Gastrointestinal stromal tumours
Jean-Yves Blay

(GIST) are sarcomas that mostly derive from precursors of the interstitial cells of Cajal (ICC), the pacemaker cells of the gastrointestinal tract responsible for its motility. GIST are the most fre- quent of all sarcomas1–6. GIST are a heterogeneous group of tumours that includes a variety of molecular entities with usually mutually exclusive activating oncogene mutations, mostly KIT or PDGFRA mutations6–11. All subtypes have different natural histories and treatments. The clinical presentation of GIST is that of a gastric or bowel tumour. Bleeding, pain and/or obstruction are fre- quent symptoms at presentation. Most GIST arise in the stomach (60–65%) and the second most common site is the small intestine (20–25%); rectum, colon, oesophagus and other sites are rare1,2,8–11 (FiG. 1). Small GIST <2 cm in size, frequently termed miniGIST (1–2 cm) or micro- GIST (<1 cm), are often incidentally discovered during gastroduodenal endoscopy. These GIST are equipped with mutations similar to those of larger GIST but are generally not resected except at primary rectal sites1,2. Localized GIST >2 cm are curable cancers and surgi- cal removal is typically the first treatment. Metastatic
relapse occurs most often in the liver and peritoneum.
GIST are resistant to standard cytotoxic treatments used for other sarcomas, but major clinical improve- ments in survival have been possible since the intro- duction of tyrosine kinase inhibitors (TKIs) that target KIT and/or PDGFRA. In the advanced disease setting, TKI treatment has dramatically improved median sur- vival from 18 months to >5 years in the past 20 years1,2.
The TKI imatinib is the mainstay of medical treatment of GIST in the first-line metastatic setting and, in the adjuvant setting, a duration of 3 years of adjuvant treat- ment with imatinib has been shown to improve overall survival (OS)1,2,12. Patients with metastatic GIST can sur- vive for longer than 20 years since the advent of TKIs1,2,13. Similar to other rare cancers and sarcomas, cure rates are optimized when patients with GIST are treated in reference centres and surgeries are performed accord- ing to oncological guidelines14–16. Treatment resistance originates mainly from the selection of multiple resist- ant tumour cell clones driven by additional mutations in KIT or PDGFRA17,18, resulting in variable sensitivity to the first generation of later-line treatments sunitinib and regorafenib. These resistance mutations can be effectively targeted with a newer generation of TKIs, such as ripretinib19,20. Rare molecular forms of GIST, for example, with NF1 inactivation, SDH loss or translo- cation of a NTRK family gene, generally have primary resistance to KIT and PDGFRA TKIs, but may respond to specific TKIs. GIST are models of the success of precision medicine in cancer.
In this Primer, we present an overview of the epi- demiology of GIST, the molecular mechanisms of tum- origenesis and the molecular subtypes of this group of tumours. We discuss state-of-the-art GIST diagnosis and treatment, incorporating the latest advances in our understanding of this disease, comment on means to improve patient quality of life (QoL) and highlight outstanding research questions.

PRIMER

Fig. 1 | Mutation prevalence in gastric, small intestine and rectal GIST. Most gastrointestinal stromal tumours (GIST) arise in the stomach and the small intestine.
GIST in the rectum, colon, oesophagus and other sites are rare. The frequency of driving mutations differs between anatomical sites. KIT exon 11 mutations are the most frequent alterations in the stomach, small intestine and rectum.

Epidemiology
The incidence of GIST ranges from 6 to 22 cases per 106 individuals per year, with differences between regions and over time3–6,21,22. Variations in incidence are likely related to differences in the completeness of registries or series between regions as well as to the inclusion of microGIST. Registries with exhaustive collections generally report an incidence of >12 cases per 106 indi- viduals per year, representing the most probable estimate. Estimates are further complicated by the large pop- ulation of individuals with microGIST, which are not included in most registries. MicroGIST are observed in up to 20% of individuals in autopsy series23,24, their natu- ral history is still unclear and most do not develop to an overtly progressive life-threatening cancer23,24.
Studies in the USA found increased incidence in Asian/Pacific Islanders or Black people compared with white people (relative risk 1.50 and 2.07, respectively)21. The sex ratio of patients with GIST ranges from 0.94 to
1.20 in different series and does not substantially fluctu- ate with patient age3,5,6,21. GIST are predominantly diag- nosed at a median age of 65 years, but can be diagnosed at any age25–27. Incidence of the different molecular subtypes of GIST varies between age groups. In children, GIST without KIT or PDGFRA mutations (also collectively termed wild-type GIST) are common, whereas GIST with KIT mutations are most common in individuals
>18 years of age25,26. Most GIST arise in the stomach (60–65%), followed by the small intestine (20–25%), rec- tum (3–5%), colon (1–2%), oesophagus (1%) and other sites (8–10%)1,2,8–11. GIST occurring in young patients, children and young adults (<30 years of age) arise mostly at gastric sites26.
Few studies have reported on the molecular epide- miology of GIST. The available data suggest that GIST

with KIT mutations have an incidence close to 8 cases per 106 individuals per year in most regions, GIST with PDGFRA mutations have an incidence of <3 cases per 106 individuals per year, and other forms of GIST have an incidence of <1 case per 106 individuals per year5,22 (TABLE 1). The exact incidence of GIST occurring in patients with germline predisposition is not known.
No environmental risk factors for GIST are currently known. Familial predisposition owing to germline muta- tions of KIT or PDGFRA are the most common but very rare known risk factors.

Mechanisms/pathophysiology
GIST mostly derive from lineage cells of ICC, but PDGFRA-mutated GIST can derive from telocytes and BRAF-mutated GIST can derive from smooth muscle cells28–30. Different types of primary GIST have almost always mutually exclusive molecular driver mutations18,28. The type of molecular tumour driver cor- relates with the primary anatomical site. GIST bearing KIT exon 9 mutations are located most often in the small intestine, colon or rectum8–11,31. Mutations of PDGFRA, the most common being D842V, are observed in primary gastric GIST. GIST with mutation or loss of expression of SDH family genes are often multifocal and located in the stomach in young adults, with a predomi- nance in women25,26. GIST in patients with germline NF1 mutations are rare (~7% of all patients with NF1 muta- tions), located in the small intestine, frequently multi- ple and typically lacking PDGFRA and KIT mutations32. In addition to being correlated with clinical and patho- logical features and outcomes, the different molecular alterations are biomarkers of sensitivity and resistance to TKIs29,30,33,34.

Activated molecular pathways in GIST
The most frequent driver mutations occur in KIT
(60–70% of cases) and PDGFRA (10–15%)7–11,17,18,31–35
(TABLE 1; FiG. 2). Around 15% of GIST have no muta- tions in either KIT or PDGFRA but have other genetic alterations, for example, in SDH family genes or loss of expression of SDH genes through epigenetic silencing, RAS family genes, BRAF, NF1, gene fusions involv- ing NTRK3 or FGFR1, or other very rare driver gene mutations25,26,32,33. These mutations or gene fusions are considered to be mutually exclusive in primary GIST, but coexistence of some of these rare mutations with KIT or PDGFRA mutations has been reported36,37.
Both KIT and PDGFRA encode type III recep- tor tyrosine kinases, and PDGFRA is a homologue of KIT with similar structure and downstream signalling pathways28,38,39. GIST express high levels of ETV1, a tran- scription factor required for ICC development and for proliferation of GIST cells40. ETV1 is regulated by the activation of the MEK–MAPK pathway, which is down- stream of activated KIT and PDGFRA (FiG. 3). Activated KIT together with ETV1 expression promotes the onco- genic transcriptional programme required to transform ICC into GIST cells, with deregulated proliferation and inhibition of apoptosis40. In addition, preclinical data suggest that activation of the PI3K pathway downstream of KIT may be required for the transformation of ICC

PRIMER

into GIST cells, as it is not activated in ICC41. KIT or PDGFRA mutations are also found in microGIST and miniGIST, as well as in familial GIST resulting from germline mutations in these genes. These patients have diffuse hyperplasia of ICC and multiple benign small GIST31,42–44. Taken together, KIT or PDGFRA muta- tions are early events and considered causative but not sufficient for the neoplastic transformation of ICC into GIST.
Additional genomic steps contributing to the transfor- mation of ICC into malignant GIST include loss or inac- tivation of MAX, which encodes a partner of the MYC transcription factor on chromosome 14 (REF.45), DEPDC5, encoding an mTORC1 inhibitor, on chromosome 22 (REF.46), DMD, encoding dystrophin, on chromosome X (REF.47) and of several cell cycle gene regulators48.

Molecular subgroups of GIST
GIST present with various molecular driver alterations and different natural histories (TABLE 2; FiG. 2). Of note, the specific molecular alterations of the various onco- genes involved in neoplastic transformation have dis- tinct effects on the efficacy of and response to different treatments, especially imatinib.

GIST with KIT mutations. KIT encodes the KIT tyros- ine kinase consisting of an extracellular domain with immunoglobulin-like motifs, a juxta-membrane (JM) domain (mostly encoded in exon 11), a cytoplasmic kinase domain consisting of an ATP-binding domain (encoded in exons 13 and 14) and an activation loop domain (encoded in exons 17 and 18)49. Normally, dimeric binding of stem cell factor (SCF) to the extra- cellular domain induces receptor dimerization, phos- phorylation and activation of downstream kinases in the MEK–MAPK and PI3K pathways. The JM domain nor- mally functions to stabilize KIT in an inactive conforma- tion and inhibits dimerization. Mutations causing loss of function of the JM domain may induce dimerization and autophosphorylation50,51.
In GIST, KIT mutations, including deletions, deletion– insertions (indels), insertions and missense mutations, occur mostly in exon 11, which encodes the JM domain. The next most frequent mutation in KIT is the duplica- tion insertion of A502–Y503 codons in exon 9 (REFs7–11). KIT protein with this AY duplication has a kinase con- formation similar to that of wild-type KIT for SCF binding, whereas KIT with mutations in the JM domain adopts a different conformation51,52. Imatinib preferen- tially binds to the kinase in the inactive conformation

induced by the JM domain mutations. Thus, KIT exon

Table 1 | Molecular subtypes of localized GIST

Molecular subtype Proportion of all GIST (%)

Netherlands (n = 166)a France (n = 106)b
KIT mutations
Overall 67.5 66.90
KIT exon 11 mutation 58.4 52.80
KIT codons 557–558 deletion NR 31.20
Other KIT exon 11 mutations NR 20.70
KIT exon 9 mutation (AY duplication) 6.6 9
KIT with exon 13 point mutations 1.2 3.60
KIT with exon 17 point mutations 0.6 1
PDGFRA mutations
Overall 16.2 16
PDGFRA exon 18 13.5 14.00
PDGFRAD842V mutation NR 9
PDGFRA exon 18 non-D842V mutations NR 5
PDGFRA exon 12 0.6 2
PDGFRA exon 14 1.8 NR
Wild-type KIT and PDGFRA
Overall 16.3 16.90
SDH-deficient GIST NR NR
Mutation of SDHA, SDHB, SHDC
or SDHD 1.8 NR
Epigenetic silencing of SDH genes NR NR
NF1 mutation 2.4 NR
BRAF mutation 0.6 NR
ETV6–NTRK3 NR NR
GIST, gastrointestinal stromal tumours; NR, not recorded. a2-year period (2011–2012) in a population of 16.7 million inhabitants (166 of 489 (33.4%) GIST had mutational testing; total incidence of GIST 14.6/106/year)5. b2-year period (2005–2006) in a population of 6.06 million inhabitants (106 of 131 (74%) GIST had mutational testing; total incidence of GIST 10.8/106/year)22.
11 mutations are more sensitive to imatinib than exon 9 mutations1,8–10,31.
GIST with KIT exon 11 mutations can be observed at any anatomical site in the gastrointestinal tract (FiG. 1) and in tumours smaller or larger than 2 cm (REFs8–11,34,53,54). The different mutations in exon 11 are associated with different clinical presentation and response to imatinib. Patients with localized GIST and KIT exon 11 dele- tion mutations have shorter recurrence-free survival (RFS) after surgery than those with exon 11 missense mutations8,34,53,54. This is the case in particular for GIST with deletions of codons 557 and 558, which are part of the code of the JM domain that contacts the activation loop. GIST with these deletions show aggressive clinico- pathological features, and patients with these tumours have an increased risk of relapse after surgery for local- ized disease54. Notably, GIST with exon 11 mutations on codons 557 and 558 show robust responses to first-line imatinib treatment9–11,53,54. GIST with duplication muta- tions in the distal portion of KIT exon 11 mainly orig- inate in the stomach and show comparatively indolent clinical behaviour8.
GIST with KIT exon 9 mutations are mainly located in the small or large intestine, and 20–25% of GIST in the small intestine have mutations in KIT exon 9 (REFs5,22) (FiG. 1). The response rate to imatinib of advanced GIST with KIT exon 9 mutations is lower than that of GIST with exon 11 mutations, but response and progression-free survival (PFS) can be improved with an increased dose of 600 mg or 800 mg imatinib per day in metastatic GIST9–11. Of note, the dose of imatinib 800 mg per day has never been tested for GIST in randomized trials in the adjuvant phase and can, therefore, not be recommended in this setting35,55,56. For sunitinib, the PFS of patients with KIT exon 9 mutations is better than that of those with KIT exon 11 mutations17,57,58.

PRIMER

⦁ KIT-mutated GIST (67%)

⦁ PDGFRA-mutated GIST (16%)

⦁ GIST without KIT or PDGFRA mutations (16–17%)

SDHC
SDHD
(?%)

(?%)

Fig. 2 | Driver mutations of molecular GIST subtypes. Gastrointestinal stromal tumours (GIST) are broadly grouped into three groups according to their driver mutations. GIST with mutations in KIT are most frequent, followed by GIST with mutations in PDGFRA and GIST without mutations in either KIT or PDGFRA. Mutations of KIT or PDGFRA and corresponding molecular GIST subtypes are distinguished by the type and location of mutations in these genes, which occur at different frequencies. GIST without mutations in KIT or PDGFRA can be differentiated into those with or without alterations to SDH genes. Driving mutations in SDH-competent GIST include alterations of genes in the
RAS–MEK–MAPK pathway, such as NF1 or BRAF, translocations involving NTRK or FGFR genes and other rare mutations. Percentages correspond to the frequency of these mutations in large series, and bold highlighted mutations confer resistance to imatinib treatment.

Primary missense mutations in KIT exon 13 and exon 17 are rare. K642E substitution is the predominant primary mutation in exon 13, and some patients with GIST with that alteration are considered to be sensitive to imatinib59. Primary and secondary KIT exon 17 muta- tions are mostly missense mutations in codon 820, 822 or 823 (REF.28). KIT with mutations of amino acids encoded by exon 17 (for example, D816V) have an activated con- formation. When exon 17 mutation is a primary muta- tion, activated KIT is generally imatinib resistant but KIT with a N822H mutation may be imatinib sensitive. However, most KIT exon 17 mutations confer secondary resistance to imatinib in vitro28.
Several groups have reported that KIT mutations in GIST induce the retention of the mutated acti- vated kinase within the cell, specifically in the Golgi apparatus60,61. KIT blockade by TKI restores the migra- tion of the KIT protein to the cell membrane. The implications of the retention of mutated activated KIT within the cell are unclear but retention of other receptor tyrosine kinases has also been observed in other cancer models, including of IGF1R60–62.

