Defining a ‘cells to society’ research framework for appendiceal tumours

Abstract

Tumours of the appendix — a vestigial digestive organ attached to the colon — are rare. Although we estimate that around 3,000 new appendiceal cancer cases are diagnosed annually in the USA, the challenges of accurately diagnosing and identifying this tumour type suggest that this number may underestimate true population incidence. In the current absence of disease-specific screening and diagnostic imaging modalities, or well-established risk factors, the incidental discovery of appendix tumours is often prompted by acute presentations mimicking appendicitis or when the tumour has already spread into the abdominal cavity — wherein the potential misclassification of appendiceal tumours as malignancies of the colon and ovaries also increases. Notwithstanding these diagnostic difficulties, our understanding of appendix carcinogenesis has advanced in recent years. However, there persist considerable challenges to accelerating the pace of research discoveries towards the path to improved treatments and cures for patients with this group of orphan malignancies. The premise of this Expert Recommendation article is to discuss the current state of the field, to delineate unique challenges for the study of appendiceal tumours, and to propose key priority research areas that will deliver a more complete picture of appendix carcinogenesis and metastasis. The Appendix Cancer Pseudomyxoma Peritonei (ACPMP) Research Foundation Scientific Think Tank delivered a consensus of core research priorities for appendiceal tumours that are poised to be ground-breaking and transformative for scientific discovery and innovation. On the basis of these six research areas, here, we define the first ‘cells to society’ research framework for appendix tumours.

Introduction

Appendix cancer is a rare malignant tumour type, with an age-adjusted incidence rate of 0.12 per million person-years¹. In the past two decades, the overall incidence of epithelial appendix tumours has increased by 232% across the USA^2,3. However, appendectomy rates have remained stable over this same time period, suggesting that the rising incidence of appendix tumours is probably not related to an increase in the diagnosis of incidental, asymptomatic tumours². However, there are no current, established appendix tumour-specific risk factors, and our understanding of the aetiology that underpins this rising surge in incidence of these tumours, and any disproportionate incidence rates across specific population groups, is largely incomplete.

The lack of standardized screening tests for primary appendiceal tumours poses a major barrier to prompt and accurate diagnoses. It is not always straightforward to identify appendix tumours on imaging, especially when disease presentation is atypical or when an abnormal appendix may not even be visualized. Colonoscopy has low sensitivity for the diagnosis of appendix tumours⁴. Data from case studies also suggest that the signs and symptoms of appendix tumours are broad and nonspecific^5,6 — these are often mistaken for more common conditions. Consequently, up to one in every two patients with appendix tumours will present with distant metastatic disease at diagnosis, most often in the peritoneum⁷. Overall, approximately 10–63% of patients with metastatic appendix tumours will survive longer than 5 years from cancer diagnosis⁸. This vast range in survival rate directly reflects the substantial tumour heterogeneity of epithelial tumours arising from the appendix and clinical outcomes that sharply vary across population groups (for example, diagnosis age, sex, and self-identified race and ethnicity)^2,9,10. Of note, approximately 20% of all appendix tumour cases are diagnosed as tumours of neuroendocrine origin¹⁰; many of these tumours are innocuous and behave differently and, thus, are staged and clinically treated according to neuroendocrine tumour guidelines^11,12. For these reasons, they are excluded from further discussion herein.

The premise of this Expert Recommendation is to briefly discuss the current state of the field before delineating the unique or persistent challenges for the study of appendiceal tumours, and to propose key priority research areas that will deliver a more complete picture of appendix carcinogenesis and metastasis. A group of diverse experts, organized by the Appendix Cancer Pseudomyxoma Peritonei (ACPMP) Research Foundation, collectively utilized the nominal group technique¹³ to gain a consensus of the core research priority areas for appendiceal tumours that are poised to be ground-breaking and transformative for scientific discovery and innovation. On the basis of these six research priorities, the authors engaged in structured, multi-level discussions to comprehensively elaborate on each concept. Here, we provide the first ‘cells to society’ research framework for appendix tumours.

An overview of appendiceal epithelial tumours

To conceptualize our cells to society research framework, we begin with an overview of the pathology and biology of appendix epithelial tumours.

Tumour histopathological subtypes

The pathological classification and staging of appendiceal tumours are challenging, attributable in part to numerous classification systems and the inconsistent terminology used. In the real-world setting of pathological review for appendix tumour cases, there is also a high discordance (~30%) between originating appendix tumour pathology — typically performed in a community hospital setting — and pathology derived at a medical centre that specializes in treating patients with appendix tumours¹⁴. Given these robust challenges and vast heterogeneity in appendix tumour histopathology^15,16, current practice is encumbered by differences in the subjective interpretation of histopathological findings and inconsistent use of recommended terminology (for example, ‘invasive’, signet ring cells (SRCs)) and uniform pathological schema, among others.

To comprehensively understand appendix tumour histopathology, the histology of the normal appendix must first be fully appreciated (Fig. 1). In the context of malignancy, epithelial tumours of the appendix can be classified as being adenocarcinomas and subtyped as having mucinous, non-mucinous (intestinal-like), goblet cell and/or SRC histology (Table 1). However, morphological classification is complicated by variable terminology and overlapping histological features of some tumours. For instance, both normal and tumour goblet cells are characterized by a large globule of mucus at the top portion of the cell. Tumours showing goblet cell differentiation represent a distinct histopathological subtype as they exhibit both endocrine and exocrine differentiation and are almost unique to the appendix. Although previously called goblet cell carcinoid tumours, and originally considered to be mixed adenoneuroendocrine carcinomas, this nomenclature has since been replaced by the World Health Organization (WHO) Classification of Tumors in favour of the term goblet cell adenocarcinomas¹⁷. This shift in tumour classification, combined with variation in histopathological appearance related to tumour grade, lends a layer of complexity to our understanding of the histopathological spectrum of appendix tumours. Within the subtype of goblet cell adenocarcinomas, high-grade goblet cell adenocarcinomas contain tumour cells called SRCs that are highly infiltrative and contain abundant intracytoplasmic mucin that pushes the nucleus to the periphery. SRC adenocarcinomas — previously defined as having more than 50% SRCs (now a 30% threshold) — are an ultra-rare histopathological subtype of tumours with considerably poorer outcomes, especially once metastasized⁸.

Fig. 1 |. Histology of the normal appendix.

The appendix is a small, blindly ending tubular organ (approximately 6–10 cm in length) in the gastrointestinal tract that is attached to the base of the caecum. The lumen-facing side of the appendix is covered by a glandular epithelium with colonic-type glands that are lined with a simple columnar epithelium and many mucin-producing goblet cells. The lamina propria between the glands contains lymphocytes and scattered eosinophils. The submucosa is separated from the mucosa by the muscularis mucosae and contains lymphoid tissue (primary and secondary lymphoid follicles). The muscularis externa is composed of an inner circular muscle layer and a thin external longitudinal muscle layer. Outside of these muscle layers is a subserosal layer of connective tissue, nerves and vasculature. The outermost element, the peritoneum, is a thin lining of mesothelial cells. Around the appendix is a triangular fold of peritoneum named the mesoappendix (appendiceal mesentery), which often extends from the terminal part of the ileum to the appendiceal tip and encloses the appendicular artery. Appendiceal epithelial neoplasms originate from the epithelial cells of the mucosa; deeply invasive tumours can extend through all layers of the appendiceal wall to involve the serosa (peritoneum).

Table 1 |.

Appendix tumour histologies and molecular features

A dash indicates that the gene was not reported in the study. Percentages are rounded to the nearest integer value. H&E, hematoxylin and eosin.

^aBased on immunohistochemistry strong positivity or complete loss, for p53.

^bBased on immunohistochemistry strong positivity, for p53.

Mucinous tumours in the appendix are further classified as adenocarcinomas or low-grade and high-grade appendiceal mucinous neoplasms (LAMN and HAMN, respectively). However, it is important to recognize that the spectrum of mucinous epithelial tumour histology for appendix tumours as collected in tumour registries may be presently skewed. Among mucinous tumours, the same International Classification of Diseases for Oncology third edition (ICD-O-3) histology code is used for LAMN, HAMN and mucinous adenocarcinomas. The WHO Classification of Tumors 2019 version classified LAMNs and HAMNs as uncertain behaviour or in situ neoplasms, respectively, and any appendiceal mucinous neoplasms with extra-appendiceal spread were behaviourally defined as a malignant neoplasm for data collection purposes¹⁷. This schema was revised in 2021 to classify HAMNs as malignant tumours (although there remains scant literature on this subtype) and LAMNs staged as T3 or T4 as malignant tumours, but the schema retained an in situ behaviour for LAMNs staged as Tis (LAMNs confined to the muscularis propria). Consequently, the completeness of data on LAMNs and HAMNs sharply varies in cancer registries — including in robust population-based cancer registries such as the National Cancer Institute’s Surveillance, Epidemiology, and End Results (SEER) Program, which has limited the possibility to deepen our understanding of these tumour types¹⁸. Overall, persistent coding issues for appendix mucinous tumour histologies preclude our ability to understand the ‘spectrum’ of appendiceal mucinous tumours (LAMN–HAMN–mucinous adenocarcinoma)¹⁹ and individual tumour pathophysiology.

