Health-Related Quality of Life and Side Effects in Gastrointestinal Stromal Tumor (GIST) Patients Treated with Tyrosine Kinase Inhibitors: A Systematic Review of the Literature

Simple Summary Gastrointestinal stromal tumors (GISTs) are the most common mesenchymal tumors of the gastrointestinal tract, mostly driven by activating mutations in KIT or PDGFRα oncogenes. The introduction of tyrosine kinase inhibitors (TKIs) has revolutionized the treatment of GIST resulting in a substantial gain in median overall survival. Nowadays, imatinib, sunitinib, regorafenib, and ripretinib are registered as first, second, third, and fourth-line therapies, and avapritinib is registered specifically for GISTs harboring a PDGFRα exon 18/D842V mutation. As a result, health-related quality of life (HRQoL) has become more relevant for this surviving population, which is increasing in number. In daily clinical practice, the side effects of TKIs and their impact on the daily lives of patients are the main reason for treatment adjustments. Currently, an overview of HRQoL issues and side effects of different TKIs registered for the treatment of GIST is lacking. Abstract Background: The introduction of tyrosine kinase inhibitors (TKIs) has revolutionized the treatment of gastrointestinal stromal tumors (GISTs), resulting in a substantial gain in median overall survival. Subsequently, health-related quality of life (HRQoL) has become more relevant. Here, we systematically review the available literature on HRQoL issues and side effects of different TKIs registered for the treatment of GIST. Methods: A search through five databases was performed. Full reports in English describing HRQoL outcomes and/or side effects in GIST patients on TKI therapy were included. Results: A total of 104 papers were included; 13 studies addressed HRQoL, and 96 studies investigated adverse events. HRQoL in patients treated with imatinib, regorafenib, and ripretinib remained stable, whereas most sunitinib-treated patients reported a decrease in HRQoL. Severe fatigue and fear of recurrence or progression were specifically assessed as HRQoL issues and had a negative impact on overall HRQoL as well as psychological and physical well-being. The majority of studies focused on physician-reported side effects. Nearly all GIST patients treated with a TKI experienced at least one adverse event, mostly mild to moderate. Conclusions: Despite the fact that almost all patients treated with a TKI experienced side effects, this did not seem to affect overall HRQoL during TKI therapy. In daily practice, it are the side effects that hamper a patient’s HRQoL resulting in treatment adjustments, suggesting that the reported side effects were underestimated by physicians, or the measures used to assess HRQoL do not capture all relevant issues that determine a GIST patient’s HRQoL.


Introduction
Gastrointestinal stromal tumors (GISTs) are the most common mesenchymal tumors of the gastrointestinal tract, affecting 10-20 people per million per year [1]. The diagnosis of GIST relies on morphology and positive immunohistochemistry for CD117 (KIT) and/or DOG1 [2,3]. Most GISTs are driven by activating mutations in KIT (75%) or PDGFRα (15%) oncogenes [4][5][6]. The mainstay of treatment for localized GISTs is complete surgical resection. However, it is not uncommon for GIST to have already metastasized to the peritoneum or liver at time of diagnosis. Before the introduction of imatinib, the median survival of patients with metastatic GIST was only a year. In 2001, imatinib was granted accelerated approval by the United States Food and Drug Administration (FDA) as first-line therapy for advanced and metastatic GIST based on response rate [7]. Imatinib significantly changed the prognosis of metastatic GIST patients to a median survival of 57 months [8]. Today, 10-15% of imatinib-treated patients with metastatic GIST are still responding to imatinib after 10 years of treatment [9]. Nevertheless, in most metastatic GIST patients, the disease progresses after 24 months [8,9], often caused by secondary mutations in the KIT gene [10,11]. In 2007, sunitinib was approved as second-line therapy for GIST based on improvement in time to progression shown in an interim efficacy analysis [12]. The FDA approved regorafenib as third-line therapy for GIST in 2013 based on significantly improved progression-free survival [13]. Recently, ripretinib was registered as fourth-line therapy [14], and avapritinib was approved, specifically for GISTs harboring the PDGFRα exon 18/D842V mutation [15,16]. After the success of imatinib in the treatment of advanced and metastatic GIST, imatinib was also approved as an adjuvant treatment for 3 years in patients with high-risk disease in 2008 [17,18].
As a result of the extended survival, aspects regarding health-related quality of life (HRQoL) have become more relevant. HRQoL is a multidimensional concept that includes the patient's perception of the impact of the disease and its treatment on physical, psychological, and social functioning [19]. Clinical outcomes (e.g., response rate, time to progression, progression-free survival) together with patient-reported outcomes are needed to determine the net clinical benefit of TKIs. Nevertheless, until recently, FDA approvals, also for imatinib in both metastatic and adjuvant setting, sunitinib and regorafenib, were only based on objective or physician-reported data.
Overall, TKIs, especially imatinib, are described as tolerable with manageable side effects [20]. Most studies present physician-reported adverse events (AEs), while tools are available to collect patient-reported side effects (e.g., patient-reported outcomes version of the Common Terminology Criteria for Adverse Events (PRO-CTCAE) or MD Anderson Symptom Inventory). In addition to the physical side effects of TKIs, psychological and social challenges may also influence the daily lives of GIST patients. As most patients with metastatic GIST eventually succumb to their disease [9,21], the fear of disease progression is undeniably a challenge for most patients [22]. Metastatic GIST patients may struggle with the side effects of TKI therapy and the consequences of living with cancer with an extended