GIST with PDGFRA mutations. Mutations in PDGFRA are mainly found in exon 18 and less frequently in exon 12 or 14 (REFs8,17,34,37–39). Most PDGFRA-mutated GIST
are located in the stomach (15–18%) followed by the small intestine (5–7%) (FiG. 1). These GIST follow a comparatively indolent clinical course. Most muta- tions in exon 12, exon 14, or indels in exon 18 result in
imatinib-sensitive PDGFRA mutants. By contrast, GIST bearing the PDGFRA exon 18 D842V missense mutation are well documented to be resistant to imatinib and other TKIs17,18,28,63. PDGFRA-D842V is observed in ~50% of GIST with PDGFRA alterations5,22 (TABLE 1). This muta- tion stabilizes the kinase conformation in an activated form and confers resistance to imatinib, sunitinib and regorafenib18,63. Conversely, the new TKIs avapritinib and ripretinib have been reported to provide objective responses and long-term tumour control in a substan- tial proportion of patients with GIST bearing the D842V mutation19,20,64–66. The molecular mechanisms of resist- ance to avapritinib in PDGFRAD842V-mutated GIST were recently reported and involve missense point mutations in PDGFRA exons 13, 14 and 15 (REFs64,65).

GIST without KIT or PDGFRA mutations. GIST without KIT or PDGFRA mutations can be divided into SDH- deficient and SDH-competent GIST18,67–71 (TABLE 2). SDH-deficient GIST include those with mutations in genes encoding SDH subunits and those with epige- netic suppression of SDH expression. SDH-competent GIST include those with mutations in genes of the RAS–MEK–MAPK pathway, those with translocations involving NTRK or FGFR genes and others with very rare mutations25,26.

GIST with mutations and alterations in the SDH com- plex. The precise oncogenic mechanisms of mutations in SDH genes in GIST have not been fully clarified.

PRIMER

SDHC
SDHD
Fig. 3 | Signalling pathways of molecular GIST subtypes. In gastrointestinal stromal tumours (GIST) with driver KIT
or PDGFRA mutations, the resulting activation of KIT and PDGFRA tyrosine kinases drives signalling via major pathways, including the RAS–MEK–MAPK and PI3K–mTOR pathways, which converge to the ETV1 signalling pathway. The activation of these pathways induces cellular proliferation and blocks apoptosis. In SDH-deficient GIST, the accumulation of succinate caused by dysfunction of the SDH complex blocks the degradation of hypoxia-inducible factor 1α (HIF1α) by prolyl hydroxylase (PHD), resulting in expression of growth factor genes, including vascular endothelial growth factors (VEGFs), platelet-derived growth factor (PDGFs), VEGF receptor (VEGFR) and insulin growth factor 1 receptor (IGF1R). Yellow
boxes indicate kinases that may have gain-of-function mutations and blue boxes indicate proteins with loss-of-function mutations in GIST.

The absence of one of the SDH-complex subunits destabilizes the complex, facilitating degradation of SDHB18,72–74. Dysfunction of the SDH complex (com- plex III in the mitochondrial electron transport chain) is thought to lead to accumulation of succinate, which inhibits prolyl-hydroxylase (PHD). Inhibition of PHD results in the accumulation of hypoxia-inducible fac- tor 1α (HIF1α) in a normoxic environment, which initiates the expression of several key genes including insulin growth factor 1 receptor (IGF1R) and vascular endothelial growth factor receptor (VEGFR), which contributes to the transformation of normal ICC into SDH-deficient GIST72–74. Succinate accumulation might also inhibit the ten-eleven translocation (TET) family of DNA hydroxylases, resulting in the genome-wide DNA hypermethylation observed in SDH-deficient GIST. SDH-deficient GIST mainly arise in the stomach of children and adolescent and young adults <30 years of age who characteristically have multinodular gas- tric masses. Their natural history is different from that of GIST with KIT or PDGFRA mutations. Although patients with SDH-deficient GIST frequently present with lymph node and hepatic metastases, the disease course is often, but not always, clinically indolent, pointing to the need for careful selection of therapy or watch-and-wait strategies in advanced disease25,26. GIST with SDH deficiency are considered to be resistant to
imatinib, but may be partly sensitive to VEGFR2 inhibi- tors, such as regorafenib and sunitinib25,67, or the IGF1R inhibitor linsitinib, which yielded stable disease in 40% of patients at 9 months75.

GIST with mutations in the RAS–MEK pathway and others. A proportion of SDH-competent GIST without KIT or PDGFRA mutations have alterations in NF1, BRAF or, more rarely, in RAS, which activate the MEK– MAPK pathway18,76–83. NF1 is a tumour suppressor gene and encodes neurofibromin, a negative regulator of RAS proteins. Germline loss of NF1 causes neurofibromatosis type 1, an autosomal dominant genetic disorder charac- terized in particular by multiple café-au-lait macules and associated cutaneous neurofibromas78. Bi-allelic inacti- vation of NF1 may induce tumour formation, and 7% of patients with germline NF1 loss develop GIST32,76–78. However, other tumours such as neurofibromas and neuroendocrine tumours may also arise in the gastro- intestinal tracts of these patients78. NF1-mutated GIST without KIT or PDGFRA mutations are resistant to cur- rently approved TKIs32. Somatic NF1 inactivation has also been reported in KIT-mutated GIST79,80.
GIST with BRAF mutations are also located in the small intestine and show spindle cell morphology81–83. Most are BRAFV600E mutations30,33,81–83. Patients with these tumours have variable prognostic outcomes75,76.

PRIMER

These GIST are resistant to imatinib but may be sen- sitive to BRAF inhibitors such as dabrafenib and MEK inhibitors33,81. GIST with other mutations in RAS genes or PIK3CA or with gene fusions involving NTRK3 or

FGFR1 are very rare but require molecular identifica- tion in case of relapse, as specific treatments targeting activated NTRKs and FGFRs, such as larorectinib, entrectinib or erdafitinib are now available25,26,84.

Table 2 | GIST driver mutations and tumour characteristics
Alterations Estimated Location Characteristics Imatinib Sunitinib Regorafenib Avapritinib Ripretinib frequency (%) sensitivity sensitivity sensitivity sensitivity sensitivity
KIT mutationa
Exon 9 5–10 Small intestine Spindle cell type; Less sensitive Sensitive Sensitive Sensitive Sensitive (or exon 8) aggressive features to imatinib
400 mg daily, higher response rates to 800mg dailyb
Exon 11 60–70
(deletions, missense, insertions or other) All sites Aggressive features Sensitive Sensitive Sensitive Sensitive Sensitive with del 557–558
Exon 13 <1 (K642E) Spindle cell features; Sensitive in Less Sensitive Sensitive Sensitive often low or some patients sensitive
intermediate grade; germline mutations described
Exon 17 1
(D820Y, N822K, Y823D) From all sites, germline Not sensitive Some Some Sensitive Sensitive mutations described (N822H Sensitive sensitive
sensitive?)
PDGFRA mutationa
Exon 12 <1 (e.g. V561D) Stomach Epithelioid cell type Probably Probably Probably Probably Probably Sensitive sensitive sensitive sensitive sensitive
Exon 14 <1 (N659K) Probably Probably Probably Probably Probably Sensitive Sensitive Sensitive Sensitive Sensitive
Exon 18 10–15
(e.g. D842V) D842V is D842V is D842V is Highly Sensitive resistant, resistant, resistant, sensitive
others may others may others may be sensitive be sensitive be sensitive
Wild-type KIT and PDGFRA, SDH-competent
NF1 1–2 Small intestine Spindle cell type; Not sensitive Probably not Unknown mutationa,c indolent clinical sensitive
course; associated with neurofibromatosis type I Unknown Unknown
BRAF <1 Small intestine/ Spindle cell type; Probably not Probably not Probably not mutation stomach VE1-positive sensitive sensitive sensitive
HRAS, NRAS Very rare Unkown Not sensitive Probably not Probably not Probably not or KRAS sensitive sensitive sensitive
mutation
Translocations Very rare Unknown Not sensitive Specific Specific Specific Specific Specific (NTRK or other) inhibitors inhibitors inhibitors inhibitors inhibitors
Wild-type KIT and PDGFRA, SDH-deficient
SDHA, ~3 Stomach>>small
SDHB, SDHC intestine or SDHD
mutation (including Carney– Stratakis syndromea) Epithelioid cell type; Not sensitive Probably not Unknown Unknown Unknown SDHB-negative; sensitive
children, adolescents and young adults; frequent lymph node metastasis; indolent clinical course
Loss of SDHB <1 Stomach expression
(including Carney triadc)
GIST, gastrointestinal stromal tumours. aSome GIST have germline mutations in a context of familial GIST (e.g. germline mutation of KIT). bPatients with GIST with KIT exon 9 mutations were reported to have a higher response rate and progression-free survival with an 800 mg daily dose compared with a 400 mg daily dose of imatinib9–11. cSome GIST have germline mutations in a context of syndromic GIST (e.g. GIST with NF1 mutations).

PRIMER

Familial GIST and multiple GIST. Most GIST are spo- radic without known risk factors; however, some con- ditions are associated with an increased risk of GIST, including germline mutations of KIT, PDGFRA or SDH family genes, or germline loss of NF1. Generally, patients with GIST with germline mutations are com- paratively young at disease onset, have multiple GIST and have ICC hyperplasia in the normal gastrointestinal tract. Patients with germline KIT mutations may have skin pigmentation, urticaria pigmentosa and dysphagia in addition to multiple GIST7,44. Patients with germline PDGFRA mutations may have inflammatory fibroid polyps in addition to multiple gastric GIST and a hand deformity85,86. Patients with germline SDH family gene mutations may have gastric GIST and paraganglioma (Carney–Stratakis syndrome)87. The Carney triad, caused by epigenetic suppression of SDH, is a distinct clinical entity of nonhereditary disease affecting children and young adult females and is associated with gastric GIST, paraganglioma, adrenal adenoma and/or pulmo- nary chondroma72 (TABLE 2). Familial GIST are generally thought to show similar TKI sensitivity to sporadic GIST with the same mutations.
Intriguing cases of coexistence of different muta- tions in different primary GIST diagnosed in the same patient have been reported. Several series have described multiple GIST with different KIT or PDGFRA mutations arising in the stomach or small intestine in the absence of any familial history88,89. The importance of these observations is unclear and may point to a multifocal regional oncogenic process, related to another genetic or epigenetic event that existed before the kinase mutations. In an unreported personal observation (J.Y.B., personal observation), a patient with two different simultaneous GIST, one of gastric location with a KIT exon 11 muta- tion and one in the small intestine with a KIT exon 9 mutation, also had pre-existing acromegaly, also a very rare condition. In addition to the above-mentioned role of IGF1 in SDH-deficient GIST, this clinical case with three very rare conditions points to the contribution of the growth hormone–IGF1R pathway in the oncogenic transformation of GIST73–75.

Molecular mechanisms of TKI resistance
Resistance to TKIs can be divided into primary resist- ance, defined as disease progression within the first 6 months of first-line treatment with imatinib, and sec- ondary resistance, defined as disease progression during imatinib treatment after >6 months of treatment. These two resistance types have different genetic mechanisms and different clinical features9,18,90. GIST with primary resistance typically show multifocal progression of all metastatic lesions on TKI therapy, whereas those with secondary resistance often show progression on one or a few of the metastases while others remain controlled by the therapy.

Primary resistance. Nearly 10% of patients with advanced GIST show primary resistance to imatinib. Primary resistance strongly correlates with the primary GIST genotype9–11,90 and is more frequent in those with KIT exon 9 mutations or GIST without KIT or PDGFRA

mutations and very rare in those with KIT exon 11 muta- tions. PDGFRA exon 18 D842V mutations also confer primary resistance to imatinib22, although patients may present with stable disease on imatinib owing to the indolent nature of this molecular subgroup63. These PDGFRAD842V-mutated GIST are, however, sensitive to avapritinib.

Secondary resistance. In the metastatic phase, most patients experience disease progression after an ini- tial response to imatinib, referred to as secondary resistance. The main causes of secondary resistance in KIT-mutated as well as in PDGFRA-mutated GIST are acquired cis-mutations in either the ATP-binding domain (encoded by exon 13 or 14 of KIT and exon 14 of PDGFRA) or the activation loop (encoded by exon 17 of KIT and exon 18 of PDGFRA)17,28,39,91,92. Most second- ary resistance mutations, acquired through a Darwinian selection process under the therapeutic pressure of a TKI, are missense mutations affecting codon V654 in exon 13, T670 in exon 14, or D816, D820, N822 or Y823 in exon 17 of KIT, or H687 in exon 14 or D842 in exon 18 of PDGFRA90–92. Patients presenting with sec- ondary resistance to imatinib may respond to sunitinib and regorafenib91–93. However, GIST cells with acquired mutations in the ATP-binding domain have differen- tial sensitivity to sunitinib, regorafenib and other TKIs. Acquired mutations in the activation loop are thought to be sensitive to regorafenib, sorafenib, nilotinib, ponati- nib and dasatinib17,92. By contrast, receptors with resist- ance mutations in the ATP-binding pocket are resistant to regorafenib and sorafenib, but moderately sensitive to sunitinib and ponatinib92. As a consequence, patients presenting with secondary resistance respond incon- sistently to sunitinib and regorafenib93–95. Both classes of resistance mutation, in the ATP-binding pocket and in the activation loop, are sensitive to ripretinib and avapritinib in preclinical studies19,20 and in clinical trials65,96,97. Of note, mechanisms of secondary resist- ance to avapritinib in PDGFRAD842V GIST were recently reported65. Resistance mutations are observed within PDGFRA exons 13, 14 and 15. These secondary PDGFRA mutations cause V658A, N659K, Y676C and G680R substitutions that impair avapritinib binding, while the GIST cells remain dependent on PDGFRA signalling65. These secondary missense mutations impair the binding of avapritinib to the mutant kinase65.
Other resistance mechanisms are infrequent and include overexpression of KIT and activation of alter- native pathways and/or downstream kinases, such as activating mutations in BRAF33. Resistant cells pro- liferate clonally, and different resistant metastatic lesions may have different mutations, but different mutations may also be observed in the same lesion, accounting for the heterogeneous patterns of progres- sion (for example, partial progression) observed during the development of secondary resistance to a TKI1,2,98,99. In this context, the detection of resistance mutations on circulating tumour DNA may enable the identifi- cation of the global spectrum of resistance mutations, but this approach has limited sensitivity with cur- rent technologies100. The molecular basis for acquired

PRIMER

resistance and the time point at which it originates have yet to be elucidated; however, a considerable number of patients who acquired mutations on progression were found to have had the same mutations before TKI treatment, which may have become progressively overrepresented by a Darwinian selection process99.

Diagnosis, screening and prevention
The clinical presentation of GIST is that of a tumour of the digestive tract. Bleeding, pain and weight loss are frequent symptoms. The presence of a perceptible abdominal mass upon clinical examination is not unu- sual, as some GIST may grow slowly with limited symp- toms for a long period of time. Some patients with GISTs present as emergencies owing to, for example, upper gastrointestinal tract bleeding, bowel occlusion or, in particular in small intestine GIST, perforation. In these patients, emergency surgery may not follow the usual oncological rules, affecting the long-term prognosis of the patient
GIST may also be diagnosed following an inciden- tal finding on ultrasonographic or CT imaging for an unrelated condition or, less frequently, on an endoscopic ultrasonography (EUS). MiniGIST or microGIST are fre- quently discovered incidentally in the stomach or the duo- denum during oesophago-gastro-duodenoscopy23,24,101 (FiG. 4). Their natural history is not precisely known.