It is valuable to highlight that there is also no consistent histopathological grading schema for appendiceal tumours to date^20–26 — the American Joint Committee on Cancer (AJCC)¹⁵, WHO¹⁷ and Peritoneal Surface Oncology Group International (PSOGI)²⁷ classification schemes all vary in definitions of cellular differentiation. Overall, current grading schemes are dominated by a peritoneal disease perspective — as the extent and type of peritoneal involvement are a strong indicator of clinical outcome — and, therefore, they work ‘backwards’ to the primary tumour (Table 2). Although this yields strong relevance to clinical management, a stark limitation of this approach is that it groups histologically and biologically distinct primary appendix tumours together into a single schema.

Table 2 |.

Grading criteria for PMP, inclusive of appendiceal origin

Peritoneal mucinous tumour grade	World Health Organization195	American Joint Committee on Cancer15
PMP, grade 1	Hypocellular mucinous deposits Neoplastic epithelial elements composed of strips of low-grade mucinous epithelium	Low cellularity (typically <20%) Tall columnar neoplastic cells with low-grade cytological atypia No SRCs No infiltrative-type invasion No LVI No PNI
PMP, grade 2	Hypercellular mucinous deposits as judged at ×20 magnification High-grade cytological features involving >10% of the tumour Infiltrative-type invasion characterized by angulated glands in a desmoplastic stroma, complex glandular growth, or a pattern of numerous mucin pools containing clusters of tumour cells	High cellularity (typically >20%) High-grade cytological atypia in ≥10% of the tumour Infiltrative-type invasion may be present LVI may be present PNI may be present No SRCs
PMP, grade 3	Mucinous tumour deposits with SRCs or sheets of tumour cells	High cellularity (typically >20%) SRCs comprising ≥10% of tumour Infiltrative-type invasion may be present LVI may be present PNI may be present

LVI, lymphovascular invasion (defined as the involvement of small lymphatic or blood vessels by the tumour); PMP, pseudomyxoma peritonei; PNI, perineural invasion (the process of neoplastic invasion of nerves); SRC, signet ring cell.

Another important consideration in the clinical management of appendiceal tumours is tumour staging, as this informs disease prognosis. Recently released by the AJCC Expert Panel on Cancers for the Lower Gastrointestinal Appendix Disease Site Working Group, results from survival analyses using the AJCC eighth edition staging criteria — which were utilized to inform stage revisions in the version 9 AJCC system for appendiceal tumours — critically revealed a lack of hierarchical prognostication in these rare tumours. This prognostic variance was most significantly observed for stage II–III non-mucinous appendiceal tumours, but also for stage III mucinous adenocarcinomas of the appendix⁸. This illustrates another substantial challenge in tumours of appendiceal origin, as to the appropriateness of applying the same stage groupings across appendix tumour histological types. Consequently, the clinical outcomes of appendix tumours vary greatly with tumour subtype and grade; whereas mucinous appendiceal tumours, goblet cell adenocarcinomas and SRCs all spread to the peritoneal lining of the abdomen, low-grade tumours (LAMNs) tend to have longer survival times than high-grade tumours such as SRCs.

The biology of appendiceal carcinomas

Evidence has accumulated in support of the molecular characteristics of appendix tumours being different from that of tumours of the colon^28–33. This has led to the general consensus that tumours originating in the appendix are a biologically distinct malignancy from colorectal cancer (CRC). However, the mechanisms involved in primary appendix carcinogenesis and metastasis remain largely unresolved.

Intrinsic features of appendix tumour cells.

Beyond the unique, emerging appendix tumour biology^34–43, there is marked heterogeneity in the genomic landscape across tumour histological subtypes (Table 1) and by age groups. Genetic variants in KRAS, GNAS (encoding Gα_s) and adenomatosis polyposis coli (APC) are less frequent in goblet cell adenocarcinomas whereas alteration rates for CDH1 (encoding E-cadherin) and fibroblast growth factor receptor 2 (FGFR2) are higher for goblet cell adenocarcinomas than for other appendix tumour subtypes⁴¹. Evidence from somatic variant panel testing in a recent single-institution study shed initial light on the potential for establishing molecular classifications of appendix tumours. Molecular subtypes of mucinous appendiceal adenocarcinomas were largely defined by RAS, GNAS and TP53 variant patterns⁴². When stratified by age, individuals with early-onset appendix tumours harboured higher odds of presenting with non-silent somatic variants in PI3K catalytic subunit-α isoform (PIK3CA), SMAD3 and tuberous sclerosis 2 (TSC2) and lower odds of presenting with GNAS variants than individuals age 50 years and over⁴³. GNAS and TP53 variants were also found to be mutually exclusive in appendiceal tumours, irrespective of age groups^30,43. Overall, notwithstanding the key challenges that remain as to the primary versus metastatic origin of tissue used for profiling in these studies and consideration of key pathological factors (for example, tumour grade) in the determination of consensus subtypes, these robust efforts strongly emphasize the importance of somatic tumour profiling for all primary and metastatic tumours of the appendix.

It has been historically believed that appendix tumours are not hereditary. Published studies screening appendix tumours for microsatellite instability (MSI) have suggested a low MSI prevalence (<4%)⁴⁴. Given the disproportionate burden of early-onset cases in this rare tumour type¹⁰ — a hallmark of inherited cancer predisposition, recent studies have begun to explore the potential of a genetic link to appendiceal tumours. Indeed, the seminal study in this field to date has evaluated clinical multi-gene panel testing results for 14 known cancer-susceptibility genes from patients with appendix tumours across the USA and discovered that one in every 10 individuals carried at least one deleterious germline variant in a cancer-susceptibility gene⁴⁵. A later single-centre study has confirmed these germline prevalence estimates in appendix tumour cases and also went on to perform somatic variant panel-based profiling of 505 genes with nearly no identification of second-hit somatic variants⁴⁶. Currently, a comprehensive, genome-wide study of the germline and somatic variant spectra in appendix tumours has not yet been completed.

Mucin has a key role both in normal tissue histology and in cancer⁴⁷, including in the appendix. Primarily expressed by epithelial cells on apical surfaces, secreted mucins form a physical barrier that protects epithelia from adverse conditions and limits exposure to commensal bacteria. However, the regulation of mucin production is not well-understood in the context of the normal appendix and appendiceal tumours. Protein expression patterns for several mucins have been examined in a single study of 108 primary appendix tumours using immunohistochemical staining, wherein it was observed that positive mucin 3 (MUC3) membrane expression may be a negative prognostic marker⁴⁸. Positive supranuclear MUC2 and cytoplasmic MUC5AC expression have been shown to be enriched in mucinous appendiceal tumours⁴⁹. In the metastatic setting, MUC2-expressing goblet cells have also been shown to contribute to the accumulation of extracellular mucin deposits.

Although understanding gene function and changes in transcriptional activity are undoubtedly key to unlocking the mechanisms of appendix carcinogenesis, existing databases, including The Cancer Genome Atlas (TCGA)^50,51, do not presently include RNA expression data for primary or metastatic appendix tumours. To our knowledge, no studies to date have characterized the global transcriptome of primary tumours of the appendix. In the metastatic setting, microarray-based profiling of peritoneal tissues from low-grade tumours of appendiceal origin defined a 139-gene cassette that distinguished two clear molecular subtypes of tumours, predictive of clinical outcomes⁵². More recently, a study of metastatic peritoneal tissues specifically from LAMN and HAMN cases undertook panel-based gene expression profiling and discerned three subtypes: immune-enriched, oncogene-enriched and mixed, which also correlated with prognostic outcomes⁵³. These data illuminate the potential value of studying the appendix tumour transcriptome, which to date has not been fully capitalized on.

Appendix tumour extrinsic features: the tumour microenvironment.

The appendix is fundamentally an immunological organ, distinguished from the remaining hindgut by its fully developed germinal centres and active immune surveillance of luminal contents⁵⁴. This richness of lymphoid tissue also suggests that its immunological function may be influenced by various metabolic processes and individual metabolites⁵⁵. Although vastly understudied⁵⁶, the metabolic and immune characteristics of the appendiceal microenvironment tend to hold important insights into the distinct pathways of carcinogenesis across subtypes of appendiceal tumours.

The appendix also contains a varied and robust microbiota that is distinct from the microbiotas present in other niches⁵⁷. The microbial composition of the appendix may have a role in modulating bowel inflammation, appendicitis and appendiceal tumours^57,58. Although no studies to date have examined the appendiceal microbiome in the context of the primary tumour, there is limited evidence of microbial dysbiosis in the setting of acute appendicitis^59–61. Small-scale studies of children and adolescents with acute appendicitis have observed microbial diversity and interindividual variability — reporting Firmicutes as a dominant phylum and the presence of oral pathogens Gemella, Parvimonas and Fusobacterium. A shift in appendix microbial communities was also reported between cases of uncomplicated appendicitis and those of complicated appendicitis⁶². This preliminary evidence raises intriguing questions as to the potential role of the appendix microbiome in overall human health and in carcinogenesis and metastasis.

Beyond the metabolome and microbiome, appendiceal tumour cells are surrounded by stromal, immune and lymphovascular cells, along with the extracellular matrix and associated soluble signalling factors. Together, this defines a dynamic and complex tumour-supportive milieu in the primary and metastatic appendix tumour setting^63–65 (Fig. 2). Although our knowledge of the primary appendix tumour microenvironment (TME) remains incomplete, a recent study of the peritoneal TME using orthotopic and flank-implanted patient-derived xenograft (PDX) models suggests that this cellular ecosystem may promote metastatic appendix adenocarcinoma growth via the upregulation of genes associated with proliferation (for example, MKI67 and EXO1) and the downregulation of components of the Hippo signalling pathway⁶⁶. This work follows a prior study of cytokine profiles in ascites or samples of peritoneal washings from cases of peritoneal mucinous neoplasms of appendiceal origin, which revealed peritoneal-specific synthesis of cytokines given the distinct cytokine patterns observed in the peritoneal microenvironment that differed from those found during infection or injury-related inflammation⁶⁷. Understanding the evolution of this cellular interplay in appendix tumours — in the context of histological subtype and patient-level characteristics — is paramount to developing therapeutic interventions in this rare disease space⁶⁸.

