Pre‐Therapeutic UGT1A1 Genotyping in Breast Cancer Patients Receiving Sacituzumab Govitecan to Improve Safety: A Meta‐Analysis and Recommendation

ABSTRACT

Pre‐therapeutic UGT1A1 genotyping is increasingly performed in patients receiving irinotecan, as its active metabolite SN‐38 is primarily cleared through UGT1A1‐mediated glucuronidation. Patients with the UGT1A1*28/*28 genotype exhibit reduced UGT1A1 activity, leading to increased SN‐38 exposure and a higher risk of adverse events such as neutropenia and diarrhea. Although sacituzumab govitecan contains the same active metabolite as irinotecan, routine UGT1A1 genotyping prior to treatment with this drug is not yet standard practice and is not included in its product information. The aim of this study was to assess whether pre‐therapeutic UGT1A1 genotyping may also benefit patients with hormone receptor‐positive, human epidermal growth factor receptor 2‐negative and triple‐negative breast cancer who are treated with sacituzumab govitecan. A literature search was conducted to identify relevant studies assessing the impact of UGT1A1 genotyping on the safety and efficacy of sacituzumab govitecan treatment. A meta‐analysis was performed on selected studies. Additionally, a pharmacological analysis was performed using public data comparing SN‐38 levels in patients treated with sacituzumab govitecan to those receiving irinotecan. The meta‐analysis shows that grade ≥ 3 adverse events, including neutropenia, febrile neutropenia, and diarrhea, occurred more frequently in patients with the *28/*28 genotype. Furthermore, a statistically significant increased risk was found for developing grade ≥ 3 diarrhea or febrile neutropenia in this group. Although the meta‐analysis was underpowered due to small sample sizes, the pharmacological analysis demonstrated higher SN‐38 levels in patients treated with sacituzumab govitecan, supporting the rationale for UGT1A1 genotyping in this context.

Study Highlights

• What is the current knowledge on the topic?

  • ○
    Pre‐therapeutic UGT1A1 genotyping is standard practice for patients receiving irinotecan, as reduced UGT1A1 activity in individuals with the 28/28 genotype can result in increased SN‐38 exposure and a higher risk of severe toxicity.
  • ○
    Although sacituzumab govitecan contains the same active metabolite, genotyping is not routinely performed before treatment and is also not mentioned in the product information.
• What question did this study address?

  • ○
    This study investigated whether UGT1A1 genotyping may be beneficial for patients with hormone receptor‐positive, human epidermal growth factor receptor 2‐negative, or triple‐negative breast cancer who are treated with sacituzumab govitecan.
• What does this study add to our knowledge?

  • ○
    This study provides the first combined analysis of clinical and pharmacological data supporting the relevance of UGT1A1 genotyping prior to sacituzumab govitecan treatment.
  • ○
    It demonstrates that patients with the 28/28 genotype are at significantly higher risk of severe toxicity and shows that SN‐38 exposure in sacituzumab govitecan therapy may be even higher than with irinotecan.
• How might this change clinical pharmacology or translational science?

  • ○
    These findings support the clinical relevance of UGT1A1 genotyping prior to sacituzumab govitecan therapy.
  • ○
    By identifying patients with the *28/*28 genotype before treatment, clinicians can adjust dosing or enhance monitoring to reduce toxicity.

1. Introduction

Breast cancer (BC) is the most frequently diagnosed cancer in women and remains one of the leading causes of cancer‐related deaths worldwide [1, 2]. BC can be classified into three primary subtypes based on the presence or absence of the following molecular markers: estrogen receptors (ER), progesterone receptors (PR), and human epidermal growth factor receptor 2 (HER2). These BC subtypes are known as hormone receptor (HR)‐positive, HER2‐positive, and triple‐negative breast cancer (TNBC), each with distinct biological features and treatment approaches [3]. Among these, TNBC represents 15%–20% of all BC cases and is the most aggressive subtype [4]. TNBC is defined by the absence of ER, PR, and HER2 expression, which makes hormone therapies and HER2‐targeted treatments ineffective. As a result, treatment options are limited to non‐specific chemotherapeutic drugs. Additionally, TNBC has a poor prognosis, with over half of patients experiencing disease recurrence within 3 to 5 years of diagnosis [5]. This high relapse rate highlights the urgent need for better therapeutic strategies to improve long‐term outcomes for TNBC patients.