Biopsy and pathological assessment
The decision to carry out a biopsy should be made on the basis of the suspected tumour type, tumour location and extent of disease1,2. Biopsy decision-making depends

Fig. 4 | Appearance of miniGIST. Small gastrointestinal stromal tumours (GIST) (1–2cm) are frequently termed miniGIST and are often incidentally discovered during gastroduodenal endoscopy. These GIST have similar driver mutations to larger GIST but are generally not resected except at primary rectal sites. a | Endoscopic image of an incidentally discovered gastric miniGIST (size 13 mm). b | Appearance of a histology sample of a miniGIST.
c | Haematoxylin and eosin stain of a miniGIST sample (magnification ×200). d | KIT expression of a miniGIST visualized by immunohistochemical stain (magnification ×200).

on the clinical presentation: when a localized lesion is accessible to an endoscopic biopsy, this will be the pre- ferred strategy; if not accessible to an endoscopic biopsy, a transparietal biopsy will be proposed; for GIST pre- senting with metastasis, transparietal biopsy of both the primary lesion and/or a metastatic lesion may be performed after a review by the multidisciplinary team deciding on patent management.
Analysis of a biopsy sample is required to confirm the diagnosis of primary GIST when the initiation of preoper- ative treatment is required following review by an expert multidisciplinary team1,2. As GIST tend to be soft and fri- able, there were concerns that a biopsy may cause tumour haemorrhage and/or rupture and intra-abdominal tumour dissemination. Thus, EUS-guided biopsy was preferred over percutaneous biopsy1,2,102,103. However, percutaneous biopsy has since not been found to be associated with an increased risk of relapse102–104. Specifically, data from the Scandinavian Sarcoma Group XVIII/Arbeitsgemeinschaft Internistische Onkologie (SSG XVIII/AIO) trial showed no detecta- ble increase in the risk of dissemination for diagnostic percutaneous biopsy in patients with localized GIST at high risk of relapse. When patients present with an abdominal nodule not amenable to endoscopic assess- ment, laparoscopic or laparotomic excision is the stand- ard approach1,2. If the tumour is strongly suggestive of GIST and resectable without extensive or mutilating (for example, requiring organ resection) surgery and without preoperative therapy, expert reference centres may reco- mmend omitting the preoperative biopsy. This should be proposed exclusively after a review by the multidiscipli- nary tumour board of an expert centre1,2,16. In the pres- ence of metastatic disease, percutaneous image-guided biopsy is appropriate for the confirmation of disease.
The morphological diagnosis of GIST is made on the basis of careful microscopic examination of ade- quate tumour tissue (FiGs 5,6,7). GIST demonstrate three different histological patterns: 70% of GIST are spindle cell type, consisting of cells with pale eosino- philic fibrillary cytoplasm, ovoid uniform nuclei and ill-defined cell borders; 20% of GIST are epithelioid type, consisting of rounded cells with eosinophilic to clear cytoplasm arranged in sheets and nests. Finally, 10% are mixed type with spindle and epithelioid cells105. The stroma of GIST is often limited and may vary from hyalinized to myxoid, although extensively myxoid GIST are rare. The morphological differential diagnosis of spindle cell GIST includes both benign and malignant diseases, such as smooth muscle tumours (leiomyoma or leiomyosarcoma), schwannoma, hae- mangioma, plexiform fibromyxoma, intra-abdominal desmoid-type fibromatosis, inflammatory myofibro- blastic tumour, solitary fibrous tumour and sarcomatoid carcinoma105,106. The differential diagnosis of epithelioid GIST includes metastatic melanoma, clear cell sarcoma, epithelioid variants of leiomyosarcoma and epithelioid haemangio-endothelioma1,2,105,106.
Tumour morphology guides the pathological diag- nosis of GIST but immunohistochemical assessment (IHC) is required to further confirm the diagnosis. The most important IHC marker is KIT (CD117), which is

PRIMER

Fig. 5 | Histopathological features of gastric GIST. a | Spindle cell type gastrointestinal stromal tumours (GIST) consist of fascicles of uniform spindle cells with eosinophilic cytoplasm. b | KIT expression visualized by immunohistochemical stain. c | Epithelioid cell type GIST consist of tumour cells with round to oval nuclei and eosinophilic cytoplasm; myxoid matrix is noted. d | Diffuse DOG1 expression visualized by immunohistochemical stain. All images with haematoxylin and eosin counterstain
and ×200 magnification.

expressed in ~95% of GIST (FiGs 5,6,7). However, pos- itive staining for KIT alone is not sufficient for diag- nosis, as other tumours can be positive for KIT on IHC, including metastatic melanoma, angiosarcoma, Ewing’s sarcoma, childhood neuroblastoma, extramed- ullary myeloid tumour, seminoma and small cell lung carcinoma106. Thus, both morphology and immuno- phenotype must be considered in the diagnosis of GIST. GIST with KIT exon 11 deletion of codons 557 and 558 are shown in FiG. 6a,b. GIST with KIT exon 9 mutations are shown in FiG. 6c,d. GIST with KIT exon 13 mutations are shown in FiG. 6e,f. Other com- monly expressed markers include CD34 (70% of GIST), smooth muscle actin (SMA; 25% of GIST) and desmin (<5% of GIST)105–107.
For ~5% of GIST that do not express KIT on IHC, establishing an accurate diagnosis is challenging. For these tumours, as well as for GIST in general, DOG1 (encoded by ANO1 and expressed in ~95% of GIST) is a useful marker. DOG1 expression does not differ between KIT-mutated, PDGFRA-mutated GIST and GIST with- out KIT or PDGFRA mutation107–110, but PDGFRA- mutated GIST often weakly express KIT (FiGs 5,6,7). Around 30% of KIT-negative GIST express DOG1 (REF.105); therefore, DOG1 IHC is helpful in the diagno- sis of these tumours. PDGFRA-mutated GIST showing typical epithelioid features with myxoid stroma and stained for KIT and DOG1 is shown in FiG. 7. PDGFRA expression can be detected in GIST IHC but its role in routine diagnosis is unclear39,111–113. Other IHC markers that may help in the diagnosis of KIT-negative GIST include CD34, PKC-θ and SMA114. S-100 and desmin staining can be performed to exclude neural tumours and

smooth muscle tumours, respectively. Genotypic analy- ses of GIST that do not express KIT are recommended to accurately classify these tumours39,110–113.
Most GIST without KIT or PDGFRA mutations are SDH-deficient with loss-of-function mutations in the SDH subunit SDHA, SDHB, SDHC or SDHD25,26,67–71.
As SDHB is ubiquitously expressed in cells, loss of SDHB in tumour cells indicates a dysfunction of the SDH complex and IHC analysis of SDHB is recommended to differentiate SDH-deficient from SDH-competent tumours67–71. Pathological assessment of SDHB-deficient GIST shows epithelioid or mixed spindle and epithelioid tumour cells separated by fibrous bands. These cells are negative for SDHB expression (FiG. 6g,h) and positive for KIT and DOG1 expression on IHC. In GIST without KIT or PDGFRA mutations and with conserved SDH expression, mutations in NF1, BRAF or RAS genes are often observed, as well as translocations involving one of the NTRK genes or, even more rarely, FGFR1.
NF1-mutated GIST present in the small intestine and usually show spindle cell morphology and indo- lent clinicopathological features32,74–76 (FiG. 6i,j). Most of them present as multiple small intestine tumours and hyperplasia of ICC31. NF1-mutated GIST express KIT and DOG1. Detection of expression of one of the NTRK proteins on IHC indicates that a NTRK fusion protein may be the oncogenic driver (FiG. 6k,l). The ETV6–NTRK3 fusion gene is one of those reported in these very rare molecular subgroups. The identifi- cation of the fusion gene requires confirmation using next-generation sequencing, as it has major therapeutic implications.

Imaging and staging
Imaging is used to assess tumours at diagnosis, initial staging and restaging after neoadjuvant treatment, as well as for therapy response monitoring and follow-up sur- veillance for possible recurrence1,2. Contrast-enhanced abdominal and pelvic CT is used for the characteriza- tion of an abdominal mass and to evaluate disease extent and the presence of metastasis. Thoracic CT is added for rectal and oesophageal lesions; chest X-ray is sufficient for other primary sites of GIST. Pelvic MRI may pro- vide better preoperative staging information for rectal GIST1,2. In addition, MRI usually yields excellent ana- tomical definition of liver metastases. PET or PET/CT may be performed when evidence of metastasis is equiv- ocal, or to assess complex metastatic disease in patients who are being considered for surgery. If a clinician considers using PET or PET/CT to monitor response to therapy, a baseline scan should be obtained before a TKI is administered1,2. PET may be recommended if neoad- juvant treatment with imatinib or avapritinib (for D842V mutations) is proposed by an expert multidisciplinary board for locally advanced disease.

Mutation analysis for diagnosis and theranostics Management of locally advanced or metastatic GIST, and decision on neoadjuvant and adjuvant treatment of localized GIST at high risk of relapse, is based on the administration of TKIs that block activated kinases. Because different mutated activated kinases

PRIMER

KIT exon 11- mutated GIST

KIT exon 9- mutated GIST

KIT exon 13- mutated GIST

SDHB-
mutated GIST

NF1 GIST

NTRK3–ETV6
GIST

Fig. 6 | Histopathological features of molecular GIST subtypes. a | KIT exon 11-mutated gastrointestinal tumour (GIST), haematoxylin and eosin stain (H&E). b | DOG1 expression in a KIT exon 11-mutated GIST, immunohistochemical stain (IHC). c | KIT exon 9-mutated GIST, H&E. d | KIT expression in a KIT exon 9-mutated GIST, IHC. e | KIT exon 13-mutated GIST, H&E. f | KIT expression in a KIT exon 13-mutated GIST, IHC. g | SDHB-mutated GIST, H&E. h | Loss of expression of SDHB in a SDHB-mutated GIST, IHC. i | NF1-mutated GIST, H&E. j | KIT expression in a NF1-mutated GIST, IHC. k | GIST
with NTRK3–ETV6 fusion, H&E. l | NTRK3 expression in a NTRK3–ETV6-mutated GIST, IHC. All images ×20 magnification. Images courtesy of Marie Karanian and Alexandra Meurgey, Centre Leon Berard.

or proteins are observed in different GIST subtypes that have different sensitivity to specific TKIs or doses, molecular characterization is essential1. Mutation profiling of metastatic GIST was demonstrated to be cost-effective115. In general, the molecular alteration
should be documented when a systemic treatment is considered, in both localized (neoadjuvant or adjuvant treatment) and advanced GIST1,2.
Of note, although molecular characterization may not be accessible for all patents in all regions, it is crucial

PRIMER

in optimal treatment decision-making. If mutation analysis is delayed or not possible, initial treatment with imatinib should not be delayed. However, imatinib treat- ment should be carefully monitored for efficacy in these cases using the described imaging tools. PET or PET/CT enables rapid assessment of response to treatment, when a rapid decrease in FDG uptake is observed.

KIT and PDGFRA. If TKIs are considered as part of the treatment plan, tumour mutation analysis is strongly recommended1,2, as the presence and nature of mutations strongly correlate with the activity of TKIs. TABLEs 1 and 2 present the frequency of the different mutually exclu- sive driver mutations in large series of GIST. Most GIST have activating mutations of the genes that encode the KIT and PDGFRA kinases, which are sensitive to the first-line TKI imatinib.
GIST types associated with primary imatinib resist- ance include those with PDGFRA exon 18 D842V mutations or GIST without KIT or PDGFRA mutations. Mutation analysis is particularly important in the neo- adjuvant or adjuvant treatment setting, as it enables inefficient treatments to be avoided in the case of an insensitive mutation or when KIT and PDGFRA are not mutated. Considering that most PDGFRA-mutated GIST are located in the stomach, gastric primary GIST should be particularly evaluated for mutational status when neoadjuvant imatinib treatment is proposed by the expert multidisciplinary board. The characterization of the driver mutation is also essential for decision-making regarding the use of adjuvant imatinib treatment.
However, in 2021, mutation analysis is not man- datory for the diagnosis of GIST at low (<10%) risk of relapse according to standard risk indexes (BOx 1) if their histology is characteristic of GIST and shows KIT and/or DOG1 positivity. Molecular analysis may be considered when KIT is not detected on IHC.
KIT and/or PDGFRA mutation testing is also reco- mmended in suspected familial GIST related to her- itable germline mutations of KIT or PDGFRA7,43,85–88. The inheritance pattern in these tumours is autoso- mal dominant. Thus, if a patient with KIT-mutated or PDGFRA-mutated GIST presents with multiple tumours and the characteristic clinical manifestations of germline KIT or PDGFRA mutations, germline testing for KIT and/or PDGFRA mutations would be indicated.

GIST without KIT or PDGFRA mutations. If the tumour is SDH-deficient, germline testing for SDH mutations should be performed to identify patients with Carney– Stratakis syndrome, which is inherited in an autosomal dominant manner87. If a germline SDH mutation is found, genetic counselling is indicated, together with mutation screening of first-degree relatives and regular screening for paraganglioma, pheochromocytoma and other tumours72,87.
A minority of GIST that lack KIT and PDGFRA muta- tions retain SDH expression, including NF1-mutated GIST, GIST with BRAF mutation and tumours with gene fusions involving NTRK3 or FGFR1 (REF.84). As these SDH-competent GIST are often less responsive or not responsive to currently approved TKIs, but may respond to other therapies, such as NTRK or BRAF inhibitors, identifying these mutations may help determine the appropriate treatment116.

Prognostic factors
Risk assessment in localized GIST aims to identify tumours that are more likely to recur at distant sites after curative surgery (BOx 1). Several classifications have been used in clinical trials and registries over the years, which account for tumour size, tumour location, mitotic index (MI) and tumour rupture before or during sur- gery. These classifications provide the basis for optimal adjuvant therapy in clinical trials.
The first two pathological features used for risk stratification were tumour size and MI, which were published in 2002 in the NIH consensus criteria117. In a long-term follow-up study of 1,765 patients with gastric GIST, only 2–3% of tumours <10 cm of largest size and MI <5 mitoses per 50 high-power fields (HPFs) but 86% of tumours >10 cm with >5 mitoses per 50 HPFs presented metastatic relapse118. Tumours >10 cm with an MI <5 per 50 HPFs and those <5 cm with an MI >5 per 50 HPFs have comparatively low relapse rates of 11% and 15%, respectively. However, in a subsequent study of 906 patients with small intestinal GIST, tumours >10 cm with an MI ≤5 per 50 HPFs and those ≤5 cm with an MI
>5 per 50 HPFs had a relapse rate of >50%, which was higher than that of gastric GIST with similar tumour parameters119. Thus, in addition to tumour size and MI, tumour site was included in the Armed Forces Institute of Pathology (AFIP) criteria105. A nomogram on the

Fig. 7 | PDGFRA-mutated GIST. a | Haematoxylin and eosin staining. b | KIT expression visualized by immunohistochemical stain. c | DOG1 expression visualized by immunohistochemical stain. All images ×200 magnification. The gastrointestinal tumour (GIST) cells have epithelioid features, and lack KIT expression but show DOG1 expression on immunohistochemistry (IHC). IHC and cell morphology suggest a PDGFRA-mutated GIST to be confirmed by molecular characterization.
The epithelioid feature of a gastric GIST suggests that this sample may be from a GIST with the PDGFRAD842V mutation.

PRIMER

basis of these three factors to predict the risk of recur- rence after surgery in the absence of adjuvant treatment has been developed to help guide patient selection for adjuvant imatinib therapy120.
Patients with tumour rupture also have a high risk of GIST recurrence121,122, and a modified NIH classification published in 2008 included tumour rupture as a pro- gnostic factor together with tumour size, MI and tumour site123 (BOx 1). Prognostic contour maps were generated using data from a pooled series of patients with GIST who did not receive adjuvant therapy, incorporating MI, tumour size, tumour rupture and primary site124. The inclusion of tumour size and MI as continuous varia- bles provides a more accurate risk range for an individ- ual patient. Together with the above mentioned AFIP classification105 and nomogram120, the maps124 are widely used to guide adjuvant treatment decisions.
Mutational status is not currently used to assess the malignant potential of a localized GIST, although some molecular subgroups, such as GIST with KIT exon 11 deletions on codons 557 and 558, have a much higher risk of relapse than others54. Two further molecular parameters have been reported as predictive of the risk of relapse for localized GIST. Genomic index was found to be effective in distinguishing GIST at high risk of relapse in an intermediate risk group125,126. In addition, a single nucleotide variant in exon 10 that encodes an M541L switch in KIT and occurs in 20% of humans was also found to be associated with an increased risk

of progression in localized GIST127,128. This variant KIT has been reported to enhance growth of the cells at low levels of SCF and increased sensitivity to imatinib128.