Sample
The literature search yielded 3455 unique hits; after screening on title and abstract, 264 papers met our criteria for full-text review. After an independent full-text review, 104 papers were included. The flow chart of the selection procedure is shown in Figure 1.

Study Characteristics
All 104 papers were published between 2002 and December 2020. Of the included studies, 8 studies addressed HRQoL, 5 studies reported on both HRQoL and adverse events, and 91 studies focused on adverse events only.

Health-Related Quality of Life (HRQoL)
Thirteen studies assessed HRQoL or a specific HRQoL issue. Studies had a prospective (n = 8), retrospective (n = 3) or qualitative (n = 2) design. The study characteristics and main findings are summarised in Table 1. Fifteen different patient-reported outcome measures (PROMs) were used, of which the EORTC QLQ-C30 (n = 9) was the most commonly used questionnaire. The characteristics of the used measures are described in Table 2.

Imatinib
Four prospective studies reported on HRQoL during imatinib therapy; one study addressed GIST patients in the adjuvant setting, and three studies GIST patients in the palliative setting. In 40 GIST patients receiving adjuvant imatinib, QoL remained stable throughout 60 months of treatment [26]. A cross-sectional study [27] evaluated imatinib adherence in 158 adjuvant GIST patients, 58% of the patients were considered nonadherent, and nonadherence was associated with a low global QoL score. A multicenter study [28] compared global health status in unresectable and metastatic GIST patients at baseline and after 12 months of imatinib therapy. Global health status did not vary significantly, with 20, 15, and 16 of the 51 patients experiencing an improvement, stable, or worsening in global health status, respectively. Subsequently, patients with controlled disease after 1 year of imatinib interrupted or continued imatinib treatment. At 6 months of follow-up, no differences were observed in global health status, functional status, or symptoms scales between the interruption and continuation group [28]. An observational study [29] in a real-world setting reported that after 18 months of imatinib, overall HRQoL was stable in 51.5% and had improved in 25.8% of the 77 GIST patients with unresectable or metastatic disease. In a prospective single-center study [30], metastatic patients received imatinib or placebo after pre-treatment with at least imatinib and sunitinib. At 8 weeks of treatment, there were no differences in global QoL and functioning scales. Cross-sectionally, the pain was significantly better, while nausea, vomiting, appetite loss, and diarrhea were worse in the imatinib group.

Sunitinib
One prospective study [31] assessed HRQoL in sunitinib-treated patients. In this study, 44 patients were treated with either sunitinib or masitinib after progression on imatinib. Global QoL was stable or improved in 5 of the 13 sunitinib-treated patients, but the timing of the longitudinal HRQoL assessment was unclear.

Regorafenib
In a multicenter randomized controlled trial [32], 122 patients received regorafenib. Health utility scores remained stable; neither cycle number nor treatment type (off-treatment vs. regorafenib; placebo vs. regorafenib) significantly influenced health utility. However, confirmed disease progression led to a significantly impaired health status.

