Unleashing the power of anti-HER2 therapies in metastatic colorectal cancer: paving the way for a brighter future

Abstract

Colorectal cancer (CRC) constitutes a significant portion of cancer-related deaths in the United States, topping cancer-related mortality among males under 50 years of age. Despite an increase in overall survival throughout the years, the prognosis of metastatic CRC remains poor. In addition to the classic molecular targets in CRC that include epidermal growth factor receptor, BRAF, and vascular endothelial growth factor pathways, human epidermal growth factor receptor 2 (HER2) amplification is an emerging target that has been successfully targeted in advanced CRC, particularly with anti-HER2 monoclonal antibodies and tyrosine kinase inhibitors. With the introduction of potent HER2-directed antibody–drug conjugates, the number of patients who might be eligible for anti-HER2 therapy is expected to increase dramatically as tumors with lower expression of HER2 (termed HER2 low) have shown relevant response rates across several histologies. Yet, several challenges remain to be addressed in this field, including the standardized method to define HER2-low disease and the most efficient payload in this setting, among others. Future studies will help to further characterize HER2-low CRCs, identify additional predictive biomarkers, and assist in the development of better treatments.

Highlights

• •HER2 amplification is detected in ∼5% of metastatic CRC with wild-type KRAS.
• •The targeting of HER2 amplification is an established approach in metastatic CRC, with promising results.
• •The emergence of highly potent agents has led to a new concept of HER2-low disease in various malignancies, including CRC.
• •The prevalence of HER2-low disease in mCRC is between 6% and 30%.
• •The future of accurate definition and targeting of HER2 in HER2-low CRC remains to be further elucidated.

Introduction

Colorectal cancer (CRC) accounts for ∼10% of cancer-related fatalities among both men and women in the United States, ranking as the third leading cause of death from cancer. Moreover, CRC is the primary cause of cancer-related death in men below the age of 50 years. Estimates suggest that in 2023, there will be ∼150 000 new cases of CRC, with roughly 10% occurring in individuals younger than 50 years. Roughly one-third of these cases involve the rectum, while 40% occur in the proximal colon.¹

Over the years, the 5-year survival rate for patients with CRC has shown improvement, rising from ∼50% in the 1970s to ∼65% during the period of 2012-2018.² This improvement can be attributed to various factors, including screening measures, advancements in surgical techniques, advancement in the use of chemotherapy and radiation, and the introduction of new drugs such as targeted therapy.

Among the predictors influencing survival in CRC, the most significant one is the stage of disease at diagnosis. According to the Surveillance, Epidemiology, and End Results (SEER) database, localized disease exhibits a 5-year survival rate of 91% whereas the 5-year survival of metastatic CRC (mCRC) is 14%.²

Molecular targets in CRC

There are several molecular targets and pathways that can be targetable in CRC.

The epidermal growth factor receptor (EGFR) is a pathway that promotes proliferation and cell survival.³ Cetuximab and panitumumab, which are Food and Drug Administration (FDA)-approved antibodies, specifically target EGFR and have demonstrated clinical advantages in KRAS wild-type (KRAS^WT) mCRC.^4,5 Indeed, KRAS mutation plays a substantial role in the development of resistance to anti-EGFR treatments.⁶

BRAF, a significant component of the mitogen-activated protein kinase (MAPK) pathway, which is successfully targeted in melanoma, is another targetable alteration in mCRC. BRAF mutations occur in ∼10% of patients with mCRC and are associated with poor prognosis. The most common BRAF mutation is BRAF^V600E.⁷ Studies have shown benefit of BRAF inhibition in mCRC.^8,9 Specifically, encorafenib combined with cetuximab is FDA approved in mCRC harboring a BRAF^V600E mutation. This was based on the BEACON CRC trial which showed that patients with a BRAF^V600E mutation receiving encorafenib plus cetuximab in the second or third line had longer median overall survival (mOS) (8.4 months versus 5.4 months), compared to chemotherapy with cetuximab.⁹

The vascular endothelial growth factor A and its receptor (VEGF-A and VEGFR) are two proteins that play a role in angiogenesis—a pathway that plays important roles in some neoplasms including CRC.³ Bevacizumab is an antibody that targets angiogenesis by binding to VEGF-A and inhibiting its interaction with VEGFR. Bevacizumab has shown clinical benefit with combined chemotherapy.10, 11, 12, 13 Currently, bevacizumab is FDA approved as first- and second-line treatments in mCRC. In addition, aflibercept is a recombinant fusion protein that inhibits angiogenesis by binding to and inhibiting the activity of VEGF-A and VEGF-B. Aflibercept, when combined with fluorouracil, leucovorin, and irinotecan, is FDA approved in mCRC that is resistant to or has progressed under a regimen containing oxaliplatin. This is based on the VELOUR study that showed an OS benefit.¹⁴