Fig. 2 |. The primary and metastatic appendix tumour ecosystems.

The primary appendiceal tumour microenvironment (TME) consists of tumour cells, connective tissue (mesenchymal cells, vascular and lymphatic tissue, extracellular matrix), infiltrating immune cells and a microbiome. Although research into this area of appendix tumours is limited and evolving, it is presumed that the TME progresses along a spectrum from normal tissue to overt carcinoma, with commensurate changes in the cellular, matriceal and microbial elements along this spectrum varying by tumour histological subtype (bottom). The peritoneal TME comprises the peritoneal tissue — including the mesothelial lining and underlying connective tissues — and peritoneal fluid (top). Peritoneal fluid in the context of carcinomatosis has a characteristic cytokine and chemokine profile that differs markedly from peripheral blood and physiological peritoneal fluid, as well as distinct populations of lymphoid and myeloid cellular constituents, making it an attractive venue for immune or other targeted therapies^63,104. The peritoneal tissue microenvironment comprises the surface layer of mesothelial cells and the submesothelial mesenchymal tissues, extracellular matrix, immune infiltrate, microbiome (not shown) and lymphovascular structures^64,65. Emerging data imply important differences in stromal and immune cellular infiltrates in peritoneal versus primary appendiceal tumour lesions, emphasizing the need to consider these microenvironments separately when developing therapeutic strategies for patients with appendix tumours and carcinomatosis¹¹⁸. DC, dendritic cell.

Research priority areas

Notwithstanding the remarkable strides in research over the past several decades, the rarity of appendix tumours presents a myriad of persistent challenges, including inaccuracies in accurately capturing disease incidence and prevalence, difficulties with the ascertainment of representative tumour tissues, limited funding opportunities, insufficiency of models and techniques to study appendix pathogenesis and metastasis, and poorly understood heterogeneity across tumour histologies. To propel novel scientific discoveries and to substantially drive the field towards ultimately reducing this disease burden, we reached consensus on six core research priorities for appendix tumours — spanning cells to society — that are comprehensively outlined below (Fig. 3).

Fig. 3 |. A ‘cells to society’ research framework for appendix tumours.

Six core areas of the appendix tumour research framework include refining tumour histology and histopathological classifications, characterizing the molecular landscape of appendix tumours and targets, defining the appendiceal primary and metastatic tumour microenvironment, developing disease-specific models and a model repository, conducting prospective investigations and trials, and understanding appendix tumours at a population level.

Refining histopathological classification

A crucial aspect of understanding and treating appendiceal tumours is an accurate classification of disease. As the gold standard for classification, histopathology is integral across the entire appendix cancer care continuum.

Continuum of tumour histopathological features.

Appendix tumours are not a single entity, and understanding the pathobiology of specific appendix tumour types has been hampered by variability in terminology to describe entities such as LAMNs and HAMNs, goblet cell adenocarcinomas, and non-mucinous appendix tumours. Given the rarity of these tumours, it is not surprising that pathologists in a general practice setting are often unfamiliar with current appendix tumour terminology, and a review by an expert gastrointestinal pathologist may be needed for accurate classification, grading and staging. Standardized approaches for macroscopic examination and sampling, consistent application of tumour classification and grading, and assessment of tumour cellularity are needed to move forward in refining histopathology for this rare malignancy.

A further complicating factor is tumour heterogeneity within appendiceal tumours. For instance, goblet cell adenocarcinoma may contain a high-grade SRC component, which dominates the clinical outcome even if it constitutes only a minor percentage of tumour cells^69,70 — highlighting the essential role of thorough tumour sampling. Although grading systems for various appendix tumours are used in clinical practice¹⁷, they are based on retrospective studies of relatively small case numbers. The optimal thresholds for classifying and grading specific appendix tumours remain unclear. For instance, in appendiceal SRC carcinoma, is there a continuum of cells detected in a primary tumour that may be optimally represented as a continuous variable rather than a threshold, and could this continuum better correlate with disease staging and outcomes?

Digitizing histology for appendix tumour detection.

Conventional brightfield microscopy remains critical in modern health care despite the inherent limitations to this practice that are largely dictated by the glass slide. For instance, the need to physically transport glass slides can limit patient access from geographically remote or resource-restricted areas to subspecialty pathologists. Consequently, digital pathology has evolved in the past decade from a potential clinical tool to a regulatory-approved means of providing diagnostic pathology services owing to advances in whole slide imaging (WSI)⁷¹. This digital conversion mitigates the physical nature of the glass slide and carries numerous benefits (for example, efficiency, ease of sharing, and accessibility). Moreover, WSI enables computational pathology, a relatively new discipline that promises automation and artificial intelligence that may potentially obviate interpretative variability. By and large, digital pathology will expand access for both patients and pathologists.

The difficulties in staffing remote laboratories or those located in resource-limited areas are well-recognized and are exacerbated by shortages of pathologists. It is further compounded by rare or uncommon diagnoses that would probably benefit from having subspecialists^72,73. The traditional approach to providing services could involve occasional on-site pathologists, typically generalists, or the transport of glass slides to a primary laboratory or consulting site. One of the most compelling use-cases for digital pathology in appendix tumours is the expansion and standardization of service to patients in these areas. These remote services include primary diagnoses, intra-operative consultations, and patient-initiated, provider-initiated or pathologist-initiated second-opinion consultations. Subspecialty expertise, either as a primary service or a consultation, can also be extended to these areas, which may otherwise not have sufficient volumes for subspecialized practice^72,74–77. Therefore, utilizing digital pathology for the provision of primary services or consultations allows both patients and pathologists to quickly access specialized knowledge that may not be available locally — particularly for rare diseases such as appendix tumours. In the future, the benefits of digital pathology may probably extend beyond a hematoxylin and eosin (H&E) diagnosis.

Computational approaches to refine diagnosis and classification.

Although digital pathology can extend the reach of subspecialty pathology practices to patients in underserved populations, the constraints of human evaluation persist. Computational pathology offers the potential to obviate subjective variability through computer vision — a subset of artificial intelligence that utilizes convolutional neural networks to analyse images⁷⁸. Applications of computer vision in pathology include automated detection, classification and grade determination, as well as the prediction of tumour mutational subtypes and expression signatures^79–84. Although no published applications of computational pathology to appendiceal tumours exist to date, it has been effectively used to identify MSI with H&E slides⁸⁵ and to predict consensus molecular subtypes in CRC⁸⁶, the latter of which currently relies on gene expression profiling for classification. In appendix tumours and pseudomyxoma peritonei (PMP; also known as mucinous carcinoma peritonei), computer vision harbours potential to support current classification and grading criteria through a similar evaluation of human-interpretable features (for example, neoplastic tissue architecture, cellular, nuclear and stromal morphologies, and shape and size of mucin pools). At this time, complex applications of computer vision on appendiceal tumours (for example, prognosis prediction and molecular prediction) are non-existent and would potentially be limited by disease rarity. However, simpler applications of computer vision might still be possible. For example, an algorithm for determining tumour cellularity (that is, identification of neoplastic epithelial elements in tissue or mucin) could be trained on appendix tumour and CRC data (as well as other cancer types) but validated on appendix tumour cases. It could have some benefit as it would impact peritoneal mucinous tumour grade (Table 2). Because human-interpretable feature-based models mirror how a pathologist approaches a histological evaluation under a microscope or on a computer screen (as these applications can be automated and deployed locally or remotely), computer vision also offers opportunities for them to validate findings and intervene when necessary. As such, it too has the potential to scale and to integrate into clinical workflows in resource-restricted areas where pathologists and/or expertise is limited to support accurate appendix tumour diagnosis and classification. Notably, computational pathology applications in this rare tumour type may also deliver cost-effective molecular predictions or novel classifications to heighten clinical accuracy and to support the delivery of tumour-specific therapeutic modalities for patients with appendix tumours.

Molecular characterization of appendix tumours

Comprehensive molecular profiling of primary and metastatic appendiceal tumours is essential to delivering a complete picture of appendix tumour pathophysiology and for downstream mechanistic and preclinical studies and therapeutic target discovery. Understanding normal appendix epithelium may also reveal why certain tumour histologies (for example, LAMNs and goblet cell adenocarcinomas) are virtually unique to the appendix.

Integration and standardization of molecular profiling into patient care.

Molecular characterization of primary, and metastatic, tumour tissues is a complementary approach to histopathology for the accurate classification of appendiceal tumours^20,21. However, the incidental discovery of primary appendix tumour cells most often occurs post-appendectomy, wherein the entire appendix has been removed ‘before’ disease diagnosis and often in the community hospital setting. For this reason, as well as owing to the very limited appendix tissue available and its priority for pathological evaluation, fresh tissues largely cannot be collected for the primary tumour.

