Significance of Microscopic SRCC Foci

The defining histologic feature of invasive cancer is the penetration of tumor cells through the basement membrane into surrounding tissues (Figure 1). For most cancers, this marks a pivotal transition from precancer to invasive cancer, with an increased potential for metastasis and a poorer prognosis. However, in carriers of pathogenic CDH1 germline variants, foci of SRCC extending into the lamina propria or muscularis mucosae—defined as T1a disease—are almost universal (Figure 1). Rates of in situ or T1a SRCC in prophylactic gastrectomy specimens from asymptomatic carriers of CDH1 pathogenic germline variants generally range from 87.5% to 97%.^10–13 The yield of SRCC is higher when a total embedding protocol is used for gastrectomy specimens. A literature review found precursor or invasive lesions in 122 of 128 (95.3%) cases using this protocol, compared with only 25 of 40 (62.5%) cases not using it.¹¹ The prevalence of SRCC in gastrectomy specimens among CDH1 carriers does not vary according to predominant family history, whether it is primarily hereditary lobular breast cancer (HLBC), HDGC, or mixed disease. Specifically, in a prospective cohort study of 283 asymptomatic patients with CDH1 pathogenic variants, those with HLBC were found to have SRCC on gastrectomy specimens just as frequently as those in the other groups. In fact, the only >T1a lesion (a T3 adenocarcinoma) found at the time of gastrectomy was in the HLBC group.¹⁴

View Full Size

Tumor staging of gastric cancer.

Citation: Journal of the National Comprehensive Cancer Network 23, 4; 10.6004/jnccn.2025.7006

The development of SRCC appears to be an early life event in patients with CDH1 pathogenic variants. Although data are limited, T1a SRCC seems to be consistently present before age 30 years.^14,15 In fact, the youngest reported patient to undergo prophylactic total gastrectomy after negative preoperative biopsy was found to have T1a SRCC at age 14 years.¹⁶ In contrast with T1a lesions, which are nearly ubiquitous in CDH1 pathogenic variant carriers, stage >T1a lesions are found in only 2% to 3% of gastrectomy specimens.^11–13 Past guidelines have recommended gastrectomy when SRCC was found on upper endoscopy,⁸ but this strategy is now in question, given that almost all patients with CDH1 pathogenic variants harbor T1a SRCC, and no more than a third develop clinically significant DGC, as discussed later.^17–19 Furthermore, the recommended reflex to gastrectomy has made it difficult to study the safety and effectiveness of endoscopic surveillance, because surveillance is often terminated in favor of gastrectomy when T1a SRCC is found.^12,20,21 A great deal of effort has been devoted to maximizing the detection of SRCC through specialized endoscopic protocols, which involve careful examination of the mucosa and numerous targeted and nontargeted biopsies.^15,21–23 Given the almost universal prevalence of SRCC in CDH1 carriers, we propose that the goal of surveillance should shift from maximizing detection of all SRCC to identifying features that predict the presence of >T1a lesions and detecting these lesions at the earliest possible stage. This strategy would represent a paradigm shift from current endoscopic surveillance, which focus on maximizing detection of endoscopically invisible or subtly visible T1a lesions. For all these reasons, there is an ongoing debate among CDH1 experts challenging the dogma that T1a SRCC are traditional invasive adenocarcinomas.

Phenotypic Spectrum of CDH1 Pathogenic Variants

Until only 5 years ago, the lifetime risk of DGC among CDH1 pathogenic variant carriers was estimated to be 40% to 70% for men and 56% to 83% for women.^24,25 These estimates were based on families preselected by strict clinical criteria for genetic testing and thus were enriched in gastric cancer cases. With the increasing use of multigene panel testing, pathogenic CDH1 mutations have been identified in many patients who do not meet HDGC clinical criteria, suggesting variable expressivity of the phenotype. Using pedigrees from 38 families with pathogenic CDH1 variants identified by multigene panel testing at a large American commercial laboratory, Xicola et al¹⁷ estimated the lifetime risk of gastric cancer to be 37% for men and 24% for women, approximately half of the prior estimates. Roberts et al¹⁸ reported remarkably similar lifetime risks of 42% for men and 33% for women, based on 41 families identified at a different commercial laboratory that were not exclusively ascertained by HDGC clinical criteria. Unlike these studies, which did not distinguish between early and advanced DGC, a recent multicenter report estimated the risk of specifically advanced (>pT1a) gastric cancer. This study, which drew on data from 7,323 individuals from 213 families and adjusted for clinic-based ascertainment bias, estimated the lifetime risk of advanced gastric cancer to be 7% to 10% for CDH1 pathogenic variant carriers, irrespective of family history. Risk was estimated to be higher for those with a strong family history of gastric cancer, reaching up to 38% for individuals with 3 affected first-degree relatives.¹⁹