One such therapy is the antibody‐drug conjugate sacituzumab govitecan (SG) (sacituzumab govitecan‐hziy, IMMU‐132, Trodelvy), which targets trop‐2, a protein widely expressed on TNBC cells (Figure 1). SG is linked to SN‐38, a topoisomerase I inhibitor. After binding of SG to the cancer cells, SN‐38 is released from a hydrolysable linker. SN‐38 then inhibits topoisomerase I, an enzyme essential for DNA replication and repair, by preventing the re‐ligation of topoisomerase I‐induced single‐strand breaks. These breaks lead to the accumulation of DNA damage, which ultimately triggers cell death. SN‐38 clearance primarily depends on glucuronidation by UGT1A1 (uridine diphosphate glucuronosyltransferase 1A1), making it more water soluble for excretion [6]. This targeted mechanism of SG improves treatment precision by specifically delivering its cytotoxic agent to cancer cells, thereby reducing systemic toxicity compared to conventional chemotherapy. Results from a phase I/II multicenter basket trial (IMMU‐132‐01, NCT01631552) have confirmed these advantages, by demonstrating significant safety and efficacy in patients with metastatic TNBC who had previously received at least two prior therapies [7, 8]. Based on these preliminary outcomes, the US Food and Drug Administration (FDA) granted accelerated approval for SG for use in patients with TNBC, on 22 April 2020 [9]. Following its approval in the US, SG was also approved by the European Medicines Agency (EMA) on 22 November 2021, making it available for use in Europe [10]. Several years later, on February 3, 2023, SG was also approved for patients with unresectable or metastatic HR‐positive, HER2‐negative breast cancer who have received endocrine‐based therapy, and at least two additional systemic therapies [10]. Since its approval, SG has proven to be an effective treatment option for patients with TNBC and HR+/HER2− mBC. Clinical studies, including the ASCENT trial (NCT02574455), have demonstrated that SG significantly improves progression‐free survival and overall survival compared to standard chemotherapy in patients with metastatic TNBC who have received at least two prior lines of treatment [11]. Similarly, the TROPICS‐02 trial (NCT03901339), in which HR+/HER2− mBC patients were assigned either SG or chemotherapy, also reported a statistically significant and clinical advantage over chemotherapy, demonstrating a 3.2‐month improvement in median overall survival while maintaining a manageable safety profile [12, 13].

FIGURE 1.

Mechanism of action of sacituzumab govitecan (SG). SG is composed of a trophoblast cell‐surface antigen 2 (Trop‐2)‐specific humanized monoclonal antibody that is linked via a CL2A linker to the cytotoxic camptothecin SN‐38. The anti‐Trop‐2 antibody binds to Trop‐2 receptors on the surface of the cancer cells, initiating the internalization of SG through endocytosis. The acidic environment in the cells causes the linker to break, resulting in free SN‐38 in the cytosol. SN‐38 then migrates to the nucleus, where it binds to topoisomerase I and inhibits the repair of DNA damage. As DNA replication continues, this blockage causes irreversible single‐stranded DNA breaks, leading to cell death. Additionally, the membrane‐permeable SN‐38 is released into the tumor microenvironment where it diffuses into surrounding cells. This phenomenon, known as the bystander effect, extends the cytotoxic effect (image created with BioRender).

However, there are still concerns about the potential toxicity of SG, particularly in patients with certain genetic variations in the gene encoding UGT1A1. In the promoter region of the UGT1A1 gene, there is a sequence of thymine‐adenine (TA) repeats that regulate gene expression. Most individuals have six TA repeats (UGT1A1*1/*1, the wild‐type allele), but the UGT1A1*28/*28 variant has seven [14]. This extra repeat reduces transcription efficiency and enzyme expression. Likewise, the UGT1A1*6 single‐nucleotide polymorphism decreases enzyme activity. Since UGT1A1 glucuronidates SN‐38, reduced enzyme function leads to accumulation of this active metabolite, increasing the risk of adverse events like neutropenia and diarrhea. Neutropenia raises infection risk and treatment complications. Febrile neutropenia is one of the most serious complications of chemotherapy, as it can result in hospitalization, delay in cancer treatment, and in case of grade 5, even death [15]. A possible solution to minimize these toxicities can be genetic screening prior to treatment. By identifying patients that are homozygous for certain polymorphisms, clinicians can adjust doses or implement closer monitoring to reduce the risk of serious adverse events, ensuring safer and more personalized care.

Genetic screening is already routinely applied in some countries for irinotecan, a chemotherapy drug commonly used for advanced pancreatic and colorectal cancer [16, 17]. Irinotecan contains the same active metabolite as SG, SN‐38. Similar to SG, irinotecan can cause significant toxicity in patients with the UGT1A1 polymorphisms. However, due to its longer market presence, irinotecan has more extensive safety data, including UGT1A1*28‐specific adverse events. Most of the national medicines authorities and guideline working groups in Europe recommend a dose reduction of 25%–30% in homozygous carriers of UGT1A1*28 [18]. The Dutch Pharmacogenetics Working Group (DPWG) has classified UGT1A1 genotyping as “essential” based on the DPWG clinical implication score, indicating it must be performed before treatment [19]. While the FDA also acknowledges the increased risk of toxicity in UGT1A1*28 carriers, genotyping is not routinely implemented in the clinic in the United States. When patients are known to be homozygous for the UGT1A1*28 allele, a reduction of at least one level in the starting dose is recommend by the FDA [20].