Screening
GIST is a submucosal tumour and, unlike adenocarci- nomas of the gastrointestinal tract that originate from mucosa, many patients with GIST are initially asymp- tomatic, which hampers early detection. No estab- lished screening methods for GIST currently exist, owing to their rarity and cost inefficiencies. MiniGIST and microGIST are frequent in the general popu- lation, but only very few progress to clinically relevant tumours129, further supporting that screening of GIST would not be cost-effective. A study conducted in Japan found that the Japanese gastric cancer screening sys- tem, which comprises barium-based radiography or oesophago-gastro-duodenoscopy, contributed to early detection of gastric GIST and favourable treatment outcomes130. However, these findings cannot be applied to other regions with a lower prevalence of gastric cancer.

Management
The therapeutic management of GIST comprises sur- gical and systemic treatments and requires multidisci- plinary collaborations involving sarcoma oncologists, surgeons, radiotherapists, teams in radiology and inter- ventional radiology, nuclear medicine and molecular pathology (FiG. 8).

Surgical treatment of localized GIST
Surgery is the primary treatment for localized GIST1,2 and is often performed using laparoscopic approaches at expert centres. A patient who presents with metastatic disease must not be operated on upfront but should receive TKI therapy as initial treatment. Large localized tumours and/or those challenging to resect because of the involvement of adjacent organs may require neoadjuvant therapy if the tumour contains a mutant kinase that is sensitive to available TKIs. Patients with localized GIST who undergo a resection considered R0 (en-bloc resection with normal tissue on all margins of resection) or R1 (en-bloc resection with tumour cells on one of the margins) have similar outcomes in reference centres131. Tumour rupture, preoperatively or during the intervention, is associated with a highly increased risk of relapse, and this parameter is, therefore, included in the most recent risk scores121–123,131. Whether a patient whose tumour has ruptured should be considered to have metastatic disease is debated.
Routine lymphadenectomy is not indicated at pri- mary presentation unless preoperative macroscopic evi- dence of lymph node involvement exists, because node involvement is very rare in localized GIST1,2. However, SDH-deficient tumours, which typically arise in the stomach, may have lymph node involvement and, in these cases, resection should focus on removal of macroscopic disease and limiting the extent of the gas- tric resection132. In this GIST subtype, R1 resection is also not associated with a worse outcome132.
MiniGIST and microGIST are often managed con- servatively, for example, with yearly EUS. This is usually

PRIMER

Fig. 8 | Management of localized and advanced or metastatic GIST. Principles of management of gastrointestinal stromal tumours (GIST) in localized (part a) and advanced (part b) phases according to clinical practice guidelines. Surgery has a central role in localized disease, adjuvant treatment is proposed for high-risk GIST, and imatinib treatment has a primary role in advanced disease. In localized GIST, surgical resection is classified as R0 (en-bloc with normal tissue on all margins of resection) or R1 (en-bloc resection with tumour cells on one of the margins). In advanced GIST, the decisions for surgery, for imatinib discontinuation and for sunitinib and regorafenib treatment depend on disease status, including progressive disease (PD), partial response (PR) and stable disease (SD). Adapted with permission from REF.1, Elsevier.

recommended in the clinical practice guidelines for lesions of <2 cm that are not located in the rectum1,2. However, some may progress and become life-threatening, as data from a large series of microGIST reported 12% mortality at 10 years101. Importantly, rectal GIST are never considered to be indolent and should be resected or proposed for neoadjuvant treatment with an active
TKI regardless of size1,2. MiniGIST and microGIST are more common than the larger GIST, which require the classical oncological management.
Patients who present with metastatic disease must not be operated on upfront and should receive TKI therapy as initial treatment, apart from specific clinical situations and after discussion in expert multidisciplinary teams.

PRIMER

Medical treatment of GIST
TKIs are the preferred first-line treatment of GIST as of 20 years ago, when a series of clinical trials demonstrated the value of this class of agents133–138. Several active treat- ments with different indications have been identified. Cytotoxic chemotherapy treatment is not active in GIST, providing median PFS of <3 months136. The mechanisms of resistance to chemotherapy have never been fully elucidated.

Imatinib mesylate. Imatinib, developed to target the chronic myelogenous leukaemia BCR–ABL transloca- tion, has efficacy against the KIT and PDGFRA tyrosine kinases133 (TABLE 3). It is the first-line drug for advanced GIST, but is also used for neoadjuvant and adjuvant ther- apy in localized GIST. Imatinib is generally well toler- ated: the most common adverse effects are rash, nausea, diarrhoea and fluid retention, in particular, periorbital oedema. Rarely (<1%) severe adverse effects, such as

Table 3 | Response rates and PFS of approved GIST treatments

Treatment (molecular GIST subtype) Treatment line Response rate (%) Median PFS (months) Refs
Imatinib
KIT exon 11 1 71.7 24.7–39.4 8–11
KIT exon 9 1 34.5 13–16.7 8–11
KIT exon 9, 400 mg daily 1 17 7–12.6 8–11
KIT exon 9, 800 mg daily 1 67 16.7–18.0 8–11
Wild-type (without precision) 1 23.1–44.6 12.3–12.8 8–11
SDH-deficient 1 2 (1/49) NA 25,26
PDGFRA 1 18–30 6.4 8,63
D842V 1 0 2.8 63
Non-D842V 1 32–84 28.5 8,63
All ≥3 0 1.8 177
Sunitinib
KIT exon 11 ≥2 5 5.1–7.0 57,151
KIT exon 9 ≥2 37.0 12.3–19.4 57,151
Wild-type (without precision) ≥2 0 19 57,151
SDH-deficient ≥2 10.5 (4/38) NA 25,26
PDGFRA (all) ≥2 0 2.8 57,151
Regorafenib
KIT exon 11 ≥3 NR 13.4 156
KIT exon 9 ≥3 NR 5.7 156
SDH-deficient ≥3 NR 10 156
Wild-type without SDH inactivation ≥3 NR 1.6 156
PDGFRA-D842V ≥3 1/1 >20 192
KIT secondary KIT exon 17 ≥2 40 22.2 95
Ripretinib
All ≥4 9 6.3 66
Avapritinib
PDGFRA-D842V ≥1 88 NR (76% at
15 months) 64
Larotrectinib
NTRK fusion protein ≥2 4/4 26.3 116
GIST, gastrointestinal stromal tumours; NA, not applicable; NR, not reported; PFS, progression-free survival.

interstitial pneumonia or liver toxicity, may preclude the use of imatinib; in this case, other options, such as sunitinib93 or nilotinib in GIST with KIT exon 11 mutations139, may be considered.
For advanced disease, the starting dose of imatinib is 400 mg daily for all patients, except for those with a documented KIT exon 9 mutation9–11. Most patients had a partial response (40–68%), up to 5% had a complete response and 14–32% had stable disease133–138. Tumours that initially progress on imatinib are those lacking KIT or PDGFRA mutations and those with PDGFRAD842V mutation9–11,63. In advanced disease, the PFS achieved with imatinib therapy for GIST with KIT exon 9 muta- tions is shorter (12.6–16.7 months) than that for GIST with KIT exon 11 mutations (typically now >2 years)9–11,53. For advanced GIST with KIT exon 9 mutations, an increased dose of 400 mg twice daily has demonstrated improved PFS compared with 400 mg daily9–11. Of note, the dose of imatinib 800 mg daily has not been tested in a prospective trial in the adjuvant phase and can, therefore, not be recommended in this setting35,55,56.
On CT imaging, therapeutic benefit is initially observed by a decrease in tumour density; the maxi- mum decrease in size may take ≥6 months to occur139,140. Median OS is estimated at 4.5–5 years9–11 and is equiv- alent in those achieving stable disease, partial response or complete response13,138. The 10-year PFS is estimated at 23%13,138, with very few progressions after 10 years.
Imatinib should be continued until disease pro- gression, even in patients with complete responses. Discontinuation after 1, 3 or 5 years of therapy results in PFS of 6, 7.1 and 10.8 months, respectively13,141–143. Poorer outcomes are seen in patients with very rapid progres- sion after discontinuation of therapy and re-challenge does not achieve equivalent disease control143.
Imatinib is also used in the adjuvant setting and improves OS when given for 3 years after resection of high-risk tumours (>50% recurrence risk)12,55,56. This therapy can also be considered in patients with an inter- mediate recurrence risk (25–50%) on a case-by-case basis. Adjuvant imatinib treatment has not demonstrated bene- fit in GIST without KIT or PDGFRA mutations or in GIST with KIT or PDGFRA mutations that confer imatinib insensitivity. These patients should undergo observation after surgical resection of localized disease1,2,35.
Currently, imatinib at 400 mg daily for ≥3 years is recommended in the adjuvant setting on the basis of the results from the SSG XVIII/AIO trial, which demon- strated an improvement in both RFS and OS in patients at high risk of recurrence defined as tumour diameter
>10.0 cm, MI >10 per 50 HPFs, diameter >5.0 cm and MI
>5 per 50 HPFs, or the presence of tumour rupture)12,35. After 3 years compared with 1 year of therapy, RFS at 5 years was 71.1% versus 52.3% and OS was 91.9% versus 82.3%, respectively12. The PERSIST-5 trial, a single-arm trial of adjuvant therapy for high-risk GIST (defined as tumours ≥2cm with MI ≥5 per 50 HPFs or nongastric pri- mary GIST ≥5 cm), demonstrated that imatinib treatment for 5 years is safe and not associated with recurrence while on therapy in patients with imatinib-sensitive tumours144. For tumours that rupture before or during surgery, result- ing in an exceedingly high rate of recurrence, therapy for

PRIMER

longer than 3 years may be considered given the >90% OS of these cases in the PERSIST-5 trial144. Two ongo- ing trials are testing 3 years versus 5 years (SSGXXII; NCT02413736) or 3 years versus 6 years (ImadGIST; NCT02260505) of adjuvant imatinib in patients with high-risk GIST (defined as in the SSG XVIII/AIO trial)12. Although surgery is the standard first-line therapy for primary GIST, neoadjuvant therapy may be required for some tumours, specifically those arising in the rectum, the small intestine requiring a Whipple procedure, the oesophagus or the oesophago-gastric junction, or for large primary tumours when resection would entail sub- stantial morbidity, such as multiple-organ resection, when adjacent organs are involved or to avoid total gastrectomy, which may result in altered imatinib pharmacokinetics145. The decision for neoadjuvant therapy requires a diagnos- tic biopsy and molecular testing to ensure the presence of an imatinib-sensitive mutation1,2. Neoadjuvant therapy has not shown a substantial increase in post-operative complications and may enable R0 or R1 resections deemed not feasible or too extensive before neoadju- vant treatment146–148. Surgery should be performed when maximal response has been achieved, which may require
≥6 months. Some advocate early fluorodeoxyglucose (FDG)–PET imaging to confirm response on the basis of decreased tumour metabolic activity149. Patients may take imatinib until the day before surgery and resume treatment when they are able to eat food1,2.
Importantly, when primary or secondary resistance to imatinib becomes evident in a patient with GIST bearing an imatinib-sensitive mutation, the physician must carefully question the patient regarding medi- cation compliance and must also consider possible malabsorption1,2. Monitoring drug levels may be useful in this situation but is not routinely used145,150.

Sunitinib mesylate. Sunitinib is approved for the treat- ment of advanced GIST following progression on imati- nib or for patients with intolerance to imatinib. It has broader kinase activity than imatinib, including VEGFR, FLT1, KDR and RET inhibition151. Sunitinib inhibits KIT with secondary resistance mutations affecting the ATP-binding pocket (exons 13 and 14), but inhibits less well KIT with mutations affecting the kinase loop domains (exons 17 and 18)17. It also has activity against exon 9 mutations and, unlike imatinib, in SDH-deficient tumours, possibly owing to inhibitory activity against VEGFR152. The toxicity profile of sunitinib includes diar- rhoea, fatigue, hypertension and cardiac toxic effects, hypothyroidism and hand–foot syndrome92,153,154.
Various doses and schedules of sunitinib have been tested, and 50 mg daily for 4 weeks followed by 2 weeks without treatment is the approved regimen93. In contrast to imatinib in the first-line setting, the objective response rate to sunitinib is low (7%) and >60% of patients expe- rience stable disease92,152–154. This low response rate is a consistent observation in second or later treatment lines of all approved or non-approved agents in GIST (TABLE 3). Compared with placebo, the PFS on sunitinib was improved from 6.4 weeks to 27.3 weeks93. Sunitinib is also commonly dosed at 37.5mg daily as a continuous regimen on the basis of a trial that did not include a randomized

comparison with the typical dosing schedule but showed comparable PFS responses and survival154. The rationale for continuous therapy originates from patients expe- riencing recurrent tumour symptoms and occasional increased tumour metabolic activity by FDG-PET when off sunitinib treatment. Efficacy is maintained on the con- tinuous schedule: partial response and stable disease rates are 13% and 44%, respectively, and the median PFS is 34 weeks. Toxic effects of continuous dosing are comparable to those of the intermittent schedule154.
Of note, the sensitivity of GIST with KIT exon 11 mutations differs from that of those with exon 9 muta- tions not only for imatinib but also for sunitinib treat- ment. The PFS of patients with exon 9-mutated GIST is better than that of patients with exon 11-mutated GIST on sunitinib therapy with median PFS of 12.3 and
7.0 months, respectively17,57,58.

Regorafenib. Regorafenib has the broadest kinase activ- ity of available agents, targeting VEGFR1–3, TEK, KIT, RET, RAF1, BRAF, PDGFR and FGFR. The agent is approved in the third-line setting after imatinib and sunitinib93,155,156. Regorafenib is used at an intermittent schedule of 160 mg for 21 days with 7 days off treatment, repeated every 28 days. Its toxicity profile is similar to that of sunitinib, but indirect comparison suggests an increased proportion of patients with hand–foot syn- drome for regorafenib. The median PFS for regorafenib treatment was 4.8–13.2 months and the response rate was 4.5–12% in phase II and phase III clinical trials94,155. The benefit of regorafenib is seen irrespective of the primary KIT mutation in exon 9 or exon 11, and regorafenib also has activity against GIST with some secondary mutations in exon 17 (REFs94,155). Activity has also been observed in SDH-deficient GIST: two patients had a partial response and four patients had stable disease156. A phase II study led by the Grupo Español de Investigación en Sarcomas (GEIS) is ongoing to confirm these results prospectively (REGISTRI; NCT02638766).

Larotrectinib and entrectinib. Translocations involv- ing NTRK3 (encoding NTRKC) have been identified in GIST without KIT or PDGFRA mutations with con- served SDHB expression76. In advanced disease, these tumours respond to the NTRK inhibitors larotrectinib and entrectinib116,157,158. Larotrectinib is a selective inhib- itor of NTRKA, NTRKB and NTRKC. Entrectinib blocks NTRKA, NTRKB, NTRKC, ROS1 and ALK. Phase I and II
trials of larotrectinib demonstrated a response rate of 75%, and 55% of patients were free from progression at 1 year116. Four patients with GIST were included, of which three had a partial response and the fourth had a complete response. This therapeutic benefit underscores the need for expanded molecular testing in GIST that are negative for KIT and PDGFRA mutations. In November 2018 and in August 2019 the FDA granted approval to larotrectinib and entrectinib, respectively, for the treatment of solid tumours that have a NTRK gene fusion.