The adoption of multigene panel testing for cancer risk evaluation has also revealed that the widely used clinical criteria for genetic testing, specifically those of the International Gastric Cancer Linkage Consortium (IGCLC)⁸ (Table 1), are insensitive for identifying individuals with pathogenic CDH1 variants. Among families with CDH1 pathogenic variants identified through multigene panel testing at commercial laboratories in the United States, the 2015 iteration of the IGCLC criteria was found to be only 19% to 33% sensitive.^18,26,27 Despite expansion of the total number of criteria from 6 to 9 and relaxation of prior age cutoffs, the 2020 IGCLC criteria were found to be equally insensitive.²⁷ In contrast, the simpler testing criteria now endorsed by the NCCN Guidelines⁹ (Table 1) had 87% sensitivity for identifying individuals with CDH1 pathogenic variants in the same cohort. Unlike prior criteria, these new criteria do not rely heavily on pathology information to confirm the diagnosis of DGC in family members, which is often unavailable.²⁷

Prophylactic Gastrectomy

Prophylactic total gastrectomy with Roux-en-Y reconstruction has traditionally been the recommended treatment approach for CDH1 pathogenic variant carriers with a family history of gastric cancer or those found to have SRCC on endoscopic surveillance.⁸ Although this procedure essentially eliminates the risk of future gastric cancer development, it is associated with significant postoperative morbidity and long-term sequelae. A retrospective study of 23 patients who underwent prophylactic total gastrectomy for truncating CDH1 mutations in Newfoundland reported that 48% experienced at least one complication in the postoperative period, and 17% had a major complication. Major complications included venous thromboembolism, anastomotic leak with abscess, and intra-abdominal abscess; the most common minor complications were subclinical leak, wound infections, urinary tract infections, and pneumonia.²⁸ A series of 101 patients with CDH1 pathogenic variants who underwent prophylactic total gastrectomy reported 1 death due to pneumonia in the early postoperative period. Additionally, 28% of patients experienced at least 1 complication within 30 days of surgery, and 20% experienced later complications. The most common early complications were wound infections requiring antibiotics or incision and drainage, anastomotic stricture requiring dilation, pleural effusion requiring drainage, and pneumothorax requiring a chest tube. Later complications mostly consisted of anastomotic strictures (75%) requiring dilation.¹³ This study included both open and minimally invasive approaches, with the minimally invasive group showing a higher number of lymph nodes retrieved and an average hospital stay that was 1 day shorter. However, the study did not compare other outcomes between the groups.¹³ A retrospective cohort study including only laparoscopic prophylactic total gastrectomies reported complications in 5 of 23 (21.7%) cases, 4 of which required reoperation.²⁹ A recent systematic review that included 353 patients who underwent risk-reducing prophylactic gastrectomy found a major complication rate of 19.2%, with the most common complications including anastomotic leaks and pulmonary issues. Five patients required reoperation due to incomplete removal of gastric tissue, and perioperative mortality was <1%.³⁰

In addition to these complications, patients undergoing total gastrectomy almost uniformly experience significant weight loss.³¹ They are also at risk for developing fatigue, reflux, fat malabsorption, nutritional deficiencies, osteopenia/osteoporosis, early and late dumping syndrome, small intestinal bacterial overgrowth, abnormal drug absorption, and body image concerns.³² In sum, these changes may have a significant impact on several aspects of patients’ lives, although QoL assessments after the first few months have shown mixed results. One prospective study of patients undergoing prophylactic total gastrectomy found that health-related QoL tended to decrease in the immediate postoperative period, return to baseline by 6 to 12 months, and then decrease again at 24 months. At 24 months, patients reported symptoms including insomnia and fatigue (88%), nausea and appetite loss (63%), pain (50%), and diarrhea (38%). Despite these symptoms, only 1 (8%) patient expressed regret about the decision to undergo gastrectomy at the end of the study period.³³ In contrast, another study found that 30 of 70 (43%) patients expressed some level of regret after prophylactic total gastrectomy. Regret was associated with postoperative complications and the absence of SRCC identified on gastrectomy explant.³⁴ In a more recent cohort of 126 individuals who underwent prophylactic gastrectomy due to a pathogenic or likely pathogenic germline CDH1 variant, 23.5% changed occupation, 3% divorced, and 1.5% became alcohol-dependent after surgery. QoL scores decreased at 1 month after gastrectomy but returned to baseline by 6 to 12 months.³¹

