Assessing Safety of the Transition to Generic Imatinib in Patients with Gastrointestinal Stromal Tumours

Abstract

Introduction

Following patent expiration of branded imatinib (Glivec®), all Dutch patients with gastrointestinal stromal tumours (GIST) switched from Glivec to generic forms. Following this switch, many patients reported new symptoms. Therefore, we conducted this observational study to assess safety of generic imatinib among patients with GIST in the Netherlands.

Methods

We included patients with GIST from four hospitals that switched from Glivec to generic imatinib. Within these patients, adverse events (AEs) without the switch to a generic were compared with AEs after the switch using a self-controlled case series design. The reference group was formed by the subset of patients who used imatinib for at least 1 year prior to the switch. As potential causes of increased AEs, we reviewed excipients and analysed plasma trough levels from 1 year prior to 1 year after the switch.

Results

In total, 201 patients switched to three generics: Accord® (n = 107), Amarox® (n = 81), and Sandoz® (n = 13). In the reference group (n = 150), 21.3% experienced new AEs, compared with 29.9–34.6% in the different generic groups. All patients that switched to Amarox (odds ratio 2.3; 95% confidence interval (CI): 1.2–4.5) and females that switched to Accord (odds ratio 2.7; 95% CI: 1.1–7.0) experienced a significant increase in AEs. Plasma trough levels were similar among all different formulations. Apart from titanium dioxide in Amarox, no additional excipients were used in any generic form.

Conclusions

The transition to generic imatinib in Dutch patients with GIST was safe. After switching to generic imatinib, up to 34.6% of patients experienced new AEs, compared with 21.3% in the reference group, indicating that many AEs may have been mistakenly attributed to the switch. The small increase in AEs that we found was unlikely due to pharmacokinetics or excipients. Therefore, we argue that the nocebo effect, where negative expectations about treatment lead to worsened symptoms, played a large role.

Supplementary Information

The online version contains supplementary material available at 10.1007/s40487-025-00353-3.

Key summary points

After switching from branded to generic imatinib, many patients with gastrointestinal stromal tumours (GIST) in the Netherlands reported new adverse events.

To objectify adverse events and assess the safety of the transition to generic imatinib, we conducted this retrospective observational study.

Many events following the switch were likely mistakenly attributed to it, as imatinib-related events also occurred frequently in our reference group, without a switch.

Between branded and generic imatinib, excipients were largely similar, and there was no difference in plasma trough levels.

We hypothesize that some adverse events after switching to generic imatinib can be attributed to the nocebo effect.

Our findings support that the switch to generic imatinib among patients with GIST in the Netherlands was generally safe.

Introduction

Imatinib is a tyrosine kinase inhibitor that is the cornerstone in the treatment of chronic myeloid leukaemia (CML) and gastrointestinal stromal tumours (GIST) [1, 2]. It was originally marketed under the brand name Glivec®, with patents lasting until 2016 for CML and 2021 for GIST [3]. After the patent expiration in 2021, all patients with GIST in the Netherlands switched to generic forms of imatinib. Shortly after, the Dutch Sarcoma Patient platform conducted a questionnaire among patients who had switched to these generic forms. Unpublished results indicated that of the 70 patients who responded, 46% experienced increased adverse events (AEs). This prompted further investigation to determine the extent of the issue and whether potential causes such as altered plasma concentrations or different excipients played a role.

In terms of plasma concentrations, the European Medicines Agency (EMA) requires generics to be bioequivalent, defined as the confidence interval of both the maximum concentration of the drug after administration (C_max) and area under the curve (AUC) being within a range of 80–125% of the reference product [4]. However, despite meeting bioequivalence criteria, a generic product can still differ significantly from the reference product, or more importantly, differences in exposure can be clinically relevant. For example, tacrolimus and anti-epileptic drugs have a small therapeutic window, wherein small changes can hamper efficacy or lead to increased AEs [5–7]. For these drugs, plasma concentrations are therefore strictly monitored. Regarding imatinib specifically, firm management of plasma levels has not yet proven necessary. However, both treatment response and toxicity are known to be exposure-related [8]. Therefore, therapeutic drug monitoring has been practised in the Netherlands for several years. Patients regularly have their imatinib plasma levels measured, allowing for dose adjustments to maintain the target trough level of ≥ 1100 ng/mL. Given the exposure–toxicity relationship, an increase in plasma concentrations following the switch to generic imatinib might have led to increased AEs.