Ripretinib
One international randomized controlled trial [33] reported on HRQoL in 129 GIST patients receiving ripretinib or a matching placebo. Overall health, role, and physical functioning remained stable in the ripretinib group compared with a decrease in the placebo group from baseline to cycle two, day 1.

Specific HRQoL Issues
Two cross-sectional studies assessed specific HRQoL issues. Fear of recurrence or disease progression occurred in 52% of the GIST patients [22]. GIST patients with high levels of fear scored lower on global QoL and the subscales of role, emotional, cognitive, and social functioning. They also experienced higher levels of psychological distress and difficulty making plans for the future. Severe fatigue occurred in 30% of the GIST patients compared to 15% in the matched healthy controls [35]. Severely fatigued patients had a lower global QoL, increased impairment in all the functional domains, lower self-efficacy, and more distress. One study [34] assessed the effect of two online interventions on fatigue, distress, and HRQoL. Patients were randomized to an internet-delivered cognitive behavior therapy or online psychoeducation program. Both interventions led to a reduction in fatigue and distress and an increase in global QoL and all functioning domains of HRQoL.

Interviews
More than half of the participants experienced side effects that influenced their daily lives in negative and challenging ways, which urged them to adapt to 'a new normal'. The majority of participants reported the well-known side effects of imatinib, such as (peri-orbital) edema, nausea, diarrhea, muscle cramps, muscle aches, joint pain, tiredness and exhaustion. Many also reported an increased need for sleep, cognitive challenges, reduced sexual desire, as well as poor stress tolerance around the intake of their GIST medication. Although participants struggled with the side effects and the consequences of living with a chronic cancer, half of them considered themselves to be healthy and able to live a normal life.

*
* Represents the overall methodological quality of the study ranging from 1 * to 5 *; 1 * indicating a study of poor quality and 5 * indicating a study of good quality.  health utility (health status/QoL) using a descriptive system that assesses five generic dimensions of health: mobility, self-care, usual activity, pain and discomfort, and anxiety and/or depression.
5 items, score ranges from 1-3 for each item, these health states can be converted to a single summary score, the EQ-5D index score. According to the EQ-5D index, 1.0 represents perfect health and 0.0 represents death.

1
The EQ-VAS records the patient's self-rated overall health on a vertical visual analogue scale.
1 item, range 0-100 from "Worst Possible" to "Best Possible" health, higher scores represent better health.
1 HADS, 1983 [43,44] Hospital Anxiety and Depression Scale To assess psychological distress and detect states of anxiety and depression.
14 items divided into 2 subscales; anxiety and depression. Score ranges from 0-3 for each item. Scores for each subscale range from 0-21, a score of 11 or higher indicates a mental disorder. Higher scores indicate more anxiety, depression, and psychological distress.

2
Mexican adaptation of the HADS, 2015 [45] Mexican adaptation of the Hospital Anxiety and Depression Scale To evaluate distress.
12 items with 6 items on anxiety and 6 on depression.
Score ranges from 0-3 for each item; total score ranges from 0 to 36, and a higher score indicates greater distress.   [36] on 50 GIST patients concluded that patients shared common experiences during each stage of disease management. Patients felt a sense of crisis during diagnosis, followed by hope upon TKI therapy initiation. Over time, they came to adapt to their new lives with GIST while acknowledging that TKI therapy could have an impact on their daily lives. With each follow-up, patients confronted the uncertainty of becoming TKI resistant and the possible need to switch therapy. Disease progression and TKI switching caused patients to revert to crisis and restart their emotional journey. Another, more recent qualitative study [23] among 20 GIST patients with metastatic disease found that more than half of the patients experienced side effects that influenced their daily lives in negative and challenging ways, which urged them to adapt to 'a new normal'. Apart from the well-known side effects of imatinib, patients also reported an increased need for sleep, cognitive challenges, reduced sexual desire, as well as poor stress tolerance around the intake of their GIST medication.

*
* Represents the overall methodological quality of the study ranging from 1 * to 5 *; 1 * indicating a study of poor quality and 5 * indicating a study of good quality.

