Clinicopathological Features and Prognosis of Primary GISTs with Tumor Rupture in the Real World

Clinicopathological Features and Prognosis of Primary GISTs with Tumor Rupture in the Real World Ann Surg Oncol (2018) 25:1961–1969 https://doi.org/10.1245/s10434-018-6505-7 OR IG INA L A R TIC L E – BO NE AN D S OFT T IS SU E S A RCOM A S Clinicopathological Features and Prognosis of Primary GISTs with Tumor Rupture in the Real World 1,2,3 4,5 6 2,3 7 8 Toshirou Nishida , Haruhiko Cho , Seiichi Hirota , Toru Masuzawa , Gaku Chiguchi , Toshimasa Tsujinaka , and the Kinki GIST Study Group 1 2 Department of Surgery, National Cancer Center Hospital, Tokyo, Japan; Department of Surgery, Osaka Police Hospital, 3 4 Osaka, Japan; Department of Surgery, Kansai Rosai Hospital, Amagasaki, Japan; Department of Gastrointestinal Surgery, Kanagawa Cancer Center, Yokohama, Japan; Department of Surgery, Tokyo Metropolitan Cancer and Infectious Disease Center Komagome Hospital, Tokyo, Japan; Department of Surgical Pathology, Hyogo College of Medicine, 7 8 Nishinomiya, Japan; Department of Gastroenterology, Yokohama Rosai Hospital, Yokohama, Japan; Department of Gastrointestinal Surgery, Kaizuka City Hospital, Kaizuka, Japan ABSTRACT Recurrence rates were higher and median recurrence-free Background. Patients with ruptured gastrointestinal stro- survival (RFS) and overall survival (OS) were shorter with mal tumor (GIST) are recommended for imatinib adjuvant ruptured than nonruptured GIST. Tumor rupture was one of therapy; however, their clinicopathological features and the independent prognostic factors for RFS, but not OS, prognosis in the era of imatinib are unknown. according to multivariate analysis. Patients and Methods. The study cohort included 665 Conclusions. Ruptured GISTs were symptomatic larger patients with histologically proven primary GISTs who tumors with high mitotic activity, frequent relapse, and underwent R0 or R1 surgery between 2003 and 2007; the shorter RFS. Tumor rupture was an independent prognostic validation cohort included 182 patients between 2000 and factor for RFS, but not for OS, in the era of imatinib. 2014. The definitions of tumor rupture in the study inclu- ded perforation at tumor site, tumor fracture, piecemeal Gastrointestinal stromal tumor (GIST) is a potentially resection including open biopsy, and macroscopic injuries to the pseudocapsule. malignant tumor and the most frequent sarcoma of the gastrointestinal tract. Currently, complete resection (R0) of Results. Tumor rupture occurred in 21 (3.2%) of 665 and 5 (2.9%) of 182 patients in the study and validation cohort, primary GIST offers the only potentially permanent cure for GIST patients. After complete resection, 60% of respectively. Ruptured GISTs were more symptomatic, patients with localized GISTs are cured, while the other were larger in size, and had higher mitotic count than 1–3 nonruptured GISTs but were not associated with tumor 40% experience disease recurrence during follow-up. Prognostic factors for recurrence after R0 surgery have location or laparoscopic surgery. GISTs with intraoperative rupture had clinicopathological features and prognostic been rigorously investigated, and four factors are recog- nized as independent prognostic factors: tumor size (cm), outcomes similar to those with preoperative rupture. mitosis (/50 HPF or/5 mm ), location (gastric versus non- 3–13 gastric), and rupture. Several risk stratifications have 4–8 been proposed based on these four factors. Among these four prognostic factors, tumor rupture is Electronic supplementary material The online version of this article (https://doi.org/10.1245/s10434-018-6505-7) contains the most ominous and is a subjective factor clinically supplementary material, which is available to authorized users. 3,10–12 judged by surgeons. Previous reports indicate that most ruptured GISTs are associated with recurrence during The Author(s) 2018. This article is an open access publication follow-up and that patients with ruptured GIST have sig- First Received: 8 February 2018; nificantly shorter recurrence-free survival (RFS) than those 14–16 Published Online: 11 May 2018 without rupture. The definition of ruptured GIST, T. Nishida however, was not consistent among researchers until a e-mail: tnishida@ncc.go.jp 1962 T. Nishida et al. recent provisional definition was proposed. The term time of diagnosis and individuals whose date of surgery, ‘‘ruptured GIST’’ may include heterogeneous entities; age, gender, and/or outcomes were missing (Supplemen- hence, in some reports, rupture may include tumor pene- tary Fig. 1). This study was reviewed and approved by the tration into the peritoneum (T4a) and microscopic Steering Committee of the Kinki GIST Study Group and by pseudocapsule injuries. Incidence of rupture has been the institutional review board (IRB) of Osaka Police 3,7,10,12,16–18 reported to vary from 2 to 22%. Some reports Hospital and Kanagawa Cancer Center. have suggested that tumor rupture is an independent Tumor rupture was considered to be subjectively 3,5,10,12,16 prognostic factor, whereas others have determined by each surgeon at that time, so we conducted a 19,20 disagreed. retrospective survey of the concept of rupture by ques- Recent clinical trials show that patients at high risk of tionnaire. Questionnaire entries included perforation at the recurrence may benefit from 3 years of adjuvant therapy tumor site, tumor fracture with blood-tinged ascites, with imatinib (Gleevec, Novartis Pharmaceuticals, Basel, piecemeal resection during surgery (including open 17,18 Switzerland) after R0 resection. It may be argued that biopsy), and macroscopic injuries to the pseudocapsule patients with ruptured GIST require prolonged adjuvant exposing tumor cells into the peritoneal cavity (Supple- therapy for longer than 3 years. Few studies have examined mentary Table 1). These were similar to the definitions of which tumors are more likely to rupture or which surgical rupture proposed by Holmebakk et al. Completeness of procedures, open or laparoscopic, may potentially increase surgical resection was assessed by local surgeons as fol- the risk of rupture. In this retrospective study, we ana- lows: R0, no detectable residual tumor; R1, microscopic lyzed real-world data obtained from two registry studies residual tumor; R2, macroscopic residual tumor. Patients conducted in Japan and clarified the clinicopathological with R2 resection were excluded from the analysis. features and prognostic effects of rupture in the era of imatinib. Statistical Analysis Statistical analyses were performed using the Chi PATIENTS AND METHODS squared test, Fisher’s exact test, Student’s t test, and Mann– Study Design Whitney U test. RFS was calculated from date of surgery to date of first tumor recurrence or date of death, excluding The study included both study and validation cohorts. living patients without recurrence at time of data collec- The former data were obtained from the GIST registry tion. Overall survival (OS) was calculated from date of study conducted by the Kinki GIST Study Group, and the surgery to date of any death, excluding living patients. RFS latter from the Kanagawa registry study. The GIST registry and OS were compared between the groups using the designed by the Kinki GIST Study Group was reported Kaplan–Meier life-table method with the log-rank test. Cumulative recurrence probability was estimated by previously. This retrospective observational study was designed to collect clinical and pathological data of cumulative incidence competing risk analysis as described previously. A forward stepwise Cox proportional hazards patients with pathologically proven GIST diagnosed in each participating hospital and treated between January model was used for multivariate analysis to identify risk 2003 and December 2007. Data on risk stratifications factors associated with RFS and OS. P values were two- classified according to the modified National Institutes of sided, and P values less than 0.05 were considered sig- Health (NIH) consensus criteria and the Armed Forces nificant. Data were analyzed using the Statistical Package Institute of Pathology (AFIP) criteria were collected, in of IBM SPSS Statistics 25, version 25.0 (IBM Corp., addition to information regarding preoperative and intra- Armonk, NY, USA). operative ruptures. The Kanagawa registry study collected similar patient data retrospectively and prospectively since RESULTS January 2001 to December 2016 from the two hospitals in Kanagawa Prefecture. The hospitals participating in each In total, 665 patients with