For generic imatinib, studies on both exposure, efficacy and toxicity in patients with CML have been published since the patent on Glivec for this indication expired in 2016. Some studies reported molecular response rates similar to those in Glivec [9–11]. However, other studies reported that complete hematologic response was lost after switching to generic imatinib, and in a few cases, re-achieved when switched back to Glivec [12–15]. This is remarkable, as a 2021 systematic review concluded that no differences in pharmacokinetics between generic imatinib and Glivec were observed [16–19]. From a safety perspective, some conflicting results were also published. Following the switch to generic imatinib, some studies reported new AEs among 17–39% of patients, while another study reported significant improvement in muscle cramps, oedema, fatigue, and diarrhoea [9–11, 20].

Given the varying results regarding AEs after switching to generic imatinib among patients with CML and concerns raised by the Dutch Sarcoma Patient platform, the safety and exposure of generic imatinib in patients with GIST deserves further evaluation. Therefore, we conducted this observational study to assess safety of the transition to generic imatinib in Dutch patients with GIST.

Methods

Study Design and Patient Selection

In this observational study, we included all patients from four Dutch hospitals (Leiden University Medical Center, Netherlands Cancer Institute, Radboud University Medical Center, and University Medical Center Groningen) who switched from Glivec to generic imatinib. For baseline characteristics, we used the Dutch GIST registry, a prospective nationwide database of five GIST expertise centres. We excluded patients if they underwent simultaneous dose adjustments, as an increased dose could introduce new AEs mistakenly attributed to the switch, while a decreased dose could mask potential AEs. The study received approval from the Internal Review Board of the Netherlands Cancer Institute (IRBd24-086) and was conducted in accordance with the principles outlined in the Declaration of Helsinki. All patients provided written informed consent for retrospective use of their data for research purposes.

We compared AEs occurring without switch with a generic form to those after a switch using a self-controlled case series design, where patients served as their own controls [21]. To assess new or worsened AEs, we reviewed the records of patients who had been on Glivec 1 year before the switch, as if they had transitioned to the generic form (Fig. 1). This approach helped distinguish whether changes in AEs were due to chance or the switch itself.

Fig. 1.

Adverse event scoring and self-controlled case series design; new or worsened AEs were assessed during 6 months after the switch, on the basis of applied criteria; start of Glivec/reference assessment period was exactly 1 year before the actual switch; AEs adverse events

AEs were classified according to the Common Terminology Criteria of Adverse Events version 5 (CTCAEv5) [22]. During the first 6 months after the switch, an AE was considered new or worsened if it met at least one of the following criteria: (1) the AE was reported within 2 months after the switch, or (2) it was documented at the first consultation after switch, or (3) the physician explicitly attributed it to the generic medication. All other events such as painful ingestion or unpleasant taste were recorded separately, as they may have influenced subsequent switches to another generic form. All AEs were retrospectively reassessed by a single physician, and subsequently, all discrepancies were independently reviewed by two additional physicians.

Plasma Concentrations

All standard of care imatinib plasma trough levels from 1 year prior to 1 year after the switch were included in the analysis. Imatinib plasma concentrations were measured with a validated liquid chromatography/tandem mass spectrometry assay [23–25]. Plasma concentrations were either drawn pre-intake as a trough level or were extrapolated to trough levels on the basis of time since last ingestion using log-linear extrapolation [26].

Statistical Analyses

Baseline characteristics were compared using the chi-squared test or Fisher’s exact test for categorical variables and the Kruskal–Wallis test for continuous variables. To account for the repeated measurements of each patient, a mixed-effects logistic regression model was employed to estimate the odds ratio of having any AE in the generic imatinib group versus the Glivec reference group. Backward stepwise model selection was performed to obtain the model with the best goodness of fit according to the Bayesian information criterion. To examine the impact of switching to generic medications on plasma trough levels, a mixed-effects linear model was utilized. Statistical analyses were performed using R version 4.2.3. All mixed effects models were performed using the package lme4 version 1.1–32 and 3.1–3.

There were no missing values in the patient characteristics or adverse event analyses. In the pharmacokinetic analyses, missing values occurred only when the time interval since the last dose intake was unknown, and as a result, no trough level could be extrapolated. These missing data were assumed to be missing at random, and no imputation was performed.