To date, SG's safety profile and clinical performance have been reviewed in multiple studies. Gui et al. analyzed real‐world safety data from the FDA Adverse Event Reporting System (FAERS) database, but did not assess adverse events in relation to UGT1A1 genotypes [21]. Sultana et al. reviewed 11 clinical trials and concluded that SG shows good clinical activity with minimal adverse events [22]. They discussed the potential role of UGT1A1 genotyping but noted that too few studies include genotype data to make predictions on this topic. Similarly, Ibrahim et al. reviewed current recommendations on UGT1A1 testing in TNBC patients treated with SG. They found it difficult to formulate uniform guidelines due to inconsistent dosing recommendations among scientific societies and they emphasized the need for prospective studies to fully assess the toxicity and safety profile of SG [23]. None of these reviews performed a pooled analysis of UGT1A1 data from all SG studies, which could address the limited power of individual trials. Moreover, no studies have compared SN‐38 exposure in SG‐treated versus irinotecan‐treated patients. If SN‐38 levels are comparable or higher with SG, this would further support the need for genotyping. This paper aims to evaluate the rationale for pre‐treatment UGT1A1 genotyping in HR+/HER2– and TNBC patients by conducting a meta‐analysis of SG trials and comparing SN‐38 pharmacokinetics between SG and irinotecan.

2. Material and Methods

2.1. Literature Search

2.1.1. Meta‐Analysis

On 12 November 2024, a literature search was conducted to collect all relevant articles related to UGT1A1 genotyping in patients receiving SG for inclusion in the meta‐analysis. First, a broad search was conducted on PubMed using the following terms: “sacituzumab govitecan,” “Trodelvy,” “IMMU‐132,” “irinotecan,” “UGT1A1,” “adverse events,” and “neutropenia.” Following this, a more targeted search focusing on sacituzumab govitecan was conducted across databases including PubMed, Web of Science, Embase, Cochrane Library, Emcare, Academic Search Premier, and Google Scholar. Articles were initially screened based on their titles and abstracts. The remaining articles were screened in full, resulting in a final selection of articles that were included in the meta‐analysis. References from these selected articles were also scanned to identify additional relevant articles that did not come up in the search.

Studies were included if they reported the following inclusion criteria:

1. The association between UGT1A1 genotypes and drug‐related adverse events in patients receiving sacituzumab govitecan.

Articles that were not written in English, reviews, case reports, preclinical (animal and in vitro), duplicates or papers considered not relevant were excluded. Articles on irinotecan were also excluded from this search due to the large number of results.

2.2. Pharmacological Analysis

For the pharmacological analysis comparing SN‐38 levels in irinotecan versus SG, multiple smaller, targeted searches were performed to retrieve relevant articles. In addition to searching databases, FDA and EMA guidelines and product information documents were screened and included when relevant. It should be noted that the irinotecan pharmacokinetic parameters in Table 4 are derived from two clinical studies summarized in the FDA Camptosar prescribing information. As these studies have different designs, this may lead to variability that should be considered when interpreting the PK comparison.

TABLE 4.

Pharmacokinetic parameters of SN‐38.

Drug	C max (ng/mL)	T 1/2 (h)	AUC0‐168h (ng × h/mL)
Sacituzumab govitecan 10 mg/kg	98	18	3696
Irinotecan 125 mg/m2	26	10	1603
Irinotecan 340 mg/m2	56	21	3318

Note: Values were obtained from FDA drug label information of sacituzumab govitecan‐hziy (Trodelvy) and irinotecan hydrochloride (Camptosar) [21,25]. The AUC_0–168h of sacituzumab govitecan was calculated by multiplying the AUC_0–24 by 7.

Abbreviations: AUC, area under the curve; C _max, maximum concentration; T _1/2, half‐life.

2.3. Statistical Analysis

A two‐proportion z‐test power analysis was performed to calculate the statistical power of the individual and combined articles post hoc. This was done by calculating the proportions of patients with a specific genotype experiencing adverse drug events. From these proportions, the effect size (Cohen's h) was calculated and used for the power analysis. A power of ≥ 80% was considered as sufficient. A fixed‐effect (FE) meta‐analysis was then performed to estimate the combined effect of the studies. This was done by calculating log odds ratio (logOR) and its standard error (SE). Finally, forest plots were created for each adverse event that visualize the logOR and confidence intervals for the individual and combined studies. Statistical differences are presented as follows: *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. Data analysis was performed using the statistical program R (version 4.3.2, Posit Software PBC, Boston, MA, USA).