Imatinib
GIST patients received imatinib in different treatment settings; neo-adjuvant (n = 6), adjuvant (n = 13), or palliative setting (n = 37). In nine studies, patients in different treatment settings were combined, or the treatment setting was unknown. Patients were mainly treated with imatinib 400 mg once a day, but other doses such as 300 mg once a day, 600 mg once a day, or 400 mg twice a day were also prescribed. Median treatment durations varied from 3 days to 9 months in the neo-adjuvant setting, from 181 days to 5 years in the adjuvant setting, and from 2 months up to 8 years in a palliative setting. Hematological adverse events, including anemia, leukopenia, and neutropenia, were common. Furthermore, dermatitis, diarrhea, fatigue, muscle cramps or spasms, nausea, periorbital edema, peripheral edema, and rash were frequently reported non-hematological adverse events. One study in particular [78] compared adverse events before and after dose escalation to 800 mg/day in 133 patients with progression on imatinib 400 mg/day and concluded that anemia and fatigue were more likely to be worse after dose escalation. In six studies, GIST patients were treated with imatinib for a median duration of at least 3 years. In an international randomized controlled trial [18], 198 patients received 3 years of adjuvant imatinib. All patients experienced at least one adverse event, mostly anemia, periorbital edema, diarrhea, nausea, and muscle cramps. Another study [26] assessed if adjuvant imatinib for 5 years was tolerable and efficacious, in which 91 patients were treated for a median duration of 55.1 months, and 46 patients completed the intended 5 years of adjuvant treatment. The most common adverse events in this study were nausea, diarrhea, fatigue, periorbital edema, and muscle spasms. A study [29] in a real-world setting reported on 151 patients with unresectable or metastatic GIST treated for a median duration of 42.6 months. Out of the 151 patients, 148 patients experienced adverse events, mostly diarrhea, asthenia, and eyelid or periorbital edema. A retrospective multicenter study [92] aimed to gain insight into GIST patients with unresectable or metastatic disease responding long-term to imatinib. Of these 58 long-term responders treated with imatinib for 5.5 to 10.4 years, 15 patients experienced new emerging adverse events after ≥ 5 years of treatment, including anemia, fatigue, renal failure, diarrhea, edema, and muscle cramps.

Sunitinib
Twenty-five studies reported on adverse events of sunitinib in patients with unresectable or metastatic GIST who progressed on or were intolerant to imatinib. In twenty-two studies, patients received sunitinib in a fractioned dose, mainly 50 mg a day in a 4-weeks-on-2-weeks-off schedule, and in eight studies, patients were treated with a continuous dose of 37.5 mg (or lower) once a day. The median treatment duration of sunitinib varied from 12.8 weeks to 60 weeks. Hematological adverse events were common, including anemia, thrombocytopenia, leukopenia, and neutropenia. Frequent non-hematological adverse events were nausea, vomiting, decreased appetite, abdominal pain, diarrhea, asthenia, fatigue, hypertension, liver dysfunction, hand-foot syndrome, rash, skin discoloration, and hair color change. In addition, mucositis and hypothyroidism were reported as grade 3 and 4 adverse events. Three studies [126][127][128] focused on the effect of sunitinib on the thyroid function. Respectively, 15 of the 42, 10 of the 24, and 2 of the 17 sunitinib treated GIST patients developed hypothyroidism. One study [136] in 75 metastatic GIST patients was conducted to determine the cardiovascular risk of sunitinib. In total, 35 patients developed hypertension, 8 patients suffered a cardiovascular event, and in 10 of the 36 patients, a left ventricular ejection fraction (LVEF) decline of ≥10% occurred. In a few studies, fractioned dosing was compared to continuous dosing of sunitinib. The overall profile of AEs was comparable, but the incidence of AEs in patients receiving continuous dosing was slightly higher [134,135]. A large treatment-use trial of 1124 sunitinib treaded GIST patients compared modified dosing schedules to the fractioned dosing schedule of 50 mg 4-weeks-on-2-weeks-off. Patients with modified dosing schedules experienced more AEs, both any grade and grade 3/4, but the number of patients that discontinued treatment was lower (26% vs 34%) [121].

