A Decreased Appetite for Prophylactic Total Gastrectomy

To prevent cancer with an operation is both personally and professionally gratifying. As surgeons, we often care for patients with cancer at advanced stages, so the opportunity to prevent future suffering can be highly motivating. Such is the case with diffuse-type gastric cancer, which is a notoriously insidious and lethal cancer. When it was revealed in 1998 that germline CDH1 gene mutations were the most common cause of hereditary diffuse gastric cancer, the practice of prophylactic total gastrectomy was adopted.¹ With our collective experience and knowledge of conditions such as familial adenomatous polyposis and hereditary breast and ovarian cancer syndrome, the application of surgery for cancer prevention in CDH1 pathogenic or likely pathogenic (P/LP) variant carriers seemed self-evident. However, we caution against a “one-size-fits-all” approach to gastric cancer prevention for individuals with CDH1 P/LP variants. Our viewpoint has evolved based on evidence that germline CDH1 P/LP variant carriers frequently harbor tiny foci of occult, indolent, pT1a gastric signet ring cell lesions and that prophylactic total gastrectomy is a life-altering operation that often results in chronic morbidity.^2–4 We contend that too many patients with CDH1 P/LP variants are subjected to prophylactic total gastrectomy, especially in adolescence and early adulthood. We will share our viewpoint through a series of clinical observations and reference the evidence that support our assertions.

When prophylactic total gastrectomy was initially reported, it was paired with the observation that stomachs from individuals with CDH1 P/LP variants could harbor many dozen, microscopic foci of intramucosal (pT1a) signet ring cell carcinomas.⁵ The number and distribution of these occult cancers was highly variable. It became clear with additional reports that 80%−100% of prophylactic total gastrectomy specimens could harbor such foci.^4,6 A reasonable conclusion from these studies was that one could expect that a prophylactic total gastrectomy specimen for CDH1 would most certainly contain occult cancer cells; this was shown to be true irrespective of the patient’s age and family history of gastric cancer.⁷ Whether or not such individuals will ultimately develop clinically actionable, advanced gastric cancer in their lifetimes is not as clear. Nevertheless, these data help inform clinical decision-making and expectation-setting prior to risk-reducing surgery.

Several data elements contribute to the ultimate decision to pursue prophylactic total gastrectomy in a patient with a germline CDH1 P/LP variant. First, what is the clinically actionable gastric cancer risk associated with CDH1 gene mutations, and how do we apply that to an individual patient? Early reports estimated the lifetime risk of diffuse gastric cancer in CDH1 variant carriers was as high as 70–83%.^1,8,9 However, separate reports published in 2019 demonstrated lower estimates of lifetime gastric cancer risk.^10,11 Of note, these studies did not distinguish between the rate of detection of early-stage (pT1a) and advanced diffuse gastric cancer lesions. More recently, a large, multi-center study that excluded pT1a (intramucosal) signet ring cell lesions demonstrated that cumulative advanced gastric cancer risk may be as low as 7% to 10%.¹² This study also provided more precise gastric cancer risk estimates that were adjusted according to the number of first degree relatives affected with hereditary diffuse gastric cancer. Taken together, these data should help both patient and surgeon understand that with expanded knowledge of this cancer predisposition syndrome our approach to clinical management has changed, also.

Next, experts in the field frequently ask what information from a screening upper endoscopy should be used to guide decision-making. Emerging data from specialty centers support that enhanced endoscopic surveillance is safe and feasible for individuals with germline CDH1 P/LP variants, at least in the short term.^2,13 Furthermore, the evidence is quite clear that occult (pT1a) signet ring cell carcinoma should be an expected finding on random endoscopic biopsy of normal appearing gastric mucosa in individuals with CDH1 P/LP variants. In asymptomatic patients, we repeat endoscopic surveillance six months after detection of occult (pT1a) signet ring cells with random mucosal biopsy, and in one year if no signet ring cell lesions are detected. This is based on International Gastric Cancer Linkage Consortium clinical management guidelines.¹⁴ Among clinicians who conduct research in gastric cancer surveillance for this unique population, our attention has shifted to identification and targeted biopsy of gross mucosal abnormalities such as ulceration, abnormal pit patterns, and mucosal thickening that may indicate a progressing signet ring cell lesion. The authors’ single-institution experience has demonstrated that advanced gastric cancer (pT2 or greater) occurs in the setting of intramucosal signet ring cells detected on the biopsy of a focal mucosal abnormality, such as thickened mucosa or ulceration, whereas pale areas of mucosa are nonspecific for occult signet ring cells.² Therefore, it is our practice to offer continued surveillance when signet ring cell lesions are detected by random mucosal biopsy up to age 70, and to recommend therapeutic total gastrectomy only when a mucosal abnormality is targeted with biopsy that demonstrates signet ring cell carcinoma.

It is common for patients to ask what the alternatives and potential consequences are if total gastrectomy is not performed in the setting of a germline CDH1 P/LP variant. To answer this question, we need to know the natural history of individuals with CDH1 P/LP variants who do not undergo prophylactic total gastrectomy. The data on endoscopic surveillance show that most patients who have undergone repeat surveillance over many years have remained alive, and free of advanced gastric cancer.^2,13 More specifically, patients should be informed of the potential disadvantage of surveillance, which is a missed diagnosis of advanced gastric cancer. Data from Asif et al. show that no cases of locally advanced or metastatic gastric cancer were missed in a series of 120 patients who underwent two or more surveillance endoscopies.² However, annual or semi-annual upper endoscopies can be emotionally taxing, time consuming, and costly. Even so, we have shown that total gastrectomy imparts life-altering chronic sequelae such as bile reflux, nutrient deficiencies, and psychosocial costs.¹⁵ These are just a few examples of how the decision-making about management of individuals with germline CDH1 P/LP variants is complex and should be deliberated over a period of time, and in consultation with subject-matter experts.

In summary, we conclude that prophylactic total gastrectomy for germline CDH1 P/LP variant carriers should be applied less often and more selectively based on accumulating evidence. First, sequential contemporary studies indicate a substantially lower lifetime advanced gastric cancer risk than initially described. Second, intramucosal (pT1a) signet ring cell lesions are ubiquitous and exhibit indolent clinical behavior in CDH1 P/LP variant carriers. Third, endoscopic surveillance appears to be a safe and effective strategy with low incidence of advanced gastric cancer in large cohort studies. We suggest this viewpoint on management of occult (pT1a) signet ring cell lesions in CDH1 variant carriers is reminiscent of the evolution in management of intraductal papillary mucinous neoplasms and that de-escalation of prophylactic surgery for CDH1 P/LP variant carriers is both evidence-based and patient-centered.