3. Results

3.1. Study Selection

The initial literature search for the meta‐analysis yielded a total of 271 articles from various databases. After removing 66 duplicates, the remaining articles were screened based on their titles and abstracts, resulting in a selection of 16 potentially relevant articles. These articles were fully screened to assess their eligibility according to the inclusion criteria which required an association between UGT1A1 genotypes and adverse events in patients receiving SG. This resulted in the exclusion of 14 articles due to a lack of relevant data on UGT1A1‐related adverse events, leaving 2 studies that met the criteria and were therefore included in the meta‐analysis (Figure 2).

FIGURE 2.

Flowchart of study selection.

3.2. Study Characteristics

In total, 2 studies were included in the meta‐analysis (Table 1). The first study is the ASCENT trial (NCT02574455), a phase 3 study in which the efficacy of SG was compared to the treatment of physician's choice [11]. The trial started in 2017 and was completed in 2020. In this trial, 267 patients with locally advanced or metastatic TNBC were assigned a dose of 10 mg/kg SG, although only 258 patients were included in the safety analysis. The majority of participants in this trial were White (81%), followed by Black (10%) and Asian (5%). For 11 (4%) participants, the ethnicity was listed as other or unknown. The most common UGT1A1 genotype was UGT1A1*1/*1 (42%), followed by UGT1A1*1/*28 (36%) and UGT1A1*28/*28 (13%).

TABLE 1.

Study characteristics.

Rugo et al., 2022	ASCENT (NCT02574455)	267	mTNBC	10	White: 215 (81), Black: 28 (10), Asian: 13 (5), other/unknown: 11 (4)	UGT1A1*1/*1: 113 (42), UGT1A1*1/*28: 96 (36), UGT1A1*28/*28: 34 (13), not done/missing: 24 (9)	Neutropenia, febrile neutropenia, diarrhea
Rugo et al., 2023	TROPICS‐02 (NCT03901339)	272	HR+/HER2– mBC	10	White: 184 (68), Black: 8 (3), Asian: 11 (4), Other/unknown: 69 (25)	UGT1A1*1/*1: 103 (38), UGT1A1*1/*28: 119 (44), UGT1A1*28/*28: 25 (9), not done/missing: 25 (9)	Neutropenia, febrile neutropenia, diarrhea

In the second study, the TROPICS‐02 trial (NCT03901339), 272 patients with HR+/HER2− mBC were assigned a dose of 10 mg/kg SG [12]. From these, only 268 patients were included in the safety analysis. This phase 3 trial, conducted between 2019 and 2023, compared the safety and efficacy of SG with a treatment of physician's choice. In this study, the majority of participants were White (68%), followed by Asian (4%) and Black (3%). The ethnicity of 69 (25%) participants was recorded as other or unknown. The most common UGT1A1 genotype was UGT1A1*1/*28 (44%), followed by UGT1A1*1/*1 (38%) and UGT1A1*28/*28 (9%).

3.3. Patient Characteristics

A total of 539 patients were included in the meta‐analysis, of whom 99% were female and 1% male (Table 2). The mean age was 56 years (±12). The majority of patients were white (74%), followed by black (7%) and Asian (4%). For 80 participants (15%), the ethnicity was unknown. Most patients had either the wild‐type UGT1A1 genotype (UGT1A1*1/*1) or the heterozygous genotype (UGT1A1*1/*28), each accounting for 40% of the patients. The remaining participants were either homozygous for the *28 allele (UGT1A1*28/*28, 11%) or had an unknown genotype. Genotyping for other variants, such as UGT1A1*6 or *93, was not performed.

TABLE 2.

Patient characteristics.

Characteristics	ASCENT (n = 267)	TROPICS‐02 (n = 272)	All patients (n = 539)
Sex, n (%)
Female	265 (99)	270 (99)	535 (99)
Male	2 (1)	2 (1)	4 (1)
Age, years
Mean (SD)	54 (11)	57 (12)	56 (12)
Ethnicity, n (%)
White	215 (81)	184 (68)	399 (74)
Black	28 (10)	8 (3)	36 (7)
Asian	13 (5)	11 (4)	24 (4)
Other/unknown	11 (4)	69 (25)	80 (15)
UGT1A1 genotype, n (%)
*1/*1	113 (42)	103 (38)	216 (40)
*1/*28	96 (36)	119 (44)	215 (40)
*28/*28	34 (13)	25 (9)	59 (11)
Other/unknown	24 (9)	25 (9)	49 (9)

3.4. Safety Analysis

3.4.1. Adverse Drug Events Per UGT1A1 Genotype

In Table 3, the most common adverse drug events are listed for all grades and for grade ≥ 3, divided per UGT1A1 genotype. The ASCENT trial reported treatment‐related adverse events (TRAEs), and the TROPICS‐02 trial reported specifically treatment‐emergent adverse events (TEAEs) when associated with UGT1A1 genotype. TEAEs are defined as adverse events that started on or after the first dose date and up to 30 days after the last dose date.