Results

Patients

A total of 201 patients were included in this study, of which 107, 81 and 13 patients switched to Accord®, Amarox® and Sandoz®, respectively (Table 1). The Amarox group consisted solely of Netherlands Cancer Institute patients, and Accord was prescribed in the other hospitals, although some patients from the University Medical Center Groningen received Sandoz. Of the total 201 patients, 150 patients already received imatinib a year prior to the switch, forming the reference group. Amarox was relatively frequently prescribed in a neoadjuvant setting and more often taken at a higher dose. All other characteristics were comparable, in particular treatment duration and historical AEs before the switch.

Table 1.

Demographics of three generic groups and the reference group

Characteristics	Accord	Amarox	Sandoz	Reference	p-value
Total patients	107	81	13	150
Centre (%)
NKI	0	81 (100)	0	60 (40.0)	< 0.01
LUMC	37 (34.6)	0	0	24 (16.0)	< 0.01
RadboudUMC	49 (45.8)	0	0	41 (27.3)	< 0.01
UMCG	21 (19.6)	0	13 (100)	25 (16.7)	< 0.01
Female sex (%)	38 (35.5)	34 (42.0)	6 (46.2)	66 (44.0)	0.56
Median age, years (IQR)	69 (61–75)	70 (59–76)	64 (62–72)	68 (59–75)	0.89
Indication (%)
Neo-adjuvant	10 (9.3)	21 (25.9)	2 (15.4)	26 (17.3)	0.01
Adjuvant	25 (23.4)	3 (3.7)	2 (15.4)	14 (9.3)	< 0.01
Palliative	72 (67.3)	57 (70.4)	9 (69.2)	110 (73.3)	0.77
Daily dosea
< 400 mg	10 (9.3)	10 (12.3)	0 (0)	16 (10.7)	0.70
400 mg	81 (75.7)	45 (55.6)	12 (92.3)	99 (66.0)	0.01
> 400 mg	16 (15.0)	26 (32.1)	1 (7.7)	35 (23.3)	0.02
Treatment duration in months (IQR)b	30 (18–71)	33 (19–91)	23 (6–66)	34 (12–81)	0.44
Total AEs before switch (IQR)c	4 (2–6)	4 (3–7)	4 (3–5)	3 (2–6)	0.31

^aDosage at date of the switch

^bRepresents the median duration of treatment until the switch

^cThe median number of imatinib-related adverse events that occurred during full treatment before the switch

NKI Netherlands Cancer Institute, LUMC Leiden University Medical Center, RadboudUMC Radboud University Medical Center, UMCG University Medical Center Groningen, AEs adverse events

Adverse Events

In the Accord group, 29.9% of patients experienced at least one new or worsened AE; this was 34.6% in the Amarox and 30.8% in the Sandoz group (Table 2). In comparison, 21.3% of patients in the reference group experienced AEs. The switch to Amarox was significantly associated with increased odds of developing AEs compared with the reference group at 2.3 (95% CI 1.2–4.5), while it was 1.7 (95% CI 0.9–3.2) for the Accord group and 1.4 (95% CI 0.4–5.7) for the Sandoz group, both of which were non-significant. Fatigue (11.2%) and muscle spasms (7.5%) were most common among patients who switched to Accord, while nausea (9.9%) and diarrhoea (8.6%) were most frequently observed in the Amarox group. Unpleasant taste was solely reported by patients in the Amarox group, which, in combination with AEs, led to a switch to a second generic in 32.1%. In the Accord and Sandoz group, 12.1% and 15.4%, respectively, had a second generic switch. When examining the switch across sexes, a significant increase in adverse events was observed exclusively in women who switched to Accord (Supplementary Table 2). However, when directly comparing both total and individual adverse events between sexes for either generic, no statistically significant differences were found.

Table 2.