Mismatch repair deficiency (MMR-D) tumors also represent a specified group that can benefit from specific treatment options. MMR-D occurs when the cells are unable to recognize and consequently repair spontaneous mutations. As a result, MMR-D is associated with two other molecular markers—microsatellite instability high (MSI-H) and high mutational burden. In the context of CRC, ∼15% of all CRC patients and 4%-5% of mCRC cases exhibit MMR-D. Data from numerous studies suggest that tumors with MMR-D display unique clinicopathological features such as location (right colon), mucinous histology, and better prognosis.¹⁵ However, in the setting of mCRC, MMR-D has been linked to a poorer response to chemotherapy and less favorable prognosis.¹⁵

In recent years, several studies have shown the benefit of targeting programmed cell death protein 1 (PD-1) in chemo-refractory mCRC with MMR-D/MSI-H.^16,17 In fact, pembrolizumab is FDA approved as a first-line treatment in MMR-D/MSI-H mCRC, based on data from KEYNOTE-177 showing that PD-1 blockade is superior to chemotherapy in the first-line setting of MSI-H mCRC.^18,19

Interestingly, in case of stage II-III rectal cancer with MSI-H, a study using another anti-PD-1 blocker, dostarlimab, as a neoadjuvant therapy resulted in 100% complete response, further strengthening the potential benefit of front-line PD-1 axis blockade in non-metastatic CRC with MMR-D∖MSI-H.²⁰

Another rare but targetable molecular feature in mCRC is the presence of fusions involving one of the neurotrophic tyrosine receptor kinase (NTRK) genes. NTRK 1-3 encode tropomyosin receptor kinase a-c, respectively, and play roles in neuronal growth and homeostasis. However, NTRK fusions result in overexpression or constitutive activation, leading to oncogenesis through several pathways, including RAS/MAPK and PLCγ, among others.²¹ Across all stages, NTRK fusions occur in <1% of cases and are associated with specific genomic features such as MSI-H and a high tumor mutational burden (TMB).^22,23 Despite their rarity, several studies have demonstrated the high efficacy of NTRK inhibitors in a tissue-agnostic manner, including in patients with mCRC.^24,25

Lastly, the proto-oncogene RET is another rare, yet targetable feature in CRC. RET is a proto-oncogene upstream to well-known oncogenic pathways such as RAS/MAPK and PI3K/AKT.²⁶ In the context of mCRC, RET alterations such as fusions occur in <1% of tumors.²⁷ Notably, RET inhibitor received an accelerated approval in a tumor-agnostic manner, based on LIBRETTO-001 which also included patients with mCRC.²⁸

HER2 Amplification

Another molecular target that has been investigated in recent years is the amplification of the human epidermal growth factor receptor (ERBB2), also known as HER2. ERBB2 is a part of the EGFR family, which consists of ERBB1 (EGFR), ERBB3, and ERBB4. ERBB2 is involved in various physiological and oncogenic processes, including proliferation and the inhibition of apoptosis.^29,30 Like other members of the EGFR family, HER2 is a receptor tyrosine kinase consisting of an extracellular domain, a transmembrane domain, and a cytoplasmic domain.³¹ However, unlike the other EGFRs, HER2 lacks a known ligand. Despite the absence of a ligand, the activity of HER2 is regulated through either homodimerization or heterodimerization with other receptors within the EGFR family.³² Dimerization of HER2 leads to the activation of downstream processes, including the MAPK and signal transducer and activator of transcription pathways.

ERBB2 amplification is found in approximately 2% of all CRCs and 5% of mCRCs with KRAS^WT.³³ There is an ongoing debate regarding its prognostic implications, as studies have shown contradicting results.^34,35 Nevertheless, ERBB2 amplification has been associated with specific molecular features, such as a low TMB and fewer KRAS mutations.³⁶

Clinically, ERBB2 amplification is more common in the distal colon than in the proximal colon. In addition, ERBB2 amplification is associated with higher rates of brain metastasis and poorer response to EGFR inhibitors such as cetuximab.37, 38, 39 This can also affect treatment outcome as many classic anti-HER2 agents are antibodies that do not fully cross the blood–brain barrier.³³

Though not fully established in CRC, several studies have shown a potential benefit using anti-HER2 treatment in CRC with HER2 amplification.40, 41, 42, 43, 44 The MyPathway study was a phase IIa study where 57 patients with refractory mCRC and HER2 amplification received treatment with trastuzumab + pertuzumab.⁴⁰ The primary endpoint was objective response rate (ORR). In this study, the ORR was 32% (18/57). The median duration of response was 5.9 months. The mOS observed in the study was 11.5 months, which aligns with the survival rates reported in other studies exploring novel treatment options for refractory mCRC.^45,46