Avapritinib. Avapritinib was designed as a highly potent and selective inhibitor of activated KIT and PDGFRA mutants, particularly those with mutations affecting

PRIMER

the activation loop19,96. Phase I testing defined the start- ing dose as 300 mg and the maximum tolerated dose as 400 mg daily64. Common toxic effects are similar to those of imatinib (anaemia, dizziness, diarrhoea, oedema, fatigue, lacrimation, nausea and vomiting), but neuro- cognitive adverse effects also occur, probably owing to the ability of avapritinib to cross the blood–brain barrier64.
Avapritinib has antitumour activity in patients with advanced GIST with the PDGFRAD842V mutation64. Responses were observed at all tested dose levels: in 56 patients, 5 patients had a complete response (9%), 44 patients had a partial response (79%) and 7 patients had stable disease64 at a dose of 300 mg daily. Initial reports also noted responses in KIT-mutated GIST at higher doses. In the patients with GIST with KIT mutation who received ≥135 mg, two had confirmed partial responses (18%) and five had stable disease64. Avapritinib is an important new agent for patients with PDGFRA-D842V tumours who have no other proven active medical therapies. Avapritinib has been approved on the basis of the phase I/II trial results64 for the treat- ment of GIST with mutations of PDGFRA exon 18, including the D842V mutation, by the FDA in January 2020 and specifically for GIST with the PDGFRAD842V mutation by the EMA in October 2020.
Avapritinib was compared with regorafenib in GIST in the third-line setting in a randomized phase III trial (VOYAGER; NCT03465722). Early top-line data indi- cated that avapritinib did not demonstrate an improve- ment over regorafenib in centrally reviewed PFS, the primary end point of the study159 (Blueprint medicines press release). The median PFS was reported to be
4.2 months for avapritinib and 5.6 months for regorafenib.

Ripretinib. Ripretinib is an inhibitor of KIT and PDGFRA that locks these kinases in an inactive con- formation20. In preclinical testing, the agent inhibited wild-type as well as KIT single and double mutants20. Of particular interest is its activity against cells with KIT exon 17 D816V, D816G, D820A, D820E, D820Y, N822K,
N822Y, N822H and Y823D primary or secondary muta- tions, as well as cells with dual mutations on exon 9 and exon 13, exon 9 and exon 14, exon 9 and exon 17 that are not well treatable with currently available agents. In an ongoing phase I trial, no clear maximum tolerated dose was identified160. The dose of 150 mg daily was used in the expansion cohort for efficacy assessment. The median PFS was 10.7, 8.3 and 5.5 months and the objective response rates were 19.4%, 14.3% and 7.2% in second-, third- and fourth-line therapy, respectively, in a GIST patient population refractory to approved agents160. The phase III INVICTUS trial evaluated ripretinib using a dose of 150 mg daily in fourth-line or later-line setting compared with placebo in patients with GIST that had progressed following all standard therapies66. In 129 patients, median PFS was 6.3 months and 1.0 months (HR 0.15, 95%CI 0.09–0.25; P < 0.0001),
median OS was 15.1 and 6.6 months and response rates were 9% and 0% in the ripretinib and placebo groups, respectively. Crossover yielded comparable PFS (median
4.6 months) and OS (median 11.6 months) in patients who were able to receive and benefit from crossover66,161.

The most common (>2%) grade 3 or 4 treatment-emergent adverse events in the ripretinib group included lipase increase, hypertension, fatigue and hypophosphataemia. Ripretinib was approved as a fourth-line or later-line treatment of GIST by the FDA in March 2020. A sec- ond phase III trial, INTRIGUE, is evaluating ripretinib compared with sunitinib as a second-line therapy162.

Other agents for rare mutations. BRAF mutations are rarely identified in GIST. The use of dabrafenib in a patient with a BRAFV600E mutation resulted in disease control for 8 months83. The use of combination thera- pies of BRAF and MEK inhibitors in GIST has not been reported to date.

Drugs in development. Crenolanib is an inhibitor of FLT3 and PDGFR family members, including the PDGFRA-D842V mutant163. In a phase I/II study in 16 patients with advanced GIST and the PDGFRAD842V mutation, treatment with 72 mg/m−2 crenolanib three times daily resulted in partial responses in two patients and stable disease in three patients164. A phase III randomized placebo-controlled trial of crenolanib in advanced GIST with the PDGFRAD842V mutation is currently ongoing (NCT02847429).
Other TKIs have been tested in clinical trials, includ- ing nilotinib, pazopanib, dasatinib, cabozantinib, and axitinib, as well as PD1 and PDL1 antibodies such as pembrolizumab, among others139,165–172. Nilotinib was compared with imatinib in a randomized phase III trial in the first-line setting139. The trial did not reach its primary end point of improved PFS. Molecular sub- group analysis indicated a lack of activity in KIT exon 9-mutated tumours; however, the effects of nilotinib were comparable to those of imatinib in patients with KIT exon 11-mutated tumours, providing a large series of patients treated with an agent other than imatinib in the first-line setting. In the very rare imatinib-intolerant patient population, this agent may be an alternative option, but must be decided by a multidisciplinary tumour board from a reference centre. Nilotinib was also compared with doctor’s choice of treatment in a rand- omized phase II study in patients with GIST progressing after imatinib and sunitinib and failed to demonstrate superiority over the control arm165.

Surgical treatment of advanced GIST
Patients who present with metastatic disease must not be operated on upfront and should receive TKI therapy as initial treatment. Resection can be performed in patients on targeted therapies, but those receiving sunitinib and regorafenib have an increased risk of complications owing to their effects on wound healing173–175. Resection, although not curative, decreases the volume of disease and the potential for resistant tumour cell clones and, theoretically, may improve survival. Retrospective data support this hypothesis but accrual of patients to prospective trials has not been feasible to date174. Only one randomized clinical trial has been reported exploring surgical resection of metastases in patients with advanced GIST receiving imatinib. The trial did not show superiority of surgery for the primary end

PRIMER

point PFS, and was discontinued prematurely owing to poor accrual. Of note, an improvement in OS in the surgery group was observed, pointing to a need for additional clinical research in this setting175.
In patients with multi-focal disease that is controlled at most metastatic sites but with limited focal progres- sion (as often observed in patients with disease that is controlled for years with imatinib), removal of the pro- gressing lesions enables continued disease stability on the current TKI173,174.
Finally, GIST molecular biology also influences a decision on surgery at advanced disease stage. Patients with neurofibromatosis typically have multifocal GIST that lack oncogenic mutations; complete resection is not necessarily indicated as these tumours may have an indolent biology176. Tumours should undergo mutational testing as they can also have typical KIT or PDGFRA mutations and may benefit from standard therapy.

Modalities for palliation
Continuing TKI therapy, even in the setting of progres- sion, mitigates rapid progression seen in the placebo arms of phase II or III trials of sunitinib87, regorafenib94, pazopanib166 and ripretinib66, and in the RIGHT trial of imatinib resumption after progression on TKI therapy177. External beam radiotherapy can be useful to control bone pain and bleeding from advanced GIST178. In addi- tion, liver-directed therapies may enable better control of pain and other symptoms related to metastatic pro- gression and can be used to prolong the TKI therapy, in the case of a single or a few progressive lesions in the liver179–181. Radiofrequency ablation can be used to con- trol a single progressive lesion. Radioembolization has also been reported in selected patients at expert centres182.

Quality of life
The evaluation of QoL is important in patients with GIST who often receive TKIs for a long period of time; however, only a few studies have addressed this question. Adverse effects related to TKIs183,184 may be limited in magni- tude, but they can impair treatment compliance in the long term, in particular, in the adjuvant phase despite a documented survival advantage. In the PERSIST-5 trial, which evaluated 5 years of imatinib treatment in the adju- vant setting, up to 49% of patients discontinued therapy transiently or permanently, increasing the risk of relapse and reducing survival144. Dose adaptation (for imatinib, sunitinib and regorafenib) with careful medical moni- toring to enable flexibility is essential to optimize treat- ment compliance184,185. Some symptomatic treatments for adverse effects such as myalgia have been proposed186.
In patients with advanced disease, the randomized trial BFR14, which evaluated treatment interruption in patients with non-progressing disease after 1, 3 and 5 years, failed to demonstrate an improvement in QoL in the interruption groups, possibly because relapses were observed quickly141,142. In the INVICTUS trial, QoL of patients assigned to ripretinib was superior to that of patients assigned to placebo, in particular for overall health and physical functioning, demonstrat- ing the importance of tumour control for symptomatic improvement187. QoL is one of the secondary end points

of the adjuvant SSGXXII and ImadGIST trials, which evaluate long treatment durations of 5–6 years.

Outlook
GIST remain paradigmatic models of cancers benefit- ing from personalized medicine approaches with TKIs. Despite the lack of formal comparative randomized studies, the long-term OS observed in the earliest stud- ies of imatinib, reporting a median survival of >6 years, is superior to that in other cancers, even compared with the newest agents in advanced disease phases (for example, osimertinib and alectinib in non-small-cell lung cancer). In GIST at high risk of relapse, adjuvant treatment with imatinib for 3 years provides a long-term survival advantage with a nearly 50% reduction of the risk of death. In the past 10 years, the treatment of GIST has moved from a one-size-fits-all approach to targeted oncogene treatments for specific molecular GIST sub- types. The natural histories of the different molecular GIST subtypes are now better understood but several important questions remain regarding the management of patients in the routine clinical setting. For example, the optimal monitoring of microGIST is still not stand- ardized and prospective studies describing their natural history with a watch-and-wait policy are needed.
Molecular characterization at diagnosis is now a standard diagnostic procedure in patients with high-risk GIST. Although expensive, molecular analysis remains much less costly than months of possibly unnecessary adjuvant or neoadjuvant treatment.
Whether the molecular characteristics of GIST should be one of the criteria to decide whether to use adjuvant treatment (for example, in intermediate-risk or low-risk GIST) is a key question that is currently being tested in the GIGIST trial (NCT02576080). This trial randomly allocates patients with intermediate-risk GIST and high-risk molecular features according to CINSARC (complexity index in sarcomas) to receive either imatinib or no treatment.
The duration of adjuvant imatinib treatment in imatinib-sensitive GIST molecular subtypes is also under investigation. After a first encouraging phase II non-controlled study144, two randomized trials are eval- uating to what extent 2 additional years (SSGXXII) or 3 additional years (ImadGIST) of imatinib treatment improve OS of patients with high-risk GIST.
It is generally recommended that patients with GIST without mutations in PDGFRA or KIT and those with a PDGFRAD842V mutation should not receive adjuvant imatinib, but no trial has yet tested the effectiveness of a 800 mg daily dose of imatinib in KIT exon 9-mutated GIST in the adjuvant setting. The subgroup analysis of SSG XVIII/AIO does not enable confirmation of a survival improvement with the 400 mg daily dose, in an underpowered subgroup analysis35. The utility of adju- vant treatment for GIST with KIT exon 9 mutation is, therefore, still unclear at the dose of 400 mg daily. A trial exploring the dose of 800 mg daily for this patient popu- lation would add important information, but requires an international collaboration to set up a worldwide trial.
The value of surgical removal of metastases in meta- static GIST has been explored in two randomized trials.

PRIMER

The first trial reported a trend towards improved RFS and OS175. The second trial was stopped owing to slow accrual (NCT00956072). Thus, whether this surgical intervention is warranted remains unclear and its pos- sible value needs to be further evaluated, possibly using registries or methodologies that require fewer patients, such as Bayesian designs, if not a randomized trial involving a global effort.
Although imatinib interruption after 1, 3 or 5 years of tumour control was associated with rapid tumour relapse53,141,142, whether relapse would also occur in patients with >10 years of tumour control needs to be evaluated in a randomized clinical trial. A randomized phase II study (GIST-TEN-01) has been initiated to explore this question in patients with stable disease or in partial or complete remission after ≥10 years of imatinib.
Several TKIs have been tested in the third-line or later-line setting with detectable clinical activity. Pazopanib was compared with placebo treatment (with crossover) in patients with advanced GIST progressing after imatinib and sunitinib166. In this trial, median PFS in the pazopanib group was superior to that with best supportive care alone (3.4 months versus 2.3 months, respectively; HR 0.59; P = 0.03). However, the difference was possibly underestimated, as pazopanib may not be absorbed in patients with gastrectomy or those receiving proton pump inhibitors (PPIs), which is common in this group of patients. Excluding those with gastrectomy or receiving PPIs, the hazard ratio was 0.31 in the subgroup of patients with mostly small intestine GIST. The effect of gastrectomy and PPI use on pazopanib activity was later confirmed by a subgroup analysis of the Palette study167. Other trials have tested masitinib168, dasati- nib in first-line and later-line settings169,170, dovitinib in second-line or later-line settings171 and cabozantinib in the third-line setting172. Dasatinib yielded a 74% met- abolic response rate and a median PFS of 13 months as first-line treatment, but only 3 months in second-line or later-line settings. Dovitinib yielded a 5% response rate with a median time to progression (TTP) of 4 months as a second-line treatment. Cabozantinib yielded a 14% response rate with a median TTP of 5.5 months as a third-line treatment. In addition, ponatinib and axitinib have been reported to inhibit TKI-resistant mutations and may require further clinical investigation91. A ran- domized phase II study is currently comparing lenvati- nib plus best supportive care with best supportive care alone (with crossover) in fourth-line or later-line settings (NCT04193553).
Immunotherapy with single-agent anti-PD1 anti- bodies has limited activity in advanced GIST188. GIST are preferentially infiltrated by macrophages whose M1 or M2 phenotype is modulated by TKIs189. Several tri- als of immunotherapy in GIST are currently ongoing. Nivolumab with or without ipilimumab is being inves- tigated in patients with advanced GIST progressing after standard TKI (NCT02880020). A randomized clinical trial that compares the addition of atezolizumab to imatinib with imatinib alone in patients who have pre- viously been treated with at least imatinib and sunitinib is ongoing. The role of immunotherapy in common and rare GIST subgroups requires further evaluation in

prospective trials, possibly guided by new biomarkers such as tertiary lymphoid structures190.
Furthermore, dedicated clinical trials are needed for GIST without PDGFRA or KIT mutations, for example, to confirm the response rates to specific TKIs in GIST with NTRK3 or FGFR1 translocations, BRAF muta- tions, and in SDH-inactivated GIST. Health technology assessment bodies must acknowledge that randomized clinical trials are not feasible in these populations, and that reimbursement and access to the drugs must be granted in these patients even without randomized trial data or precise knowledge of their natural history. These cancers are exceedingly rare and it is important to avoid discrimination owing to the rarity of the cancer type.
In addition, the clinical activity of new TKIs that block a wider range of secondary resistance mutation, including ripretinib, avapritinib and AZD3229 (REF.191), should be explored in earlier settings than fourth-line advanced disease; in particular, for ripretinib, in the first-line neoadjuvant setting for locally advanced, sensitive tumours.
Consequently and given the increasing complexity of nosological classifications of GIST, all GIST should be preferentially managed in expert sarcoma centres. Several reports have shown that this care setting is asso- ciated with improved OS, reduced risk of relapse and better compliance to standard of care, in addition to a reduced overall cost of management for rare cancers15,16. Finally, the optimal modalities and duration of follow-up monitoring of patients with GIST require additional prospective work. The analysis of circulating tumour DNA in patients in remission and the duration of systematic follow-up CT require careful utility and medical–economic evaluation for those patients who are
now very often long-term survivors.
In conclusion, GIST were the first solid tumour type for which it was demonstrated that single-agent oncogene-targeted treatment against mutated activated proteins could replace poorly effective cytotoxic chemo- therapies. In the past 20 years, considerable progress has been made in the routine management of patients with GIST and their survival has increased dramatically com- pared with that in historical series. No randomized studies were needed to demonstrate the superiority of imatinib over standard doxorubicin. Randomized clinical trials may not be mandatory to demonstrate the added value of a new treatment guided by the understanding of the underlying biology of the disease in ultrarare malignancies, such as molecular subsets of GIST, as shown by the examples of avapritinib and NTRK inhibitors. Single-arm studies with innovative designs are now required to address the chal- lenges of developing new treatments in very rare malignant conditions with specific molecular alterations.
Published online xx xx xxxx
This orphan disease now has seven approved agents worldwide following successful randomized trials in the adjuvant setting and in advanced GIST. Translational research studies coupled to some of these clinical trials have enabled refinement of the dose and the drugs given to each subtype. To date, GIST remain models for the development of precision medicine in cancer.