Regorafenib
Ten studies investigated patients with advanced or metastatic GIST treated with regorafenib, mostly after failure of imatinib and sunitinib. In eight studies, patients were treated with regorafenib for the first 3-weeks in a 4-week cycle, and in two studies, patients received a continuous dose once a day. The median treatment duration of regorafenib varied from 20 weeks to 60 weeks. The most common adverse event was hand-foot skin reaction (56-92%). Other frequently reported adverse events were diarrhea, fatigue, hoarseness, hypertension, and oral mucositis. One study [147] specifically investigated the incidence of regorafenib-associated hepatic toxicity, in which 5 of the 21 metastatic GIST patients developed (laboratory) hepatic toxicity. In a single-center retrospective study of 28 patients, toxicity and efficacy of regorafenib in fractioned dosing, 160 mg a day for the first 3 weeks of every 4-week cycle, was compared to continuous dosing of 120 mg a day. Despite the small numbers, the study concluded that continuous dosing was better tolerated with comparable efficacy [146].

Ripretinib
One international randomized, placebo-controlled trial [33] reported adverse events in 85 GIST patients with advanced disease who received ripretinib as a fourth-line therapy. The most common grade 1 and 2 adverse events were alopecia, myalgia, nausea, fatigue, and hand-foot syndrome. Grade 3 or 4 adverse events were rare, but lipase increase, hypertension, and anemia of these grades were reported.