Like the MyPathway study, the TAPUR study aimed to evaluate the efficacy of trastuzumab and pertuzumab in patients with mCRC who had HER2 amplification and no other available standard treatment options.⁴¹ Notably, this study also examined the impact of trastuzumab and pertuzumab in HER2-mutated tumors. The primary endpoint of the study was disease control (DC), which was assessed at the 16-week mark. Overall, a total of 28 patients with HER2 amplification were enrolled in the study. The results of this trial are considered positive in HER2-amplified tumors, with a DC rate in this group determined to be 54%. The median progression-free survival (mPFS) was 17.2 weeks, and the mOS was 60 weeks. It is noteworthy that, in the population of ERBB2/3-mutated tumors, the trial is considered negative, with a DC rate of 10%, an mPFS of 9.6 weeks, and an mOS of 28.8 weeks.

The TRIUMPH study examined the effects of trastuzumab and pertuzumab in patients who were intolerant or with mCRC refractory to fluoropyrimidine, irinotecan, oxaliplatin, and an anti-EGFR agent.⁴² Overall, the study enrolled 30 patients. HER2 status was evaluated using both immunohistochemistry (IHC) and circulating tumor DNA testing, and there was a significant overlap between the groups. The primary endpoint of the study was the ORR, which was found to be 30% in patients with HER2 positivity based on IHC. The PFS was 3.1 months, while the mOS was 8.8 months.

Zanidatamab is a novel bispecific anti-HER2 antibody that targets the extracellular and dimerization domains of HER2. In a recent phase I study, patients with different types of locally advanced or metastatic HER2-amplified or HER2-expressing cancers were enrolled to receive zanidatamab.⁴⁷ This study included two phases: a dose escalation and a dose expansion. In the second phase, patients with CRC, biliary, and other cancers with HER2 amplification were enrolled to receive zanidatamab. A total of 28 CRC patients were included, of which 71% had HER2 +3 on IHC, and 29% had HER2 +2 on IHC/in situ hybridization (ISH) positivity. Two patients were still on treatment at the time of analysis and therefore were not included in the tumor response analysis. The ORR was 38%, the DC rate was 77%, and the PFS was 6.8 months. Thus, the potential benefit of anti-HER2 antibodies may not be limited to trastuzumab and pertuzumab alone. The potential benefit of anti-HER2 agents is not limited solely to antibodies. Several studies involving tyrosine kinase inhibitors also showed some potential. For instance, HERACLES A was a phase II study that tested the efficacy of lapatinib + trastuzumab in 27 patients with mCRC KRAS^WT who were chemo-refractory. The primary endpoint was ORR which was 30%. The mPFS and mOS were 21 weeks and 46 weeks, respectively.⁴³

The more recent MOUNTAINEER trial was a phase II study that compared tucatinib + trastuzumab to tucatinib only in patients with chemo-refractory, KRAS^WT locally advanced or metastatic CRC. The primary endpoint was ORR which was 38.1% and 3.3% in patients receiving tucatinib + trastuzumab and tucatinib, respectively.⁴⁴ DC was achieved in 16 out of 27 (59%) patients who received trastuzumab plus tucatinib. The mPFS and mOS among patients receiving tucatinib + trastuzumab were 7 months and 24.1 months, respectively.

Several important considerations arise from the aforementioned studies. Firstly, it is worth noting that all the studies discussed are classified as phase II trials, and phase III trials are currently underway and actively enrolling participants. Secondly, it is important to acknowledge that none of the studies included a control arm for comparison. Consequently, additional research comparing the treatment outcomes of these anti-HER2 agents with the current best available treatment options will be instrumental in determining their efficacy and benefits within this context.

It is worth noting that the MOUNTAINEER trial demonstrated a numerically superior mOS compared to the MyPathway, TAPUR, and TRIUMPH trials, although OS was not the primary endpoint in these studies. Furthermore, in the MOUNTAINEER trial, the combination of tucatinib and trastuzumab was not directly compared to trastuzumab alone.

Interestingly, the MOUNTAINEER findings contrast with the results of the HERACLES A trial, which did not show an OS benefit when comparing tucatinib plus trastuzumab to trastuzumab alone (though not in a direct comparison). This contradiction may be explained by the superiority of tucatinib over lapatinib. This hypothesis is further strengthened by the increase in mOS and mPFS in the TRIUMPH trial compared to the HERACLES A trial. Furthermore, tucatinib exhibited a better anti-proliferative profile in HER2-overexpressing cell lines (though from breast origin).^48,49 However, inferring tucatinib’s superiority over lapatinib is limited without a direct comparison.

These points underscore the necessity for further investigations and direct comparisons of various treatment modalities within the same study. This approach is crucial to acquire a comprehensive understanding of their respective efficacies.

