatment has stopped, follow-up every 3 months during the first 2 years is advisable (since the risk of relapse is considerably higher) and then every 6 months until the 5th year, after which it may be annual.

Induction therapy

In patients in whom exeresis of the primary tumour is not clearly indicated initially, due to the size of the tumour, its

Table 3 Recommendations for adjuvant therapy with imatinib according to risk

Risk group Recommendation

High risk Adjuvant therapy with imatinib 400 mg/d for 3 yearsIntermediate risk Insufficient evidence to put forward a recommendation. Consider that non-gastric patients with mitosis between 6 and 10 are at high risk with any sizelow–very low risk Adjuvant therapy not recommendedCapsule rupture Treatment with imatinib recommended as advanced disease

Clin Transl Oncol (2012) 14:536-540 539

location or a high risk of major functional loss, neoadju-vant therapy may be evaluated individually and within a multidisciplinary context. In these cases, a tumour biopsy with a TRu-CuT needle is required to confirm diagnosis and a mutational study prior to neoadjuvant therapy. Also, it can be very useful to conduct a baseline PET and again after a month of treatment to assess early response to drug treatment. Once neoadjuvant therapy with imatinib has been started, we recommend maintaining it until maxi-mum response is achieved (between 6 and 12 months) before proceeding to final radical surgery, to avoid the risk of progression due to the development of new tumour mutations.

Treatment of locally advanced unresectable or metastatic disease

GIST is a tumour resistant to chemotherapy (response rate 0–5%, median survival 14–18 months). Clinical studies conducted since 2001 confirm the significant activity of imatinib mesylate in metastatic and inoperable locally ad-vanced GIST.

The usual dose of 400 mg/day of imatinib was es-tablished based on two randomised Phase III trials with patients with c-kit-positive metastatic GIST (EORTC-ISG-AGITG [18] and NASG-S0033 [19]), comparing 400 vs. 800 mg/day, without finding any differences in overall survival (OS) in either study, and with a favourable toxicity profile for the daily 400 mg dose. The clinical benefit rates (CR+PR+SD) were 90% and 88% in the NASG-S0033 study and 91% and 87% in the EORTC one with the 800 and 400 mg doses respectively.

A meta-analysis including data from both Phase III studies showed that the genotype in GIST represents an independent prognostic factor for both time to progression (TTP) and OS. In fact, patients with a mutation affecting exon 11 had better TTP and OS compared to GIST mu-tated in exon 9 or wild-type GIST. This meta-analysis also confirmed that GISTs mutated in exon 9 were a dose-de-pendent predictive factor that achieved significantly higher progression-free survival if treated with imatinib at a dose of 400 mg twice a day [20].

Given all of the above, the following recommendations have been established:

1. Start treatment with 400 mg/day. 2. For patients with documented exon 9 mutation, a

starting dose of 800 mg may be recommended.3. Duration of treatment with imatinib should be indefi-

nite according to the studies conducted by the French

Sarcoma Group (BFR14) [21, 22]. In patients with advanced disease treated with surgical resection for metastatic lesions, treatment should be maintained in order to prevent rapid disease progression.

4. Regular monitoring of the response is essential given the persistent risk of secondary resistance.

5. In the event of first-line progression with 400 mg, there is the possibility of escalating the dose to 800 mg/day with close monitoring of adverse events.

Toxicity of imatinib with a dose of 400 mg/day is mild or moderate. The most common adverse events are oedema (70%) (especially periorbital), nausea (50%), diarrhoea (45%), myalgia (40%), fatigue (35%), dermatitis or ery-thema (30%), headache (25%) and abdominal pain (25%). Only 20% presented severe adverse events (grades 3 and 4).

Some drugs may alter the metabolism of imatinib, due to inhibition or induction of cytochrome CyP3A4, altering the drug’s plasma concentrations.

Patients with advanced GIST progressing under treat-ment with imatinib 800 mg should be offered treatment with sunitinib 50 mg/day for 4 weeks every 6 weeks (level of evidence IIB) or alternatively 37.5 mg/day with no inter-ruption.

Sunitinib provides multi-targeted inhibition in com-parison with the selectivity provided by imatinib. A 2:1 randomised study compares sunitinib with placebo regard-ing TTP and intolerance of imatinib. This study showed the advantages of sunitinib in TTP: 28.9 vs. 7 weeks. The response rate was 7% and stabilisation 58%. Sunitinib showed higher probability of diarrhoea but fewer oedemas than imatinib. Genotypes with mutated exon 9 or wild type had limited possibilities of response to imatinib but they showed the best clinical efficacy with sunitinib [23].

Following further progression, participation in clinical trials should be considered [24, 25]. Other options include surgery and other drugs that have shown certain activity, such as sorafenib or dasatinib (so far with a clinical evi-dence of response to D842V). Regorafenib seems to offer very good PFS although its indication has not been ap-proved yet.

Conflict of interest A. Poveda and J. Martín have participate as advi-sors in Novartis and Pfizer Adv Boards.

Clinical Guideline Working Group on behalf of the Spanish Society of Medical Oncology (SEOM) Executive Committee 2011-2013: Juan Jesús Cruz, Pilar Garrido, Agustí Barnadas, Pablo Borrega, Francisco Javier Barón, Elvira del Barco, Rocío García-Carbonero, Jesús García-Mata, Encarnación González, Pilar lianes, Antonio llombart and Fernando Rivera.