In the research space, these practical tissue limitations point to a timely consideration of the utility of spatial and single-cell genomic and/or transcriptomic approaches in primary appendix tumours. Although costly, these relatively newer approaches are able to preserve local tumour architecture, dissect cellular heterogeneity that cannot be determined with bulk sequencing, comprehensively characterize the TME, explore intratumoural and histopathological heterogeneity, and overcome challenges with low DNA input, some of which are already compatible with formalin-fixed, paraffin-embedded tumour tissues. In the clinical setting, however, optimization and partnership with clinical testing laboratories will be a requisite to standardizing sample selection, delivering scalable profiling strategies, and considering distinct strategies specific to histological features of each appendix tumour that can be widely implemented.

Particularly for mucinous tumours of the appendix, another key challenge in delivering a comprehensive molecular characterization is limited tumour cellularity¹⁵. Low-cellularity tumours are defined as those with neoplastic epithelium present within mucin, which accounts for typically less than 20% of the mucinous component²⁴. Because DNA yield is proportionate to cellularity⁸⁷, low cellularity often leads to sequencing ‘failures’ and a gap in the molecular profile for this histological subtype. However, an attractive opportunity for clinical-grade RNA sequencing arises, given that this approach can overcome tumour cellularity-based limitations and deliver a wealth of data for biological interrogation⁸⁸. Together, these efforts will yield high-quality molecular data in appendiceal tumours — to support new opportunities for targeted drug development and for the standardization of appendix tumour molecular profiling as an integral component to oncology care.

Beyond profiling of the primary tumour, there is also substantive clinical value to the parallel characterization of metastatic appendix tumour tissues. For cases with distant disease spread, these paired tissues offer a robust opportunity to elucidate key drivers of appendiceal metastasis on a biological level and to identify novel druggable targets for potential therapeutic development. However, meaningful consideration in this approach is warranted — given the potential clonal heterogeneity and tumour evolution by anatomical location within the peritoneal cavity, as discussed later.

From the lens of genetics, initial data reporting a relatively high prevalence (10%) and spectrum of germline genetic variants among patients with appendix tumours support that genetic evaluation may be warranted for this patient population⁴⁵. An advantage to genetic testing is that it can deliver potentially life-saving cancer prevention and surveillance strategies — considering both a potential risk of a second primary cancer for patients — and the benefits of cascade testing for family members at risk of hereditary cancer syndromes. It also yields additional, valuable data for a deeper understanding of whether genetics has a role in appendix carcinogenesis. Moreover, to optimize this testing strategy in the clinical setting via appropriate gene selection specific to this patient population, comprehensive genome-wide discovery and functional validation of novel appendix cancer-susceptibility variants is warranted as a next step. Such efforts would be well-positioned to yield paradigm-shifting evidence in our understanding of a potential causal role of genetic variants in an unbiased manner and novel mechanistic insights into appendix carcinogenesis.

Understanding the normal appendix epithelium.

A complete understanding of appendix tumour pathophysiology relies on the exhaustive characterization of the normal appendix epithelium. However, a unique complexity associated with normal-to-tumour comparisons in the appendix is the absence of any residual normal epithelial cells in many cases, owing to complete obliteration of normal mucosa by the appendix tumour. Although histologically similar to normal colonic mucosa, the normal appendiceal mucosa is largely uncharacterized on a molecular level, and the reason that certain tumour types (for example, LAMN and goblet cell adenocarcinoma) are virtually unique to the appendix remains an enigma. This lack of insight highlights a timely need for the establishment of a molecular atlas of normal appendix epithelium and will be a key requisite to not only define tumour initiation and progression but to also facilitate the development of new drugs for appendix tumour treatment.

Establishing a composite multi-omics view of appendiceal tumours.

While embarking on comprehensive molecular profiling of primary and metastatic appendiceal tumours, our ability to computationally overlay data from individual molecular landscapes (for example, the genome, transcriptome, proteome and epigenome)⁸⁹ and understand the complex interplay of molecules is poised to fast-forward making novel predictions about appendiceal carcinogenesis for further mechanistic and preclinical studies, for novel target discovery, and for precision-based medicine approaches. To establish such a composite multi-omics landscape, there is a need to overcome the substantial costs for multi-omics assays and a need for computing power, data integration approaches, the emergence of novel omics technologies, and high-quality paired clinical and demographic characteristics, as well as for large sample sizes to inform measurable results. Nevertheless, the benefits outweigh the costs in the rare disease setting, wherein such data remain sparse.

As the composite of every human environmental exposure from conception to death, the biological response to these exposures — the exposome — is an emerging omics in the field of oncology⁹⁰. Thus, it is not surprising that advancements in our understanding of exposure biology in the context of appendix carcinogenesis is also an intriguing area of interest to explore moving forward. Given that environmental exposures leave many types of ‘fingerprints’ on human cells⁹¹, these exposome efforts may support defining how the environment could contribute to appendix tumour development for the first time.

Addressing appendix tumour disparities.

Cancer health disparities are a constellation of biology, genetics, behaviour and social determinants of health⁹². However, the interplay of biological and social factors that actively contribute to appendix tumour disparities are unknown; studies have yet to explore the biological features and/or molecular mechanisms contributing to inequities across population groups⁹ (for example, genetic ancestry and sex) with appendix tumours. A major and additional benefit to defining comprehensive molecular profiles in appendiceal tumours is our ability to perform global ancestry estimation⁹³ using these data and to evaluate molecular (and clinical) signatures across genetic ancestral groups. Ensuring that future composite, multi-level views of appendix tumour biology are diverse and representative of all patients diagnosed with this tumour type is a challenging, yet vital, task ahead.

The assessment of tumour molecular profiles is an important component to precision oncology and clinical care, including for appendiceal tumours. Despite the exponential growth of clinical tumour sequencing, there remains a tumour-agnostic gap in equitable access to high-quality cancer care and to this technology among historically marginalized groups^94,95. Therefore, as we emphasize the need to standardize molecular profiling into clinical care for patients with appendiceal tumours, it is underscored with the parallel need to ensure the timely and continued expansion of equitable access to these clinical assays and to support effective communication that addresses any medical mistrust. These efforts are aimed to mitigate, and not exacerbate, the disparities in outcomes across diverse populations.

Defining the appendiceal TME

Disentangling the primary appendix tumour ecosystem.

Surprisingly, there are minimal clinical data demonstrating the activity of immune checkpoint inhibitors in appendix tumours to date, relative to more common tumour types such as CRC and ovarian cancer. However, a recent phase II clinical trial has shown promising early results with a combination of atezolizumab (a programmed cell death 1 ligand 1 (PDL1) monoclonal antibody) and bevacizumab (a vascular endothelial growth factor receptor (VEGF) antibody) with a high rate of disease stability among patients with unresectable appendiceal adenocarcinomas⁹⁶ — supporting further clinical trials of this and other immune checkpoint inhibitor combinations, as discussed later. Furthermore, the early signs of activity with this combination suggest that immune profiling in appendiceal tumours could yield actionable targets for rational combinatorial approaches.

The formation and evolution of the appendiceal TME raise an important question as to how to sample and include a representative set of tumour tissues for immune phenotyping. With the development of numerous computational methods to infer the abundance of multiple cell types in the TME from bulk RNA-sequencing data, we suggest that a reasonable approach in appendix tumours is to leverage these deconvolution pipelines⁹⁷ and apply them to clinical-grade RNA-sequencing data from appendiceal tumours. The use of spatial and/or single-cell transcriptomic technologies on archival tumour tissues will also deliver an in-depth characterization of the total and specific compositions of immune cell types in the TME. In parallel, proteomics can shed light on the extracellular matrix microenvironment⁹⁸, and microbiomics can yield clues into potential bacterial modulators of immune response⁹⁹. The accurate annotation of tumour features, centralized expert pathology review, consistency in biospecimen collection and processing, high-quality multi-omics assays, robust patient-level data, and appropriate selection and refinement of computational techniques will also be critical factors to elucidating ‘hot’, ‘altered’ and ‘cold’ immune tumours¹⁰⁰ of the appendix across histological subtypes and to deliver on this strategy.

Understanding mechanisms of metastasis.

Profiling a single patient’s spectra of appendiceal tumours is integral to classify modulators of the TME and how this evolution may inform metastatic disease spread. In the study of appendiceal metastasis, a unique question remains to be answered: why is it that most primary tumours do not have a propensity to spread via the lymphatic system or hematogenously? The rupture of an appendix obstructed by tumour leads to the spread of tumour cells to the surfaces of organs in the abdominal cavity, wherein the peritoneum is a vascular ‘desert’. Among females, tumour cells may also spread to the ovaries¹⁰¹. Over time, peritoneal disease develops variable levels of fibrosis that renders resection impossible and may resemble or result from epithelial-to-mesenchymal transition in appendiceal tumours¹⁰². Therefore, in the study of appendix metastasis — particularly through the peritoneal cavity (‘peritoneal carcinomatosis’) — a key priority emerges in identifying shared biological mechanisms across disease sites with histological similarities. For example, what biological processes also drive peritoneal spread in select gastric, pancreatic and colon tumours? Another priority area is in understanding how a subset of appendix tumours metastasize to the small bowel, as small bowel involvement in peritoneal carcinomatosis remains a significant prognostic indicator¹⁰³. We call on the development of a peritoneal molecular atlas of appendiceal origin to address these (and other valuable) questions that aim to disentangle the complex biological processes driving appendiceal tumour metastasis.

The influence of the peritoneal microenvironment.