TABLE 3.

Adverse drug events per UGT1A1 genotype.

Adverse events, n (%)	All patients (n = 530)	UGT1A1 *1/*1 (n = 216)	UGT1A1 *1/*28 (n = 215)	UGT1A1 *28/*28 (n = 59)
All grades
Neutropenia	333 (63)	149 (69)	141 (66)	43 (73)
Febrile neutropenia	29 (5)	9 (4)	13 (6)	7 (12)
Diarrhea	297 (56)	125 (58)	134 (62)	38 (64)
Grade > 3
Neutropenia	255 (48)	106 (49)	113 (53)	36 (61)
Febrile neutropenia	29 (5)	9 (4)	13 (6)	7 (12)
Diarrhea	52 (10)	17 (8)	24 (11)	11 (19)

Note: The severity of the adverse events was graded based on the National Cancer Institute's Common Terminology Criteria for Adverse Events (CTCAE, version 5.0).

In the overall cohort, neutropenia was the most frequently reported adverse event (63%), followed by diarrhea (56%) and febrile neutropenia (5%). In the grade ≥ 3 cohort, neutropenia was again the most frequently reported adverse event (48%), followed by diarrhea (10%) and febrile neutropenia (5%). Since febrile neutropenia starts at grade 3, the frequency of grade ≥ 3 febrile neutropenia was the same as that for all‐grade febrile neutropenia. When divided per genotype, all‐grade neutropenia was most common in the *28/*28 group (73%) followed by the *1/*1 group (69%) and the *1/*28 group (66%). Similarly, grade ≥ 3 neutropenia was most common in the *28/*28 group (61%), followed by *1/*28 (53%) and *1/*1 (49%). Febrile neutropenia of all grades and for grade ≥ 3 was most often reported in the *28/*28 group (12%), compared to the *1/*28 group (6%) and the *1/*1 group (4%). The incidence of all‐grade diarrhea was also higher in the *28/*28 group (64%), followed by *1/*28 (62%) and *1/*1 (58%). For diarrhea grade ≥ 3, most cases were again in the *28/*28 group (19%) followed by *1/*28 (11%) and *1/*1 (8%).

Among all 530 patients included in the safety analysis of both trials, treatment‐emergent adverse events (TEAEs) leading to dose reduction were observed in different proportions based on the UGT1A1 genotype. Specifically, in the *1/*1 group, 21% experienced TEAEs leading to dose reduction. In patients with the *1/*28 genotype, this percentage increased to 31%. The highest frequency of TEAEs requiring dose reduction was reported by patients in the *28/*28 group, with 37%. In the TROPICS‐02 trial, TEAEs that led to treatment disruption or continuation were also reported. In this trial, TEAEs leading to treatment interruption were reported in 68% of patients with the *1/*1 genotype, 64% of those with *1/*28, and 76% in the *28/*28 group. TEAEs leading to treatment discontinuation were relatively rare, occurring in 5%, 6%, and 12% of patients with the *1/*1, *1/*28, and *28/*28 genotypes, respectively.

3.5. Meta‐Analysis

The post hoc power analysis demonstrated limited statistical power for all assessed adverse events (Supporting Information). In ASCENT, power ranged from 14% (diarrhea) to 73% (febrile neutropenia), and in TROPICS from 8% (febrile neutropenia) to 47% (diarrhea). In the combined analysis, power values ranged from 32% (neutropenia) to 48% (febrile neutropenia). In Figure 3, forest plots are shown of all three adverse events for grade ≥ 3. For neutropenia (Figure 3A), both individual studies reported positive logORs. The ASCENT trial observed a logOR of 0.35 (95% CI: −0.39, 1.08), and the TROPICS‐02 trial observed a logOR of 0.52 (95% CI: −0.34, 1.38). These positive logORs suggest a potential risk of developing grade ≥ 3 neutropenia for individuals with the *28/*28 genotype. However, neither logOR was statistically significant. The combined fixed‐effect model resulted in a positive logOR of 0.42 (95% CI: 0.14, 0.98). Although this could again indicate a possible higher risk of neutropenia grade ≥ 3 for the *28/*28 group, this observation was not statistically significant.

FIGURE 3.

Forest plots of grade ≤ 3 adverse drug events. Forest plots of (A) neutropenia, (B) febrile neutropenia, and (C) diarrhea in the ASCENT and TROPICS‐02 trials showing logORs, 95% confidence intervals, and pooled fixed‐effects (FE) models.