primary GIST who underwent registry are listed at the end of the manuscript. R0 or R1 surgery were included in the study cohort and 172 Using data obtained from the GIST registry database, patients in the validation cohort. The clinicopathological eligible patients were selected (Supplementary Fig. 1). features of the patients analyzed in the study and validation Eligibility criteria included tumor morphology compatible cohorts are presented in Table 1. The study cohort included with GIST, KIT positivity on immunohistochemistry, and 506 gastric, 119 small intestinal, 26 colorectal, 10 eso- macroscopically complete resection of the primary tumor. phageal, and 4 extra-gastrointestinal GISTs. The study We excluded patients with metastatic or recurrent GIST at sample included 339 males and 326 females with median Features of Ruptured GISTs in Clinical Practice 1963 TABLE 1 Patient Factor Study cohort (N = 665) Validation cohort (N = 172) characteristics Age (Median; years) 66.0 (18–93) 62.5 (17–89) Gender Male 339 (51.0%) 100 (58.1%) Female 326 (49.9%) 72 (41.9%) Cancer association Cancer history 113 (17.0%) 37 (21.5%) No cancer history 536 (80.6%) 135 (78.5%) Unavailable 16 (2.4%) 0 (0%) Location Esophagus 10 (1.5%) 2 (1.2%) b b b Stomach 506 (76.1%) 120 (69.4%) b b b Small intestine 119 (17.9%) 37 (21.5%) Colon and rectum 26 (3.9%) 14 (8.1%) Others 4 (0.6%) 0 (0%) Tumor size (median; cm) 4.0 (0.1–35) 5.0 (1.1–25) Mitosis (median;/50 HPF) 2.6 (0.0–250.0) 5.0 (0.0–250.0) Symptoms at diagnosis Yes 257 (38.6%) No 407 (61.2%) Unavailable 1 (0.2%) Neoadjuvant Yes 9 (1.3%) 15 (8.7%) No 648 (97.4%) 157 (91.3%) Unavailable 8 (1.3%) 0 (0%) Adjuvant Yes 37 (5.6%) 30 (17.4%) No 627 (94.3%) 142 (82.6%) Unavailable 1 (0.1%) 0 (0%) Surgery Open 459 (69.0%) 116 (67.4%) Laparoscopy 200 (30.1%) 51 (29.7%) Local 6 (0.9%) 5 (2.9%) Completeness of surgery R0 661 (99.4%) 162 (94.2%) R1 4 (0.6%) 10 (5.8%) Tumor rupture No 644 (96.8%) 167 (97.1%) Yes 21 (3.2%) 5 (2.9%) Preoperative 12 3 Intraoperative 9 2 Histological type Spindle 538 (80.9%) 83 (48.3%) Epithelioid 22 (3.3%) 2 (1.2%) Mixed 35 (5.3%) 7 (4.1%) Unavailable 70 (10.5%) 80 (46.4%) Recurrence No recurrence 570 (87.5%) 124 (72.1%) Recurrence 95 (12.5%) 47 (27.9%) Estimated 5-year RFS (median ? SE) 78.6 ? 1.8% 72.5 ? 3.9% Overall survival Alive 600 (90.2%) 151 (87.8%) Death 65 (9.8%) 21 (12.2%) Estimated 5-year OS (median ? SE) 91.5 ? 1.2% 92.2 ? 2.3% Median follow-up: 4.67 years for study cohort and 5.12 years for validation cohort RFS recurrence-free survival, OS overall survival, SE standard error Not available in validation cohort One duplicated patient with gastric and small intestinal GISTs in each cohort age of 66 years. Median tumor size was 4.0 cm, and the (N = 12) or intraoperatively (N = 9). In the validation median mitotic count was 2.6/50 HPF; 21 (3.2%) patients cohort, there were 120 gastric, 37 small intestinal, 14 were reported to have tumor rupture, either preoperatively colorectal, and 2 esophageal GISTs. Median age was 1964 T. Nishida et al. 62.5 years, with 100 male and 72 female patients in the nonrupture group (Supplementary Fig. 2). RFS of (Table 1). Median tumor size was 5.0 cm, the median patients with intraoperative ruptured GIST was not differ- mitotic count was 5.0/50 HPF, and only five patients ent from that of patients with preoperative rupture (Fig. 2; (2.9%) had tumor rupture, of whom three had preoperative P = 0.6709). In the study cohort, recurrence in patients and two intraoperative rupture. Most patients in both with ruptured GIST was more frequent in the peritoneum cohorts underwent R0 surgery, by either open or laparo- and local lesions compared with those in patients with scopic procedure. Nearly all patients received imatinib nonruptured GIST (Table 2). Median OS of patients with therapy after relapse. A small fraction of patients with ruptured GIST (6.4 years; 95% CI 5.5–7.3 years) was high-risk features received neoadjuvant (1.3 and 8.7%), significantly shorter than that of those with nonruptured while a relatively larger number of patients received GIST in the study cohort (11.9 years; 95% CI adjuvant therapy (5.6 and 17.4%) in the study and valida- 10.7–13.0 years; P = 0.0218); this was not confirmed in tion cohort, respectively. the validation cohort, likely because of the low statistical A questionnaire to participating surgeons indicated the power and higher rate of imatinib adjuvant therapy. following results (Supplementary Table 1): most surgeons Multivariate analyses using the Cox proportional haz- considered that tumor fracture and perforation, piecemeal ards model indicated that location, tumor size, mitotic resection, open biopsy, and macroscopic injuries to the count, and rupture were independent prognostic factors for pseudocapsule exposing tumor cells represented rupture; in RFS and that age, gender, and mitotic count were inde- contrast, they considered that core and needle biopsy pendent prognostic factors for OS in the study cohort without complications, luminal perforation of tumors, (Table 3). In the validation cohort, independent prognostic peritoneal tumor penetration, and microscopic injuries to factors for recurrence included tumor size, mitotic count, the pseudocapsule did not represent rupture. rupture, and use of adjuvant therapy, and those for OS were Tumor size was larger and mitotic count was higher in gender, tumor size, and mitotic count (Supplementary ruptured GISTs than in nonruptured tumors in the study Table 4). In the combined analysis of both cohorts, tumor cohort (Table 2). Patients with ruptured GIST were more size, mitotic count, location, rupture, and gender were symptomatic at admission, regardless of preoperative or independent prognostic factors for RFS, and age, gender, intraoperative rupture (Table 2, Supplementary Table 3). and mitotic count were independent prognostic factors for Among the nine patients who had neoadjuvant therapy, one OS. patient experienced tumor rupture during surgery. Inci- dence of rupture was not different between open and DISCUSSION laparoscopic surgery in either group. Location (gastric and nongastric) was not correlated with GIST rupture in either This study found that ruptured GIST was seen in nearly the study or validation cohort. There was no significant 3% of primary GISTs, being more symptomatic and difference in terms of location, symptoms, tumor size, exhibiting aggressive features of larger size and higher mitotic count, or recurrence between preoperative and mitotic count compared with nonruptured tumors. The intraoperative rupture (Supplementary Table 2). reported frequency of tumor rupture varies depending on During median follow-up of 4.67 years, there were 95 the study; population-based studies indicated that it was 7 10 3 11 (12.5%) relapses and 65 (9.8%) deaths in the study cohort, less than 10% (1%, 4.0%, 5.9%, and 7.1% ), while in compared with 47 (27.9%) recurrences and 21 (12.2%) clinical trials, it was higher than 10% (20% of high-risk 17 18 deaths in the validation cohort with median follow-up of GISTs, 11% of intermediate- and high-risk GISTs, and 5.12 years. Recurrence was more frequent for patients with 17% ). The true incidence of tumor rupture is speculated ruptured GISTs than those without rupture in both cohorts to be several percent in clinical practice. Ruptured GISTs (Table 2, Supplementary Table 4). Median RFS of patients were shown to be more symptomatic with high-risk fea- with ruptured GIST [2.4 years; 95% confidence interval tures, including larger tumor size and higher mitotic count, (CI) 1.4–3.4 years in the study cohort; P \ 0.0001, and and were treated by emergency surgery in previous, as well 11,14,15 3.2 years in the validation cohort; 95% CI 1.3–5.8 years; as present, studies. P = 0.0392] was significantly shorter than that of patients Tumor rupture occurred both before and during surgery. with nonruptured GIST (8.4 years; 95% CI 8.0–8.8 years, The frequency of preoperative and intraoperative rupture 17,24 and 8.4 years; 95% CI 7.4–9.3 years) in the study and was similar in this registry study. Clinicopathological validation cohort, respectively (Fig. 1). Cumulative inci- features and prognostic outcomes of GISTs with preoper- dence analysis of the study cohort indicated that all events ative rupture were similar to those with intraoperative were recurrence of GIST in the rupture group, whereas rupture, although the sample size was small (Table 2, one-third of events were recurrence of GIST, and deaths Supplementary Table 3). The laparoscopic approach was due to other diseases might account for the other two-thirds