A key aspect to our peritoneal atlas for appendiceal tumours is comprehensive characterization of the peritoneal microenvironment — inclusive of the peritoneal lining (mesothelial cells and underlying mesenchymal and visceral structures), peritoneal fluid and free-floating cells within the peritoneal fluid. This is because physiological peritoneal fluid harbours a cytokine signature that is distinct from plasma and transforms in the setting of carcinomatosis or malignant ascites into an overwhelmingly maladaptive, tumour-promoting cytokine milieu¹⁰⁴. Biological interrogation of this peritoneal fluid together with multiple peritoneal metastatic tissues from an individual patient is considered the optimal approach to dramatically augment our understanding of tumour evolution and disease progression across appendix tumour histologies.

The mesothelial lining¹⁰⁵ and connected adipose, skeletal muscle, and other tissue types also tend to be important contributors to appendix tumour progression. Adipose tissue–tumour crosstalk in other gastrointestinal malignancies¹⁰⁶ — including in primary CRC¹⁰⁷ and metastatic gastric cancers^108,109 — suggests that adipocytes could similarly contribute to appendix tumour progression and peritoneal dissemination. Thus, the molecular interrogation of adipose and other underlying tissues in proximity to metastatic lesions (but also primary appendix tumours) will be a vital step in understanding the peritoneal microenvironment in appendix tumours and selecting rational therapies.

Building appendiceal tumour biobanks.

To support the integration of molecular profiling into clinical care, to discover new mechanisms involved in appendix tumorigenesis and metastasis, and to define the appendiceal TME, a robust collection of high-quality, diverse and representative biological samples with accompanying clinical elements stands as an integral research resource for forward progress in this field¹¹⁰. A large biorepository of pathologically annotated specimens — inclusive of tumour and normal appendiceal tissues alike — is acknowledged as a considerable financial and physical space investment, but it can also become a turning point in optimizing specimen collections, processing and validation and in extraction for tissues and peritoneal fluid. Appendix-specific biobanks, including ‘living cell biobanks’, may also deliver unprecedented opportunities for experimental and translational research studies in the current era of personalized or precision medicine.

Development of disease-specific models

Appendiceal tumour models — in vitro, in vivo and computational — are integral tools to comprehensively interrogate tumour biology and to enable preclinical evaluation of therapeutics^111–114. This is of heightened importance in the rare cancer setting, wherein the number of patients for clinical trial enrolment is rate-limiting. Currently, the number of appendix tumour models is extremely limited for several reasons, the most obvious being that these are uncommon tumours with no murine tissue counterpart, thus precluding the possibility of genetically engineered model systems. Other practical barriers include the following: the research community that is focused on appendix tumours is relatively small and the resources to support model development have been scarce; both mucinous and non-mucinous appendiceal tumours, particularly LAMNs, have not been successfully established in 2D culture, and in 3D culture, they rarely survive beyond a few passages; and primary appendix tumours are generally diagnosed incidentally, therefore negating the opportunity for tissue collection. Fortunately, in the past few years, the evolution of patient-derived organoids (PDOs) and PDXs in other distinct tumour types has provided a methodological framework for multiple groups to develop appendix tumour models. Several PDX models established from metastatic mucinous appendiceal tumours have been reported as detailed later in this section (Table 3). Below, we discuss the current state of appendix tumour model development along the continuum from in vitro to in vivo.

Table 3 |.

Inventory of existing patient-derived xenograft models of appendix tumours

Line or model designation	Appendix carcinoma histological subtype	Origin	Patient age and sexa	Histology described within the animal	Ref.
PMP1	Mucinous	Peritoneal metastasis	83 years old F	PMCA-1	128
PMP2			56 years old F	PMCA-1
DPAM1	Mucinous	Peritoneal metastasis	66 years old F	DPAM	121
DPAM2			72 years old M	DPAM
DPAM3			56 years old M	DPAM
PMCA-3	Mucinous	Peritoneal metastasis	62 years old M	PMCA-signet ring cell	127
PMCA1	Mucinous	Peritoneal metastasis	Not provided	PMCA	122
PMCA2				PMCA
PMCA3				PMCA
PMCA4				PMCA
PMCA5				PMCA
PMCA6				PMCA
PMCA7				PMCA
DPAM1				DPAM
PMP1	Mucinous	Peritoneal metastasis	Not provided	DPAM	126
PMP2				DPAM
PMP3				PMCA-signet ring cell
PMP4				DPAM
PMP5				DPAM
PMP6				DPAM
TM00351	Mucinous	Peritoneal metastasis	49 years old F	MAC	125
PMP-754	Mucinous	Peritoneal metastasis	58 years old F	PMCA	124
HG-PMP	Mucinous	Peritoneal metastasis	Not provided	PMCA-signet ring cell	123

DPAM, disseminated peritoneal adenomucinosis; F, female; HG, high-grade; M, male; MAC, mucinous adenocarcinoma; PMCA, peritoneal mucinous carcinomatosis; PMP, pseudomyxoma peritonei.

^aSelf-identified race and ethnicity was not provided for any lines or models.

In vitro model development.

Typically, the culturing of appendix tumour cell lines in 2D is a starting point for model development. However, to date, no such models have been successfully established. This has also been a striking hurdle in other solid malignancies, as even aggressive malignancies such as pancreatic cancer can be difficult to culture in 2D. For appendix tumours, the barriers to 2D cell line development are probably multifactorial and may include the presence of mucin and an incomplete understanding of the biological requirements for particular growth factors. Another key practical issue is that the sheer magnitude of the effort being applied in this space has been small, as many times, continued trial and error are required to overcome these barriers. Although the ease of manipulation of traditional 2D models provides substantial advantages for mechanistic studies, 2D cell lines also have well-documented limitations. They represent a highly selected tumour cell population which may poorly represent features of the primary disease due to clonal selection¹¹⁵. Three-dimensional PDOs can overcome some of these 2D model pitfalls, given their ability to capture intratumoural heterogeneity and recapitulate complete or partial features of parental tumour physiology^116,117. Because organoid growth and maintenance are heavily influenced by the media formulation, there is also a need for the development of specific media formulations for future appendix tumour models. Although formulations may differ based on tumour histological subtypes, the standardization of these formulations is essential to support an accurate comparison of results across investigators and institutions. This also aligns with our recommendation to prioritize the sharing of protocols, including those for viable tissue collections, as well as the development of appendix tumour-specific models, to deliver rigorous and reproducible research.

The collection of fresh tissues from primary appendiceal tumours is often impossible — as discussed in earlier sections — creating paramount challenges in the establishment of patient-derived models from all types of primary appendix tumours. Therefore, every effort should be made for elective surgery cases (wherein both primary and metastatic appendix tumour deposits are present) to be prospectively captured and used for tissue modelling. It will be vital to develop innovative and creative approaches for models that can recapitulate the phenotype of the primary tumour and be durably passaged to understand the principles of carcinogenesis within the appendix organ. Nevertheless, in the setting of appendiceal metastases, the recent development of patient-derived organotypic slice culture models of metastatic tissues from appendiceal adenocarcinomas harbour the advantage of preserving the tumour microenvironment via the creation of a 3D ‘tissue slice’ from patient samples, and recent data suggest that drug responses in these models may mirror clinical activity in patients^118,119. This novel approach can support preclinical evaluation of candidate therapeutics, yet as this new ex vivo slice model can only be maintained for a duration of 7–10 days, it also highlights the difficulty in establishing and durably passaging patient-derived models that must be addressed moving forward.

When considering other models that can recapitulate the appendiceal TME, the co-culture of appendix tumour cells with adipocytes, fibroblasts and/or immune cells will be an important line of investigation to comprehensively characterize tumour and stromal and/or immune cell crosstalk. The development of novel methodologies to co-culture multiple cell types (that is, epithelial, immune and stromal) and the generation of biomedically engineered platforms (that is, organ-on-a-chip)¹²⁰ will be important areas to be explored in appendiceal tumours. These co-culture models are particularly relevant given the unique TME within each appendix tumour subtype. For example, if we can understand how tumour–immune–stromal cell interactions impact responses to immunotherapy, we can develop rationale strategies to improve outcomes by modulating these interactions via targeted therapies. The establishment of a living biobank of tumour organoids established with standardized protocols is an important priority to drive such investigations because it would also provide the unique opportunity to screen both existing and novel drugs and therapeutic combinations that may be relevant for specific patient subgroups and ultimately moved into clinical trials.

Preclinical animal models for therapeutic advances.

A corner-stone of the drug development process is preclinical testing using in vivo model systems. The spectrum of models typically includes cell line-derived xenografts (which are typically in immunodeficient mice), PDXs and genetically engineered mouse models (GEMMs). As mentioned above, the development of GEMMs for appendix tumours is not possible given the absence of an appendix in mice. Although it may be feasible to genetically engineer ‘appendiceal-like’ tumour development in the mouse colon based on the genetics of appendix tumour subtypes (for example, KRAS and GNAS mutations), this has not yet been achieved. If successful, the extent to which this will mimic human biology is unclear. Nevertheless, such efforts are important for the field as developing syngeneic models will allow for the study of immunomodulatory approaches that are not possible using PDX models established in immunodeficient mice. Although rabbits have a distinct appendix, the higher maintenance costs, breeding challenges, lower transgenic efficiency and production of knockout rabbits (owing to the lack of embryonic stem cells for gene targeting in rabbits) present practical limitations for their application and use in preclinical research into appendix tumours.

