Japanese gastric cancer treatment guidelines 2014 (ver. 4)

This English edition was made based on the Japanese version published as a book in 2014. Our policy in compiling this edition was to attempt not to include new evidence that emerged since the publication of the Japanese version so as to maintain consistency of the two editions. However, for some particularly important issues, we provided additional comments and new references reflecting the new evidence.

Version 4 of the Japanese Gastric Cancer Treatment Guidelines was completed in May 2014, incorporating new evidence that includes those delivered as a quick bulletin in the website of the Japan Gastric Cancer Association after publication of the previous version. It remains largely conformed to the textbook style, but a new section consisting of clinical questions and answers (Q&A) was added to address some important clinical issues for which hard evidence is unavailable.

To compile this version, the guideline committee nominated several working groups, each assigned to make relevant contributions to unsolved issues on the following topics: (1) surgery and lymphadenectomy for junctional cancer, (2) clinical pathway, (3) follow-up after curative surgery, (4) treatment of technically resectable metastatic cancer, (5) risk calculation for surgical intervention and (6) treatment of cancer of the gastric remnant. Of these, tentative consensuses were reached on the first three topics that were included as new sections in the text, whereas further discussion was deemed necessary for the last two topics. The clinical importance of the fourth topic and lack of hard evidence related to that topic prompted the committee to establish a Q and A section to provide tentative best answers to important clinical questions on technically resectable metastatic cancer.

Major points of revision in the current version are listed below:

The description of tumor status (T/N/M and stage) in this guideline remains to be based on the third English edition of the Japanese Classification of Gastric Carcinoma [1], which is identical to that in the 7th edition of the International Union Against Cancer (UICC)/TNM.

Although recent advances in chemotherapy have achieved considerable tumor regression in many cases of unresectable/recurrent gastric cancer, these responses have not ultimately led to complete cure. The median survival time achieved in clinical trials for the disease at this stage remains 6–13 months. The current goal of chemotherapy therefore is to delay the manifestation of disease-related symptoms and/or to prolong survival. Survival benefit of chemotherapy has been proven in randomized controlled trials comparing chemotherapy with best supportive care in patients with unresectable gastric cancer with performance status (PS) of 0–2. Although very rare, some patients with advanced disease even survive more than 5 years by chemotherapy alone. Thus, chemotherapy is the treatment to be primarily considered for unresectable/recurrent gastric cancer among patients with sufficiently good PS.

Palliative care is an approach that improves the quality of life of patients and their families facing the problems associated with life-threatening illness through the prevention and relief of suffering by means of early identification and impeccable assessment and treatment of pain and other problems, physical, psychosocial and spiritual (WHO Definition of Palliative Care, 2002). The importance of palliative care increases incrementally as cancer progresses. The knowledge and technique to cope with pain, to communicate and to manage symptoms are required. Methods to accomplish these aims include radiotherapy and psychotherapy in addition to medication. Various clinical studies are on-going with particular emphasis on pain control.

It is extremely difficult to establish a clinical pathway for patients undergoing gastric cancer surgery that is widely applicable to various surgical procedures and institutions. However, it is possible to propose some core items based on which individual pathways could be constructed, and these could contribute to reducing disparities in surgical management for gastric cancer. A basic pathway has been constructed around the timing of some core items such as removal of the nasogastric tube, initiation of oral fluid intake, initiation of solid food intake, administration of antibiotics, stoppage of intravenous fluid administration and discharge from the hospital (Table 1). This clinical pathway is applicable to all surgical procedures including distal, total and proximal gastrectomy regardless of whether the surgery was performed laparoscopically or by open approach. However, postoperative management should be individualized for high-risk patients with severe comorbidities that include impaired cardiac, pulmonary, hepatic or renal functions. Recently, investigators have been inclined to aim for further shortening of postoperative hospital stays through the concept of ERAS (enhanced recovery after surgery), but the value of such programs in gastric cancer surgery is yet to be defined.

Follow-up at the outpatient clinic could be helpful so that the patients can readjust to their lives at home, cope with postgastrectomy symptoms and overcome the nutritional issues. In addition, surveillance for early detection of recurrence and secondary cancer is usually conducted according to the level of risk for recurrence, estimated based on the clinical stages. However, evidence that such surveillance actually improves survival is lacking. Due to the paucity of prospective studies that explored follow-up programs after gastrectomy, it is not possible to make any recommendation on how often the examinations should be performed, or even on which examination to perform. However, some retrospective studies suggest that CT, measurement of tumor markers (CEA and CA19-9) and endoscopy are effective to detect recurrence, gastric remnant cancer and metachronous multiple cancer. Tumor markers, when applicable, are apt to rise 2–3 months before metastatic lesions become detectable by imaging modalities. Models of follow-up programs for early-stage cancer and advanced cancer are shown in Tables 2 and 3.