Lastly, a more recently developed drug class is the antibody–drug conjugates (ADCs). These drugs consist of an antibody targeting a tumor-related antigen, a cytotoxic drug called a payload, and a linker connecting the two.⁵⁰ With this design, ADCs work through several mechanisms. Firstly, the antibodies have antitumor properties by inhibiting various cellular pathways. Additionally, the cytotoxic drugs can be delivered with higher specificity to the tumor while sparing other tissues. Thus, ADCs represent a more complex class of drugs and may be referred to as targeted chemotherapy.⁵¹ There are several effective targets for this drug class, including HER2 with significant benefit in some histological types of metastatic breast cancer. There are several trials which showed benefit of this class of drugs in other malignancies.52, 53, 54, 55, 56

In the context of CRC, there is a lack of data; however, a case report highlighted the experience of a 58-year-old patient with HER2-amplified mCRC harboring a KRAS mutation.⁵⁷ The patient had progressed on multiple treatment lines, including trastuzumab + pertuzumab. As a last resort, the patient was administered trastuzumab emtansine (T-DM1), an antibody conjugated to an anti-microtubule agent targeting HER2. Interestingly, the patient achieved a PFS of 4 months, which was longer than the PFS reported in the MyPathway trial for patients with mutated KRAS (4 months versus 1.4 months).

However, in a larger study, the HERACLES B study, the primary endpoint of an ORR of 30% or greater was not achieved.⁵⁸ This study included 31 patients with KRAS^WT, BRAF wild-type (BRAF^WT) HER2-positive mCRC who had tumors that were refractory to chemotherapy. The patients received T-DM1 + pertuzumab. In the study, the ORR was 9.7%, the DC rate was 77.1%, and the mPFS was 4.1 months.

Another ADC that has been investigated in the context of HER2-positive mCRC is trastuzumab deruxtecan (T-DXd). The DESTINY-CRC01 study, a phase II trial, was conducted to assess the efficacy of T-DXd in patients with HER2-expressing mCRC who had undergone at least two lines of therapy.⁵⁹ In this study, a total of 86 patients were enrolled and assigned to three cohorts, with cohort A consisting of 53 patients with HER2-positive tumors. A subset of these patients had previously received anti-HER2 treatment (30.2%). Cohort B consisted of patients with IHC +2 and FISH negative, and cohort C included patients with IHC +1. The primary endpoint of the study was the ORR, which was reported as 45.3% in cohort A. The mPFS and mOS were reported as 6.9 months and 15.5 months, respectively.

Interestingly, data suggest that lower doses of T-DXd may also be useful in mCRC. In the DESTINY-CRC02 trial, 82 patients received T-DXd at a lower dose of 5.4 mg/kg, and they were compared to 40 patients who received T-DXd at a regular dose of 6.4 mg/kg.⁶⁰ All patients in this trial were HER2 positive, which was similar to group A in the DESTINY-CRC01 trial.⁵⁹ There was no statistically significant difference between the groups in terms of duration of response (5.5 months for both groups) or mPFS (5.8 months for the low-dose group and 5.5 months for the regular-dose group).

Additionally, the DESTINY-CRC02 trial enrolled more patients with mutated RAS, with ∼15% of patients having mutated RAS, in contrast to <3% in DESTINY-CRC01 (cohort A). T-DXd has demonstrated antitumor activity in almost 30% of patients with mutated RAS. These data align with the case report described above and suggest that T-DXd may also be beneficial in RAS-mutated HER2-positive mCRC.

Altogether, these findings suggest that ADCs may hold potential in the treatment of HER2-positive mCRC. In fact, T-DXd has received Breakthrough Therapy designation (BTD) from the FDA for HER2-positive mCRC patients who have undergone at least two lines of therapy, based on the results of the DESTINY-CRC01 and DESTINY-CRC02 trials. Interestingly, this BTD applies exclusively to patients with an IHC score of +3, excluding those with +2/ISH-positive results, despite both trials classifying IHC +2/ISH-positive cases as HER2-positive tumors. However, further studies with larger cohorts and appropriate control arms are needed to gain a better understanding of the therapeutic benefits of these drugs and their suitability for specific patient populations. A schematic representation of potential treatments for HER2-positive tumors is shown in Figure 1.

Figure 1.

A schematic presentation of the prevalence of HER2 expression in CRC. In the blue box are HER2-directed treatments that have shown potential benefit in CRC.

CRC, colorectal cancer; HER2, human epidermal growth factor receptor 2; ISH, in situ hybridization; T-DXd, trastuzumab deruxtecan.

HER2 Low—a distinct entity in CRC?