Features of Ruptured GISTs in Clinical Practice 1965 TABLE 2 Background of GIST patients with and without tumor rupture (study cohort) Nonruptured (N = 644) Ruptured (N = 21) P value Age (years) 66 (18–93) 68 (55–90) 0.2236 Gender Male 326 (50.6%) 13 (61.9%) 0.3087 Female 318 (49.4%) 8 (38.1%) Primary location Gastric 493 (76.6%) 13 (61.9%) 0.1215 Nongastric 151 (23.4%) 8 (38.1%) Association of cancer No 518 (80.4%) 18 (85.7%) 0.7098 Yes 110 (17.1%) 3 (14.3%) Unavailable 16 (2.5%) 0 (0%) Median tumor size (cm) 4.0 (0.1–35.0) 9.6 (2.6–30.0) 0.0008 Symptoms No 406 (63.0%) 1 (4.7%) \ 0.0001 Yes 237 (36.8%) 20 (95.2%) Unavailable 1 (0.2%) 0 (0%) Neoadjuvant No 629 (97.7%) 19 (90.5%) 0.1561 Yes 8 (1.2%) 1 (5%) Unavailable 7 (1.1%) 1 (5%) Adjuvant therapy No 613 (95.2%) 14 (66.7%) \ 0.0001 Yes 30 (4.7%) 7 (23.3%) Unavailable 1 (0.1%) 0 (0%) Surgery Open 441 (68.5%) 18 (85.7%) 0.2383 Laparoscopic 197 (30.6%) 3 (14.3%) Local 6 (0.9%) 0 (0%) R R0 642 (99.7%) 19 (90.5%) \ 0.0001 R1 2 (0.3%) 2 (9.5%) Median mitosis (/50 HPF) 2.5 (0.0–250) 13.0 (0.0–115) 0.0004 Cell type Spindle 518 (80.4%) 20 (95.2%) 0.3347 Epithelioid 22 (3.4%) 0 (0%) Mixed 35 (5.4%) 0 (0%) Unavailable 69 (10.7%) 1 (4.8%) Median RFS (95% CI; years) 8.4 (8.0–8.9) 2.4 (1.4–3.4) \ 0.0001 Estimated 5-year RFS (median ? SE) 80.7 ? 1.7% 16.4 ? 8.6% Recurrence No 565 (87.7%) 5 (23.8%) \ 0.0001 Yes 79 (12.3%) 16 (76.2%) a b b Recurrence site Liver 53 (67.1%) 6 (37.5%) 0.0108 b b Lung 2 (2.5%) 0 (0%) b b Local 8 (10.1%) 4 (25%) b b Peritoneum 24 (30.4%) 14 (87.5%) Median OS (95% CI; years) 11.9 (10.7–13.0) 6.4 (5.6–7.3) 0.0218 Estimated 5-year OS (median ? SE) 88.9 ? 7.4% 91.6 ? 1.2% Overall survival Alive 584 (90.7%) 16 (76.2%) 0.0452 Dead 60 (9.3%) 5 (23.8%) c c Death due to GIST 21 (35%) 5 (100%) c c Death due to other diseases 39 (65%) 0 (0%) RFS recurrence-free survival, OS overall survival, SE standard error, CI confidence interval Duplicated number % of total recurrence in each group % of total death in each group 1966 T. Nishida et al. Recurrence-free Survival Overall Survival A B Non-ruptured (N=643; estimated RFS=8.4 yrs) Non-ruptured (N=643; estimated OS=11.9 yrs) P<0.0001 P=0.0218 Ruptured Ruptured (N=21; estimated median OS=6.4 yrs) (N=21; estimated median RFS=2.4 yrs) 02468 02468 Years after surgery (years) Years 0 2 4 6 8 0 2 4 6 8 Non- 644 499 377 138 31 644 536 422 163 41 ruptured Ruptured 21 7 4 1 0 21 18 16 3 0 Recurrence-free Survival Overall Survival C D Non-ruptured (estimated median RFS=8.3 yrs) P=0.4305 P=0.0392 Non-ruptured (estimated median OS: not reached) Ruptured (estimated median OS: not Ruptured reached) (estimated median RFS=3.2 yrs) 02 4 6 8 10 0 2 4 68 10 Years after surgery (years) 0 2 4 6 8 10 Years 0 2 4 6 8 10 Non-ruptured 167 122 76 38 21 9 167 146 103 64 41 25 5 3 3 2 0 0 Ruptured 5 2 1 1 0 0 FIG. 1 Recurrence-free (a, c) and overall survival (b, d) after surgery for patients with or without tumor rupture in the study (a, b) and validation cohort (c, d) not considered to increase incidence of rupture, nor did Tumor rupture was an independent prognostic factor for neoadjuvant therapy in this study (Table 2, Supplementary both RFS and OS before imatinib. We showed that rup- Table 3). Taken together, rupture might occur in GISTs ture remains an important prognostic factor of RFS, but not with high-risk features regardless of rupture timing; intra- OS, in the era of imatinib. This is likely due to the activity operative rupture might not be due to surgical techniques of imatinib, sunitinib, and/or regorafenib used following but rather due to tumor factors, such as fragility, size, and/ recurrence. In fact, most patients received imatinib and or adhesion to adjacent organs. subsequently sunitinib after recurrence, although patients Tumor rupture may result in peritoneal seeding of tumor receiving adjuvant therapy represented a small fraction. All cells, hence surgery may be considered R1 even if guidelines suggest that patients with high-risk GISTs achieving macroscopic complete resection. This study, should have adjuvant therapy, but only 7 of 21 patients however, revealed that some surgeons considered such (33%) with ruptured GIST received adjuvant therapy in surgery to be R0 after macroscopic complete resection this study. This low rate of adjuvant therapy reflects the during the study period (Table 2, Supplementary Table 3). historical background of the registries. Taken together, Survival Rate (%) Survival Rate (%) Features of Ruptured GISTs in Clinical Practice 1967 was viewed as tumor rupture in this study. In our definition 1.0 of tumor rupture, there was no prognostic difference between preoperative and intraoperative rupture, suggest- P=0.6709 0.8 ing that our definition might be acceptable. There are some limitations to consider. The study is Preoperative Rupture 0.6 (N=15; estimated median RFS=2.7 yrs) retrospective, and the number of patients was limited, especially in the events of recurrence and death; however, 0.4 it is based on two multi-institutional registry studies. The use of two different cohorts may help to check repro- 0.2 ducibility to confirm the obtained results. The median Intra-operative Rupture (N=11; follow-up of the registry studies was 4.7 and 5.1 years for estimated median RFS=2.4yrs) 0.0 the study and validation cohort, respectively, which may be 0 24 68 insufficient to evaluate OS in the era of imatinib, although Years after surgery (years) it may be long enough to determine RFS. Historically, imatinib has been the standard therapy for recurrent dis- Years 0 2 4 6 8 Preoperative 15 5 3 1 0 eases. However, adjuvant therapy was not used sufficiently; Operative 11 5 2 1 0 thus, adjuvant imatinib was not an independent prognostic factor as previously suggested. Finally, the definition of FIG. 2 Recurrence-free survival in patients with pre- or tumor rupture was subjective among surgeons; however, as intraoperative rupture in the study cohort; there is no significant mentioned above, participating investigators shared similar difference between them views regarding tumor rupture, indicating that this multi- institutional study is valid. tumor rupture is an independent prognostic factor of recurrence, but not OS, after complete resection in the era This study focused on ruptured GISTs. Tumor rupture may be defined by tumor fracture and perforation at the of imatinib. The definition of tumor rupture was subjectively deter- tumor site, piecemeal resection, open biopsy, and macro- scopic injuries to the pseudocapsule, whereas core and mined and was not yet agreed upon when patients registered in this study underwent surgery. However, the needle biopsy without complications, luminal perforation of tumors, microscopic peritoneal breaks on tumors, or survey showed that the results appeared to be similar to the definition of tumor rupture recently proposed by Holme- microscopic breaks of the pseudocapsule on pathological bakk et al. There are some differences; half of surgeons examination are not considered to be tumor rupture. By this did not consider microscopic infiltration into neighboring definition, GISTs with tumor rupture were seen in several structures as tumor rupture when they performed en bloc percent of GISTs in clinical practice, showed aggressive features of larger size and higher mitotic count regardless resection, while macroscopic injury to the pseudocapsule TABLE 3 Multivariate Independent prognostic factor HR (95% CI) P value analysis for RFS and OS (study cohort) Recurrence-free survival (study cohort): Location (Ref: gastric) 1.637 (1.339–2.002) 0.0140 Size (cm) 1.070 (1.055–1.085) \ 0.0001 Mitotic count (/50 HPF) 1.012 (1.010–1.014) \ 0.0001 Rupture (Ref: nonrupture) 4.545 (3.307–6.234) \ 0.0001 Overall survival (study cohort): Age (years) 1.033 (1.018–1.047) 0.0168 Gender (Ref: female) 2.347 (1.738–3.168) 0.0045 Mitotic count (/50 HPF) 1.014 (1.011–1.017) \ 0.0001 Other factors included in the analysis for RFS using a forward stepwise Cox proportional hazards model were age (P = 0.157), gender (P = 0.086), symptoms (P = 0.551), association of NF1 (P = 0.733), neoadjuvant therapy (P = 0.454), adjuvant therapy (P = 0.453), histology (P = 0.177), and R (complete- ness of surgery) (P = 0.887) Other factors included in the analysis for OS were rupture (P = 0.251), tumor location (P = 0.743), symptoms (P = 0.475), association of NF1 (P = 0.311), neoadjuvant therapy (P = 0.821), adjuvant therapy (P = 0.637), histology (P = 0.696), and R (completeness of surgery) (P = 0.763) Survival Rate (%) 1968 T. Nishida et al. 2. Nishida T, Blay JY, Hirota S, et al. The standard diagnosis, of time of rupture, and had poor prognosis even in the era treatment, and follow-up of gastrointestinal stromal tumors based of imatinib; nevertheless, general standard oncologic on guidelines. Gastric Cancer. 