Thus far, the development of preclinical animal models of appendix tumours has been limited primarily to PDXs (Table 3). Several groups have examined a range of therapeutic approaches from chemotherapy to small molecules in such models^121–123. However, it is of great importance that models are shared, extensively characterized for their preclinical growth kinetics, benchmarked for response to conventional chemotherapy, and genomically profiled. Creation of a model repository (discussed below) replete with such information will accelerate progress by providing investigators worldwide with common ground for studies, thus allowing valid comparison of results and nomination of therapeutic candidates that should advance to the clinic. Although several groups have successfully established models from LAMNs and HAMNs^124–128, models from more clinically aggressive histologies such as high-grade colonic type (non-mucinous) adenocarcinoma, SRC carcinomas and goblet cell adenocarcinomas have not been reported and, thus, remain a colossal unmet need. The unique biology of each histology will probably demand distinct methodologies to successfully produce such models.

As new preclinical animal and human tissue-based models of appendix tumours are generated, this will deliver additional opportunities to conduct preclinical trials of novel therapeutics. Overall, the development of representative preclinical models in appendix tumours will also support our ability to define the molecular basis of resistance to specific therapies and, thereby, accelerate clinical development of more effective agents for patients.

Establishing a model repository.

From in vitro to in vivo models, moving forward, it is essential to establish these platforms to be representative of all appendix tumour histology subtypes, patients, and to best recapitulate the original tumour tissue both genotypically and phenotypically. Towards the development of a much-needed centralized repository of appendix tumour models, a key first step is to define the gaps that exist in our current models (for example, histologies, race and ethnicity, genetic ancestral groups, and sex) (Table 3). Additionally, defining and characterizing the appendix TME from patient tissues will instruct which other cell types should be obtained and preserved as part of the repository. Established methodologies to recreate the multicellular interactions ex vivo will be imperative for future mechanistic studies and for the development of novel therapeutic strategies.

In parallel, we acknowledge the real-time infrastructure costs associated with repository activities, including collection, maintenance, storage, distribution and analyses. This places an emphasis on the need for monetary investment from funding agencies, participating academic medical centres, and philanthropy to consistently support broad collection and rapid dissemination and accessibility of these resources, including protocols. Without a doubt, this network and resource will become a nucleus for catalysing appendix tumour research and collaborations across the field.

Clinical studies of appendix tumours

Clinical research — which includes treatment trials and cohort, behavioural and health services studies — is fundamental to advancements in our understanding of appendiceal tumours and to informing clinical practice patterns and evidence-based medicine.

Prospective, patient-partnered clinical cohorts.

With a rare malignant tumour type, single-institution-based studies deliver limited impact. Selection bias may exist at large centres owing to referral patterns. Consequently, these studies may not represent the overall patient population and inherently diminish our ability to derive equal benefits to all patients with appendix tumours. This places an emphasis on collaborative, multi-centre efforts to yield evidence-based, clinically impactful findings and conclusions. As we work to advance our fundamental understanding of appendix carcinogenesis, there remains an attractive benefit to large-scale, patient-partnered or community-engaged efforts — such as the Genetics of Appendix Cancer (GAP) study (NCT05734430)^129,130 — that can overcome real-world challenges (for example, sufficient case numbers), as well as limited data and representative tissue collections, to establish a robust appendix tumour atlas and to drive clinical and translational research discoveries for the field.

Interventional clinical trials.

Current treatment recommendations for appendiceal tumours are primarily based on retrospective analyses and extrapolation of treatments used for CRC owing to anatomical proximity¹³¹ — even considering the body of evidence that posits fundamental differences between these two tumour types as described earlier. Despite the high unmet medical need, there has been little interest from major pharmaceutical companies in drug development for appendix tumours. This is partly attributable to the rarity of this cancer type, but also to the unique biological behaviour of mucinous tumours of the appendix, which demonstrate a slow growth rate, potential challenges of drug penetration, and complexity in measuring highly mucinous lesions for disease response assessment. These characteristics limit the extrapolation of standard drug development approaches that are used for other solid tumours. This lack of investigation represents a major challenge in standardizing appendix tumour treatment, in bringing precision medicine to this rare malignancy, and in improving patient quality-of-life and outcomes for patients with appendix tumours.

To date, no phase III clinical trials for appendix tumours have been conducted, and only three phase II trials of systemic therapy have been completed^96,132,133. The definitive treatment for early-stage appendix tumours is surgery with or without regional lymphadenectomy. However, for low-grade appendix tumours, retrospective data sets have suggested no benefit of lymphadenectomy — which aligns with the predominant biological pattern of peritoneal dissemination over lymphatic spread^134,135. This is reflected in the rarity of well-differentiated mucinous stage III disease from national registries^22,136. The role of adjuvant chemotherapy in this setting has not been prospectively studied, although registry data has also suggested a benefit from adjuvant therapy in stage II or stage III appendix tumours¹³⁶. Owing to the predilection for peritoneal-only dissemination and an indolent growth rate, metastatic mucinous appendix tumours are treated with cytoreductive surgery (CRS) and hyperthermic intraperitoneal chemotherapy (HIPEC) when a complete cytoreduction is achievable²⁰. Although neither treatment has been prospectively studied in randomized trials, a complete CRS is strongly correlated with improved survival in retrospective series^135,137. Despite a number of negative randomized phase III trials investigating the use of HIPEC in CRC, HIPEC is still utilized in appendiceal tumours and its use represents a space of investigational need^138–140. Moreover, the higher rate of peritoneal-only dissemination of appendix tumours may limit the extrapolation of CRS ± HIPEC (CRS with or without HIPEC) data from CRC to appendix tumours. Specific questions that require investigation relate to whether there is an added benefit of HIPEC on top of CRS and to the optimal methodology of HIPEC (for example, temperature, duration, chemotherapy agent and chemotherapy dosing). Although retrospective series have primarily demonstrated a role for CRS ± HIPEC for low-grade mucinous tumours, the benefit of CRS ± HIPEC for high-grade mucinous appendix tumours or non-mucinous appendix tumours is less well-demonstrated.

Patients with appendix tumours and extensive peritoneal disease may not be suitable candidates for a complete CRS, and the benefit of incomplete CRS ± HIPEC is controversial — partly related to the high morbidity and mortality rate associated with CRS. However, owing to the bias towards peritoneal-only dissemination and indolent disease biology of low-grade mucinous appendix tumours, palliative incomplete CRS ± HIPEC is often used in clinical practice, despite the lack of evidence supporting an incomplete cytoreduction. Limited studies have suggested a benefit of palliative HIPEC in the control of refractory malignant ascites^141,142. Two critical prognostic factors post-CRS ± HIPEC are the completeness of cytoreduction (CC) score and the peritoneal cancer index (PCI)^143,144. Because these are both operative predictors of outcome, efforts to preoperatively assess the ability to achieve a complete CRS are warranted. Although preoperative radiographical scores have been identified, validation and optimization of these preoperative imaging scoring systems are still needed^145,146. Furthermore, the use of novel radiomic and volumetric radiological assessments should be explored in this setting.

Another area for clinical trial investigation relates to the role of systemic therapy for mucinous appendiceal tumours and, in particular, LAMNs. Although the use of systemic chemotherapy has been extrapolated from its use in CRC¹³¹, the benefit from such therapy is uncertain because many fundamental challenges exist — especially for LAMNs. First, the indolent and mucinous nature of these tumours poses challenges to the assessment of tumour efficacy. Second, the low rate of cell division in these indolent tumours suggests that cytotoxic systemic chemotherapy predicated on targeting cell division is poorly suited for this cancer type, and therefore, alternative systemic therapy approaches are needed. Third, the highly mucinous nature of these tumours suggests that a drug that targets mucin — the predominant component of these tumours — may be the optimal therapeutic target.

Despite a phase II trial of the chemotherapy combination capecitabine and mitomycin-C showing a clinical benefit from therapy¹³³, this was based primarily on tumour marker improvement and semi-quantitative radiographical changes. By contrast, a recent small randomized clinical trial of fluoropyrimidine-based chemotherapy compared to observation alone for low-grade mucinous appendiceal adenocarcinomas has not demonstrated a statistically significant difference in the rate of tumour growth between observation and chemotherapy¹³². The divergent outcomes of these two trials demonstrate the inherent challenges of investigating systemic therapy in this indolent disease and the need for a consensus approach for efficacy assessment. The use of the standardized tumour assessment methodology known as Response Evaluation Criteria in Solid Tumors (RECIST) is of limited utility in LAMNs because changes in tumour size are limited and most commonly stable disease will be present with or without a therapeutic intervention. The duration of stability or progression-free survival are probably better determinants of efficacy but are confounded by the naturally indolent nature of this disease. Eligibility criteria requiring prior progression of disease within the 12 months before study entry, as utilized in clinical trials for low-grade neuroendocrine tumours, should be considered to address the variable natural history of this disease¹⁴⁷. All these challenges relating to radiographical assessment suggest that randomized clinical trial design may be needed to robustly demonstrate efficacy in appendix tumours.

The biology of LAMNs strongly suggests a need for a non-chemotherapy-based systemic approach such as targeted therapy, anti-mucin therapy or immune-based therapy^96,148. Although LAMNs tend to have limited genomic alterations, the common co-alteration of mutations in KRAS and GNAS suggest a unique opportunity for targeted therapy^28,43,149. (Fig. 4). Recently, inhibitors targeting the KRAS oncoprotein have shown clinical activity and should be explored in appendix tumours, along with the development of agents targeting Gα_s(ref. 150). Additionally, a cancer therapy trial that included 13 patients with GNAS-mutant peritoneal mucinous carcinomatosis originating from a primary appendiceal tumour has observed that single-agent pal-bociclib — a cyclin-dependent kinase 4/6 (CDK4/6) inhibitor — had clinical activity that was superior to that previously reported with cytotoxic chemotherapy and, therefore, should be further explored as a novel therapeutic strategy in the setting of appendiceal tumours with peritoneal spread¹¹⁹.