The definition of HER2 positivity, though varied from different malignancies, was based on clinical data which showed that anti-HER2 treatments like trastuzumab and pertuzumab provided a clinical benefit in the context of breast cancer.^61,62 However, it is noteworthy to mention that this finding is not well established in CRC.

Interestingly, in the context of breast cancer, the newly developed ADCs have demonstrated promising results in a subgroup of patients who were initially classified as HER2 negative.⁶³ Though conducted on a smaller cohort using a surrogate endpoint, promising results were also shown in the context of gastric cancer.⁵⁹

Additionally, the DESTINY-PanTumor02 trial enrolled patients with various types of locally advanced or metastatic cancer to receive T-DXd.⁶⁴ In this trial, patients with HER2 of +3 or +2 on IHC were included and ISH was not reported. Patients with CRC were excluded from this trial. The ORR of the entire patient population was 37.1% and it increased to 61.3% in tumors with HER2 +3 IHC. In the HER2 +2 group, the ORR was 27.2%. These data suggest that T-DXd may also provide some benefit in tumors that are not HER2 positive. It is noteworthy that, based on this trial, T-DXd has received BTD for tumors with HER2 +3.

These observations have led to the emergence of a new concept called HER2-low disease. As a result, the classification of HER2 status has expanded to three categories: HER2 positive, HER2 low, and HER2 0. HER2 positive is defined by IHC +3 or IHC +2 with a positive ISH test, HER2 low is defined as IHC +1 or IHC +2 with a negative ISH result, and HER2 0 is defined as HER2 IHC 0. With these revised definitions, the proportion of patients classified as HER2 0 has decreased (discussed below).

The utility of HER2 low as a predictive biomarker to T-DXd in breast cancer has prompted investigations into its potential applicability in CRC. Over the past few years, several studies have aimed to determine the prevalence, molecular signature, prognosis, and response of these tumors to anti-HER2 treatments.

Yagisawa and colleagues were the pioneers in exploring these inquiries.⁶⁵ In their research, they conducted an analysis of 370 patients with mCRC in Japan, revealing that the proportion of HER2-low tumors was 6%. HER2 levels were defined in accordance with the international harmonization on provisional diagnostic criteria. Their findings indicated that the prognosis of HER2-low tumors resembled that of HER2-negative tumors, which was superior to that of HER2-high tumors. Additionally, molecular characteristics such as BRAF and KRAS alterations were comparable to the HER2-negative group. Lastly, HER2-low tumors, like HER2-negative tumors, exhibited a positive response to anti-EGFR therapy, distinguishing them from HER2-high tumors. These results suggest that within the context of mCRC, HER2-low tumors bear similarities to HER2-negative tumors.

The research conducted by Wu et al. further strengthens the evidence supporting the similarities between HER2-low and HER2-negative tumors.⁶⁶ Their study focused on examining the clinicopathological characteristics of patients with localized CRC who underwent surgical resections in China. The study revealed that ∼37% of patients had HER2-low CRC, a significantly higher percentage compared to the findings of Yagisawa et al. In this study, HER2 levels were determined using the HERACLES criteria, which will be discussed later. Notably, HER2-low tumors in this study exhibited greater similarities to HER2-negative tumors than to HER2-positive tumors. This was evident in the prevalence of BRAF/RAS mutations, with HER2-low tumors showing a rate of ∼50% compared to 20% in HER2-high tumors. Additionally, the percentage of node metastasis was higher in HER2-negative and HER2-low tumors (∼65%) compared to HER2-high tumors (50%). Moreover, the OS of HER2-low tumors was comparable to that of HER2-negative tumors.

A recent study conducted on an Israeli population yielded similar findings.⁶⁷ In this study, 30% of the patients were identified as having HER2-low tumors, and no discernible differences compared to HER2-negative tumors were observed in terms of TP53, BRAF, KRAS, or mismatch repair status. HER2 levels were determined according to the American Society of Clinical Oncology/College of American Pathologists (ASCO/CAP) 2018 guidelines that are used in breast cancer.

Consequently, the cumulative evidence from the aforementioned studies indicates that HER2-low tumors exhibit similarities to HER2-negative (HER2 0) tumors, implying that they may not be considered as a distinct entity.

However, despite the lack of molecular-level differences, there may exist variations in other characteristics. For instance, Lang-Schwarz et al. conducted a comprehensive study on HER2-low tumors in patients with CRC across all stages. Their objective was to examine and characterize the tumor microenvironment of HER2-low tumors relative to HER2-negative (HER2-0) tumors.⁶⁸ In their study, the incidence rate of HER2-low tumors among the analyzed tumors was found to be 47.1%. Notably, they observed a significant difference in the percentage of tumors with >5% tumor-infiltrating lymphocytes (TILs) between the HER2-low group and the HER2-negative group. Key clinicopathological features of HER2-low CRC found in the various studies are summarized in Table 1.