PRIMER

⦁ Casali, P. G. et al. ESMO Guidelines Committee and EURACAN. Gastrointestinal stromal tumours: ESMO-EURACAN Clinical Practice Guidelines for diagnosis, treatment and follow-up. Ann. Oncol. 29 (Suppl. 4), iv267 (2018).
⦁ von Mehren, M. et al. Gastrointestinal stromal tumors, version 2.2014. J. Natl Compr. Canc Netw. 12, 853–862 (2014).
⦁ Søreide, K. et al. Global epidemiology of gastrointestinal stromal tumours (GIST): a systematic review of population-based cohort studies. Cancer Epidemiol. 40, 39–46 (2016).
⦁ Yang, Z. et al. Incidence, distribution of histological subtypes and primary sites of soft tissue sarcoma in China. Cancer Biol. Med. 16, 565–574 (2019).
⦁ Verschoor, A. J. et al. The incidence, mutational status, risk classification and referral pattern of gastro-intestinal stromal tumours in the Netherlands: a nationwide pathology registry (PALGA) study. Virchows Arch. 472, 221–229 (2018).
⦁ de Pinieux, G. et al. Nationwide incidence of sarcomas and connective tissue tumors of intermediate malignancy over four years using an expert pathology review network. PLOS ONE 16, e0246958 (2021).
⦁ Hirota, S. et al. Gain-of-function mutations of c-kit in human gastrointestinal stromal tumors. Science 279, 577–580 (1998).
The first article showing the presence of KIT
mutations in sporadic and familial GIST.
⦁ Wozniak, A. et al. Prognostic value of KIT/PDGFRA mutations in gastrointestinal stromal tumours (GIST): Polish Clinical GIST Registry experience. Ann. Oncol. 23, 353–360 (2012).
⦁ Debiec-Rychter, M. et al. KIT mutations and dose selection for imatinib in patients with advanced gastrointestinal stromal tumours. Eur. J. Cancer 42, 1093–1103 (2006).
The first article showing differential responses to imatinib 400 mg per day and 800 mg per day in patients with KIT exon 9-mutated GIST.
⦁ Heinrich, M. C. et al. Correlation of kinase genotype and clinical outcome in the North American Intergroup Phase III Trial of imatinib mesylate for treatment of advanced gastrointestinal stromal tumor: CALGB 150105 Study by Cancer and Leukemia Group B
and Southwest Oncology Group. J. Clin. Oncol. 26, 5360–5367 (2008).
⦁ Gastrointestinal Stromal Tumor Meta-Analysis Group (MetaGIST). Comparison of two doses of imatinib for the treatment of unresectable or metastatic gastrointestinal stromal tumors:
a meta-analysis of 1,640 patients. J. Clin. Oncol. 28, 1247–1253 (2010).
⦁ Joensuu, H. et al. Survival outcomes associated with
3 years vs 1 year of adjuvant imatinib for patients with high-risk gastrointestinal stromal tumors: an analysis of a randomized clinical trial after 10-year follow-up. JAMA Oncol. 29, e202091 (2020).
The most recent update of the randomized trial that demonstrated an improvement in survival with 3 years of imatinib in patients with high-risk GIST.
⦁ Casali, P. G. et al. Ten-year progression-free and overall survival in patients with unresectable or metastatic GI Stromal tumors: long-term analysis of the European Organisation for research and treatment of cancer, Italian Sarcoma Group, and Australasian Gastrointestinal Trials Group Intergroup Phase III Randomized Trial on imatinib at two dose levels.
J. Clin. Oncol. 35, 1713–1720 (2017).
⦁ Martin-Broto, J. et al. Relevance of reference centers in sarcoma care and quality item evaluation: results from the prospective registry of the spanish group for research in sarcoma (GEIS). Oncologist 24, e338–e346 (2019).
⦁ Blay, J. Y. et al. Surgery in reference centers improves survival of sarcoma patients: a nationwide study. Ann. Oncol. 30, 1143–1153 (2019).
⦁ Blay, J. Y. et al. Improved survival using specialized multidisciplinary board in sarcoma patients.
Ann. Oncol. 28, 2852–2859 (2017).
⦁ Heinrich, M. C. et al. Primary and secondary kinase genotypes correlate with the biological and clinical activity of sunitinib in imatinib-resistant gastrointestinal stromal tumor. J. Clin. Oncol. 26, 5352–5359 (2008).
⦁ von Mehren, M. & Joensuu, H. Gastrointestinal stromal tumors. J. Clin. Oncol. 36, 136–143 (2018).
⦁ Evans, E. K. et al. A precision therapy against cancers driven by KIT/PDGFRA mutations. Sci. Transl. Med. 9, eaao1690 (2017).
⦁ Smith, B. D. et al. Ripretinib (DCC-2618) is a switch control kinase inhibitor of a broad spectrum of
oncogenic and drug-resistant KIT and PDGFRA variants. Cancer Cell 35, 738–751 (2019).
⦁ Ma, G. L., Murphy, J. D., Martinez, M. E. & Sicklick, J. K. Epidemiology of gastrointestinal
stromal tumors in the era of histology codes: results of a population-based study. Cancer Epidemiol.
Biomarkers Prev. 24, 298–302 (2015).
⦁ Cassier, P. A. et al. A prospective epidemiological study of new incident GISTs during two consecutive years in Rhône Alpes region: incidence and molecular distribution of GIST in a European region. Br. J. Cancer 103, 165–170 (2010).
⦁ Agaimy, A. et al. Minute gastric sclerosing stromal tumors (GIST tumorlets) are common in adults and frequently show c-KIT mutations. Am. J. Surg. Pathol. 31, 113–120 (2007).
⦁ Kawanowa, K. et al. High incidence of microscopic gastrointestinal stromal tumors in the stomach. Hum. Pathol. 37, 1527–1535 (2006).
The first study showing the frequency of microGIST.
⦁ Janeway, K. A. et al. Defects in succinate dehydrogenase in gastrointestinal stromal tumors lacking KIT and PDGFRA mutations. Proc. Natl Acad. Sci. USA 108, 314–318 (2011).
⦁ Boikos, S. A. et al. Molecular Subtypes of KIT/PDGFRA wild-type gastrointestinal stromal tumors: a report from the National Institutes of Health gastrointestinal
stromal tumor clinic. JAMA Oncol. 2, 922–928 (2016).
⦁ Basse, C. et al. Sarcomas in patients over 90: natural history and treatment — a nationwide study over
6 years. Int. J. Cancer 145, 2135–2143 (2019).
⦁ Corless, C. L., Barnett, C. M. & Heinrich, M. C. Gastrointestinal stromal tumours: origin and molecular oncology. Nat. Rev. Cancer 11, 865–878 (2011).
⦁ Ricci, R. et al. Telocytes are the physiological counterpart of inflammatory fibroid polyps and PDGFRA-mutant GISTs. J. Cell Mol. Med. 22, 4856–4862 (2018).
⦁ Kondo, J. et al. A smooth muscle-derived, Braf-driven mouse model of gastrointestinal stromal tumor (GIST): evidence for an alternative GIST cell-of-origin. J. Pathol. 252, 441–450 (2020).
⦁ Joensuu, H., Hohenberger, P. & Corless, C. L. Gastrointestinal stromal tumour. Lancet 382, 973–983 (2013).
⦁ Nishida, T. et al. Gastrointestinal stromal tumors
in Japanese patients with neurofibromatosis type I.
J. Gastroenterol. 51, 571–578 (2016).
⦁ Agaram, N. P. et al. Novel V600E BRAF mutations in imatinib-naive and imatinib-resistant gastrointestinal stromal tumors. Genes Chromosomes Cancer 47, 853–859 (2008).
⦁ Joensuu, H. et al. KIT and PDGFRA mutations and the risk of GI stromal tumor recurrence. J. Clin. Oncol. 33, 634–642 (2015).
⦁ Joensuu, H. et al. Effect of KIT and PDGFRA mutations on survival in patients with gastrointestinal stromal tumors treated with adjuvant imatinib: an exploratory analysis of a randomized clinical trial. JAMA Oncol. 3, 602–609 (2017).
An important analysis of the SSG XVIII/AIO study describing the benefits of 3-year adjuvant imatinib in patients with GIST bearing different mutations.
⦁ Brcic, I., Kashofer, K., Skone, D. & Liegl-Atzwanger, B. KIT mutation in a naïve succinate dehydrogenase- deficient gastric GIST. Genes Chromosomes Cancer 58, 798–803 (2019).
⦁ Wu, J. et al. Targeted deep sequencing reveals unrecognized KIT mutation coexistent with NF1 deficiency in GISTs. Cancer Manag. Res. 13, 297–306 (2021).
⦁ Heinrich, M. C. et al. PDGFRA activating mutations in gastrointestinal stromal tumors. Science 299, 708–710 (2003).
The first article showing mutually exclusive PDGFRA
mutations in GIST without KIT mutations.
⦁ Corless, C. L. et al. PDGFRA mutations in gastrointestinal stromal tumors: frequency, spectrum and in vitro sensitivity to imatinib. J. Clin. Oncol. 23, 5357–5364 (2005).
⦁ Chi, P. et al. ETV1 is a lineage survival factor that cooperates with KIT in gastrointestinal stromal tumours. Nature 467, 849–853 (2010).
⦁ Bosbach, B. et al. Direct engagement of the PI3K pathway by mutant KIT dominates oncogenic signaling in gastrointestinal stromal tumor. Proc. Natl Acad. Sci. USA 114, E8448–E8457 (2015).
⦁ Rossi, S. et al. KIT, PDGFRA, and BRAF mutational spectrum impacts on the natural history of imatinib-naive localized GIST: a population-based study. Am. J. Surg. Pathol. 39, 922–930 (2015).
⦁ Nishida, T., Goto, O., Raut, C. P. & Yahagi, N. Diagnostic and treatment strategy for small gastrointestinal stromal tumors. Cancer 122, 3110–3118 (2016).
⦁ Nishida, T. et al. Familial gastrointestinal stromal tumours with germline mutation of the KIT gene. Nat. Genet. 19, 323–324 (1998).
⦁ Schaefer, I. M. et al. MAX inactivation is an early event in GIST development that regulates p16 and cell proliferation. Nat. Commun. 8, 14674 (2017).
⦁ Pang, Y. et al. Mutational inactivation of mTORC1 repressor gene DEPDC5 in human gastrointestinal stromal tumors. Proc. Natl Acad. Sci. USA 116, 22746–22753 (2019).
⦁ Wang, Y. et al. Dystrophin is a tumor suppressor in human cancers with myogenic programs. Nat. Genet. 46, 601–606 (2014).
⦁ Heinrich, M. C. et al. Genomic aberrations in cell cycle genes predict progression of KIT-mutant gastrointestinal stromal tumors (GISTs). Clin. Sarcoma Res. 9, 3 (2019).
⦁ Roskoski, R. Jr. Structure and regulation of Kit protein-tyrosine kinase — the stem cell factor receptor. Biochem. Biophys. Res. Commun. 338, 1307–1315 (2005).
⦁ Mol, C. D. et al. Structural basis for the autoinhibition and STI-571 inhibition of c-Kit tyrosine kinase.
J. Biol. Chem. 279, 31655–31663 (2004).
⦁ Yuzawa, S. et al. Structural basis for activation
of the receptor tyrosine kinase KIT by stem cell factor.
Cell 130, 323–334 (2007).
⦁ Pierotti, M. A., Tamborini, E., Negri, T., Pricl, S. & Pilotti, S. Targeted therapy in GIST: in silico modeling for prediction of resistance. Nat. Rev. Clin. Oncol. 8, 161–170 (2011).
⦁ Patrikidou, A. et al. Long-term outcome of molecular subgroups of GIST patients treated with standard- dose imatinib in the BFR14 trial of the French Sarcoma Group. Eur. J. Cancer 52, 173–180 (2016).
⦁ Martin-Broto, J. et al. Prognostic time dependence of deletions affecting codons 557 and/or 558 of KIT gene for relapse-free survival (RFS) in localized GIST: a Spanish Group for Sarcoma Research (GEIS) Study. Ann. Oncol. 21, 1552–1557 (2010).
⦁ Dematteo, R. P. et al. Adjuvant imatinib mesylate after resection of localised, primary gastrointestinal stromal tumour: a randomised, double-blind, placebo- controlled trial. Lancet 373, 1097–1104 (2009).
The first randomized trial of adjuvant imatinib treatment in GIST.
⦁ Casali, P. G. et al. Final analysis of the randomized trial on imatinib as an adjuvant in localized gastrointestinal stromal tumors (GIST) from the EORTC Soft Tissue
and Bone Sarcoma Group (STBSG), the Australasian Gastro-Intestinal Trials Group (AGITG), UNICANCER, French Sarcoma Group (FSG), Italian Sarcoma Group (ISG), Spanish Group for Research on Sarcomas (GEIS). Ann. Oncol. https://doi.org/10.1016/j.annonc.2021. 01.004 (2010).
⦁ Reichardt, P. et al. Correlation of KIT and PDGFRA mutational status with clinical benefit in patients with gastrointestinal stromal tumor treated with sunitinib in a worldwide treatment-use trial. BMC Cancer 16, 22 (2015).
⦁ Reichardt, P. et al. Clinical outcomes of patients with advanced gastrointestinal stromal tumors: safety and efficacy in a worldwide treatment-use trial of sunitinib. Cancer 121, 1405–1413 (2015).
⦁ Bachet, J. B. et al. Diagnosis, prognosis and treatment of patients with gastrointestinal stromal tumour (GIST) and germline mutation of KIT exon 13. Eur. J. Cancer 49, 2531–2541 (2013).
⦁ Tabone-Eglinger, S. et al. KIT mutations induce intracellular retention and activation of an immature form of the KIT protein in gastrointestinal stromal tumors. Clin. Cancer Res. 14, 2285–2294 (2008).
⦁ Obata, Y. et al. Oncogenic signaling by Kit tyrosine kinase occurs selectively on the Golgi apparatus
in gastrointestinal stromal tumors. Oncogene 36, 3661–3672 (2017).
⦁ Asmane, I. et al. Insulin-like growth factor type 1 receptor (IGF-1R) exclusive nuclear staining:
a predictive biomarker for IGF-1R monoclonal antibody (Ab) therapy in sarcomas. Eur. J. Cancer 48, 3027–3035 (2012).
⦁ Cassier, P. A. et al. Outcome of patients with
platelet-derived growth factor receptor alpha-mutated gastrointestinal stromal tumors in the tyrosine kinase inhibitor era. Clin Cancer Res. 18, 4458–4464 (2012).
An international retrospective analysis showing the activity of tyrosine kinase inhibition in GIST with different types of PDGFRA mutation.