Follow-up should continue for no longer than 5 years after which patients should be referred to regional general physicians or should be encouraged to undergo surveillance examinations provided as a part of health care programs in their districts or at their places of work. In that aspect, collaboration among various levels of medical facilities is needed to provide comprehensive care to gastric cancer survivors. Ultimately, there remains a need to scientifically verify the prognostic relevance of postoperative follow-up programs.

Para-aortic metastases from gastric cancer are classified as M1, and surgery with curative intent is not indicated according to the treatment algorithm of the current guidelines (refer to “Algorithm of standard treatments to be recommended in clinical practice”).

Systemic para-aortic lymph node dissection (PAND) had been attempted in Japan as clinical studies until its survival benefit was denied in a randomized trial in which only patients without lymphadenopathy in the para-aortic region were eligible [5]. However, numerous retrospective data from patients who underwent PAND are available in Japan and these almost invariably indicate that (1) metastases to the para-aortic nodes are pathologically confirmed in a certain percentage of these patients and (2) cure was achieved in approximately 10–20 % of the patients who harbored metastases to the para-aortic nodes. A similar result was recently reported from a Western country [36]. Thus, it is not possible to totally deny the survival benefit of PAND when lymphadenopathy restricted to the No. 16 a2–b1 region is found by preoperative imaging studies.

As a multidisciplinary treatment for this population, a treatment strategy of two courses of neoadjuvant chemotherapy with S-1 + cisplatin followed by PAND was explored in a phase II trial. Patients with bulky nodal disease with or without lymphadenopathy restricted to the No. 16 a2–b1 region were eligible, and peritoneal metastasis as well as the CY1 status had to be ruled out by staging laparoscopy prior to registration. Since a 5-year survival rate of 53 % was reported in this trial [37], this treatment strategy could be recommended for institutions with sufficient expertise in PAND. On the other hand, there are arguments that a 5-year survival rate of around 10 % can be achieved by chemotherapy when the para-aortic lymph node metastasis is the only factor that renders patients incurable [38, 39]. However, these retrospective studies are known to include patients who eventually underwent surgery after responding to the chemotherapy and may at least partially reflect the benefit of the multidisciplinary approach.

Hepatic metastases from gastric cancer are classified as M1, and surgery with curative intent is not indicated according to the treatment algorithm of the current guidelines (refer to “Algorithm of standard treatments to be recommended in clinical practice”).

Hepatic metastases from gastric cancer are often deemed unresectable since they are liable to be found as multiple nodules distributed to both hepatic lobes and are likely to be accompanied with metastatic lesions outside of the liver. No prospective trial exploring a benefit of hepatectomy has been conducted, and only retrospective analyses of small cohorts collected over several decades mostly as single-institution studies [40‐42] are available. However, 5-year survival rates ranging from 10 to 40 % have been reported from these studies, and one cannot deny a possibility that hepatectomy results in long-term survival among highly selected patients. Solitary metastasis or a small number of metastatic nodules has been highlighted as a favorable prognostic factor in most of the studies [43]. Given the recent advances in imaging studies, and the fact the diagnosis of solitary metastasis could be unreliable in older cases, hepatectomy may be considered for patients with a small number of metastatic nodules, and not restricted to a solitary tumor, provided that there is no other non-curative factor. Since there was no agreement on whether the synchronous metastases fare better than metachronous metastases, surgery could also be considered for recurrences in the liver if they fulfill the conditions mentioned above. Most patients eventually suffer from recurrences, however, and perioperative chemotherapy could be recommended for the population that had not been treated by adjuvant chemotherapy prior to detection of the hepatic metastases. Evidence on which chemotherapeutic regimen can be recommended in this particular setting, however, is totally lacking.

In Japan, peritoneal washing samples are usually collected during surgery for cytologic examination to detect free cancer cells. Free cancer cells in the peritoneal cavity (CY1) are classified as M1, and surgery with curative intent is not indicated according to the treatment algorithm of the current guidelines (refer to “Algorithm of standard treatments to be recommended in clinical practice”). However, patients with CY1 status are often treated by standard gastrectomy in the absence of other no-curative factors. The outcome of these patients had originally been dismal with a median survival time of approximately 12 months and 5-year survival rate of 7.8 %, but such data often included patients who were treated with surgery alone [44].