For febrile neutropenia (Figure 3B), a positive logOR of 1.68 (95% CI: 0.55, 2.81) was observed in the ASCENT trial. This logOR is statistically significant and therefore indicates an increased risk of neutropenia grade ≥ 3 for individuals with the *28/*28 genotype. In the TROPICS‐02 trial, a negative logOR of −0.48 (95% CI: −2.55, 1.59) was observed. This was however not significant. The pooled fixed‐effect model yielded in a statistically significant positive logOR of 1.19 (95% CI: 0.20, 2.18), again suggesting an increased risk for neutropenia grade ≥ 3 for the *28/*28 genotype compared to the other UGT1A1 genotypes. Looking at diarrhea (Figure 3C), both the ASCENT and TROPICS‐02 trials have a positive logOR of 0.49 (95% CI: 0.57, 1.54) and 1.11 (95% CI: 0.08, 2.13), respectively. The logOR of the TROPICS‐02 trial was statistically significant. The fixed‐effect model also showed a statistically significant positive logOR of 0.81 (95% CI: 0.07, 1.54) indicating an increased risk of developing diarrhea grade ≥ 3 for individuals with the *28/*28 genotype.

Overall, these findings indicate an increased risk of grade ≥ 3 adverse events in individuals with the 28/28 genotype. However, the limited statistical power (< 80%) of both the individual studies and pooled analysis should be considered when interpreting these results.

3.6. SN‐38 Exposure in Patients Treated With SG Versus Irinotecan

The pharmacokinetic parameters of SN‐38 were compared between SG and irinotecan at the recommended dose of 10 mg/kg for SG, and at the weekly dose of 125 mg/m² and the 3‐weekly dose of 340 mg/m² for irinotecan (Table 4). The maximum plasma concentration (C _max) of SN‐38 for SG is 98 ng/mL, which is almost four times that of irinotecan at 125 mg/m² (26 ng/mL) and almost twice that of irinotecan at 340 mg/m² (56 ng/mL). The half‐lives of SN‐38 in SG and the higher irinotecan dose (340 mg/m²) were comparable, while the half‐life for the lower irinotecan dose (125 mg/m²) was approximately half as long. The area under the curve (AUC) over a 1‐week period for SN‐38 was substantially higher in patients treated with SG, at 3696 ng × h/mL compared to only 1603 ng × h/mL for irinotecan at 125 mg/m² and 3318 ng × h/mL for irinotecan at 340 mg/m². This data shows that patients treated with irinotecan, at clinically used doses, have significantly lower SN‐38 levels compared to patients treated with SG. The higher C _max and AUC associated with SG suggest more sustained and effective systemic exposure to the SN‐38, which may contribute to the clinical efficacy. Additionally, the comparable half‐lives between SG and the higher dose of irinotecan also confirm the pharmacokinetic advantage of SG over irinotecan.

4. Discussion

In this meta‐analysis, all available evidence for UGT1A1 genotyping in HR+/HER2– mBC and TNBC patients was collected and assessed. This was done by combining the findings from the meta‐analysis and a pharmacological analysis. Based on the results of these, it can be concluded that pre‐therapeutic UGT1A1 genotyping in HR+/HER2– mBC and TNBC patients receiving SG has the potential to improve safety.

The individual studies included in this analysis lacked sufficient statistical power (< 80%) on their own to draw definitive conclusions. Although the meta‐analysis improved the overall power, it still remained insufficient to provide conclusive evidence (Supporting Information). Despite the limitations, the analysis showed a clear association between grade ≥ 3 adverse events and UGT1A1 genotype. In particular, significant differences were observed for febrile neutropenia and diarrhea, suggesting that individuals with the *28/*28 genotype have a higher risk of developing these adverse events compared to those with other genotypes. These findings are consistent with data from the Pharmacogenomics Knowledge Base (PharmGKB), which notes that individuals homozygous for the UGT1A1*28 allele are at increased risk for neutropenia and diarrhea when treated with SG [24]. Additionally, the EMA and FDA also both indicate that patients with a *28/*28 genotype are at increased risk for neutropenia, febrile neutropenia, and anemia when treated with SG [10, 25].