Adverse events per generic form and the reference group

Events	Accord (n = 107)	Amarox (n = 81)	Sandoz (n = 13)	Reference (n = 150)
Any new or worsened AE (%)a	32 (29.9)	28 (34.6)	4 (30.8)	32 (21.3)
Odds ratio (95% CIb)	1.7 (0.9–3.2)	2.3 (1.2–4.5)	1.4 (0.4–5.7)	N/A
Individual AEs (%)
Fatigue	12 (11.2)	4 (4.9)	0	6 (4.0)
Skin toxicity	8 (7.5)	3 (3.7)	1 (7.7)	1 (0.7)
Muscle spasms	8 (7.5)	4 (4.9)	0	5 (3.3)
Diarrhoea	6 (5.6)	7 (8.6)	1 (7.7)	5 (3.3)
Flushing	5 (4.7)	0	1 (7.7)	0
Nausea	3 (2.8)	8 (9.9)	0	6 (4.0)
Swollen eyes	0	0	2 (15.4)	0
Other	18 (16.8)	23 (28.4)	2 (15.4)	18 (12.0)
Other events (%)
Painful ingestion	0	1 (1.2)	0	0
Unpleasant taste	0	23 (28.4)	0	0
AEs and other events combined (%)	32 (29.9)	42 (51.9)	4 (30.8)	32 (21.3)
Switch to second generic (%)	13 (12.1)	26 (32.1)	2 (15.4)	N/A

^aRepresents the amount of patients that experienced at least one AE. Some patients had multiple individual AEs

^bThe odds ratio was calculated paired, relative to the reference group; “log(Duration)” was used as a mixed effect in the model

AE(s) adverse event(s), CI confidence interval, N/A not available

Plasma Concentrations

For pharmacokinetic analysis, 390 plasma samples were available during intake of Glivec, 128 for Accord, 144 for Amarox and 57 for Sandoz. The median trough concentration for Glivec was 1016 ng/mL, compared with 1000 ng/mL for Accord, 1102 ng/mL for Amarox and 1048 ng/mL for Sandoz. As both dosage and generic form changed in several patients during the inclusion period, a median of the plasma levels before and after the switch could not accurately be determined. Instead, we applied a linear mixed model, which also allowed us to account for daily dose, age, treatment duration and sex. Among these variables, only dosage was significantly correlated with higher plasma levels. The switch to Accord resulted in a 3 ng/mL decrease, while Amarox and Sandoz showed increases of 5.5 ng/mL and 40 ng/mL, respectively, all of which were not significant nor clinically relevant (Table 3).

Table 3.

Linear mixed effect model for drug plasma trough levels

Covariates	Change in ng/mL	Confidence interval	p-value
Daily dose in mg	1.21	(0.95–1.47)	0.00
Age in years	1.12	(−4.13–6.38)	0.68
Glivec duration in months	0.38	(−0.81–1.57)	0.53
Sex (male)	−54.33	(−171.97–63.31)	0.37
Accord	−2.64	(−81.46–76.18)	0.95
Amarox	5.49	(−61.49–72.47)	0.87
Sandoz	40.41	(−72.08–152.90)	0.48

In this model, 390 samples for Glivec, 128 for Accord, 144 for Amarox and 57 for Sandoz were used

Excipients

Regarding the excipients in the drug formulation, no differences were noted between Glivec, Accord and Sandoz (Table 4). For Amarox, titanium dioxide was used in the tablet coating, which was the only excipient not used in the other formulations. In the core of Amarox, fewer excipients were used compared with the other forms.

Table 4.

Excipients per formulation

Excipient	Amarox	Accord	Sandoz	Glivec
Core
Hypromellose		X	X	X
Cellulose microcrystalline		X	X	X
Crospovidone		X	X	X
Colloidal anhydrous silica		X	X	X
Magnesium stearate	X	X	X	X
Coating
Talc	X	X	X	X
Macrogol 4000	X	X	X	X
Yellow iron oxide	X	X	X	X
Red iron oxide	X	X	X	X
Titanium dioxide	X

All excipients were based on patient information leaflet; X indicates that the excipient was used in the formulation; concentrations are unknown

Discussion

The aim of our study was to assess the safety of the transition to generic imatinib in Dutch patients with GIST. Therefore, we analysed 201 patients who switched from Glivec to generic imatinib. We demonstrated that, depending on the generic form, 29.9–34.6% of patients experienced new or worsened adverse events following this switch, compared with 21.3% in the reference group. Patients who switched to Amarox were significantly more likely to experience new or worsened adverse events compared with the reference group. In addition, female patients who switched to Accord showed a significantly increased risk of adverse events. We demonstrated that plasma imatinib trough concentrations were not affected by the switch, and excipients were largely similar.