More recently, a prospective phase II study was conducted in which technically resectable cancer with CY1 as the only non-curative factor (patients with minimal and resectable peritoneal deposits included) was treated by standard gastrectomy followed by S-1 monotherapy until disease progression. The median recurrence-free and overall survival time in this study were 376 and 705 days, and 5-year recurrence-free and overall survival rates were 21 and 26 %, respectively [45]. In addition, a single-institution retrospective study of 120 CY1 patients who underwent surgery followed by S-1 monotherapy revealed a 5-year survival rate of 26.6 % [46], which was compatible with the trial result. These results are far better than the results obtained before S-1 became available and are equivalent to that of a series of curatively resected linitis plastica-type cancers, which often recur as peritoneal disease [47]. Furthermore, CY1 patients are deemed eligible for JCOG0501, a phase III trial to explore neoadjuvant chemotherapy by S-1 + cisplatin for scirrhous type gastric cancer in which the standard treatment arm consists of standard gastrectomy followed by S-1.

These facts indicate that CY1 patients could be indicated for the strategy consisting of standard gastrectomy and perioperative chemotherapy. In addition, S-1 monotherapy could be recommended for patients whose CY1 status was informed after gastrectomy. On the other hand, if the information on CY status was available prior to surgery, a chemotherapy-first strategy could be taken whereby only patients whose cytology status turned negative could be indicated for surgery [48, 49]. However, details of the optimal multidisciplinary treatment strategy in this setting, including the chemotherapeutic regimen to be used and the number of cycles to be delivered, remain to be elucidated in future clinical trials.

Postoperative adjuvant chemotherapy with S-1 has been established as a standard of care for p-Stage II/III gastric cancer by the ACTS-GC trial. However, the treatment for patients who had recurrent disease after the adjuvant treatment remains to be elucidated.

The response rate of treatment by S-1 + cisplatin is reportedly low (5 %) for patients who had recurrence within 6 months from completion of the S-1 adjuvant therapy when compared with the response rate for recurrences after 6 months from the completion (37.5 %) [50]. This result, found in a multi-institutional retrospective analysis, suggests that cancers that recur during or early after completion of an adjuvant chemotherapy are resistant to the drug used in that chemotherapy. On the other hand, a retrospective analysis of the patients registered for the ACTS-GC study revealed that patients who received S-1 among other drugs in salvage line treatments survived longer after recurrence than those who did not receive S-1, regardless of the time interval between the adjuvant chemotherapy and recurrence. However, results of this study will have to be interpreted with caution since the study suffers from several biases in the background of the patients such as whether oral food intake was possible (patients who did not receive S-1 after recurrence might have been those with bowel obstruction who were unable to eat and suffered from poor performance status).

In treatment for colorectal cancer during the era of adjuvant chemotherapy with 5FU alone, drugs used in the salvage line treatment depended on the time interval between the completion of the adjuvant chemotherapy and recurrence. New regimens have been developed as first-line therapy for patients who had recurrence more than 6 months after completion of the adjuvant treatment and as second line for those who had recurrence during or within 6 months of the adjuvant therapy.

The same rule has been applied for gastric cancer, and patients with late recurrence after adjuvant treatment have been deemed eligible for clinical trials exploring a first-line treatment, whereas those with early recurrences were registered in clinical trials for the second-line treatment.

Thus, patients with recurrences during or early after completion of the adjuvant treatment are considered as targets of second-line treatments, and S-1 monotherapy is usually avoided for this population. However, there is currently no evidence to recommend any specific regimen for this setting.

Standard of care has not been established for this population since the patients have not been eligible for most clinical trials for advanced/metastatic gastric cancer. Most patients in this population suffer from poor general status and will not tolerate the S-1 + cisplatin combination. Benefit for delivering chemotherapy should be weighed carefully against the risk, and best supportive care should be considered as an alternative.

The JCOG0106 study was one of the few in which only patients with peritoneal metastases detected by imaging studies such as CT and barium enema were eligible. In this trial, sequential therapy combining methotrexate + 5FU, which had been considered promising in this setting, was explored with continuous intravenous administration of 5FU (5FUci) as a control, but failed to show a survival benefit, while infusional 5FU was found to be less toxic [51]. Moreover, 5FUci enabled oral food intake in 41 % (7/27) of patients who had been unable to eat at the time of entry to the trial. Thus, 5FUci will be the current first choice for patients with bowel obstruction due to peritoneal metastases, but its effect on massive ascites remains elusive. On the other hand, another domestic phase III trial (ISO-5FU10) has shown non-inferiority of the 5FU + leucovorin (LV) combination against S-1 [52]. Thus, 5FU + LV, which can be delivered in the outpatient clinic, is another option for patients with relatively good general status.