To further validate and support these findings from the meta‐analysis, the issue was evaluated from pharmacological perspective by comparing SN‐38 levels in patients treated with SG versus irinotecan. The results demonstrate a clear difference in pharmacokinetics between SG and irinotecan in terms of systemic exposure to SN‐38, as shown by the significantly higher C _max and AUC for SG. Specifically, the C _max of SN‐38 for SG was approximately four times higher than that of irinotecan at 125 mg/m² and almost double that of irinotecan at 340 mg/m². Additionally, the higher AUC observed for SG further underscores its potential for sustained therapeutic activity compared to irinotecan. Furthermore, the comparable half‐lives between SG and the higher dose of irinotecan (340 mg/m²) suggest that SG provides consistent drug levels without requiring higher or more frequent dosing. These findings align with a pre‐clinical study by Sharkey et al., which demonstrated that SG delivers up to 136‐fold more SN‐38 to tumors than irinotecan, with tumor‐to‐blood concentration ratios favoring SG by 20‐ to 40‐fold. The study also noted that intestinal exposure to SN‐38 and its metabolite SN‐38G was nine‐fold lower with SG, further supporting its potential to enhance therapeutic efficacy while minimizing off‐target toxicity [26]. Goldenberg et al. also emphasized the clinical relevance of prolonged SN‐38 exposure and an extended half‐life in optimizing therapeutic efficacy for targeted cancer treatments. Their work shows how sustained SN‐38 levels improve anti‐tumor activity by allowing continuous interaction with tumor cells while reducing the need for frequent dosing [27]. Although direct comparisons of SN‐38 levels between SG and irinotecan were not possible, the similar incidence of grade ≥ 3 adverse events in *1/*1 patients receiving SG and *28/*28 patients receiving irinotecan suggests that SG‐treated patients, regardless of genotype, are likely exposed to equal or even higher SN‐38 concentrations than patients treated with irinotecan.

The observed differences in SN‐38 exposure between SG and irinotecan can be largely explained by their distinct mechanisms of action and delivery. Irinotecan acts as a prodrug that needs to be activated in the liver by carboxylesterases, a process that is highly inefficient, with only 2%–5% of the administered dose converted to SN‐38 [17]. This results in relatively low systemic SN‐38 concentrations. In contrast, SG delivers SN‐38 directly to tumor cells via its Trop‐2 antibody, ensuring targeted delivery and gradual release through a CL2A linker with a half‐life of approximately one day [7, 28, 29]. While SN‐38 in irinotecan is rapidly inactivated via glucuronidation, SN‐38 bound to the antibody in SG remains stable in the bloodstream, allowing for prolonged exposure and a higher AUC [30]. Furthermore, the formation of SN‐38G is much lower in SG‐treated patients. With irinotecan, widespread distribution of SN‐38 leads to high levels of SN‐38G, which are 4.5‐ to 32‐fold higher than free SN‐38 [31, 32]. This contributes to enterohepatic recirculation and is associated with delayed diarrhea [33]. In contrast, SG limits systemic SN‐38 exposure, and clinical data shows that only 20%–40% of SN‐38 is found as SN‐38G [27]. Additionally, the bystander effect, where SN‐38 diffuses into nearby cells, increases therapeutic efficacy but may also lead to toxicities such as neutropenia and diarrhea, as surrounding healthy tissues are also exposed [34].

In addition to the pharmacological characteristics of the drugs, individual patient factors, such as genetic variations, can further influence SN‐38 exposure and the associated toxicity profiles. Patients with the *28/*28 genotype have reduced UGT1A1 activity, resulting in higher systemic SN‐38 exposure and increased risk of neutropenia and diarrhea, as observed in the ASCENT and TROPiCS‐02 trials [11, 12]. Bardia et al. and Loriot et al. similarly reported higher toxicity and more dose interruptions in *28/*28 patients treated with SG [8, 35]. This aligns with findings in irinotecan, where multiple meta‐analyses confirmed a higher risk of grade ≥ 3 neutropenia and diarrhea in *28/*28 patients [36, 37, 38, 39]. These toxicity differences may relate to SN‐38 exposure per genotype. If SN‐38 levels in *1/*1 or *1/*28 SG patients match or exceed those in *28/*28 irinotecan patients, UGT1A1 genotyping could be useful for SG treatment. However, direct comparisons were limited. Santi et al. compared SG and irinotecan SN‐38 levels but without genotype stratification [40]. Sathe et al. found no significant correlation, yet *28/*28 patients still had higher AUC and C _max values [41]. In contrast, irinotecan data showed significantly higher SN‐38 exposure and lower glucuronidation in *28/*28 patients across multiple studies [42, 43, 44]. Although direct comparisons of SN‐38 levels between genotypes were not possible, the incidence of grade 3–4 neutropenia is well documented for both drugs and can therefore be compared. In our analysis, grade ≥ 3 neutropenia occurred in 49% of *1/*1 SG patients, a percentage comparable to the 48% observed in *28/*28 irinotecan patients. This suggests that SN‐38 exposure in SG‐treated patients may already be elevated regardless of genotype, which supports the potential value of UGT1A1 genotyping in patients treated with SG.