References

1. Miettinen M, lasota J (2001) Gastrointestinal stromal tumors definition, clinical, histological, im-munohistochemical, and molecular genetic features and differential diagnosis. Virchows Arch 438:1

2. Miettinen M, lasota J (2006) Gastrointestinal

stromal tumors: pathology and prognosis at differ-ent sites. Semin Diagn Pathol 23:70–83

3. Rutkowski P, Bylina E, Wozniak A et al (2011) Validation of the Joensuu risk criteria for primary resectable gastrointestinal stromal tumour: the im-pact of tumour rupture on patient outcomes. Eur J Surg Oncol 37:890–896

4. Gold JS, Gonen M, Gutierrez A et al (2009) De-

velopment and validation of a prognostic nomo-gram for recurrence-free survival after complete surgical resection of localised primary gastroin-testinal stromal tumour: a retrospective analysis. lancet Oncol 10:1045–1052

5. Fletcher CD, Berman JJ, Corless C et al (2002) Diagnosis of gastrointestinal stromal tumors: a consensus approach. Hum Pathol 33:459–465

540 Clin Transl Oncol (2012) 14:536-540

6. Edge SB, Byrd DR, Carducci MA, Compton CC (eds) (2009) AJCC Cancer Staging Manual, 7th edn. Springer, New york, Ny

7. Hohenberger P, Ronellenfitsch u, Oladeji O et al (2010) Pattern of recurrence in patients with ruptured primary gastrointestinal stromal tumour. Br J Surg 97:1854–1859

8. Martin J, Poveda A, llombart-Bosch A et al (2005) Deletions affecting codons 557–558 of the c-KIT gene indicate a poor prognosis in pa-tients with completely resected gastrointestinal stromal tumors: a study by the Spanish Group for Sarcoma Research (GEIS). J Clin Oncol 23: 6190–6198

9. lau S, Tam KF, Kam CK et al (2004) Imaging of gastrointestinal stromal tumor (GIST). Clin Ra-diol 59:487–498

10. Demetri G, Benjamin R, Blanke CD et al (2007) NCCN Task Force report: optimal management of patients with gastrointestinal stromal tumor (GIST). update of the NCCN Clinical Practice Guidelines. J Natl Compr Canc Netw 5[Suppl 2]: S1–29

11. lee CM, Chen HC, leung TK, Chen yy (2004) Gastrointestinal stromal tumor: computed tomo-graphic featrures. World J Gastroenterol 10:2417–2418

12. Casali PG, Jost l, Reichardt P et al; ESMO Guidelines Working Group (2008) Gastrointesti-nal stromal tumors: ESMO clinical recommenda-tions for diagnosis, treatment and follow-up. Ann Oncol 19[Suppl 2]:ii35–38

13. Bensimhon D, Soyer P, Boudiaf M et al (2009) Imaging of gastrointestinal stromal tumors. J Ra-diol 90:469–480

14. Miettinen M, Majidi M, lasota J (2002) Pathol-ogy and diagnostic criteria of gastrointestinal stromal tumors (GISTs): a review. Eur J Cancer 38[Suppl 5]:S39–51

15. liegl B, Hornick Jl, Corless Cl, Fletcher CD (2009) Monoclonal antibody DOG1.1 shows higher sensitivity than KIT in the diagnosis of gastrointestinal stromal tumors, including unusual subtypes. Am J Surg Pathol 33:437–446

16. Dematteo RP, Ballman KV, Antonescu CR et al (2009) Adjuvant imatinib mesylate after resec-tion of localised, primary gastrointestinal stromal tumour: a randomised, double-blind, placebo-controlled trial. lancet 373:1097–1104

17. Joensuu H, Eriksson M, Hatrmann J et al (2011) Twelve versus 36 months of adjuvant imatinib (IM) as treatment of operable GIST with a high risk of recurrence: final results of a ran-domized trial (SSGXVIII/AIO). J Clin Oncol 29[Suppl]:lBA1

18. Zalcberg JR, Verweij J, Casali PG et al (2005) Outcome of patients with advanced gastro-in-testinal stromal tumours crossing over to a daily Imatinib dose of 800 mg after progression on 400 mg. Eur J Cancer 41:1751–1757

19. Blanke CD, Rankin C, Demetri GD et al (2008) Phase III randomized, intergroup trial assessing Imatinib mesylate at two dose levels in patients with unresectable or metastatic gas-

trointestinal stromal tumors expressing the kit receptor tyrosine kinase: S0033. J Clin Oncol 26:626–632

20. Gastrointestinal Stromal Tumor Meta-Analysis Group (2010) Comparison of two doses of Ima-tinib for the treatment of unresectable or metastat-ic gastrointestinal stromal tumors: a meta-analysis of 1,640 patients. J Clin Oncol 28:1247–1253

21. le Cesne A, Ray-Coquard I, Nguyen Bui B et al (2010) Discontinuation of imatinib in patients with advanced gastrointestinal stromal tumours after 3 years of treatment: an open-label multi-centre randomised phase 3 trial. lancet Oncol 11:942–949

22. Blay Jy, Bonvalo S, Casali O et al (2005) Consen-sus meeting for the management of gastrointesti-nal stromal tumors. Report of the GIST Consen-sus Conference of 20–21 March 2004, under the auspices of ESMO. Ann Oncol 16:566–578

23. Demetri D, Van Oosterom AT, Garret CR et al (2006) Efficacy and safety of sunitinib in patients with advanced gastrointestinal stromal tumours after failure of Imatinib: a randomised controlled trial. lancet 368:1329–1338

24. Poveda A, Artigas V, Cervera J et al (2011) Guía de práctica clínica en los tumores estromales gastrointestinales (GIST): actualización 2010. Cir Esp 86[Suppl 1]:1–21

25. Poveda A, Artigas V, Cervera J et al (2011) Guía de práctica clínica en los tumores estromales gas-trointestinales (GIST). Clin Transl Oncol [Suppl 1]:1–20