2016;19:3–14. principle of avoiding surgical tumor rupture is critically 3. Joensuu H, Vehtari A, Riihimaki J, et al. Risk of recurrence of important in surgery. To improve the prognosis of patients gastrointestinal stromal tumour after surgery: an analysis of pooled population-based cohorts. Lancet Oncol. 2012;13:265–74. with ruptured GISTs, more prolonged imatinib adjuvant 4. Fletcher CD, Berman JJ, Corless C, et al. Diagnosis of gas- therapy may be required. trointestinal stromal tumors: a consensus approach. Hum Pathol. 2002;33:59–465. ACKNOWLEDGMENT The authors appreciate the kind statisti- 5. Joensuu H. Risk stratification of patients diagnosed with gas- cal suggestions of Dr. Taro Shibata and Dr. Junki Mizusawa in the trointestinal stromal tumor. Hum Pathol. 2008;39:1411–9. Biostatistics Division of the Center for Research Administration and 6. Miettinen M, Lasota J. Gastrointestinal stromal tumors: pathol- Support, National Cancer Center, Tokyo, Japan. ogy and prognosis at different sites. Semin Diagn Pathol. 2006;23:70–83. COLLABORATORS FOR THE KINKI GIST REGISTRY 7. Gold JS, Go ¨ nen M, Gutie ´rrez A, et al. Development and vali- GROUPS NHO Osaka National Hospital (Kazuhiro Nishikawa), dation of a prognostic nomogram for recurrence-free survival Osaka Hospital of Japan Seafarers Relief Association (Tetsuji after complete surgical resection of localised primary gastroin- Sawada), NHO Osaka Minami Medical Center (Tsunehiro Maeda), testinal stromal tumour: a retrospective analysis. Lancet Oncol. Osaka Medical College (Masaru Kawai), Nissay Hospital (Hirofumi 2009;10:1045–52. Ikushima), Kansai Medical University Medical Center (Kouji Nakai), 8. Rossi S, Miceli R, Messerini L, et al. Natural history of imatinib- Kansai Electric Power Hospital (Izumi Kawamoto), Kitano Hospital naive GISTs: a retrospective analysis of 929 cases with long-term and The Tazuke Kofukai Medical Research Institute (Syugo Ueda), follow-up and development of a survival nomogram based on Kindai University Hospital (Takuya Nakai), Shiga General Hospital mitotic index and size as continuous variables. Am J Surg Pathol. (Masazumi Zaima), Shizuoka General Hospital (Shinsuke Sato), 2011;35:1646–56. Tokyo Metropolitan Cancer and Infectious Diseases Center Koma- 9. Goh BK, Chow PK, Yap WM, et al. Which is the optimal risk gome Hospital (Kazuto Yajima), Osaka University Hospital stratification system for surgically treated localized primary (Tsuyoshi Takahashi), Osaka City General Hospital (Hisashi Kubo), GIST? Comparison of three contemporary prognostic criteria in Osaka General Medical Center (Kazumasa Fujitani), Osaka Rosai 171 tumors and a proposal for a modified Armed Forces Institute Hospital (Junichi Hasegawa), Suita Municipal Hospital (Chikara of Pathology risk criteria. Ann Surg Oncol. 2008;15:2153–63. Ebisu), Sakai City Medical Center (Junya Fujita), Toyonaka Munic- 10. Takahashi T, Nakajima K, Nishitani A, et al. An enhanced risk- ipal Hospital (Hiroshi Imamura), Sumitomo Hospital (Atsushi group stratification system for more practical prognostication of Yamamoto), Rinku General Medical Center (Noriyuki Yamamura), clinically malignant gastrointestinal stromal tumors. Int J Clin Tenri Hospital (Tsunehiro Yoshimura), Hoshigaoka Medical Center Oncol. 2007;12:369–74. (Mitsutoshi Tatsumi), Matsushita Memorial Hospital (Akinori 11. Rutkowski P, Bylina E, Wozniak A, et al. Validation of the Noguchi), Kanazawa Medical University Hospital (Miki Noguchi), Joensuu risk criteria for primary resectable gastrointestinal stro- Kyoto City Hospital (Hirokazu Matsuo), Yodogawa Christian mal tumour—the impact of tumour rupture on patient outcomes. Hospital (Akihiro Toyokawa), Osaka Red Cross Hospital (Seiichiro Eur J Surg Oncol. 2011;37: 890–6. Kanaya), Hyogo Cancer Center (Takashi Yasuda), Osaka Interna- 12. Yanagimoto Y, Takahashi T, Muguruma K, et al. Re-appraisal of tional Cancer Institute (Takeshi Omori), Osaka Police Hospital, risk classifications for primary gastrointestinal stromal tumors Department of Surgery (Toru Masuzawa, Toshirou Nishida), Kaizuka (GIST) after complete resection: indications for adjuvant therapy. City Hospital (Toshimasa Tsujinaka). The Kanagawa GIST registry: Gastric Cancer. 2015;18:426–33. Kanagawa Cancer Center (Haruhiko Cho), Yokohama Rosai Hospital 13. Joensuu H, Eriksson M, Hall KS, et al. Risk factors for gas- (Manabu Chigaku). trointestinal stromal tumor recurrence in patients treated with adjuvant imatinib. Cancer. 2014;120:2325–33. FUNDING This work is supported in part by a Grant-in-Aid 14. Hølmebakk T, Bjerkehagenc, B, Boye K, et al. Definition and (16H05419 and 16K15600) for Scientific Research from the Japanese clinical significance of tumour rupture in gastrointestinal stromal Ministry of Education, Culture, Sports, Science, and Technology, and tumours of the small intestine. Br J Surg. 2016;103:684–691. by a grant (28-A-16) from the National Cancer Center Research and 15. Hohenberger P, Ronellenfitsch U, Oladeji O, et al. Pattern of Development Fund. recurrence in patients with ruptured primary gastrointestinal stromal tumour. Br J Surg. 2010;97:1854–9. OPEN ACCESS This article is distributed under the terms of the 16. Rutkowski P, Nowecki ZI, Michej W, et al. Risk criteria and Creative Commons Attribution 4.0 International License (http://crea prognostic factors for predicting recurrences after resection of tivecommons.org/licenses/by/4.0/), which permits unrestricted use, primary gastrointestinal stromal tumor. Ann Surg Oncol. distribution, and reproduction in any medium, provided you give 2007;14:2018–27. appropriate credit to the original author(s) and the source, provide a 17. Joensuu H, Eriksson M, Sundby Hall K, et al. One versus three link to the Creative Commons license, and indicate if changes were years of adjuvant imatinib for operable gastrointestinal stromal made. tumor: a randomized trial. JAMA. 2012;307:1265–72. 18. Casali PG, Le Cesne A, Poveda Velasco A, et al. Time to definitive failure to the first tyrosine kinase inhibitor in localized GI stromal tumors treated with imatinib as an adjuvant: a Euro- REFERENCES pean Organisation for Research and Treatment of Cancer Soft Tissue and Bone Sarcoma Group intergroup randomized trial in 1. Joensuu H, Hohenberger P, Corless CL. Gastrointestinal stromal collaboration with the Australasian Gastro-Intestinal Trials tumour. Lancet. 2013;382:973–83. Group, UNICANCER, French Sarcoma Group, Italian Sarcoma Group, and Spanish Group for Research on Sarcomas. J Clin Oncol. 2015;33:4276–83. Features of Ruptured GISTs in Clinical Practice 1969 19. Rutkowski P, Bylina E, Wozniak A, et al. Validation of the 23. Verduijn M, Grootendorst DC, Dekker FW, et al. The analysis of Joensuu risk criteria for primary resectable gastrointestinal stro- competing events like cause-specific mortality-beware of the mal tumour–the impact of tumour rupture on patient outcomes. Kaplan–Meier method. Nephrol Dial Transplant. 2011;26:56–61. Eur J Surg Oncol. 2011;37:890–6. 24. DeMatteo RP, Ballman KV, Antonescu CR, et al. Long-term 20. Bischof DA, Kim Y, Behman R, et al. A nomogram to predict results of adjuvant imatinib mesylate in localized, high-risk, disease-free survival after surgical resection of GIST. J Gas- primary gastrointestinal stromal tumor: ACOSOG Z9000 (Al- trointest Surg. 2014;18:2123–9. liance) intergroup phase 2 trial. Ann Surg. 2013;258:422–9. 21. Kim JS, Kim HJ, Park SH, et al. Computed tomography features 25. Ng EH, Pollock RE, Munsell MF, et al. Prognostic factors and predictive findings of ruptured gastrointestinal stromal influencing survival in gastrointestinal leiomyosarcomas. Impli- tumours. Eur Radiol. 2017;27:2583–90. cations for surgical management and staging. Ann Surg. 22. Yamamoto K, Tsujinaka T, Takahashi T, et al. Impact of the 1992;215:68–77. Japanese gastric cancer screening system on treatment outcomes 26. Joensuu H, Eriksson M, Hall KS, et al. Risk factors for gas- in gastric gastrointestinal stromal tumor (GIST): an analysis trointestinal stromal tumor recurrence in patients treated with based on the GIST registry. Ann Surg Oncol. 2015;22:232–9. adjuvant imatinib. Cancer. 2014;120:2325–33. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Annals of Surgical Oncology Springer Journals

Clinicopathological Features and Prognosis of Primary GISTs with Tumor Rupture in the Real World

Free
9 pages
Loading next page...