Table 1.

Clinicopathological features of HER2 low in CRC

Study	Number of patients	Female gender	Primary tumor site (left∖right)	HER2-low frequency (%)	KRASWT (% of patients examined)	BRAFWT (% of patients examined)	MMR-D (% of patients examined)
Yagisawa et al.65	370	40	71.4/28.5	6	38 (n = 21)	0 (n = 21)a	N/a
Wu et al.66	2768	39.4	61.9/38.1	37	50.1 (n = 585)b	50.1 (n = 585)b	12.5 (n = 1031)
Lang-Schwarz et al.68	304	51.7	28.6/71.3	47	58.3 (n = 25)	N/a	20 (n = 143)
Babkoff et al.67	100	40	60/40	30	42.3 (n = 26)	88.5 (n = 26)	3.3 (n = 30)

CRC, colorectal cancer; HER2, human epidermal growth factor receptor 2; MMR-D, mismatch repair deficient; n/a, not available.

^aOnly V600E mutation was tested.

^bData refer to RAS or BRAF mutational status as no data are given regarding KRAS or BRAF alone.

Therefore, although HER2-low and HER2-negative tumors may exhibit similar molecular profiles, they can diverge in various aspects, such as the composition of the tumor microenvironment. These distinctions hold significant importance, as a higher presence of TILs has been associated with improved responses to immune checkpoint inhibitors in numerous tumor types, including gastrointestinal malignancies.^69,70

The emergence of ADCs and their demonstrated efficacy in breast and gastric cancer have prompted investigations into their use in CRC. The DESTINY-CRC01 trial, which is mentioned above, was a phase II study conducted to evaluate the effectiveness of T-DXd in patients with mCRC expressing HER2 who had undergone at least two lines of therapy.⁵⁹ As mentioned, in addition to cohort A which was HER2 positive, there were two other cohorts that fall within the definition of HER2 low (see Table 2). In contrast to HER2-positive tumors (cohort A), no responses were observed in cohorts B and C. These results may suggest that HER2 low may not be a distinct entity from HER2 negative. However, future pathological, molecular, and clinical studies including response to various treatments may suggest otherwise.

Table 2.

HER2 status determination for different cancer types

HER2 status	Breast71	Gastric—surgical specimen70	Gastric—biopsy specimen70	Colon (HERACLES)70
Negative	a. No staining is observed or membrane staining that is incomplete and is faint/barely perceptible and within <10% of tumor cells (0). b. Incomplete membrane staining that is faint/barely perceptible and in >10% of tumor cells (+1). c. Equivocal + ISH negative.	a. No reactivity or membranous reactivity in <10% of cancer cells (0). b. Faint or barely perceptible membranous reactivity in ≥10% of cancer cells; cells are reactive only in part of their membrane (+1). c. Equivocal + ISH negative.	a. No reactivity or no membranous reactivity in any cancer cell (0). b. Cluster of five or more cancer cells with a faint or barely perceptible membranous reactivity irrespective of percentage of cancer cells positive (+1). c. Equivocal + ISH negative.	a. No staining (0). b. Faint staining (+1), any cellularity. c. Moderate staining (+2) in <50% of cells. d. Intense stain (+3) in ≤10% of cells. e. Equivocal + ISH negative.
Equivocal—ISH test needed	Weak to moderate complete membrane staining observed in >10% of tumor cells (+2).	Weak to moderate complete, basolateral, or lateral membranous reactivity in ≥10% of cancer cells.	Cluster of five or more cancer cells with a weak to moderate complete, basolateral, or lateral membranous reactivity irrespective of percentage of cancer cells positive.	Moderate stain (+2) in ≥50% of cells.
Positive	a. Circumferential membrane staining that is complete, intense, and in >10% of tumor cells (+3). b. Equivocal + ISH positive.	a. Strong complete, basolateral, or lateral membranous reactivity in ≥10% of cancer cells. b. Equivocal + ISH positive.	a. Cluster of five or more cancer cells with a strong complete, basolateral, or lateral membranous reactivity irrespective of percentage of cancer cells positive (+3). b. Equivocal + ISH positive.	a. Intense stain (+3) in >10% and <50% of cells.a b. Intense stain (+3) in ≥50% of cells. c. Equivocal + ISH positive.

ASCO, American Society of Clinical Oncology; CAP, College of American Pathologists; HER2, human epidermal growth factor receptor 2; ISH, in situ hybridization; T-DXd, trastuzumab deruxtecan; WT, wild type.

^aAlthough considered positive, an ISH test was required for enrollment in the HERACLES trial. Additionally, all patients were required to have RAS^WT (KRAS and NRAS) status. It is worth noting that, despite new treatments such as T-DXd now being available for HER2 (+1) and HER2 (+2) cases, the 2023 ASCO/CAP guidelines update does not further divide HER2 categories into HER2 low and HER2 negative.