Discussion
Imatinib, sunitinib, and regorafenib were all approved for the treatment of GIST based on studies without any HRQoL data. In the two decades that GIST patients have been treated with TKIs, the number of studies addressing HRQoL is remarkably low. Available literature showed that HRQoL in patients responding to imatinib, regorafenib, and ripretinib remained stable, while most sunitinib-treated patients reported a decrease in HRQoL. Imatinib, sunitinib, and regorafenib are also registered for the treatment of other cancer types. Before its approval as a treatment for advanced and metastatic GIST, imatinib was approved for the treatment of chronic myeloid leukemia (CML). In imatinib-treated CML patients, physical functioning and well-being remained stable during 18 months of treatment [148]. Another study investigated whether patients with CML treated with long-term imatinib had a different HRQoL compared to their respective peers without cancer in the general population. The HRQoL of CML patients on imatinib age 60 years or older was comparable with that of their peers, while younger patients and women reported the largest HRQoL differences compared to their peers [149]. Most studies included in this review did not compare HRQoL of GIST patients to a normative population, but one study compared the prevalence and severity of fatigue with matched healthy controls. Severe levels of fatigue were found in 30% of the GIST patients compared with 15% in matched healthy controls and were associated with worse HRQoL. In CML patients treated with long-term imatinib, chronic fatigue was found to be the most important factor limiting HRQoL [150]. Sunitinib is also used in the treatment of metastatic renal cell carcinoma (RCC) and metastatic gastroenteropancreatic neuroendocrine tumors (GEP-NETs). Patients with metastatic RCC treated with sunitinib had a stable overall HRQoL during treatment, but their physical well-being worsened over time [151]. The health status of sunitinibtreated patients with GEP-NETs remained stable during the first six cycles (50 mg a day 4-weeks-on-2-weeks-off) of treatment [152]. These results differ from a study that assessed HRQoL in GIST patients, in which most patients reported a decrease in HRQoL; however, we need to take into account the small number of patients (n = 13), and the timing of HRQoL assessment being unclear. An explanation might be that RCC patients received sunitinib as first-line treatment and therefore have no comparison with other TKI therapies, while GIST patients had prior treatment with imatinib, which is well tolerated in general. A frequently reported side effect in sunitinib-treated patients is hand-foot skin reaction, which negatively impacts HRQoL [153]. Regorafenib is a third-line treatment for GIST patients but is also registered as a treatment for hepatocellular carcinoma (HCC) and metastatic colorectal cancer (mCRC). Hofheinz et al. pooled HRQoL data of patients treated with regorafenib in four trials, including GIST, HCC, and mCRC patients [154]. Across all tumor types, regorafenib significantly delayed the patient's first clinical important deterioration in HRQoL score, with the median time to HRQoL deterioration ranging from 16-24 weeks for regorafenib compared to 8-12 weeks for placebo.
Studies included in this review showed that nearly all patients treated with a TKI experienced at least one adverse event, mostly mild to moderate. Severe adverse events were uncommon but did occur. TKI therapies had different side effects related to agent, dose, treatment duration, age, and ethnicity. A review of the safety data of imatinib in CML and GIST patients showed that imatinib treatment led to similar side effects in both diseases [155]. However, severe nausea and diarrhea were more frequent in GIST than in CML patients when treated with the same dose, and this may be due to the origin of GIST and the fact that GIST patients often had previous gastrointestinal surgery. In line with the studies on GIST patients receiving long-term imatinib, imatinib in CML patients was not associated with unacceptable cumulative or late toxic effects [156]. Although imatinib is generally well tolerated, nausea, edema, and fatigue are the main reasons for dose reductions in imatinib-treated GIST patients [72,76]. Furthermore, in daily clinical practice, dose reductions are applied in case of intolerable side effects to continue treatment and maintain a patient's HRQoL. Sunitinib for GIST can be prescribed in two different schedules, a fractioned dosing schedule of 50 mg a day for 4 weeks followed by 2 weeks off or a continuous dosing schedule of 37.5 mg a day. As previously discussed in the results, patients receiving modified dosing schedules experienced more AEs, but fewer patients discontinued treatment, resulting in a longer median overall survival of 23.5 months compared to 11.1 months in patients receiving a fractioned dosing schedule of 50 mg 4-weeks-on-2-weeks-off [121]. This study underlines the importance of appropriate dose adjustments, resulting in a tolerable prolonged treatment with beneficial clinical outcomes. Alternative dosing schedules of sunitinib were also assessed in patients with RCC. Results suggest that a 2-weeks-on-1-week-off schedule is less toxic with similar efficacy, while there was no benefit in safety or efficacy for continuous dosing compared to a 4-weeks-on-2weeks-off schedule [157]. The toxicity profiles of regorafenib in the treatment of GIST, HCC, and mCRC patients were comparable [158]. Among GIST patients, hand-foot skin reaction (HFSR) was the most common adverse event. The incidence of HFSR varied significantly per tumor type and was 60.2% for GIST, 50.0% for HCC, and 46.6% for mCRC [159]. There is no evidence for a relationship between the incidence of HFSR and previous TKI use or duration of TKI use, the exact molecular mechanisms behind the increased incidence are poorly understood, and the occurrence of HFSR seems rather dose-related. Clinicians mostly prescribe regorafenib in a fractioned dosing schedule of 160 mg a day in a 3-weekson-1-week-off schedule, which often leads to unacceptable toxicity resulting in lowering the dose, intermittent drug withdrawal, or complete drug withdrawal. A meta-analysis of studies focusing on regorafenib-associated AEs reported a significant correlation between the occurrence of adverse events and the recommended dose of 160 mg (3 weeks on-1 week off), while no significant correlation was found at a dose of 120 mg with a similar schedule [160]. Data on the optimal dosing of regorafenib are limited, but these results suggest that a dose of 120 mg might be a better fit. Only one small single-center study in GIST patients compared different dosing schedules of regorafenib and concluded that continuous dosing of 120 mg daily was better tolerated with comparable efficacy [146]. Another study pointed out that dosing of regorafenib and toxicity management is critical, as the median duration of treatment was longer [145], which may lead to a durable clinical benefit. Furthermore, less toxicity during treatment will probably result in a better HRQoL, as patients experience fewer physical complaints and uncertainty of needing to interrupt or stop treatment due to intolerable side effects.