Recently, a systemic review and meta‐analysis including 999 patients across multiple tumor types further confirmed that UGT1A1*28/*28 homozygotes are at significantly increased risk of severe SG‐related toxicity (OR for grade ≥ 3 events: 7.03; 95% Cl: 3.41–14.50), with higher rates of dose reductions and treatment interruptions [45]. The authors recommend pre‐treatment UGT1A1 genotyping, suggesting an upfront 25% dose reduction in *28/*28 patients. This is in line with the current label, which already advises a similar reduction after toxicity occurs. This study also aligns with our findings in breast cancer and further supports the implementation of genotype‐guided dosing. Evidence from irinotecan treatment further illustrates how UGT1A1 genotyping can improve treatment safety. UGT1A1 genotype‐guided dosing in irinotecan treatment has been supported by several prospective studies. A dose‐finding study by Innocenti et al. showed that the maximum tolerated dose varied by genotype, with *28/*28 patients tolerating lower doses than *1/*1 or *1/*28 patients, yet achieving similar SN‐38 exposure across genotypes [46]. Hulshof et al. observed a 17.5% reduction in febrile neutropenia in *28/*28 patients following dose reduction, along with a cost saving of €183 per patient [16]. Importantly, systemic SN‐38 exposure in the reduced‐dose group remained slightly higher than in patients receiving the full dose, alleviating concerns about reduced efficacy and demonstrating that genotype‐guided dosing improves patient safety while maintaining therapeutic efficacy.

This study has several limitations that should be considered. First, there is insufficient data on SG‐related adverse events divided by the different genotype in HR+/HER2– mBC and TNBC patients. In particular, the *28/*28 genotype group was consistently small in the included studies, emphasizing the need for larger studies to achieve adequate statistical power. Another limitation is the imbalance in ethnicity among the participants, with a significantly higher number of white individuals compared to Asians. Given that the *28/28 genotype is significantly less common in Asian populations compared to White and Black populations, the relevance of these findings to Asian patient populations may be limited. Moreover, this study did not include UGT1A1*6 genotyping, which is more prevalent and clinically relevant in Asian populations, thereby further limiting the applicability of the findings for this group. Furthermore, limited information is currently available on SN‐38 concentrations per UGT1A1 genotype in patients receiving SG. Therefore, more studies investigating the pharmacokinetic profile of SG are needed. Such research could allow direct comparisons of SN‐38 levels between SG and irinotecan, which could lead to more precise recommendations regarding pre‐therapeutic genotyping. A review by Ibrahim et al. on integrating UGT1A1 genotyping into clinical practice also emphasized that UGT1A1 testing could help predict treatment‐related toxicities. However, they also pointed out that the lack of sufficient data leads to inconsistent recommendations and that prospective studies on the toxicity risk and safety profile of SG are therefore warranted to optimize the personalized treatment [23]. Similarly, a retrospective real‐world analysis by Wong et al. emphasized the need for more future trials to conform the association between genotypes and treatment outcomes [47].

Overall, the results of the meta‐analysis and pharmacological analysis support pre‐therapeutic UGT1A1 genotyping as a routine part of clinical practice. The meta‐analysis showed a clear increase in adverse events associated with the UGT1A1 *28/*28 genotype, with 37% of these patients requiring dose reductions due to TEAEs. Starting treatment with a lower dose in this group could effectively reduce adverse events while maintaining therapeutic efficacy. Furthermore, pre‐therapeutic genotyping has already proven effective in genotype‐guided dosing irinotecan treatment. Although SG and irinotecan differ in their mechanisms of action, both are metabolized via UGT1A1, having the same risk on adverse events. And since the SN‐38 levels and frequency of adverse events were much higher in SG compared to irinotecan, it would be a logical step to also perform pre‐treatment genotyping in patients treated with SG. Patients are longer and more exposed to SN‐38 and have therefore a high risk on adverse events. This risk is well recognized by regulatory agencies like the FDA and EMA, who therefore already recommend lower starting doses for *28/*28 patients treated with SG. These recommendations reflect a consensus on the clinical importance of genotyping to minimize risks. Implementing routine genotyping as part of standard care would therefore be essential to optimize treatment safety and efficacy. However, more prospective studies investigating the pharmacokinetics of SG in all UGT1A1 genotypes are needed to provide better recommendations on starting doses.

In summary, it can be concluded that pre‐therapeutic UGT1A1 genotyping followed by genotype‐guided dosing is likely to improve patient safety. Therefore, we recommend that pre‐therapeutic UGT1A1 genotyping for patients treated with SG should be considered as standard care. However, further research with larger patient cohorts investigating the pharmaco‐kinetic and ‐dynamic properties of SG is needed to optimize the personalized treatment recommendations.

Author Contributions

T.G. wrote the manuscript. H.‐J.G. designed the research. T.G. performed the research and analyzed the data.

Funding

The authors have nothing to report.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Data S1: cts70462‐sup‐0001‐supinfo.pdf.

Goedhart T. and Guchelaar H.‐J., “Pre‐Therapeutic UGT1A1 Genotyping in Breast Cancer Patients Receiving Sacituzumab Govitecan to Improve Safety: A Meta‐Analysis and Recommendation,” Clinical and Translational Science 19, no. 1 (2026): e70462, 10.1111/cts.70462.