The optimal timing of prophylactic gastrectomy remains unclear. Previous guidelines have generally recommended surgery in early adulthood, typically between 20 and 30 years of age.^7,8 Recent studies of CDH1 pathogenic variant carriers report a mean age at gastric cancer diagnosis of 46 to 47 years, though there is significant variability (SD, 14–17 years).^13,17,18 One study found that the cumulative risk of gastric cancer remains relatively low until age 40 years (2.8% for women and 1.3% for men), then nearly doubles with each successive decade.¹⁷ A limitation of these studies is that they do not distinguish between early (T1a), asymptomatic SRCC and advanced, symptomatic DGC. Using a Markov model, Laszkowska et al³⁵ found that quality-adjusted life-years would be optimized if prophylactic gastrectomy were performed at age 39 years for men and age 30 years for women. However, this model was based on prior higher, likely overestimated, lifetime gastric cancer mortality rates of 62% for men and 54% for women. Ultimately, the timing of prophylactic gastrectomy, if being pursued, should be a shared decision between the patient and a multidisciplinary care team, and should take into consideration additional factors such as age of onset of gastric cancer in family members, nutritional status, family planning, and psychological readiness.

Active Endoscopic Surveillance

We refer to surveillance as “active” endoscopic surveillance, based on evidence suggesting that stage T1a SRCC is ubiquitous among CDH1 carriers. Active surveillance should be undertaken with consideration of patient preferences, as well as available evidence on the risk of progression and outcomes. The concept of active surveillance in individuals with known cancer is not new and has been shown to be a viable alternative to treatment in low-risk prostate cancer.³⁶

Recent studies suggest that active endoscopic surveillance may be a viable alternative to prophylactic gastrectomy for patients with pathogenic CDH1 variants. However, these findings are limited by short follow-up periods and frequent reflex to gastrectomy upon identifying SRCC on endoscopy. In a retrospective analysis of CDH1 pathogenic germline variant carriers at the Netherlands Cancer Institute, 20 patients underwent at least 2 endoscopies. Of these patients, 8 ultimately opted for gastrectomy after a median of 3 endoscopies; 71% made this decision after having SRCC detected on endoscopic biopsies. In the remaining 12 patients, there were no advanced interval cancers during a median follow-up of 19.2 months. Only 1 patient continued endoscopic surveillance after identification of SRCC; this patient had been under surveillance for 5 years at the time of publication. In this study, 4 patients were found to have advanced (>T1a) SRCC, all of which were recognized as advanced lesions on baseline endoscopy.²⁰

In a different retrospective review of 48 patients with a CDH1 pathogenic germline variant, 22 underwent >1 surveillance endoscopy (mean, 4.4 endoscopies; range, 2–12) over a period of 3.6 years (range, 0.6–9.4). Twelve of these patients ultimately pursued gastrectomy, and 10 remained under surveillance at the time of publication. There was no mortality in the surveillance group. One patient was diagnosed with stage II (T3N0M0) gastric cancer after 9.4 years of surveillance, subsequently undergoing gastrectomy and chemotherapy before being deemed disease-free. In this study, all patients with SRCC identified on endoscopy pursued immediate gastrectomy. Among the 25 patients who underwent gastrectomy despite having no SRCC identified on endoscopy, foci of in situ or T1a SRCC were identified in 24 cases.¹² Similarly, another retrospective analysis of 48 patients with CDH1 pathogenic variants who had at least 1 surveillance endoscopy found that all patients with SRCC detected on endoscopy (n=17) underwent immediate gastrectomy, and another 7 underwent gastrectomy despite having no SRCC identified on endoscopy. In 17 patients, no SRCC was identified on annual endoscopic surveillance over a median follow-up of 34.6 months.¹⁰ Another study followed 32 CDH1 pathogenic variant carriers who had undergone at least 2 surveillance endoscopies over a median 17.9 months (range, 4.4–75.4). All patients with SRCC detected on endoscopy either underwent gastrectomy or were lost to follow-up because they pursued further management at a different institution. Of the 25 patients with no SRCC detected on endoscopy, 8 elected for gastrectomy and 17 chose to continue surveillance; none of the patients developed advanced gastric cancer during the study period.²¹