PRIMER

⦁ Heinrich, M. C. et al. Avapritinib in advanced PDGFRA D842V-mutant gastrointestinal stromal tumour (NAVIGATOR): a multicentre open-label, phase 1 trial. Lancet Oncol. 21, 935–946 (2020).
The first study demonstrating the activity of avapritinib in PDGFRA-mutated GIST.
⦁ Grunewald, S. et al. Resistance to avapritinib in PDGFRA-driven GIST is caused by secondary mutations in the PDGFRA kinase domain.
Cancer Discov. https://doi.org/10.1158/2159-8290. CD-20-0487 (2020).
⦁ Blay, J. Y. et al. Ripretinib in patients with advanced gastrointestinal stromal tumours (INVICTUS):
a double-blind, randomised, placebo-controlled, phase 3 trial. Lancet Oncol. https://doi.org/10.1016/ S1470-2045(20)30168-6 (2020).
The randomized trial showing the activity of ripretinib in advanced GIST progressing after imatinib, sunitinib and regorafenib.
⦁ Italiano, A. et al. SDHA loss of function mutations in a subset of young adult wild-type gastrointestinal stromal tumors. BMC Cancer 12, 408 (2012).
⦁ Oudijk, L. et al. SDHA mutations in adult and pediatric wild-type gastrointestinal stromal tumors. Mod. Pathol. 26, 456–463 (2013).
⦁ Pantaleo, M. A. et al. Analysis of all subunits, SDHA, SDHB, SDHC, SDHD, of the succinate dehydrogenase complex in KIT/PDGFRA wild-type GIST. Eur. J. Hum. Genet. 22, 32–39 (2014).
⦁ Doyle, L. A., Nelson, D., Heinrich, M. C., Corless, C. L. & Hornick, J. L. Loss of succinate dehydrogenase subunit B (SDHB) expression is limited to a distinctive subset of gastric wild-type gastrointestinal stromal tumours: a comprehensive genotype-phenotype correlation study. Histopathology 61, 801–809 (2012).
⦁ Gill, A. J. et al. Immunohistochemistry for SDHB divides gastrointestinal stromal tumors (GISTs) into 2 distinct types. Am. J. Surg. Pathol. 34, 636 (2010).
⦁ Boikos, S. A. & Stratakis, C. A. The genetic landscape of gastrointestinal stromal tumor lacking KIT
and PDGFRA mutations. Endocrine 47, 401–408
(2014).
⦁ Tarn, C. et al. Insulin-like growth factor 1 receptor is a potential therapeutic target for gastrointestinal stromal tumors. Proc. Natl Acad. Sci. USA 105, 8387–8392 (2008).
⦁ Nannini, M., Biasco, G., Astolfi, A., Urbini, M.
& Pantaleo, M. A. Insulin-like growth factor (IGF) system and gastrointestinal stromal tumours (GIST): present and future. Histol. Histopathol. 29, 167–175 (2014).
⦁ von Mehren, M. et al. Linsitinib (OSI-906) for the treatment of adult and pediatric wild-type gastrointestinal stromal tumors, a SARC phase II study. Clin. Cancer Res. 26, 1837–1845 (2020).
⦁ Andersson, J. et al. NF1-associated gastrointestinal stromal tumors have unique clinical, phenotypic, and genotypic characteristics. Am. J. Surg. Pathol. 29, 1170–1176 (2005).
⦁ Miettinen, M., Fetsch, J. F., Sobin, L. H. & Lasota, J. Gastrointestinal stromal tumors in patients with neurofibromatosis 1: a clinicopathologic and molecular genetic study of 45 cases. Am. J. Surg. Pathol. 30, 90–96 (2006).
⦁ Agaimy, A., Vassos, N. & Croner, R. S. Gastrointestinal manifestations of neurofibromatosis type 1 (Recklinghausen’s disease): clinicopathological spectrum with pathogenetic considerations.
Int. J. Clin. Exp. Pathol. 5, 852–862 (2012).
⦁ Burgoyne, A. M. et al. Duodenal-jejunal flexure gi stromal tumor frequently heralds somatic NF1 and notch pathway mutations. JCO Precis Oncol. 17, 00014 (2017).
⦁ Mühlenberg, T. et al. KIT-Dependent and
KIT-independent genomic heterogeneity of resistance in gastrointestinal stromal tumors – TORC1/2 inhibition as salvage strategy. Mol. Cancer 18, 1985–1996 (2019).
⦁ Agaimy, A. et al. V600E BRAF mutations are alternative early molecular events in a subset of KIT/ PDGFRA wild-type gastrointestinal stromal tumours. J. Clin. Pathol. 62, 613–616 (2009).
⦁ Huss, S. et al. Clinicopathological and molecular features of a large cohort of gastrointestinal stromal tumors (GISTs) and review of the literature: BRAF mutations in KIT/PDGFRA wild-type GISTs are rare events. Hum. Pathol. 62, 206–214 (2017).
⦁ Falchook, G. S. et al. BRAF mutant gastrointestinal stromal tumor: first report of regression with BRAF inhibitor dabrafenib (GSK2118436) and whole
exomic sequencing for analysis of acquired resistance.
Oncotarget 4, 310–315 (2013).
⦁ Shi, E. et al. FGFR1 and NTRK3 actionable alterations in “wild-type” gastrointestinal stromal tumors. J. Transl. Med. 14, 339 (2016).
⦁ Ricci, R. E. PDGFRA-mutant syndrome. Mod. Pathol.
28, 954–964 (2015).
⦁ Manley, P. N. et al. Familial PDGFRA-mutation syndrome: somatic and gastrointestinal phenotype. Hum. Pathol. 76, 52–57 (2018).
⦁ McWhinney, S. R., Pasini, B. & Stratakis, C. A. International carney triad and carney-stratakis syndrome consortium. familial gastrointestinal stromal tumors and germ-line mutations. N. Engl. J. Med. 357, 1054–1056 (2007).
⦁ Agaimy, A. et al. Multiple sporadic gastrointestinal stromal tumors (GISTs) of the proximal stomach are caused by different somatic KIT mutations suggesting a field effect. Am. J. Surg. Pathol. 32, 1553–1559 (2008).
⦁ Shen, Y. Y. et al. Clinicopathologic characteristics, diagnostic clues, and prognoses of patients with multiple sporadic gastrointestinal stromal tumors:
a case series and review of the literature. Diagn. Pathol.
15, 56 (2020).
⦁ Heinrich, M. C. et al. Molecular correlates of
imatinib resistance in gastrointestinal stromal tumors.
J. Clin. Oncol. 24, 4764–4774 (2006).
⦁ Nishida, T. et al. Secondary mutations in the kinase domain of the KIT gene are predominant in imatinib- resistant gastrointestinal stromal tumor. Cancer Sci. 99, 799–804 (2008).
⦁ Serrano, C. et al. Complementary activity of tyrosine kinase inhibitors against secondary kit mutations in imatinib-resistant gastrointestinal stromal tumours. Br. J. Cancer 120, 612–620 (2019).
⦁ Demetri, G. D. et al. Efficacy and safety of sunitinib in patients with advanced gastrointestinal stromal tumour after failure of imatinib: a randomised controlled trial. Lancet 368, 1329–1338 (2006). The randomized trial demonstrating the activity of sunitinib in advanced GIST progressing after imatinib.
⦁ Demetri, G. D. et al. GRID study investigators. Efficacy and safety of regorafenib for advanced gastrointestinal stromal tumours after failure of imatinib and sunitinib (GRID): an international, multicentre, randomised, placebo-controlled, phase 3 trial. Lancet 381, 295–302 (2014).
The randomized trial demonstrating the activity of regorafenib in advanced GIST progressing after sunitinib.
⦁ Yeh, C. N. et al. A phase II trial of regorafenib in patients with metastatic and/or a unresectable gastrointestinal stromal tumor harboring secondary mutations of exon 17. Oncotarget 8, 44121–44130 (2017).
⦁ Gebreyohannes, Y. K. et al. Robust activity of avapritinib, potent and highly selective inhibitor of mutated KIT,
in patient-derived xenograft models of gastrointestinal stromal tumors. Clin. Cancer Res. 25, 609–618 (2019).
⦁ Heinrich, M. C. et al. Clinical activity of avapritinib in
≥fourth-line (4L+) and PDGFRA exon 18 gastrointestinal stromal tumors (GIST). J. Clin. Oncol. https://doi.org/ 10.1200/JCO.2019.37.15_suppl.11022 (2020).
⦁ Desai, J. et al. Clonal evolution of resistance to imatinib in patients with metastatic gastrointestinal stromal tumors. Clin. Cancer Res. 13, 5398–5405 (2007).
⦁ Wardelmann, E. et al. Polyclonal evolution of multiple secondary KIT mutations in gastrointestinal stromal tumors under treatment with imatinib mesylate.
Clin. Cancer Res. 12, 1743–1749 (2006).
⦁ Serrano, C. et al. Clinical value of next generation sequencing of plasma cell-free DNA in gastrointestinal stromal tumors. BMC Cancer 20, 99 (2020).
⦁ Coe, T. M. et al. Population-based epidemiology and mortality of small malignant gastrointestinal
stromal tumors in the USA. J. Gastrointest. Surg. 20, 1132–1140 (2016).
⦁ Sepe, P. S., Moparty, B., Pitman, M. B., Saltzman, J. R. & Brugge, W. R. EUS-guided FNA for the diagnosis of GI stromal cell tumors: sensitivity and cytologic yield. Gastrointest. Endosc. 70, 254–261 (2009).
⦁ van Houdt, W. J. et al. Oncological outcome after diagnostic biopsies in gastrointestinal stromal tumors: a retrospective cohort study. Ann Surg. https://doi.org/10.1097/SLA.0000000000003744 (2019).
⦁ Eriksson, M. et al. Needle biopsy through the abdominal wall for the diagnosis of gastrointestinal
stromal tumour – does it increase the risk for tumour cell seeding and recurrence? Eur. J. Cancer 59, 128–133 (2016).
⦁ Miettinen, M. & Lasota, J. Gastrointestinal stromal tumors: Pathology and prognosis at different sites. Semin. Diagn. Pathol. 23, 70–83 (2006).
⦁ Miettinen, M. & Lasota, J. KIT (CD117): a review on expression in normal and neoplastic tissues,
and mutations and their clinicopathologic correlation.
Appl. Immunohistochem. Mol. Morphol. 13, 205–220
(2005).
⦁ Miettinen, M., Wang, Z. F. & Lasota, J. DOG1 antibody in the differential diagnosis of gastrointestinal stromal tumors: a study of 1840 cases. Am. J. Surg. Pathol. 33, 1401–1408 (2009).
⦁ West, R. B. et al. The novel marker, DOG1, is expressed ubiquitously in gastrointestinal stromal tumors irrespective of KIT or PDGFRA mutation status. Am. J. Pathol. 165, 107–113 (2004).
⦁ Liegl, B., Hornick, J. L., Corless, C. L. & Fletcher, C. D. M. 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 (2009).
⦁ Nishida, T. et al. Adherence to the guidelines and the pathological diagnosis of high-risk gastrointestinal stromal tumors in the real world. Gastric Cancer 23, 118–125 (2020).
⦁ Rossi, G. et al. PDGFR expression in differential diagnosis between KIT-negative gastrointestinal stromal tumours and other primary soft-tissue tumours of the gastrointestinal tract. Histopathology 46, 522–531 (2005).
⦁ Miselli, F. et al. PDGFRA immunostaining can help in the diagnosis of gastrointestinal stromal tumors. Am. J. Surg. Pathol. 32, 738–743 (2008).
⦁ Peterson, M. R., Piao, Z., Weidner, N. & Yi, E. S.
Strong PDGFRA positivity is seen in GISTs but
not in other intra-abdominal mesenchymal tumors: Immunohistochemical and mutational analyses.
Appl. Immunohistochem. Mol. Morphol. 14, 390–396
(2006).
⦁ Koo, D. H. et al. Asian consensus guidelines for the diagnosis and management of gastrointestinal stromal tumor. Cancer Res. Treat. 48, 1155–1166 (2016).
⦁ Banerjee, S. et al. Cost-effectiveness analysis of genetic testing and tailored first-line therapy for patients with metastatic gastrointestinal stromal tumors. JAMA Netw. Open 3, e2013565 (2020).
⦁ Hong, D. S. et al. Larotrectinib in patients with TRK fusion-positive solid tumours: a pooled analysis of three phase 1/2 clinical trials. Lancet Oncol. 21, 531–540 (2020).
⦁ Fletcher, C. D. et al. Diagnosis of gastrointestinal stromal tumors: a consensus approach. Hum. Pathol. 33, 459–465 (2002).
This paper describes the first prognostic classification of localized GIST.
⦁ Miettinen, M., Sobin, L. H. & Lasota, J. Gastrointestinal stromal tumors of the stomach: a clinicopathologic, immunohistochemical, and molecular genetic study
of 1765 cases with long-term follow-up. Am. J. Surg. Pathol. 29, 52–68 (2005).
⦁ Miettinen, M., Makhlouf, H., Sobin, L. H. & Lasota, J. Gastrointestinal stromal tumors of the jejunum and ileum: a clinicopathologic, immunohistochemical, and molecular genetic study of 906 cases before imatinib with long-term follow-up. Am. J. Surg. Pathol. 30, 477–489 (2006).
⦁ Gold, J. S. et al. Development and validation of a prognostic nomogram for recurrence-free survival after complete surgical resection of localised primary gastrointestinal stromal tumour: a retrospective analysis. Lancet Oncol. 10, 1045–1052 (2009).
⦁ Takahashi, T. et al. An enhanced risk-group stratification system for more practical prognostication of clinically malignant gastrointestinal stromal tumors. Int. J. Clin. Oncol. 12, 369–374 (2007).
⦁ Rutkowski, P. et al. Risk criteria and prognostic factors for predicting recurrences after resection of primary gastrointestinal stromal tumor. Ann. Surg. Oncol. 14, 2018–2027 (2007).
⦁ Joensuu, H. Risk stratification of patients diagnosed with gastrointestinal stromal tumor. Hum. Pathol. 39, 1411–1419 (2008).
⦁ Joensuu, H. et al. Risk of recurrence of gastrointestinal stromal tumour after surgery: an analysis of pooled population-based cohorts. Lancet Oncol. 13, 265–274 (2012).
This paper describes one of the key prognostic classifications of localized GIST.