Fig. 4 |. Crosstalk between Gα_s and KRAS signalling.

Low-grade appendiceal mucinous neoplasms (LAMNs) are uniquely characterized by the common occurrence of co-mutations in KRAS and GNAS (which encodes Gα_s). The antagonistic crosstalk between KRAS and Gα_s signalling works in opposite directions, such that one is inhibitory as the other is agonistic. These co-mutations can deregulate oncogenic signalling pathways, including WNT, Hedgehog and transforming growth factor-β (TGFβ)–SMAD. The potential interactions between these two pathways suggest that targeting these in LAMNs may have an impact on both cellular growth and mucin production. Figure adapted from ref. 28, CC BY 4.0.

Beyond the call to action for developing clinical trials in appendix tumours, it is imperative to emphasize the importance of inclusive clinical trials for this patient population. With the worse outcomes among historically marginalized groups with appendix tumours, it will be important to meaningfully consider the barriers to awareness, recruitment and participation for these patients in future clinical trials^9,151. As an integral part of holistic health and care, it is also relevant to call out the need for behavioural interventional studies and programmes in this setting to be able to minimize treatment side effects, reduce symptoms (for example, depression), and support improvements in overall quality-of-life (for example, exercise interventions)¹⁵². Incorporation of patient-reported outcome measurement tools in clinical trial design¹⁵³ should also be explored to deepen our understanding of individual perspectives related to the health, quality-of-life and functional status of patients¹⁵⁴. Overall, the development of a robust clinical-trial pipeline investigating a spectrum of novel drugs identified from rigorous molecular and preclinical investigations of appendix tumours cannot be executed without substantive collaboration and investments from financial stakeholders. This includes support from pharmaceutical companies and federal funding agencies for investigator-initiated trials.

Novel blood-based biomarkers for surveillance and outcomes.

Blood-based tumour markers specific to appendix tumours do not currently exist. Markers that are utilized in the management of other gastrointestinal and gynaecological malignancies¹⁵⁵, including CA-125, CA19-9 and carcinoembryonic antigen (CEA), are also commonly elevated in patients with appendiceal carcinomas and associated with poorer outcomes^156–161. Consequently, US national consensus working groups, including the Peritoneal Surface Malignancies (PSM) Consortium Consensus Guidelines^20,21, have endorsed the measurement of all three markers in patients with appendix tumours. However, the use of these serum tumour markers is not universal in the clinical setting for this population. This may be owing to concerns with lack of payer coverage in the USA (for example, Medicare does not currently cover CA-125 or CA19-9 testing for patients with appendiceal carcinomas) or limited knowledge of the treating providers.

Novel markers — including circulating tumour DNA (ctDNA)^162–165 — have started to be tested in appendix tumours. Initial evidence from retrospective studies indicates that the detection of ctDNA during surveillance may be associated with shorter recurrence-free survival. In addition, the observation that lower ctDNA levels were detected in patients with grade 2–3 appendiceal carcinomas with peritoneal metastases for whom chemotherapy was administered 6 weeks prior hints that ctDNA kinetics may be an early indicator of therapy response¹⁶⁶. Future prospective studies and trials are of considerable value to correlate serum tumour markers, ctDNA and/or other novel biomarkers to radiographical, as well as clinical, response in patients with appendix tumours. Moreover, the discovery and validation of effective blood-based biomarkers specific to appendix tumours will be fundamental to clinical assessment for early evidence of pharmacological response.

Diagnostic and imaging modalities for appendiceal tumours.

Computed tomography is the primary diagnostic modality used to stage patients with appendix tumours. However, this is fraught with challenges in this patient population because peritoneal disease does not necessarily provide definitive borders of involvement and does not fulfill standard RECIST 1.1 criteria. Thus, the assessment of treatment response is also difficult to define. Furthermore, most of the literature to date are small case cohorts and/or retrospective analyses^167–169. Increased vigilance in the radiological setting is one important strategy to reduce missed or inaccurate appendix tumour diagnosis. Another strategy is the development of a strong partnership with diagnostic imaging technology industries to improve the application and efficacy of imaging modalities, including dual-modality imaging, for appendix tumours. These are integral to disease detection, monitoring and surveillance.

Quality of and equity in health-care delivery and outcomes.

Examining access to and the use, quality, delivery, costs and outcomes of health-care services for patients with appendix tumours is a valuable yet understudied area of research^170–172. For example, effectiveness studies of the impact of systems of care on patient outcomes, including preferred provider organizations and referral patterns¹⁷³, can support efforts to refine referral practices and recommendations for these patients, especially in the community practice setting, and to improve the awareness and reach of future clinical trials. As we work towards the development of novel appendix tumour therapies, health services research will remain a key component to support optimal health-care delivery and outcomes for this population.

Appendix cancer on a population level

Understanding the appendix tumour burden at a population level is aimed at primary cancer prevention, risk assessment and early detection, health outcomes, survivorship and the delivery of cancer care to all communities. These collective efforts across the cancer care continuum harness the ability to translate findings to clinical trials, improve clinical practice and impact public policy.

Elucidating exposures, risk factors and patterns.

Given the rising incidence of epithelial tumours of the appendix and their clinically ‘silent’ behaviour in the early stages, identifying potential risk factors and/or exposures associated with appendix tumours may allow for the identification of high-risk populations and for earlier diagnosis. Furthermore, identification of these risk factors may have important public health implications, allowing for cancer prevention and detection strategies at a community level¹⁷⁴. As with any rare disease, identifying potential risk factors and exposures requires a tremendous, years-long undertaking and undoubtedly relies on pooling existing resources in accomplishing this goal. An example is the Appendiceal Cancer Consortium (APPECC) — an effort recently established to characterize modifiable factors (as well as biomarkers) of appendix tumour risk¹⁷⁵. Given the inherent limitations of cohort-based strategies, including the need to often rely on retrospective data with potential inaccuracies¹⁷⁶, there is also a need to pursue contemporary case–control or case–cohort studies to identify risk factors for appendix tumours. Future studies investigating birth cohort, period, geographical and age-specific patterns in appendix tumour incidence will also augment our understanding of what life-course exposures (for example, environmental, geographical and in utero)¹⁷⁷ are relevant for further study in this rare tumour type. These interrogations may also help to shed light on what the distinct and shared pathways of appendix carcinogenesis are compared with other well-characterized distinct tumour types, such as CRC.

An evidence-based survivorship care plan.

The burgeoning appendix tumour burden with unexplained aetiologies is also inevitably leading to a growing population of survivors with a unique cancer journey. This is partly attributable to the trajectory in appendix tumour diagnosis, its rarity, and the aggressive CRS ± HIPEC treatment for eligible patients with peritoneal disease. For example, malignant bowel obstruction is a common presentation of patients who have peritoneal metastases¹⁷⁸ and is associated with substantial symptoms that impact quality-of-life and lead to poor prognostic outcomes¹⁷⁹. With the majority of patients presenting with advanced disease and owing to the rarity of this malignancy, survivorship remains an understudied area of clinical care for patients with appendix tumours^180–187. Currently, survivorship guidelines are largely adopted from CRC. However, the increased recognition that appendix tumours are a separate entity from CRC supports the development of an appendix tumour-specific conceptual framework and an evidence-based survivorship care plan that outlines the distinct needs of patients with appendix tumours (Fig. 5). This is strongly recommended to support the delivery of high-quality and continuous clinical care, especially when this care is coordinated across multiple practice settings. Doing so requires recognition and comprehensive research of the issues faced by all patients with appendix tumours from diagnosis to treatment and into surveillance.

Fig. 5 |. Care framework for patients with appendix tumours.

Clinical care for patients with appendix tumours is complex and multifaceted. These domains of care after an appendix tumour diagnosis are inclusive of disease information and education, health-care delivery and practice, prevention and surveillance for disease recurrence and second primary cancers, and surveillance and management of the physical, psychosocial, emotional and spiritual effects of an appendix tumour diagnosis and its treatment(s). AJCC, American Joint Committee on Cancer.

Another key consideration for all cancer survivors is the risk of developing a second primary cancer¹³¹. As a leading cause of morbidity and mortality among cancer survivors¹⁸⁸, second cancers may become a central issue over time for patients with appendix tumours — particularly those diagnosed with early-onset disease¹⁸⁹. Whereas the cumulative incidence of second cancers varies by patient age and first primary cancer type, presently, there is a dearth of evidence on second cancer risk within the population of patients with appendix tumours. Further research in this space will be critical to help guide high-quality, holistic clinical management of survivors.

Ascertaining real-world patient data for new clinical insights.

Beyond the prospective clinical studies, trials and consortium-based efforts acknowledged thus far, there is also a valuable role for research in appendix tumours utilizing multi-institutional databases. A precedent for this recommendation comes from prior studies that have included patients with appendix tumours and peritoneal metastases who underwent CRS ± HIPEC^190–192 and studies evaluating the role of ctDNA in appendix tumours¹⁶⁶, discordance in appendix tumour pathology¹⁹³, and the natural history of this disease¹⁹⁴. With careful consideration from disease experts for the curation and standardization of appendix tumour-specific data fields, a multi-institutional database model delivers more timely access to real-world data and may yield valuable clinical insights that drives future research and innovation in this rare malignancy.

Utilizing cancer registries to understand disease trends and control.