Unlike these studies in which essentially all patients pursued immediate gastrectomy upon SRCC diagnosis, 2 prospective studies have included a substantial number of patients who opted for surveillance after SRCC detection on endoscopy.^15,37 In the Cambridge study, 36 patients meeting HDGC clinical criteria underwent at least one follow-up endoscopy after SRCC detection. Over a median surveillance period of 22 months (IQR, 9–39), none of these patients showed clinical signs of cancer progression beyond T1a disease.¹⁵ The NIH study included 120 patients with pathogenic CDH1 variants who had ≥2 endoscopies over a median 31.1 months (IQR, 17.1–42.1). While under surveillance, 2 patients developed focal gastric ulcerations with biopsies positive for SRCC and subsequently underwent therapeutic gastrectomy, at which point they were diagnosed with stage II (T3N0M0) disease. Among 74 patients with occult SRCC detected, 58 continued with endoscopic surveillance, and none developed advanced (stage III or IV) gastric cancer during the median 31.1-month follow-up.³⁷

Although no cases of advanced gastric cancer or gastric cancer–related mortality were reported in these studies,^{10,12,15,20,21,37} limitations include short follow-up duration and potential selection bias, because patients choosing to continue with surveillance may differ from those opting for gastrectomy. Additionally, these data come from specialized centers (in resource-rich Western nations) with rigorous endoscopic protocols and experienced endoscopists. Endoscopic procedures in these studies involved assessment of gastric distensibility, clearance of all mucus and debris, careful examination of 5 to 22 predefined anatomic sites using high-definition white light endoscopy with or without narrow-band imaging, and a total of 25 to 88 nontargeted biopsies, in addition to biopsies of any mucosal abnormalities.^15,20,21,37 The importance of focal T1a versus multifocal finding of T1a lesions is uncertain. Data on the endoscopic appearance of >T1a lesions are very limited. Unlike T1a lesions, which may manifest as pale round lesions, >T1a lesions have been associated with erosive/ulcerated lesions, disturbed vascular patterns, coarse pit patterns, and thickened, rigid gastric folds.^8,15,20,37 Given that nearly all CDH1 pathogenic variant carriers have at least stage pT1a SRCC, biopsy protocols such as the Cambridge protocol, which requires 25 to 30 random biopsies, should only be used in clinical practice if the presence of stage pT1a SRCC would influence patient decision-making regarding gastrectomy. However, it should be considered that extensive sampling, especially repeatedly, may cause gastric scarring, which can potentially make it more difficult to identify clinically significant gastric cancers in the future.

Conclusions

One of the most difficult aspects of practicing medicine is the need to make life-altering decisions with incomplete information. Much remains unknown about the natural history of CDH1-associated gastric cancer. Nevertheless, when patients with this condition seek guidance, we must support them in choosing the best option based on the most pertinent and up-to-date data available. The purpose of this review was to highlight a few key recent developments in our understanding of CDH1-associated gastric cancer: (1) the penetrance of advanced gastric cancer in this condition is lower than previously estimated; (2) gastric T1a SRCC is nearly ubiquitous in CDH1 pathogenic variant carriers, meaning detection of these lesions on endoscopy does not necessarily warrant immediate gastrectomy; (3) total gastrectomy carries risks of major complications and significant impact on QoL; (4) emerging evidence suggests that active endoscopic surveillance every 6 to 12 months may be a viable alternative to prophylactic total gastrectomy in the absence of indicators of >T1a SRCC, though current data are limited by short follow-up time in select patient cohorts (Table 2). Given these areas of uncertainty, shared decision-making is especially important between the patient and a multidisciplinary team including experts in genetics, endoscopic surveillance, psychology, and surgical oncology, with consideration of the patient’s risk tolerance, values, and family experiences. Future research is needed to (1) identify the mechanisms/risk factors driving variable penetrance of T1a progression; (2) identify and validate endoscopic and histopathologic features that predict >T1a disease; (3) comprehend the value or lack thereof of the intense biopsy protocols for informing on clinical management; and (4) clarify what role family history should play in the choice between active endoscopic surveillance and gastrectomy. Prospective studies of active endoscopic surveillance with longer follow-up are needed to further validate alternative approaches to prophylactic gastrectomy as well as active surveillance following detection of SRCC lesions without worrisome endoscopic appearance. Attention should also be given to novel detection strategies and potential chemoprevention, as well as factors that influence variable penetrance of advanced SRCC.