In a phase II trial exploring weekly administration of paclitaxel in gastric cancer patients with ascites, improvement in the volume of ascites evaluated by a five-point measurement using the CT image was seen in 39 % (25/64) of the patients [53]. In a randomized phase II trial comparing second-line treatment by weekly administration of paclitaxel with the best available 5FU (either 5FUci or MTX + 5FU, which was not used in the first-line treatment) in patients with peritoneal metastases, a benefit in progression-free survival was proven, but no difference was detected in overall survival. However, paclitaxel was associated with a more favorable toxicity profile [54]. These results indicate that weekly paclitaxel can be considered for patients with severe peritoneal disease in both the first- and second-line setting. In addition, a phase II trial of the FLTAX regimen, which is a combination of paclitaxel with 5FU + LV, has shown that this combination reduced ascites in 44 % of patients [55]. Further evidence through a randomized comparison of the combination with a single-agent treatment is awaited.

Although S-1 + cisplatin is the standard first-line treatment for unresectable/recurrent gastric cancer in Japan, only patients up to 74 years of age were eligible for the SPIRITS trial that generated this evidence, and only 17 % (50/298) of patients registered for this trial were actually 70 years of age or older. In a subset analysis stratified by age, the hazard ratio of treatment by S-1 + cisplatin versus S-1 monotherapy was 0.75 (95 % CI 0.61–0.92) for patients under 60 years of age (n = 111) as opposed to 0.98 (95 % CI 0.82–1.17) for those between 60 and 69 and 0.95 (95 % CI 0.71–1.27) for those between 70 and 74 [15]. Thus, a benefit of adding cisplatin is unclear for the elderly population. In fact, there was no difference in survival between patients treated by S-1 + cisplatin and those treated by S-1 monotherapy in a retrospective study of the elderly population of ≥70 years of age, despite the apparent bias that the seemingly more fit patients were selected to receive the combination therapy [56]. Furthermore, a rather favorable outcome through S-1 monotherapy was reported in a cohort of patients ≥75 years of age in a prospective phase II trial focusing on elderly patients [57].

In short, regarding chemotherapy for advanced/metastatic gastric cancer, the evidence generated by the general population is unlikely to be directly applicable to elderly patients. That said, it may still be inadequate to estimate the tolerability of elderly patients to chemotherapy based only on chronological age without taking into account the major organ functions, comorbidities and past history. Unfortunately, however, a method to comprehensively evaluate the vulnerability of each aged individual has not been established.

Further evidence through clinical trials is needed for various decision-makings when treating elderly patients with gastric cancer. Until then, whether or not to deliver S-1 + cisplatin to these patients will have to be decided on a patient-by-patient basis based on the experience of each physician. Such decision will have to be based on the general condition of the patient with particular attention to the renal and cardiac function, always bearing in mind that S-1 monotherapy is quite reasonable as an alternative. Even after the treatment has started, the patient will have to be monitored with upmost care with attention paid not only to severe adverse events but also to anorexia, stomatitis and diarrhea, which could be particularly debilitating for elderly patients.

A trastuzumab-containing regimen is recommended for the first-line treatment of HER2-positive gastric cancer as a result of the ToGA trial [17]. There is no evidence to recommend any specific regimen for the specific cohort of HER2-positive patients who progressed during or after the trastuzumab-containing regimen. Either the taxanes or irinotecan could be selected as in the case of second-line treatment for HER2-negative gastric cancer.

On the other hand, a promising response rate of 37.0 % (95 % CI 23.2–52.5) and disease control rate of 82.6 % (95 % CI 68.2–92.2) were reported in a phase II trial exploring paclitaxel (weekly administration) + trastuzumab (JFMC45-1102 trial) for 46 evaluable patients with HER2-positive gastric cancer who were pretreated with a regimen that did not contain trastuzumab [58]. However, results of this trial will have to be interpreted with care for the following reasons: (1) Patients registered for this trial were not a typical cohort that receives second-line chemotherapy after the first-line treatment with a combination of fluorouracil and platinum agent in that 12 patients (26 %) were pretreated only with postoperative adjuvant chemotherapy while 5 patients (11 %) had already been treated with two lines of treatment; (2) although post-treatment cardiac function tests revealed that only one patient showed >10 % reduction in the left ventricular ejection fraction, other safety data are currently under analysis and have not been published.

There is currently no evidence in support of efficacy or safety for continuing with trastuzumab in case the patient was pretreated with a trastuzumab-containing regimen (trastuzumab beyond progression).