 
/lp/springer_journal/clinicopathological-features-and-prognosis-of-primary-gists-with-tumor-aRu0gH2qoG
Publisher
Springer International Publishing
Copyright
Copyright © 2018 by The Author(s)
Subject
Medicine & Public Health; Surgical Oncology; Oncology; Surgery
ISSN
1068-9265
eISSN
1534-4681
D.O.I.
10.1245/s10434-018-6505-7
Publisher site
See Article on Publisher Site

Abstract

Ann Surg Oncol (2018) 25:1961–1969 https://doi.org/10.1245/s10434-018-6505-7 OR IG INA L A R TIC L E – BO NE AN D S OFT T IS SU E S A RCOM A S Clinicopathological Features and Prognosis of Primary GISTs with Tumor Rupture in the Real World 1,2,3 4,5 6 2,3 7 8 Toshirou Nishida , Haruhiko Cho , Seiichi Hirota , Toru Masuzawa , Gaku Chiguchi , Toshimasa Tsujinaka , and the Kinki GIST Study Group 1 2 Department of Surgery, National Cancer Center Hospital, Tokyo, Japan; Department of Surgery, Osaka Police Hospital, 3 4 Osaka, Japan; Department of Surgery, Kansai Rosai Hospital, Amagasaki, Japan; Department of Gastrointestinal Surgery, Kanagawa Cancer Center, Yokohama, Japan; Department of Surgery, Tokyo Metropolitan Cancer and Infectious Disease Center Komagome Hospital, Tokyo, Japan; Department of Surgical Pathology, Hyogo College of Medicine, 7 8 Nishinomiya, Japan; Department of Gastroenterology, Yokohama Rosai Hospital, Yokohama, Japan; Department of Gastrointestinal Surgery, Kaizuka City Hospital, Kaizuka, Japan ABSTRACT Recurrence rates were higher and median recurrence-free Background. Patients with ruptured gastrointestinal stro- survival (RFS) and overall survival (OS) were shorter with mal tumor (GIST) are recommended for imatinib adjuvant ruptured than nonruptured GIST. Tumor rupture was one of therapy; however, their clinicopathological features and the independent prognostic factors for RFS, but not OS, prognosis in the era of imatinib are unknown. according to multivariate analysis. Patients and Methods. The study cohort included 665 Conclusions. Ruptured GISTs were symptomatic larger patients with histologically proven primary GISTs who tumors with high mitotic activity, frequent relapse, and underwent R0 or R1 surgery between 2003 and 2007; the shorter RFS. Tumor rupture was an independent prognostic validation cohort included 182 patients between 2000 and factor for RFS, but not for OS, in the era of imatinib. 2014. The definitions of tumor rupture in the study inclu- ded perforation at tumor site, tumor fracture, piecemeal Gastrointestinal stromal tumor (GIST) is a potentially resection including open biopsy, and macroscopic injuries to the pseudocapsule. malignant tumor and the most frequent sarcoma of the gastrointestinal tract. Currently, complete resection (R0) of Results. Tumor rupture occurred in 21 (3.2%) of 665 and 5 (2.9%) of 182 patients in the study and validation cohort, primary GIST offers the only potentially permanent cure for GIST patients. After complete resection, 60% of respectively. Ruptured GISTs were more symptomatic, patients with localized GISTs are cured, while the other were larger in size, and had higher mitotic count than 1–3 nonruptured GISTs but were not associated with tumor 40% experience disease recurrence during follow-up. Prognostic factors for recurrence after R0 surgery have location or laparoscopic surgery. GISTs with intraoperative rupture had clinicopathological features and prognostic been rigorously investigated, and four factors are recog- nized as independent prognostic factors: tumor size (cm), outcomes similar to those with preoperative rupture. mitosis (/50 HPF or/5 mm ), location (gastric versus non- 3–13 gastric), and rupture. Several risk stratifications have 4–8 been proposed based on these four factors. Among these four prognostic factors, tumor rupture is Electronic supplementary material The online version of this article (https://doi.org/10.1245/s10434-018-6505-7) contains the most ominous and is a subjective factor clinically supplementary material, which is available to authorized users. 3,10–12 judged by surgeons. Previous reports indicate that most ruptured GISTs are associated with recurrence during The Author(s) 2018. This article is an open access publication follow-up and that patients with ruptured GIST have sig- First Received: 8 February 2018; nificantly shorter recurrence-free survival (RFS) than those 14–16 Published Online: 11 May 2018 without rupture. The definition of ruptured GIST, T. Nishida however, was not consistent among researchers until a e-mail: tnishida@ncc.go.jp 1962 T. Nishida et al. recent provisional definition was proposed. The term time of diagnosis and individuals whose date of surgery, ‘‘ruptured GIST’’ may include heterogeneous entities; age, gender, and/or outcomes were missing (Supplemen- hence, in some reports, rupture may include tumor pene- tary Fig. 1). This study was reviewed and approved by the tration into the peritoneum (T4a) and microscopic Steering Committee of the Kinki GIST Study Group and by pseudocapsule injuries. Incidence of rupture has been the institutional review board (IRB) of Osaka Police 3,7,10,12,16–18 reported to vary from 2 to 22%. Some reports Hospital and Kanagawa Cancer Center. have suggested that tumor rupture is an independent Tumor rupture was considered to be subjectively 3,5,10,12,16 prognostic factor, whereas others have determined by each surgeon at that time, so we conducted a 19,20 disagreed. retrospective survey of the concept of rupture by ques- Recent clinical trials show that patients at high risk of tionnaire. Questionnaire entries included perforation at the recurrence may benefit from 3 years of adjuvant therapy tumor site, tumor fracture with blood-tinged ascites, with imatinib (Gleevec, Novartis Pharmaceuticals, Basel, piecemeal resection during surgery (including open 17,18 Switzerland) after R0 resection. It may be argued that biopsy), and macroscopic injuries to the pseudocapsule patients with ruptured GIST require prolonged adjuvant exposing tumor cells into the peritoneal cavity (Supple- therapy for longer than 3 years. Few studies have examined mentary Table 1). These were similar to the definitions of which tumors are more likely to rupture or which surgical rupture proposed by Holmebakk et al. Completeness of procedures, open or laparoscopic, may potentially increase surgical resection was assessed by local surgeons as fol- the risk of rupture. In this retrospective study, we ana- lows: R0, no detectable residual tumor; R1, microscopic lyzed real-world data obtained from two registry studies residual tumor; R2, macroscopic residual tumor. Patients conducted in Japan and clarified the clinicopathological with R2 resection were excluded from the analysis. features and prognostic effects of rupture in the era of imatinib. Statistical Analysis Statistical analyses were performed using the Chi PATIENTS AND METHODS squared test, Fisher’s exact test, Student’s t test, and Mann– Study Design Whitney U test. RFS was calculated from date of surgery to date of first tumor recurrence or date of death, excluding The study included both study and validation cohorts. living patients without recurrence at time of data collec- The former data were obtained from the GIST registry tion. Overall survival (OS) was calculated from date of study conducted by the Kinki GIST Study Group, and the surgery to date of any death, excluding living patients. RFS latter from the Kanagawa registry study. The GIST registry and OS were compared between the groups using the designed by the Kinki GIST Study Group was reported Kaplan–Meier life-table method with the log-rank test. Cumulative recurrence probability was estimated by previously. This retrospective observational study was designed to collect clinical and pathological data of cumulative incidence competing risk analysis as described previously. A forward stepwise Cox proportional hazards patients with pathologically proven GIST diagnosed in each participating hospital and treated between January model was used for multivariate analysis to identify risk 2003 and December 2007. Data on risk stratifications factors associated with RFS and OS. P values were two- classified according to the modified National Institutes of sided, and P values less than 0.05 were considered sig- Health (NIH) consensus criteria and the Armed Forces nificant. Data were analyzed using the Statistical Package Institute of Pathology (AFIP) criteria were collected, in of IBM SPSS Statistics 25, version 25.0 (IBM Corp., addition to information regarding preoperative and intra- Armonk, NY, USA). operative ruptures. The Kanagawa registry study collected similar patient data retrospectively and prospectively