Population-based and hospital-based cancer registries are essential to understanding cancer trends and control, patient burden, treatments and outcomes, and even health disparities — especially in the context of rare tumour types. However, the standardized data fields used in these robust cancer registries (for example, SEER¹⁸ and the National Cancer Database (NCDB)) do not all fit the appendix tumour continuum. For example, surgical resection of the primary tumour — a routine variable captured in registries — often occurs before disease diagnosis when the tumour is discovered at appendectomy for presumed appendicitis. Many of these patients will later undergo CRS and HIPEC for peritoneal metastases, but the current NCDB system combines all individual surgeries for a patient into one procedure code and reports only the most extensive surgery. The predominant spread of appendiceal tumours to the peritoneal cavity is also not uniformly captured because it is not a registry-specified site for metastasis.

In addition, current morphology codes used for capturing histological types in registries are suboptimal for appendiceal tumours — especially for LAMNs and mucinous adenocarcinomas. Use of the same morphology code for LAMN, HAMN and mucinous adenocarcinoma, with differences only in the behaviour code, make it impossible to retrieve these cases as separate entities from the registries, thereby severely hampering analyses of clinical outcomes. This, together with the recent shift from a four-category to a three-category grading variable in SEER, leads to potential misclassification or misuse of tumour grade in current registry data and in downstream secondary analyses. This inherently precludes the potential to consider tumour grade appropriately in analyses of existing cancer registry data. Together, these limitations point to the need for meaningful discussions among site-specific experts, registrars and registry staff to acknowledge these appendix tumour-specific nuances and consider what can be addressed on a practical level for moving forward in developing robust registry data.

Provider and public education, and resources.

Failure to recognize chronic, nonspecific gastrointestinal symptoms as a possible appendix tumour is a longstanding problem, and the risks associated with this are exacerbated by the trend towards more non-operative management of appendicitis. Thus, an important aspect of defining appendix tumours on a population level is provider and public education. For clinical care providers, this includes the importance of keeping occult appendiceal tumours in the differential diagnosis for patients presenting in this manner — particularly those under the age of 50 years¹⁰. Effective education and resources focused on (1) the signs and symptoms of appendix tumours, (2) the clinical and biological distinction of appendiceal tumours from cancers originating in the colon, (3) the changes in behaviour codes used to classify LAMN and HAMNs, (4) the independent International Classification of Diseases, Tenth Revision (ICD-10) diagnosis code that is specific to appendiceal tumours, and (5) up-to-date treatment modalities for appendiceal tumours that are vital for high-quality patient care (for example, to support correct and timely diagnoses) and for robust cancer registry data and research to guide evidence-based practices.

Conclusions

Research in the field of appendiceal tumours has grown over the past several decades. This is evident from an increase in high-quality publications and in the exciting discoveries that have begun to disentangle the intricate complexities of this distinct tumour type. It is, therefore, timely that defining a cells to society research framework for appendix tumours presented for the first time herein by a group of leading experts is poised to revolutionize this field, including by characterizing appendix tumour pathogenesis, metastasis and the tumour microenvironment, refining tumour histopathological subtypes, establishing novel appendix tumour model systems for laboratory-based studies, developing disease-specific therapeutic modalities and clinical management guidelines, and defining the overall population burden for appendix tumours and health inequities. An indispensable step in achieving these advancements requires robust investments from funding agencies and organizations. A recurrent theme in the rare cancer space is that this work cannot be done independently: the collaboration of researchers across numerous disciplines, including basic and translational scientists, clinicians, computational and systems biologists, and population scientists is poised to yield major breakthroughs for appendix tumours. Irrefutably, continued progress in this important and exciting field will impact our fundamental understanding of rare appendix tumours, drive the field forward to improve treatments and outcomes for patients, and sketch a ‘blueprint’ for promising strategies to adapt and to apply in the setting of other rare malignancies.

Key points.

• The rarity of appendiceal tumours present a plethora of unique challenges, inclusive of the following: inaccuracies in accurately capturing disease incidence and prevalence, difficulties with the ascertainment of representative tumour tissues, limited funding opportunities, insufficiency of models and techniques to study appendix pathogenesis and metastasis, and poorly understood heterogeneity across tumour histologies.

• Defining the continuum of appendiceal tumour histopathological features — utilizing both digital and computational pathology approaches — is poised to expand access for both patients and pathologists, to support current classification and grading criteria, and to potentially deliver molecular predictions and/or novel classifications in appendiceal tumours that can heighten clinical accuracy.

• Comprehensive molecular profiling of primary and metastatic appendiceal tumours is essential to delivering a complete picture of appendix tumour pathophysiology and for downstream mechanistic and preclinical studies and therapeutic target discovery. Understanding normal appendix epithelium may also reveal why certain tumour histologies (for example, low-grade mucinous neoplasms and goblet cell adenocarcinomas) are virtually unique to the appendix.

• Evolution of the dynamic and complex ecosystem surrounding appendiceal tumour cells in the primary and metastatic disease setting — the tumour microenvironment (TME) — largely remains an enigma. Disentangling this complex biological interplay within the TME may deliver a new array of rational and potentially combinatorial therapeutic strategies for targeting appendix tumours.

• The interrogation of appendix tumour biology and advancement in the rate of clinical translation in this disease space is a requisite research need reliant on the establishment of appropriate in vitro, in vivo and computational model systems that are representative of diverse tumour histologies, patient characteristics and the TME.

• Informing clinical practice patterns and evidence-based medicine for appendiceal tumours is consequent on the conduct of prospective clinical investigations — including cohort, behavioural and health services studies — and clinical trials that address unique considerations of this tumour type.

• Characterizing the appendix tumour burden on a population level is a priority area that will, in time, deliver evidence to support primary cancer prevention and risk assessment, early detection strategies, improved survivorship and health outcomes, and optimal cancer care delivery to diverse communities.

• Revolutionizing collaborative research in rare appendiceal tumours — inclusive of robust funding investments and transdisciplinary scientific partnerships — will undoubtedly drive this field towards improving therapies and outcomes for this growing patient population and will also chart a ‘blueprint’ for promising strategies that can be extended into other rare cancer types.

Glossary

Appendectomy

A surgical procedure to remove the appendix, a small, finger-shaped organ originating in the colon (large bowel), located in the lower right side of the abdomen.

Ascites

Swelling in the peritoneal cavity caused by an abnormal accumulation of abdominal fluid.

Colonoscopy

A procedure in which a flexible instrument is inserted through the anus to examine the inner lining of the colon after cleansing it of stool using a variety of laxative preparative regimens.

Completeness of cytoreduction (CC) score

Assessment of the postoperative extent of peritoneal disease removal.

Complicated appendicitis

Appendicitis associated with necrosis leading to perforation or a periappendicular abscess.

Convolutional neural networks

A mathematical algorithm that analyses visual images by processing data in multiple layers to detect and classify objects in an image.

Cytoreductive surgery

(CRS). A surgical procedure to remove all visible tumours and diseased tissue in the peritoneum.

Dual-modality imaging

The use of two complementary imaging techniques to improve diagnostic accuracy and assessments.

Early-onset appendix tumours

Tumours in individuals between 18 and 49 years of age.

Genetic variants

Changes in DNA sequence between individuals within a population.

Goblet cell

Intestinal epithelial cell that synthesizes and secretes mucus and mucins, named for its goblet cup-like appearance formed by mucin granulae that fill up the cytoplasm.

Hyperthermic intraperitoneal chemotherapy

(HIPEC). A procedure wherein a catheter containing chemotherapeutic drugs is inserted into the abdominal cavity. The catheter is connected to a perfusion machine, which heats the chemotherapy drugs and pumps them through the abdomen.

Intra-operative consultations

With respect to frozen sections. Immediate ad hoc pathologist interpretations that guide surgical management.

Mesothelial cells

A thin layer of cells present on the surface of the peritoneum that allows internal organs to move freely, that secretes lubricants for tissue protection, and that initiates an immune response when encountering tumour cells or foreign organisms.

Microsatellite instability

(MSI). Regions of repeated DNA that change in length when mismatch repair is defective. These deficiencies in DNA mismatch repair can be caused by hereditary, germline mutations or epigenetic silencing by hypermethylation.

Mucin

A family of glycoproteins that are secreted by epithelial cells and form a major component of mucus.

Peritoneal cancer index

(PCI). A scoring system (from 0–39) used to quantify the extent of disease spread into the peritoneal cavity.

Peritoneal washings

A procedure wherein a salt–water solution is used to wash the peritoneal cavity. This solution is then removed to check for cancer cells.

Peritoneum

A protective membrane lining the abdominal cavity that also extends to cover most of the organs in the abdomen.

Pseudomyxoma peritonei

(PMP). Also known as mucinous carcinoma peritonei. A clinical entity characterized by diffuse intra-abdominal gelatinous ascites with mucinous implants on peritoneal surfaces. Mucinous neoplasms of the appendix are the most common, but not the sole, tumour type that can give rise to PMP.

Signet ring cells

(SRCs). Glandular epithelial cells that line digestive organs. Their respective nucleus is shifted to one side by a large cytoplasmic vacuole.

Tumour cellularity

The overall percentage of a tumour that contains neoplastic epithelium with mucinous deposits by visual pathological estimation.

Tumour grade

A qualitative assessment of the degree of tumour differentiation. Grade may reflect the extent to which a tumour resembles normal tissue.

Whole slide imaging

(WSI). The microscopic scanning of whole glass tissue slides to convert them into digital equivalents.