To our knowledge, this study is the first to assess the transition to generic imatinib in patients with GIST, but studies on this transition in patients with CML have been published. Compared with these studies, the new AEs reported in 29.9–34.6% of our patients appear high. For example, Scalzulli et al. reported a 20% increase among 168 patients who switched to Accord, and Bonifacio et al. found new or worsened adverse events in 17% of 294 patients who switched to generic Sandoz [9, 10]. A possible explanation for this variation in reported AEs could be differences in the definition of new or worsened events. While we included all reported events regardless of their potential correlation with the switch, other authors might have required possible causation as a condition when documenting new AEs. In addition, the duration of prior Glivec use may have played a role, as most AEs occur during the first 1–3 years of imatinib treatment [27]. While the median duration of Glivec use was 12 and 7.4 years in the two CML studies, respectively, it was only 2–3 years in our patients.

Furthermore, the results of the reference group provide important context for the frequency of AEs we observed. These results demonstrate that new or worsening AEs naturally occur over time, regardless of a formulation change. When patients are aware of a recent switch, they may mistakenly attribute these AEs to the new formulation. The reference group, however, suggests that most of these events are unrelated to the switch.

Nonetheless, we observed a substantial difference in AEs between the reference and generic groups, which was significant among all patients switching to Amarox and females switching to Accord. This indicates that the transition itself had an effect on adverse events. We believe several aspects contributed to this difference in AEs.

First to address is a potential difference in pharmacokinetics, as an increase in drug exposure can lead to increased AEs [8]. However, we demonstrated that after correcting for dose, age, treatment duration and sex, there was no difference in trough levels between Glivec and all three generic forms (Table 3). This is in line with findings in patients with CML , where both drug levels and molecular response were similar [9–11, 16, 17]. It should be noted, however, that we only measured trough levels in this study. Differences in excipients between formulations may have affected absorption kinetics, potentially leading to higher peak plasma concentrations, which could in turn contribute to an increase in adverse events. We conclude that on the basis of trough levels alone, we could not identify a pharmacokinetic explanation for the observed increase in AEs.

Second, a cause of new AEs could be a difference in excipients. For example, some severe anaphylactic cases have been described after a switch from branded to generic treatment, even though literature is in general reassuring [28, 29]. Because the excipients used in Glivec, Accord and Sandoz are identical, only the quantity of these excipients might differ, which seems an unlikely cause of new events. Unlike the other generics, Amarox contains titanium dioxide, which is an excipient that is widely used in several thousands of authorized products according to the EMA [30]. The highest quantities of titanium dioxide are found in candies, sweets and pastries [31]. As an excipient in drug formulations, the used quantities and exposure levels are considerably less than those used in foods [32]. Therefore, the use of titanium dioxide in Amarox tablets seems unlikely to have contributed to the reported AEs. Amarox tablets also differed in the absence of several excipients in the core. Absence of, for example, microcrystalline cellulose, a binding agent, could have led to rapid oral disintegration of the tablet, resulting in an unpleasant taste, which was reported in approximately one third of the patients in the Amarox group [33]. The impact of this should not be underestimated, because the high percentage of these patients who switched to another generic may be attributed to this unpleasant taste. Although we attempted to separate the unpleasant taste from the adverse events, it likely caused nausea in some patients, which could explain why it was more frequently reported compared with the other generics. In conclusion, while most generics used the same excipients, we think that the lack rather than the presence of certain excipients in Amarox tablets contributed to some new AEs.

Lastly, a potential cause of the increase in AEs is the nocebo effect, which describes worsened symptoms caused by negative expectations about the treatment. A 2024 systematic review by Car et al. investigated the nocebo effect across several biosimilar switching studies. Although biosimilar and generics are not the same, this review describes interesting findings regarding the nocebo effect: while all studies observed no changes in immunogenicity, efficacy nor safety, 4.8–28.2% of patients discontinued treatment [34]. Awareness of receiving a different formulation may provoke new symptoms or heighten patients’ attention towards symptoms they already have [35]. Regarding generic imatinib, one would expect that some of the most common imatinib-related adverse events, such as diarrhoea, nausea, muscle cramps and skin toxicity, are also the most reported new AEs after switching to a generic, which is exactly what we observed [36]. These are also the most frequently reported adverse events in studies assessing the same generics among patients with CML [9–11, 20]. Car et al. also describe how communication between patients on social media can play a pivotal role in shaping patients’ experience [34]. Within the Netherlands, patients with GIST maintain an active patient forum, on which patients might have talked about the switch to generic imatinib. Moreover, the questionnaire performed by the patient organization in March 2021, approximately 3 months after switch, while raising valid issues, might also have raised expectations about the treatment being poorly tolerated.