since RESULTS January 2001 to December 2016 from the two hospitals in Kanagawa Prefecture. The hospitals participating in each In total, 665 patients with primary GIST who underwent registry are listed at the end of the manuscript. R0 or R1 surgery were included in the study cohort and 172 Using data obtained from the GIST registry database, patients in the validation cohort. The clinicopathological eligible patients were selected (Supplementary Fig. 1). features of the patients analyzed in the study and validation Eligibility criteria included tumor morphology compatible cohorts are presented in Table 1. The study cohort included with GIST, KIT positivity on immunohistochemistry, and 506 gastric, 119 small intestinal, 26 colorectal, 10 eso- macroscopically complete resection of the primary tumor. phageal, and 4 extra-gastrointestinal GISTs. The study We excluded patients with metastatic or recurrent GIST at sample included 339 males and 326 females with median Features of Ruptured GISTs in Clinical Practice 1963 TABLE 1 Patient Factor Study cohort (N = 665) Validation cohort (N = 172) characteristics Age (Median; years) 66.0 (18–93) 62.5 (17–89) Gender Male 339 (51.0%) 100 (58.1%) Female 326 (49.9%) 72 (41.9%) Cancer association Cancer history 113 (17.0%) 37 (21.5%) No cancer history 536 (80.6%) 135 (78.5%) Unavailable 16 (2.4%) 0 (0%) Location Esophagus 10 (1.5%) 2 (1.2%) b b b Stomach 506 (76.1%) 120 (69.4%) b b b Small intestine 119 (17.9%) 37 (21.5%) Colon and rectum 26 (3.9%) 14 (8.1%) Others 4 (0.6%) 0 (0%) Tumor size (median; cm) 4.0 (0.1–35) 5.0 (1.1–25) Mitosis (median;/50 HPF) 2.6 (0.0–250.0) 5.0 (0.0–250.0) Symptoms at diagnosis Yes 257 (38.6%) No 407 (61.2%) Unavailable 1 (0.2%) Neoadjuvant Yes 9 (1.3%) 15 (8.7%) No 648 (97.4%) 157 (91.3%) Unavailable 8 (1.3%) 0 (0%) Adjuvant Yes 37 (5.6%) 30 (17.4%) No 627 (94.3%) 142 (82.6%) Unavailable 1 (0.1%) 0 (0%) Surgery Open 459 (69.0%) 116 (67.4%) Laparoscopy 200 (30.1%) 51 (29.7%) Local 6 (0.9%) 5 (2.9%) Completeness of surgery R0 661 (99.4%) 162 (94.2%) R1 4 (0.6%) 10 (5.8%) Tumor rupture No 644 (96.8%) 167 (97.1%) Yes 21 (3.2%) 5 (2.9%) Preoperative 12 3 Intraoperative 9 2 Histological type Spindle 538 (80.9%) 83 (48.3%) Epithelioid 22 (3.3%) 2 (1.2%) Mixed 35 (5.3%) 7 (4.1%) Unavailable 70 (10.5%) 80 (46.4%) Recurrence No recurrence 570 (87.5%) 124 (72.1%) Recurrence 95 (12.5%) 47 (27.9%) Estimated 5-year RFS (median ? SE) 78.6 ? 1.8% 72.5 ? 3.9% Overall survival Alive 600 (90.2%) 151 (87.8%) Death 65 (9.8%) 21 (12.2%) Estimated 5-year OS (median ? SE) 91.5 ? 1.2% 92.2 ? 2.3% Median follow-up: 4.67 years for study cohort and 5.12 years for validation cohort RFS recurrence-free survival, OS overall survival, SE standard error Not available in validation cohort One duplicated patient with gastric and small intestinal GISTs in each cohort age of 66 years. Median tumor size was 4.0 cm, and the (N = 12) or intraoperatively (N = 9). In the validation median mitotic count was 2.6/50 HPF; 21 (3.2%) patients cohort, there were 120 gastric, 37 small intestinal, 14 were reported to have tumor rupture, either preoperatively colorectal, and 2 esophageal GISTs. Median age was 1964 T. Nishida et al. 62.5 years, with 100 male and 72 female patients in the nonrupture group (Supplementary Fig. 2). RFS of (Table 1). Median tumor size was 5.0 cm, the median patients with intraoperative ruptured GIST was not differ- mitotic count was 5.0/50 HPF, and only five patients ent from that of patients with preoperative rupture (Fig. 2; (2.9%) had tumor rupture, of whom three had preoperative P = 0.6709). In the study cohort, recurrence in patients and two intraoperative rupture. Most patients in both with ruptured GIST was more frequent in the peritoneum cohorts underwent R0 surgery, by either open or laparo- and local lesions compared with those in patients with scopic procedure. Nearly all patients received imatinib nonruptured GIST (Table 2). Median OS of patients with therapy after relapse. A small fraction of patients with ruptured GIST (6.4 years; 95% CI 5.5–7.3 years) was high-risk features received neoadjuvant (1.3 and 8.7%), significantly shorter than that of those with nonruptured while a relatively larger number of patients received GIST in the study cohort (11.9 years; 95% CI adjuvant therapy (5.6 and 17.4%) in the study and valida- 10.7–13.0 years; P = 0.0218); this was not confirmed in tion cohort, respectively. the validation cohort, likely because of the low statistical A questionnaire to participating surgeons indicated the power and higher rate of imatinib adjuvant therapy. following results (Supplementary Table 1): most surgeons Multivariate analyses using the Cox proportional haz- considered that tumor fracture and perforation, piecemeal ards model indicated that location, tumor size, mitotic resection, open biopsy, and macroscopic injuries to the count, and rupture were independent prognostic factors for pseudocapsule exposing tumor cells represented rupture; in RFS and that age, gender, and mitotic count were inde- contrast, they considered that core and needle biopsy pendent prognostic factors for OS in the study cohort without complications, luminal perforation of tumors, (Table 3). In the validation cohort, independent prognostic peritoneal tumor penetration, and microscopic injuries to factors for recurrence included tumor size, mitotic count, the pseudocapsule did not represent rupture. rupture, and use of adjuvant therapy, and those for OS were Tumor size was larger and mitotic count was higher in gender, tumor size, and mitotic count (Supplementary ruptured GISTs than in nonruptured tumors in the study Table 4). In the combined analysis of both cohorts, tumor cohort (Table 2). Patients with ruptured GIST were more size, mitotic count, location, rupture, and gender were symptomatic at admission, regardless of preoperative or independent prognostic factors for RFS, and age, gender, intraoperative rupture (Table 2, Supplementary Table 3). and mitotic count were independent prognostic factors for Among the nine patients who had neoadjuvant therapy, one OS. patient experienced tumor rupture during surgery. Inci- dence of rupture was not different between open and DISCUSSION laparoscopic surgery in either group. Location (gastric and nongastric) was not correlated with GIST rupture in either This study found that ruptured GIST was seen in nearly the study or validation cohort. There was no significant 3% of primary GISTs, being more symptomatic and difference in terms of location, symptoms, tumor size, exhibiting aggressive features of larger size and higher mitotic count, or recurrence between preoperative and mitotic count compared with nonruptured tumors. The intraoperative rupture (Supplementary Table 2). reported frequency of tumor rupture varies depending on During median follow-up of 4.67 years, there were 95 the study; population-based studies indicated that it was 7 10 3 11 (12.5%) relapses and 65 (9.8%) deaths in the study cohort, less than 10% (1%, 4.0%, 5.9%, and 7.1% ), while in compared with 47 (27.9%) recurrences and 21 (12.2%) clinical trials, it was higher than 10% (20% of high-risk 17 18 deaths in the validation cohort with median follow-up of GISTs, 11% of intermediate- and high-risk GISTs, and 5.12 years. Recurrence was more frequent for patients with 17% ). The true incidence of tumor rupture is speculated ruptured GISTs than those without rupture in both cohorts to be several percent in clinical practice. Ruptured GISTs (Table 2, Supplementary Table 4). Median RFS of patients were shown to be more symptomatic with high-risk fea- with ruptured GIST [2.4 years; 95% confidence interval tures, including larger tumor size and higher mitotic count, (CI) 1.4–3.4 years in the study cohort; P \ 0.0001, and and were treated by emergency surgery in previous, as well 11,14,15 3.2 years in the validation cohort; 95% CI 1.3–5.8 years; as present, studies. P = 0.0392] was significantly shorter than that of patients Tumor rupture occurred both before and during surgery. with nonruptured GIST (8.4 years; 95% CI 8.0–8.8 years, The frequency of preoperative and intraoperative rupture 17,24 and 8.4 years; 95% CI 7.4–9.3 years) in the study and was similar in this registry study. Clinicopathological validation cohort, respectively (Fig. 1). Cumulative inci- features and prognostic outcomes of GISTs with preoper- dence analysis of the study cohort indicated that all events ative rupture were similar to those with intraoperative were recurrence of GIST in the rupture group, whereas rupture, although the sample size was small (Table 2, one-third of events were recurrence