PRIMER

⦁ Lartigue, L. et al. Genomic index predicts clinical outcome of intermediate-risk gastrointestinal stromal tumours, providing a new inclusion criterion for imatinib adjuvant therapy. Eur. J. Cancer 51, 75–83 (2015).
⦁ Chibon, F. et al. Validated prediction of clinical outcome in sarcomas and multiple types of cancer on the basis of a gene expression signature related to genome complexity. Nat. Med. 16, 781–787 (2010).
⦁ Brahmi, M. et al. KIT exon 10 variant (c.1621 A>C) single nucleotide polymorphism as predictor of GIST patient outcome. BMC Cancer 15, 780 (2015).
⦁ Foster, R. et al. Association of paediatric mastocytosis with a polymorphism resulting in an amino acid substitution (M541L) in the transmembrane domain of c-KIT. Br. J. Dermatol. 159, 1160–1169 (2008).
⦁ Schaefer, I. M., Mariño-Enríquez, A. & Fletcher, J. A. What is new in gastrointestinal stromal tumor?
Adv. Anat. Pathol. 24, 259–267 (2017).
⦁ Yamamoto, K. et al. Impact of the Japanese gastric cancer screening system on treatment outcomes
in gastric gastrointestinal stromal tumor (GIST): an analysis based on the GIST registry. Ann. Surg. Oncol. 22, 232–239 (2015).
⦁ Gronchi, A. et al. Quality of surgery and outcome in localized gastrointestinal stromal tumors treated
within an international intergroup randomized clinical trial of adjuvant imatinib. JAMA Surg. 155, e200397 (2020).
⦁ Weldon, C. B. et al. Surgical management of wild-type gastrointestinal stromal tumors: a report from the national institutes of health pediatric and wildtype GIST clinic. J. Clin. Oncol. 35, 523–528 (2017).
⦁ Heinrich, M. C. et al. Inhibition of c-kit receptor tyrosine kinase activity by STI 571, a selective tyrosine kinase inhibitor. Blood 96, 925–932 (2000).
⦁ van Oosterom, A. et al. Safety and efficacy of imatinib (STI571) in metastatic gastrointestinal stromal tumours: a phase I study. Lancet 358, 1421–1423 (2001).
⦁ Demetri, G. et al. Efficacy and safety of imatinib mesylate in advanced gastrointestinal stromal tumors. N. Engl. J. Med. 347, 472–480 (2002).
One of two key randomized trials to establish the activity of imatinib in advanced GIST.
⦁ Verweij, J. et al. Progression-free survival in gastrointestinal stromal tumours with high-dose imatinib: randomised trial. Lancet 364, 1127–1134 (2004).
One of two key randomized trials to establish the activity of imatinib in advanced GIST.
⦁ Blanke, C. D. et al. Phase III randomized, intergroup trial assessing imatinib mesylate at two dose
levels in patients with unresectable or metastatic gastrointestinal stromal tumors expressing the kit receptor tyrosine kinase: S0033. J. Clin. Oncol. 26, 626–632 (2008).
⦁ Blanke, C. D. et al. Long-term results from a randomized phase II trial of standard- versus higher-dose imatinib mesylate for patients with unresectable or metastatic gastrointestinal stromal
tumors expressing KIT. J. Clin. Oncol. 26, 620–625 (2008).
⦁ Blay, J. Y. et al. Nilotinib versus imatinib as first-line therapy for patients with unresectable or metastatic gastrointestinal stromal tumours (ENESTg1):
a randomised phase 3 trial. Lancet Oncol. 16, 550–560 (2015).
⦁ Choi, H. et al. Correlation of computed tomography and positron emission tomography in patients with metastatic gastrointestinal stromal tumor treated
at a single institution with imatinib mesylate: proposal of new computed tomography response criteria.
J. Clin. Oncol. 25, 1753–1759 (2007).
⦁ Blay, J. Y. et al. Prospective multicentric randomized phase III study of imatinib in patients with advanced gastrointestinal stromal tumors comparing interruption versus continuation of treatment beyond 1 year: the French Sarcoma Group. J. Clin. Oncol. 25, 1107–1113 (2007).
⦁ Le Cesne, A. et al. Discontinuation of imatinib in patients with advanced gastrointestinal stromal tumours after 3 years of treatment: an open-label multicentre randomised phase 3 trial. Lancet Oncol. 11, 942–949 (2010).
⦁ Patrikidou, A. et al. French Sarcoma Group. Influence of imatinib interruption and rechallenge on the residual disease in patients with advanced GIST: results of the BFR14 prospective French Sarcoma Groups randomize, phase III trial. Ann. Oncol. 24, 1087–1093 (2013).
⦁ Raut, C. P. et al. Efficacy and tolerability of 5-year adjuvant imatinib treatment for patients with resected intermediate- or high-risk primary gastrointestinal stromal tumor: the PERSIST-5 clinical trial. JAMA Oncol. 4, e184060 (2018).
⦁ Demetri, G. D. et al. Imatinib plasma levels are correlated with clinical benefit in patients with unresectable/metastatic gastrointestinal stromal tumors. J. Clin. Oncol. 27, 3141–3147 (2009).
⦁ Blesius, A. et al. Neoadjuvant imatinib in patients with locally advanced non metastatic GIST in the prospective BFR14 trial. BMC Cancer 11, 72 (2011).
⦁ Rutkowski, P. et al. Neoadjuvant imatinib in locally advanced gastrointestinal stromal tumors (GIST): the EORTC STBSG experience. Ann. Surg. Oncol. 20, 2937–2943 (2013).
⦁ Wang, D. et al. Phase II trial of neoadjuvant/adjuvant imatinib mesylate for advanced primary and metastatic/recurrent operable gastrointestinal stromal tumors: long-term follow-up results of Radiation Therapy Oncology Group 0132. Ann. Surg. Oncol. 19, 1074–1080 (2012).
⦁ Holdsworth, C. H. et al. CT and PET: early prognostic indicators of response to imatinib mesylate in patients with gastrointestinal stromal tumor. Am. J. Roentgenol. 189, W324–W330 (2007).
⦁ Bouchet, S. et al. Relationship between imatinib trough concentration and outcomes in the treatment of advanced gastrointestinal stromal tumours in a real-life setting. Eur. J. Cancer 57, 31–38 (2016).
⦁ Demetri, G. D. et al. Molecular target modulation, imaging, and clinical evaluation of gastrointestinal stromal tumor patients treated with sunitinib malate after imatinib failure. Clin. Cancer Res. 15, 5902–5909 (2009).
⦁ Janeway, K. A. et al. Sunitinib treatment in pediatric patients with advanced GIST following failure
of imatinib. Pediatr. Blood Cancer 52, 767–771 (2009).
⦁ Demetri, G. D. et al. Complete longitudinal analyses of the randomized, placebo-controlled, phase III trial of sunitinib in patients with gastrointestinal stromal tumor following imatinib failure. Clin. Cancer Res. 18, 3170–3179 (2012).
⦁ George, S. et al. Clinical evaluation of continuous daily dosing of sunitinib malate in patients with advanced gastrointestinal stromal tumour after imatinib failure. Eur. J. Cancer 45, 1959–1968 (2009).
⦁ George, S. et al. Efficacy and safety of regorafenib in patients with metastatic and/or unresectable GI stromal tumor after failure of imatinib and
sunitinib: a multicenter phase II trial. J. Clin. Oncol.
30, 2401–2407 (2012).
⦁ Ben-Ami, E. et al. Long-term follow-up of the multicenter phase II trial of regorafenib in patients with metastatic and/or unresectable GI stromal tumor after failure of standard tyrosine kinase inhibitor therapy. Ann. Oncol. 27, 1794–1799 (2016).
⦁ Drilon, A. et al. Entrectinib in ROS1 fusion-positive non-small-cell lung cancer: integrated analysis of three phase 1-2 trials. Lancet Oncol. 21, 261–270 (2019).
⦁ Wilding, C. P., Loong, H. H., Huang, P. H. & Jones, R. L. Tropomyosin receptor kinase inhibitors in the management of sarcomas. Curr. Opin. Oncol. 32, 307–313 (2020).
⦁ Farag, S., Smith, M. J., Fotiadis, N., Constantinidou, A. & Jones, R. L. Revolutions in treatment options
in gastrointestinal stromal tumours (GISTs): the latest updates. Curr. Treat. Options Oncol. 21, 55 (2020).
⦁ Janku, F. et al. Switch control inhibition of KIT and PDGFRA in patients with advanced gastrointestinal stromal tumor: a phase I study of ripretinib.
J. Clin. Oncol. 38, 3294–3303 (2020).
⦁ Serrano, C. et al. Efficacy and safety of ripretinib as
≥4th-line therapy for patients with gastrointestinal stromal tumor (GIST) following crossover from placebo: analyses from INVICTUS. Ann. Oncol. 31 (Suppl. 3), S236 (2020).
⦁ Nemunaitis, J. et al. Intrigue: phase III study of ripretinib versus sunitinib in advanced gastrointestinal stromal tumor after imatinib. Future Oncol. 16, 4251–4264 (2020).
⦁ Heinrich, M. C. et al. Crenolanib inhibits the drug- resistant PDGFRA D842V mutation associated with imatinib-resistant gastrointestinal stromal tumors. Clin. Cancer Res. 18, 4375–4384 (2012).
⦁ von Mehren, M. et al. Dose escalating study of crenolanib besylate in advanced GIST patients with
PDGFRA D842V activating mutations. J. Clin. Oncol.
34, Abstr. 11010 (2016).
⦁ Reichardt, P. et al. Phase III study of nilotinib versus best supportive care with or without a TKI in patients with gastrointestinal stromal tumours resistant to or intolerant of imatinib and sunitinib. Ann. Oncol. 23, 1680–1687 (2012).
⦁ Mir, O. et al. PAZOGIST study group of the French Sarcoma Groupe–Groupe d’Etude des Tumeurs Osseuses (GSF-GETO). Pazopanib plus best supportive care versus best supportive care alone in advanced gastrointestinal stromal tumours resistant to imatinib and sunitinib (PAZOGIST): a randomised, multicentre, open-label phase 2 trial. Lancet Oncol. 17, 632–641 (2016).
⦁ Mir, O. et al. Impact of concomitant administration of gastric acid-suppressive agents and pazopanib on outcomes in soft-tissue sarcoma patients treated within the EORTC 62043/62072 trials. Clin. Cancer Res. 25, 1479–1485 (2019).
⦁ Adenis, A. et al. Masitinib in advanced gastrointestinal stromal tumor (GIST) after failure of imatinib:
a randomized controlled open-label trial. Ann. Oncol.
25, 1762–1769 (2014).
⦁ Montemurro, M. et al. Long-term outcome of dasatinib first-line treatment in gastrointestinal stromal tumor: a multicenter, 2-stage phase 2 trial (Swiss Group for Clinical Cancer Research 56/07). Cancer 124, 1449–1454 (2018).
⦁ Zhou, Y. et al. A prospective multicenter phase II study on the efficacy and safety of dasatinib in the treatment of metastatic gastrointestinal stromal tumors failed
by imatinib and sunitinib and analysis of NGS in peripheral blood. Cancer Med. 9, 6225–6233 (2020).
⦁ Joensuu, H. et al. Dovitinib in patients with gastrointestinal stromal tumour refractory and/or intolerant to imatinib. Br. J. Cancer 117, 1278–1285 (2017).
⦁ Schöffski, P. et al. Activity and safety of the
multi-target tyrosine kinase inhibitor cabozantinib in patients with metastatic gastrointestinal stromal tumour after treatment with imatinib and sunitinib:
European Organisation for Research and Treatment
of Cancer phase II trial 1317 ‘CaboGIST’. Eur. J. Cancer
134, 62–74 (2020).
⦁ Raut, C. P. et al. Surgical management of advanced gastrointestinal stromal tumors after treatment with targeted systemic therapy using kinase inhibitors.
J. Clin. Oncol. 24, 2325–2331 (2006).
⦁ Bauer, S. et al. Long-term follow-up of patients with GIST undergoing metastasectomy in the era of imatinib – analysis of prognostic factors
(EORTC-STBSG collaborative study). Eur. J. Surg. Oncol. 40, 412–419 (2014).
⦁ Du, C. Y. et al. Is there a role of surgery in patients with recurrent or metastatic gastrointestinal stromal tumours responding to imatinib: a prospective randomised trial in China. Eur. J. Cancer 50, 1772–1778 (2014).
⦁ Mussi, C. et al. Therapeutic consequences from molecular biology for gastrointestinal stromal
tumor patients affected by neurofibromatosis type 1.
Clin. Cancer Res. 14, 4550–4555 (2008).
⦁ Kang, Y. K. et al. Resumption of imatinib to control metastatic or unresectable gastrointestinal stromal tumours after failure of imatinib and sunitinib (RIGHT): a randomised, placebo-controlled, phase 3 trial. Lancet Oncol. 14, 1175–1182 (2013).
⦁ Joensuu, H. et al. Radiotherapy for GIST progressing during or after tyrosine kinase inhibitor therapy:
a prospective study. Radiother. Oncol. 116, 233–238
(2015).
⦁ Cao, G., Li, J., Shen, L. & Zhu, X. Transcatheter arterial chemoembolization for gastrointestinal stromal tumors with liver metastases. World J. Gastroenterol. 18, 6134–6140 (2012).
⦁ Yamanaka, T. et al. Radiofrequency ablation for
liver metastasis from gastrointestinal stromal tumor.
J. Vasc. Interv. Radiol. 24, 341–346 (2013).
⦁ Hakimé, A. et al. A role for adjuvant RFA in managing hepatic metastases from gastrointestinal stromal tumors (GIST) after treatment with targeted systemic therapy using kinase inhibitors. Cardiovasc. Intervent Radiol. 37, 132–139 (2014).
⦁ Rathmann, N. et al. Radioembolization in patients with progressive gastrointestinal stromal tumor liver metastases undergoing treatment with tyrosine kinase
inhibitors. J. Vasc. Interv. Radiol. 26, 231–238 (2015).
⦁ Poort, H. et al. Prevalence, impact, and correlates of severe fatigue in patients with gastrointestinal stromal tumors. J. Pain Symptom Manage. 52, 265–271 (2016).

PRIMER

⦁ Joensuu, H., Trent, J. C. & Reichardt, P. Practical management of tyrosine kinase inhibitor-associated side effects in GIST. Cancer Treat. Rev. 37, 75–88 (2011).
⦁ Wang, Y. et al. Adherence to adjuvant imatinib therapy in patients with gastrointestinal stromal tumor in clinical practice: a cross-sectional study. Chemotherapy 64, 197–204 (2019).
⦁ Chae, H., Ryu, M. H., Ma, J., Beck, M. & Kang, Y. K. Impact of L-carnitine on imatinib-related muscle cramps in patients with gastrointestinal stromal tumor. Invest. N. Drugs 38, 493–499 (2020).
⦁ Heinrich, M. C. et al. Quality of life (QoL) and self-reported function with ripretinib in ≥4th-line
therapy for patients with gastrointestinal stromal tumors (GIST): analyses from Invictus. J. Clin. Oncol. 38 (Suppl. 15), 11535–11535 (2020).
⦁ Toulmonde, M. et al. Use of PD-1 targeting, macrophage infiltration, and IDO pathway activation in sarcomas: a phase 2 clinical trial. JAMA Oncol. 4, 93–97 (2018).
⦁ Balachandran, V. P. et al. Imatinib potentiates antitumor T cell responses in gastrointestinal stromal tumor through the inhibition of Ido. Nat. Med. 17, 1094–1100 (2011).
⦁ Petitprez, F. et al. B cells are associated with survival and immunotherapy response in sarcoma. Nature 577, 556–560 (2020).
⦁ Banks, E. et al. Discovery and pharmacological characterization of AZD3229, a potent KIT/PDGFRα inhibitor for treatment of gastrointestinal stromal tumors. Sci. Transl. Med. 12, eaaz2481 (2020).
⦁ Grellety, T. et al. Clinical activity of regorafenib in PDGFRA-mutated gastrointestinal stromal tumor. Future Sci. OA 1, FSO33 (2015).

Acknowledgements
T.N. is supported by a Grant-in-Aid (19H03722) for Scientific Research from the Japanese Ministry of Education, Culture, Sports, Science and Technology, and by a Grant (31-A-14) from the National Cancer Center Research and Development Fund. J.-Y.B. holds grants from NetSARC (INCA & DGOS) and RREPS (INCA & DGOS), RESOS (INCA & DGOS), LYRICAN (INCA-DGOS-INSERM 12563), Association DAM’s, Eurosarc (FP7-278742), Fondation ARC, Infosarcome, InterSARC (INCA), LabEx DEvweCAN (ANR-10-LABX0061), PIA Institut Convergence François Rabelais PLAsCAN (PLASCAN, 17-CONV-0002), La Ligue de L’Ain contre le Cancer, La Ligue contre le Cancer, and EURACAN (EC 739521), RHU4 DEPGYN (ANR-18-RHUS-0009).
Author contributions
Introduction (J.-Y.B.); Epidemiology (J.-Y.B.); Mechanisms/ pathophysiology (T.N.); Diagnosis, screening and prevention (Y.-K.K., T.N., M.v.M.); Management (Y.-K.K., T.N., M.v.M.);
Quality of life (J.-Y.B.); Outlook (J.-Y.B., Y.-K.K., T.N., M.v.M.).

Competing interests
J.-Y.B.: research support and honoraria from Novartis, GSK, Bayer, Roche, Deciphera and Ignyta. T.N.: honoraria from Pfizer, Novartis, Bayer, Taiho, Eli Lilly outside the submitted work. M.V.M.: honoraria from Deciphera, Blueprint, Exelexis; research support from Novartis. Y.-K.K. declares no competing interests.

Peer review information
Nature Reviews Disease Primers thanks M. Heinrich,
P. Hohenberger, J. Sicklick and the other, anonymous, reviewer(s) for their contribution to the peer review of this work.

Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

© Springer Nature Limited 2021