The majority of studies included in this review used a physician-reported measure, mostly the CTCAE, to rate and grade side effects. The CTCAE might not be the appropriate measure to report TKI-related side effects due to the subacute and persisting nature of these side effects. The needed treatment adjustments during TKI therapy in clinical practice underscore the limitations of the CTCAE as a measure for AE reporting. The short and severe toxicities due to conventional chemotherapy differ from the daily and long-lasting lower grade toxicities of TKI therapy [161]. In addition, previous research has shown a gap between physician-reported and patient-reported outcomes. Physicians tend to underreport symptoms, as patients report symptoms earlier and more frequently with worse symptom severity than physicians [162,163]. This phenomenon is also observed in the treatment with TKIs; physicians underestimated the severity of long-term side effects of imatinib in CML patients, in particular for muscle cramps and musculoskeletal pain [164]. In order to create a more complete overview of side effects and symptoms, the patient's perspective is needed. With the continuing improvements in cancer treatment, the use of PROMs in cancer research is considered more important, for example, the use of the PRO-CTCAE to capture patient-reported side effects. It becomes relevant to not only assess treatment effectiveness in terms of objective clinical outcomes (e.g., response, recurrence, and survival) but also in terms of patient-reported outcomes, to determine the net clinical benefit of treatments. This suggests that time to HRQoL improvement and time to sustained HRQoL improvement are potentially important outcomes [165].
In this review, only 13 of the 104 included studies used PROMs. The EORTC QLQ-C30 was the most frequently used PROM, while FACT-G, EQ-5D, and SF-36 were sporadically used. These PROMS are population-generic (e.g., SF-36, EQ-5D) or cancer-generic (e.g., EORTC QLQ-C30, FACT-G) and have the disadvantage that they do not cover all relevant issues for GIST patients on TKI therapy. For instance, (periorbital) edema, muscle pain and cramps, and hand-foot skin reaction are not included in these PROMs, indicating that they may lack content validity. On the other hand, the MDASI-GIST, a PROM that was developed particularly for GIST, was never used. An explanation could be that this questionnaire focuses on nine imatinib-induced side effects and therefore does not cover side effects for other lines of TKI therapy. Recently, the EORTC Symptom Based Questionnaire (EORTC-SBQ) was developed for patients receiving targeted therapy [166]. In the EORTC-SBQ, many TKI-related side effects are included, i.e., periorbital edema, peripheral edema, muscle cramps or pains, pain or soreness in the mouth, and skin problems. The EORTC is also developing a survivorship core questionnaire (EORTC-SURV100) to assess the late effects of cancer diagnosis, treatment, and HRQoL in cancer survivors [167]. Both questionnaires can be used in future HRQoL research in combination with a still to be developed questionnaire addressing GIST-specific issues, such as problems because of changed appearance or the feeling that the impact of having a GIST and the side effects of treatment are not understood by friends or family. By combining these different PROMs, a new measurement strategy is applied to cover all relevant issues of GIST patients on (long-term) TKI therapy.
The strength of this review is that it is the first to provide an overview of the available literature on HRQoL and the side effects of different TKIs used in the treatment of GIST patients. With this review, awareness of potential side effects and their impact on HRQoL is raised. Both health care professionals and GIST patients are provided with information that can have important implications for patients' HRQoL, (shared) decision-making, treatment strategies, and clinical outcomes. Because avapritinib and ripretinib were approved recently, this review does not include studies on avapritinib and only one study on ripretinib. In 2021, the results of the VOYAGER, a phase III study comparing avapritinib with regorafenib as third-line or later treatment in patients with unresectable or metastatic GIST, were published. There was no significant difference in median progression-free survival or treatment-related adverse events between both treatments, but the type of adverse events differ with avapritinib inducing more cognitive effects (25.9% vs 3.8%). Due to the low number of studies investigating HRQoL and incorporating PROMs, a limited overview of HRQoL issues for GIST patients on TKI therapy is presented. Furthermore, presented HRQoL data need to be interpreted with caution, as assessment of HRQoL was often conducted in small samples, after a relatively short duration of treatment, or stopped after disease progression or cross-over. Therefore, important aspects of long-term TKI treatment or treatment discontinuation could be missing.

Conclusions
In conclusion, this review showed that most TKI treated GIST patients experience side effects, mostly mild to moderate, which did not seem to affect overall HRQoL. However, in daily clinical practice, side effects and their impact on the daily lives of patients are the main reason for dose reductions, dose interruptions, and schedule modifications. Treatment adjustments are needed in order to maintain a patient's HRQoL, risking worse clinical outcomes, but pre-emptive toxicity management can result in a longer duration of therapy, hence the importance of HRQoL. On the one hand, this suggests that the reported side effects were underestimated, as most studies used the CTCAE, a physician-reported measure, to rate and grade side effects. Apart from the fact that the CTCAE might not be the appropriate measure to report TKI-related side effects, previous research has shown a gap between physician-reported and patient-reported outcomes. On the other hand, using cancer-generic PROMs might not capture all relevant issues that determine a GIST patient's HRQoL. Therefore, a new measurement strategy should be applied to detect, with more sensitivity, patient-reported side effects, symptoms, and HRQoL issues relevant to GIST patients on TKI therapy.