of GIST, and deaths Supplementary Table 3). The laparoscopic approach was due to other diseases might account for the other two-thirds Features of Ruptured GISTs in Clinical Practice 1965 TABLE 2 Background of GIST patients with and without tumor rupture (study cohort) Nonruptured (N = 644) Ruptured (N = 21) P value Age (years) 66 (18–93) 68 (55–90) 0.2236 Gender Male 326 (50.6%) 13 (61.9%) 0.3087 Female 318 (49.4%) 8 (38.1%) Primary location Gastric 493 (76.6%) 13 (61.9%) 0.1215 Nongastric 151 (23.4%) 8 (38.1%) Association of cancer No 518 (80.4%) 18 (85.7%) 0.7098 Yes 110 (17.1%) 3 (14.3%) Unavailable 16 (2.5%) 0 (0%) Median tumor size (cm) 4.0 (0.1–35.0) 9.6 (2.6–30.0) 0.0008 Symptoms No 406 (63.0%) 1 (4.7%) \ 0.0001 Yes 237 (36.8%) 20 (95.2%) Unavailable 1 (0.2%) 0 (0%) Neoadjuvant No 629 (97.7%) 19 (90.5%) 0.1561 Yes 8 (1.2%) 1 (5%) Unavailable 7 (1.1%) 1 (5%) Adjuvant therapy No 613 (95.2%) 14 (66.7%) \ 0.0001 Yes 30 (4.7%) 7 (23.3%) Unavailable 1 (0.1%) 0 (0%) Surgery Open 441 (68.5%) 18 (85.7%) 0.2383 Laparoscopic 197 (30.6%) 3 (14.3%) Local 6 (0.9%) 0 (0%) R R0 642 (99.7%) 19 (90.5%) \ 0.0001 R1 2 (0.3%) 2 (9.5%) Median mitosis (/50 HPF) 2.5 (0.0–250) 13.0 (0.0–115) 0.0004 Cell type Spindle 518 (80.4%) 20 (95.2%) 0.3347 Epithelioid 22 (3.4%) 0 (0%) Mixed 35 (5.4%) 0 (0%) Unavailable 69 (10.7%) 1 (4.8%) Median RFS (95% CI; years) 8.4 (8.0–8.9) 2.4 (1.4–3.4) \ 0.0001 Estimated 5-year RFS (median ? SE) 80.7 ? 1.7% 16.4 ? 8.6% Recurrence No 565 (87.7%) 5 (23.8%) \ 0.0001 Yes 79 (12.3%) 16 (76.2%) a b b Recurrence site Liver 53 (67.1%) 6 (37.5%) 0.0108 b b Lung 2 (2.5%) 0 (0%) b b Local 8 (10.1%) 4 (25%) b b Peritoneum 24 (30.4%) 14 (87.5%) Median OS (95% CI; years) 11.9 (10.7–13.0) 6.4 (5.6–7.3) 0.0218 Estimated 5-year OS (median ? SE) 88.9 ? 7.4% 91.6 ? 1.2% Overall survival Alive 584 (90.7%) 16 (76.2%) 0.0452 Dead 60 (9.3%) 5 (23.8%) c c Death due to GIST 21 (35%) 5 (100%) c c Death due to other diseases 39 (65%) 0 (0%) RFS recurrence-free survival, OS overall survival, SE standard error, CI confidence interval Duplicated number % of total recurrence in each group % of total death in each group 1966 T. Nishida et al. Recurrence-free Survival Overall Survival A B Non-ruptured (N=643; estimated RFS=8.4 yrs) Non-ruptured (N=643; estimated OS=11.9 yrs) P<0.0001 P=0.0218 Ruptured Ruptured (N=21; estimated median OS=6.4 yrs) (N=21; estimated median RFS=2.4 yrs) 02468 02468 Years after surgery (years) Years 0 2 4 6 8 0 2 4 6 8 Non- 644 499 377 138 31 644 536 422 163 41 ruptured Ruptured 21 7 4 1 0 21 18 16 3 0 Recurrence-free Survival Overall Survival C D Non-ruptured (estimated median RFS=8.3 yrs) P=0.4305 P=0.0392 Non-ruptured (estimated median OS: not reached) Ruptured (estimated median OS: not Ruptured reached) (estimated median RFS=3.2 yrs) 02 4 6 8 10 0 2 4 68 10 Years after surgery (years) 0 2 4 6 8 10 Years 0 2 4 6 8 10 Non-ruptured 167 122 76 38 21 9 167 146 103 64 41 25 5 3 3 2 0 0 Ruptured 5 2 1 1 0 0 FIG. 1 Recurrence-free (a, c) and overall survival (b, d) after surgery for patients with or without tumor rupture in the study (a, b) and validation cohort (c, d) not considered to increase incidence of rupture, nor did Tumor rupture was an independent prognostic factor for neoadjuvant therapy in this study (Table 2, Supplementary both RFS and OS before imatinib. We showed that rup- Table 3). Taken together, rupture might occur in GISTs ture remains an important prognostic factor of RFS, but not with high-risk features regardless of rupture timing; intra- OS, in the era of imatinib. This is likely due to the activity operative rupture might not be due to surgical techniques of imatinib, sunitinib, and/or regorafenib used following but rather due to tumor factors, such as fragility, size, and/ recurrence. In fact, most patients received imatinib and or adhesion to adjacent organs. subsequently sunitinib after recurrence, although patients Tumor rupture may result in peritoneal seeding of tumor receiving adjuvant therapy represented a small fraction. All cells, hence surgery may be considered R1 even if guidelines suggest that patients with high-risk GISTs achieving macroscopic complete resection. This study, should have adjuvant therapy, but only 7 of 21 patients however, revealed that some surgeons considered such (33%) with ruptured GIST received adjuvant therapy in surgery to be R0 after macroscopic complete resection this study. This low rate of adjuvant therapy reflects the during the study period (Table 2, Supplementary Table 3). historical background of the registries. Taken together, Survival Rate (%) Survival Rate (%) Features of Ruptured GISTs in Clinical Practice 1967 was viewed as tumor rupture in this study. In our definition 1.0 of tumor rupture, there was no prognostic difference between preoperative and intraoperative rupture, suggest- P=0.6709 0.8 ing that our definition might be acceptable. There are some limitations to consider. The study is Preoperative Rupture 0.6 (N=15; estimated median RFS=2.7 yrs) retrospective, and the number of patients was limited, especially in the events of recurrence and death; however, 0.4 it is based on two multi-institutional registry studies. The use of two different cohorts may help to check repro- 0.2 ducibility to confirm the obtained results. The median Intra-operative Rupture (N=11; follow-up of the registry studies was 4.7 and 5.1 years for estimated median RFS=2.4yrs) 0.0 the study and validation cohort, respectively, which may be 0 24 68 insufficient to evaluate OS in the era of imatinib, although Years after surgery (years) it may be long enough to determine RFS. Historically, imatinib has been the standard therapy for recurrent dis- Years 0 2 4 6 8 Preoperative 15 5 3 1 0 eases. However, adjuvant therapy was not used sufficiently; Operative 11 5 2 1 0 thus, adjuvant imatinib was not an independent prognostic factor as previously suggested. Finally, the definition of FIG. 2 Recurrence-free survival in patients with pre- or tumor rupture was subjective among surgeons; however, as intraoperative rupture in the study cohort; there is no significant mentioned above, participating investigators shared similar difference between them views regarding tumor rupture, indicating that this multi- institutional study is valid. tumor rupture is an independent prognostic factor of recurrence, but not OS, after complete resection in the era This study focused on ruptured GISTs. Tumor rupture may be defined by tumor fracture and perforation at the of imatinib. The definition of tumor rupture was subjectively deter- tumor site, piecemeal resection, open biopsy, and macro- scopic injuries to the pseudocapsule, whereas core and mined and was not yet agreed upon when patients registered in this study underwent surgery. However, the needle biopsy without complications, luminal perforation of tumors, microscopic peritoneal breaks on tumors, or survey showed that the results appeared to be similar to the definition of tumor rupture recently proposed by Holme- microscopic breaks of the pseudocapsule on pathological bakk et al. There are some differences; half of surgeons examination are not considered to be tumor rupture. By this did not consider microscopic infiltration into neighboring definition, GISTs with tumor rupture were seen in several structures as tumor rupture when they performed en bloc percent of GISTs in clinical practice, showed aggressive features of larger size and higher mitotic count regardless resection, while macroscopic injury to the pseudocapsule TABLE 3 Multivariate Independent prognostic factor HR (95% CI) P value analysis for RFS and OS (study cohort) Recurrence-free survival (study cohort): Location (Ref: gastric) 1.637 (1.339–2.002) 0.0140 Size (cm) 1.070 (1.055–1.085) \ 0.0001 Mitotic count (/50 HPF) 1.012 (1.010–1.014) \ 0.0001 Rupture (Ref: nonrupture) 4.545 (3.307–6.234) \ 0.0001 Overall survival (study cohort): Age (years) 1.033 (1.018–1.047) 0.0168 Gender (Ref: female) 2.347 (1.738–3.168) 0.0045 Mitotic count (/50 HPF) 1.014 (1.011–1.017) \ 0.0001 Other factors included in the analysis for RFS using a forward stepwise Cox proportional hazards model were age (P = 0.157), gender (P = 0.086), symptoms (P = 0.551), association of NF1 (P = 0.733), neoadjuvant therapy (P = 0.454), adjuvant therapy (P = 0.453), histology (P = 0.177), and R (complete- ness of surgery) (P = 0.887) Other factors included in the analysis for OS were rupture (P = 0.251), tumor location (P = 0.743), symptoms (P = 0.475), association of NF1 (P = 0.311), neoadjuvant therapy (P = 0.821), adjuvant therapy (P = 0.637), histology (P = 0.696), and R (completeness of surgery) (P = 0.763) Survival Rate (%) 1968 T. Nishida et al. 2. Nishida T, Blay JY, Hirota S, et al. The standard diagnosis, of time of rupture, and had poor prognosis even in the era treatment, and follow-up of gastrointestinal stromal tumors based of imatinib; nevertheless, general standard oncologic on guidelines. Gastric Cancer. 