Interestingly, despite the reduced statistical power due to conducting sex-specific analyses, the switch to Accord in women specifically resulted in a significant increase in AEs. Although we were unable to demonstrate a significant difference between men and women, this finding is nonetheless noteworthy. In the hypothetical situation where an increase in statistical power would reveal a true sex-related difference, we consider it unlikely that this would be due to biological factors such as hormonal influences, given that Accord and Glivec contain the same excipients. Instead, we hypothesize that such an effect would more plausibly be attributed to the nocebo effect, which has been shown to be more pronounced in women [37].

We acknowledge certain limitations in our study, primarily due to its retrospective design. The assessment of AEs relied on reports from various physicians, which varied in detail from brief to extensive. As a result, some low-grade adverse events may not have been documented. Furthermore, being aware of AEs of other patients, the physician may have been more proactive in asking patients about specific symptoms. In addition, regarding the high switching rate among patients using Amarox, we could not explore whether this was driven by taste, tolerability or prescriber perceptions due to our retrospective study design. Finally, the exact influence of the nocebo effect could not be determined within this study design, and its magnitude can only be estimated by ruling out alternative explanations.

Conclusions

Driven by patient reports of new AEs after switching to generic imatinib, we conducted this observational study to evaluate the safety of this transition. Over 30% of patients reported new or worsened AEs post-switch. However, given the similarity in excipients and the absence of detectable differences in drug exposure, these AEs can only be partially attributed to formulation differences. Instead, we argue that most resulted from random occurrence misattributed to the switch. In addition, we hypothesize that the nocebo effect, where negative expectations about treatment lead to worsened symptoms, played a pivotal role. On the basis of these findings, we conclude that the transition to generic imatinib among Dutch patients with GIST was safe.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

Joris van Sabben was responsible for the conceptualization, data collection, data analysis, and drafting of the initial manuscript. Maud van der Kleij and Evelyne Roets contributed to the conceptualization and independently reviewed the process during data collection. Renaud Tissier provided support with statistical analyses. Dorieke van Balen, Alwin Huitema, Ingrid Desar, Anna Reyners and Hans Gelderblom contributed to data acquisition. Neeltje Steeghs contributed to conceptualization, data acquisition and supervised the overall project. All authors critically reviewed and approved the final version of the manuscript.

Funding

A research grant for the Dutch GIST registry was received from Novartis, Pfizer, Bayer, Deciphera and Blueprint Medicines, from which the rapid service publication fee of Oncology and Therapy was paid. These funding sources were not involved in the conduct of this research.

Data availability

The datasets generated and analysed during the current study are available from the corresponding author upon reasonable request.

Declarations

Conflicts of interest

Neeltje Steeghs provided consultation or attended advisory boards for Boehringer Ingelheim, Bristol-Myers Squibb, Ellipses Pharma, GlaxoSmithKline, and Incyte. N. Steeghs received research grants from Abbvie, Actuate Therapeutics, Amgen, Anaveon, AstraZeneca, Bayer, Blueprint Medicines, Boehringer Ingelheim, Bristol-Myers Squibb, CellCentric, Cogent Biosciences, Cresecendo Biologics, Daiichi Sankyo, Deciphera, Exelixis, Genentech, GlaxoSmithKline, Iambic, IDRx, Immunocore, Incyte, Janssen, Kling Biotherapeutics, Lixte, Merck, Merck Sharp & Dohme, Merus, Molecular Partners, Novartis, Pfizer, Revolution Medicin, Roche, Sanofi, and Zentalis—all outside the submitted work, with all payment to the Netherlands Cancer Institute. Joris van Sabben, Maud van der Kleij, Evelyne Roets, Renaud Tissier, Dorieke van Balen, Alwin Huitema, Ingrid Desar, Anna Reyners and Hans Gelderblom declare no conflicts of interest.

Ethical approval

The study received approval from the Internal Review Board of the Netherlands Cancer Institute (IRBd24-086) and was conducted in accordance with the principles outlined in the Declaration of Helsinki. All patients provided written informed consent for retrospective use of their data for research purposes.