Yet, despite the disappointing results, several points warrant consideration. Firstly, all patients enrolled in the study, across all cohorts, had received prior irinotecan therapy before T-DXd. It is plausible that HER2-low tumors may have acquired resistance to the payload. Conversely, cohort A, which also received irinotecan before T-DXd, demonstrated a favorable response. Conducting further studies, such as investigating the response of in vitro irinotecan-resistant models and patient-derived xenografts to T-DXd, could provide insights into whether prior irinotecan treatment plays a pivotal role in resistance. A summary of major clinical studies conducted on HER2-positive and HER2-low tumors and their results is presented in Table 3.

Table 3.

Major clinical trials in patients with HER2-positive and HER2-low mCRC

Study	HER2 status	Number of patients	Female gender (%)	Treatment	RR% (CR%)	DCR%	mPFS	mOS
MyPathway40	HER2 amplification	57	28 (59%)	Trastuzumab + pertuzumab	32% (2%)	44%	2.9 months	11.5 months
TAPUR41	HER2 amplificationa	28	10 (36%)	Trastuzumab + pertuzumab	25% (0)	54%	17.2 weeks	60 weeks
TRIUMPH42	HER2 amplificationb	27	16 (59%)	Trastuzumab + pertuzumab	30% (3.7%)	N/a	4 months	10.1 months
HERACLES A43	HER2 amplification	27	4 (15%)	Trastuzumab + lapatinib	30% (4%)	59%	21 weeks	46 weeks
MOUNTAINEER44	HER2 amplification	86c	33 (39%)	Trastuzumab + tucatinib	38.1% (4%)	71%	8.2 months	24.1 months
HERACLES B58	HER2 amplification	31	7 (23%)	Pertuzumab + trastuzumab emtansine	9.7% (0)d	77.4%	4.1 months	N/a
NCT0289212347	HER2 amplification	28e	14 (50%)	Zanidatamab	38%	77%	6.8 months	N/a
DESTINY-CRC0159	HER2 amplification (cohort A) and HER2 low (cohorts B + C)	HER2 amplification—53 HER2 low: Cohort B—15 Cohort C—18	HER2 amplification—28 (52.8%) HER2 low: Cohort B—5 (33.3%) Cohort C—7 (38.9%)	Trastuzumab deruxtecan to all groups	HER2 amplification—45.3 (0) low: Cohort B—0 Cohort C—0	HER2 amplification—83% low: Cohort B—NE Cohort C—NE	HER2 amplification—6.9 months HER2 low: Cohort B—2.1 months Cohort C—1.4 months	HER2 amplification—15.5 HER2 low: Cohort B—7.3 months Cohort C—7.7 months
DESTINY-CRC0260	HER2 amplification. Two arms with different dosages (regular and low)	Standard dose—40 Low dose—82	Standard dose—n/a Low dose—n/a	Trastuzumab deruxtecan. One arm received standard dose and the other arm received a lower dose	Standard dose—27.5% Low dose—37.8%	Standard dose—n/a Low dose—n/a	Standard dose—5.5 months Low dose—5.8 months	Standard dose—n/a Low dose—n/a

DCR, disease control rate; HER2, human epidermal growth factor receptor 2; mCRC, metastatic colorectal cancer; mPFS, median progression-free survival; mOS, median overall survival; n/a, not available; NE, not evaluable; RR, response rate.

^aThe study contained a cohort of HER2 mutations that is not reported here.

^bOnly patients with HER2 positive in immunohistochemistry.

^cIn this study, cohort A received trastuzumab + tucatinib with no randomization whereas cohort B received a similar treatment after randomization. In their analysis, the authors related to both cohorts.

^dDid not reach its primary endpoint.

^eTwenty-eight patients were enrolled but 26 were included in the response analysis.

It is noteworthy that a recent study found that lung cancer cells (non-small-cell lung carcinoma, Ba/F3) with acquired resistance to T-DXd also presented reduced sensitivity to irinotecan [concentration that causes 50% inhibition of growth (IC₅₀) =200 nM in the resistant cells compared to the parental cells’ IC₅₀ = 66 nM]. Additionally, they demonstrated a reduced sensitivity, albeit to a lesser extent, to poziotinib—a pan-ERBB inhibitor (IC₅₀ = 14 nM in the resistant cells compared to the parental cells’ IC₅₀ = 7.5 nM). These results may suggest that reduced sensitivity to the payload is part of the mechanism of resistance. Further investigations using CRC cell lines and a variety of assays would enhance our understanding of the resistance mechanism(s).⁷²

Another aspect worth exploring is the influence of the linker on different tissues. It is conceivable that colon-derived carcinomas may not efficiently cleave the linker. Conducting additional studies utilizing different linkers could provide insights into whether distinct tissues necessitate specific linkers and whether this mechanism contributes significantly to resistance against T-DXd. Lastly, it is also noteworthy that this was a phase II study with a low number of patients (84 in all three groups).