2016;19:3–14. principle of avoiding surgical tumor rupture is critically 3. Joensuu H, Vehtari A, Riihimaki J, et al. Risk of recurrence of important in surgery. To improve the prognosis of patients gastrointestinal stromal tumour after surgery: an analysis of pooled population-based cohorts. Lancet Oncol. 2012;13:265–74. with ruptured GISTs, more prolonged imatinib adjuvant 4. Fletcher CD, Berman JJ, Corless C, et al. Diagnosis of gas- therapy may be required. trointestinal stromal tumors: a consensus approach. Hum Pathol. 2002;33:59–465. ACKNOWLEDGMENT The authors appreciate the kind statisti- 5. Joensuu H. Risk stratification of patients diagnosed with gas- cal suggestions of Dr. Taro Shibata and Dr. Junki Mizusawa in the trointestinal stromal tumor. Hum Pathol. 2008;39:1411–9. Biostatistics Division of the Center for Research Administration and 6. Miettinen M, Lasota J. Gastrointestinal stromal tumors: pathol- Support, National Cancer Center, Tokyo, Japan. ogy and prognosis at different sites. Semin Diagn Pathol. 2006;23:70–83. COLLABORATORS FOR THE KINKI GIST REGISTRY 7. Gold JS, Go ¨ nen M, Gutie ´rrez A, et al. Development and vali- GROUPS NHO Osaka National Hospital (Kazuhiro Nishikawa), dation of a prognostic nomogram for recurrence-free survival Osaka Hospital of Japan Seafarers Relief Association (Tetsuji after complete surgical resection of localised primary gastroin- Sawada), NHO Osaka Minami Medical Center (Tsunehiro Maeda), testinal stromal tumour: a retrospective analysis. Lancet Oncol. Osaka Medical College (Masaru Kawai), Nissay Hospital (Hirofumi 2009;10:1045–52. Ikushima), Kansai Medical University Medical Center (Kouji Nakai), 8. Rossi S, Miceli R, Messerini L, et al. Natural history of imatinib- Kansai Electric Power Hospital (Izumi Kawamoto), Kitano Hospital naive GISTs: a retrospective analysis of 929 cases with long-term and The Tazuke Kofukai Medical Research Institute (Syugo Ueda), follow-up and development of a survival nomogram based on Kindai University Hospital (Takuya Nakai), Shiga General Hospital mitotic index and size as continuous variables. Am J Surg Pathol. (Masazumi Zaima), Shizuoka General Hospital (Shinsuke Sato), 2011;35:1646–56. Tokyo Metropolitan Cancer and Infectious Diseases Center Koma- 9. Goh BK, Chow PK, Yap WM, et al. Which is the optimal risk gome Hospital (Kazuto Yajima), Osaka University Hospital stratification system for surgically treated localized primary (Tsuyoshi Takahashi), Osaka City General Hospital (Hisashi Kubo), GIST? Comparison of three contemporary prognostic criteria in Osaka General Medical Center (Kazumasa Fujitani), Osaka Rosai 171 tumors and a proposal for a modified Armed Forces Institute Hospital (Junichi Hasegawa), Suita Municipal Hospital (Chikara of Pathology risk criteria. Ann Surg Oncol. 2008;15:2153–63. Ebisu), Sakai City Medical Center (Junya Fujita), Toyonaka Munic- 10. Takahashi T, Nakajima K, Nishitani A, et al. An enhanced risk- ipal Hospital (Hiroshi Imamura), Sumitomo Hospital (Atsushi group stratification system for more practical prognostication of Yamamoto), Rinku General Medical Center (Noriyuki Yamamura), clinically malignant gastrointestinal stromal tumors. Int J Clin Tenri Hospital (Tsunehiro Yoshimura), Hoshigaoka Medical Center Oncol. 2007;12:369–74. (Mitsutoshi Tatsumi), Matsushita Memorial Hospital (Akinori 11. Rutkowski P, Bylina E, Wozniak A, et al. Validation of the Noguchi), Kanazawa Medical University Hospital (Miki Noguchi), Joensuu risk criteria for primary resectable gastrointestinal stro- Kyoto City Hospital (Hirokazu Matsuo), Yodogawa Christian mal tumour—the impact of tumour rupture on patient outcomes. Hospital (Akihiro Toyokawa), Osaka Red Cross Hospital (Seiichiro Eur J Surg Oncol. 2011;37: 890–6. Kanaya), Hyogo Cancer Center (Takashi Yasuda), Osaka Interna- 12. Yanagimoto Y, Takahashi T, Muguruma K, et al. Re-appraisal of tional Cancer Institute (Takeshi Omori), Osaka Police Hospital, risk classifications for primary gastrointestinal stromal tumors Department of Surgery (Toru Masuzawa, Toshirou Nishida), Kaizuka (GIST) after complete resection: indications for adjuvant therapy. City Hospital (Toshimasa Tsujinaka). The Kanagawa GIST registry: Gastric Cancer. 2015;18:426–33. Kanagawa Cancer Center (Haruhiko Cho), Yokohama Rosai Hospital 13. Joensuu H, Eriksson M, Hall KS, et al. Risk factors for gas- (Manabu Chigaku). trointestinal stromal tumor recurrence in patients treated with adjuvant imatinib. Cancer. 2014;120:2325–33. FUNDING This work is supported in part by a Grant-in-Aid 14. Hølmebakk T, Bjerkehagenc, B, Boye K, et al. Definition and (16H05419 and 16K15600) for Scientific Research from the Japanese clinical significance of tumour rupture in gastrointestinal stromal Ministry of Education, Culture, Sports, Science, and Technology, and tumours of the small intestine. Br J Surg. 2016;103:684–691. by a grant (28-A-16) from the National Cancer Center Research and 15. Hohenberger P, Ronellenfitsch U, Oladeji O, et al. Pattern of Development Fund. recurrence in patients with ruptured primary gastrointestinal stromal tumour. Br J Surg. 2010;97:1854–9. OPEN ACCESS This article is distributed under the terms of the 16. Rutkowski P, Nowecki ZI, Michej W, et al. Risk criteria and Creative Commons Attribution 4.0 International License (http://crea prognostic factors for predicting recurrences after resection of tivecommons.org/licenses/by/4.0/), which permits unrestricted use, primary gastrointestinal stromal tumor. Ann Surg Oncol. distribution, and reproduction in any medium, provided you give 2007;14:2018–27. appropriate credit to the original author(s) and the source, provide a 17. Joensuu H, Eriksson M, Sundby Hall K, et al. One versus three link to the Creative Commons license, and indicate if changes were years of adjuvant imatinib for operable gastrointestinal stromal made. tumor: a randomized trial. JAMA. 2012;307:1265–72. 18. Casali PG, Le Cesne A, Poveda Velasco A, et al. Time to definitive failure to the first tyrosine kinase inhibitor in localized GI stromal tumors treated with imatinib as an adjuvant: a Euro- REFERENCES pean Organisation for Research and Treatment of Cancer Soft Tissue and Bone Sarcoma Group intergroup randomized trial in 1. Joensuu H, Hohenberger P, Corless CL. Gastrointestinal stromal collaboration with the Australasian Gastro-Intestinal Trials tumour. Lancet. 2013;382:973–83. Group, UNICANCER, French Sarcoma Group, Italian Sarcoma Group, and Spanish Group for Research on Sarcomas. J Clin Oncol. 2015;33:4276–83. Features of Ruptured GISTs in Clinical Practice 1969 19. Rutkowski P, Bylina E, Wozniak A, et al. Validation of the 23. Verduijn M, Grootendorst DC, Dekker FW, et al. The analysis of Joensuu risk criteria for primary resectable gastrointestinal stro- competing events like cause-specific mortality-beware of the mal tumour–the impact of tumour rupture on patient outcomes. Kaplan–Meier method. Nephrol Dial Transplant. 2011;26:56–61. Eur J Surg Oncol. 2011;37:890–6. 24. DeMatteo RP, Ballman KV, Antonescu CR, et al. Long-term 20. Bischof DA, Kim Y, Behman R, et al. A nomogram to predict results of adjuvant imatinib mesylate in localized, high-risk, disease-free survival after surgical resection of GIST. J Gas- primary gastrointestinal stromal tumor: ACOSOG Z9000 (Al- trointest Surg. 2014;18:2123–9. liance) intergroup phase 2 trial. Ann Surg. 2013;258:422–9. 21. Kim JS, Kim HJ, Park SH, et al. Computed tomography features 25. Ng EH, Pollock RE, Munsell MF, et al. Prognostic factors and predictive findings of ruptured gastrointestinal stromal influencing survival in gastrointestinal leiomyosarcomas. Impli- tumours. Eur Radiol. 2017;27:2583–90. cations for surgical management and staging. Ann Surg. 22. Yamamoto K, Tsujinaka T, Takahashi T, et al. Impact of the 1992;215:68–77. Japanese gastric cancer screening system on treatment outcomes 26. Joensuu H, Eriksson M, Hall KS, et al. Risk factors for gas- in gastric gastrointestinal stromal tumor (GIST): an analysis trointestinal stromal tumor recurrence in patients treated with based on the GIST registry. Ann Surg Oncol. 2015;22:232–9. adjuvant imatinib. Cancer. 2014;120:2325–33.

Journal

Annals of Surgical OncologySpringer Journals

Published: May 11, 2018

References

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off