An important aspect to highlight is the considerable variability in the prevalence of HER2-low levels observed across different studies. As previously mentioned, the study conducted by Yagisawa et al. reported a HER2-low percentage of 6%, whereas in all other studies, the percentage exceeded 30%. One possible explanation for this disparity is the variation in the studied populations. However, a crucial factor to consider is the methodology used to determine HER2 levels, which differed among the various studies.

HER2 Status determination

A crucial aspect for future management of HER2 status in CRC revolves around defining its expression status. In modern times, the optimal biomarker is IHC and ISH as needed, especially when HER2 IHC yields equivocal results. The IHC scoring system classifies expression levels into four categories ranging from 0 to 3. However, despite similar IHC categories, the specific expression levels used to define each HER2 expression category vary across different cancer types. This discrepancy arises due to the diverse patterns of HER2 expression observed in different tumor types. For example, HER2 expression in gastric tumors is predominantly cytoplasmic, while in breast cancer, it is primarily membrane-bound.⁷³ Additionally, gastric tumors often exhibit incomplete HER2 staining, unlike breast cancer. Consequently, circumferential staining is considered on IHC in breast cancer but not in gastric cancer.⁷⁴ Another significant distinction is intratumoral heterogeneity, where different regions of a gastric tumor may express varying levels of HER2. This phenomenon is less common in breast cancer. Consequently, the method of specimen acquisition, whether through biopsy or surgery, can also influence the IHC scoring process in gastric cancer.⁷⁵

In the context of HER2 in CRC, there is a lack of consensus regarding the scoring of IHC. This variability stems from several factors. On one hand, HER2 staining in CRC primarily occurs on the cell membrane, like breast tissue. On the other hand, there are certain features that resemble gastric cancer. For example, in tumors with weaker HER2 expression, categorized as +2, the distribution of HER2 is predominantly basolateral. This differs from breast cancer, where staining is circumferential but weaker in intensity.⁷³

Consequently, there is a lack of standardized criteria for HER2 scoring in CRC across different studies. For instance, Ingold Heppner et al. employed the criteria used for gastro-esophageal carcinoma when examining the prevalence of HER2-positive mCRC.⁷⁶ In contrast, Yagisawa et al. adopted the international harmonization on provisional diagnostic criteria, which aligns more closely with the criteria used for breast carcinoma.⁷⁷

To ensure accurate identification of patients for anti-HER2 therapies, Valtorta et al. established a specific set of criteria for the HERACLES study.⁷³ These criteria differ from those used for breast and gastric carcinoma and are summarized in Table 2. Subsequently, many studies, such as the one conducted by Wu et al. as discussed earlier, have utilized the HERACLES criteria. It is of utmost importance that future investigators adhere to the same set of criteria in their studies.

It is worth mentioning that other methods besides IHC may be beneficial. For instance, various protein-based methods such as reverse phase protein arrays,⁷⁸ automated quantitative analysis,⁷¹ and mass spectrometry⁷⁹ have shown potential as tools for predicting the efficacy of anti-HER2 in breast cancer. Additionally, non-protein methods such as analyzing the transcription levels of HER2 have also demonstrated promise in predicting the response to anti-HER2 agents in breast cancer.⁸⁰ Further studies specifically in CRC may assist in selecting patients who will benefit from anti-HER2 treatments.

Future directions

Despite the findings presented above suggesting that HER2 low may not significantly differ from HER2 negative, further research is warranted to explore additional aspects. Firstly, while key molecular pathways have been investigated and no significant differences found between HER2 low and HER2 negative, it remains possible that other molecular pathways are altered. Next-generation sequencing, transcriptome, and proteome can help answer those questions. Secondly, additional research could aid in characterizing alternative biomarkers for anti-HER2 treatment beyond ISH and IHC. This could involve examining correlations between specific outcomes and factors such as HER2 transcript levels and HER2 protein levels.

Conclusions

The HER2 pathway plays a crucial role in different malignancies, influencing both prognosis and treatment strategies. Recent advancements in drug development have expanded treatment options, leading to improved survival rates and more patients eligible for anti-HER2 treatment.⁸¹ Nevertheless, the utilization of these advancements in the context of CRC is still under investigation. Supplementary Table S1, available at https://doi.org/10.1016/j.esmogo.2023.100032, lists ongoing clinical trials, enrolling trials, and trials that are not recruiting yet. Further molecular, pre-clinical, and clinical studies will help us in developing new drugs for more patients with this common malignancy.