Gastrectomy after neoadjuvant chemotherapy in five cases of locally advanced gastric cancer with pancreatic head invasion

Abstract

Locally advanced gastric cancer (LAGC) with pancreatic head invasion (T4b) carries a poor prognosis despite radical surgery. Herein, we report the effectiveness of neoadjuvant chemotherapy (NAC) in improving the resectability of advanced gastric cancer with pancreatic invasion. A total of 2191 cases of gastric cancer were retrospectively analyzed from 13 institutions within the Hiroshima Surgical Study Group of Clinical Oncology (Hisco) database from 2018 to 2020. Among them, 5 of the 24 patients with Stage cT4b gastric cancer underwent NAC for three-to-eight cycles. Following chemotherapy, three patients underwent total gastrectomy, two patients underwent distal gastrectomy, and no patient underwent pancreaticoduodenectomy (PD). All five patients achieved a chemotherapeutic response of Grade 1b or higher, and only one case showed residual pancreatic invasion on pathology. This study suggests that NAC for Stage T4b LAGC with pancreatic head invasion may have the potential to obviate the need for PD.

Introduction

Diagnosing locally advanced gastric cancer (LAGC) with pancreatic head invasion (T4b) remains challenging due to the difficulty in accurately evaluating invasion depth and lymph-node metastasis. Moreover, such cases may present with distant metastases or dissemination, rendering curative resection unsuitable. Accordingly, LAGC carries a poor prognosis in affected patients, even with radical gastrectomy and D2 lymphadenectomy followed by adjuvant chemotherapy [1].

Pancreaticoduodenectomy (PD) is considered the gold standard for achieving R0 resection in LAGC with pancreatic invasion. Recent studies suggest that patients undergoing resection combined with PD for LAGC resulted in improved survival [2–5]. However, given the high morbidity, mortality, and unfavorable prognosis associated with PD, this procedure has been rarely performed in general hospitals for gastric cancer management [6]. In addition, these prior studies included cases without pancreatic involvement, potentially leading to unnecessarily extensive surgery for some patients.

Recent randomized-controlled studies have demonstrated the effectiveness of adjuvant chemotherapy in improving outcomes for advanced gastric cancer. Neoadjuvant chemotherapy (NAC) is also showing promise as an effective approach with ongoing clinical trials [7]. However, there are limited data on the therapeutic impact of radical resection for LAGC with pancreatic head invasion after NAC [8, 9] (Table 1).

Table 1.

Previous case reports of neoadjuvant chemotherapy for locally advanced gastric cancer with pancreatic head invasion

Report	Author, published year	Age, gender	Type	Tumor locationa	Histological type	NACb	Course of NAC	Surgical methodc	Rd
1 [8]	Yura, 2021	66, Male	3	LD,Circ	pap, tub1	T-mab +  S-1 +  L-OHP	4	PD	R0
2 [9]	Kondo, 2020	74, Male	3	L, Circ	por	S-1 +  L-OHP	3	DG	R0

Report	Surgical time	Bleeding	Depth of invasione (pT)	Lymph-node metastasise (pN)	Pathological stage	Patho-logical effective grade	Mobidityf	ACg	Recurrence	OSh (months)
1 [8]	676	490	T4b (panc)	N0	IIIA	1a	Aspiration pneumonia, pancreatic fistula, intra-abdominal abscess	S-1 + L-OHP	None	Alive, 8
2 [9]	316	208	T0	N0	0	3	None	Unknown	Unknown	Unknown

^aTumor location: L (antrum and pylorus), D (duodenum), Circ (circle)

^bNAC: neoadjuvant chemotherapy; S-1 (Tegafur Gimeracil Oteracil Potassium), L-OHP (Oxaliplatin), T-mab (Trastuzumab)

^cSurgical method: DG (distal gastrectomy), PD (gastrectomy combined with pancreatoduodenectomy)

^dR: residual tumor

^eStaging: according to Japanese Classifcation of the15th edition

^fMobidity: according to Clavien-Dindo classification

^gAC: postoperative adjuvant chemotherapy

^hOS: overall survival

Therefore, this study aimed to investigate the effectiveness of NAC in improving the resectability of LAGC with pancreatic head invasion based on the Hiroshima Surgical Study Group of Clinical Oncology (Hisco) database.

Case series

We retrospectively analyzed data from 2191 patients who underwent gastric cancer surgery at 13 institutions between 2018 and 2020; the data were obtained from the Hisco database. We identified 24 cases of clinical T4b (cT4b) LAGC with pancreatic head invasion, which could possibly be cured by resection. Patients with evident non-curative factors were excluded from the study (Fig. 1). For each case, patient demographics (age, sex, comorbidity, and performance status), surgical details (operative method, surgical time, and complication), pathological findings, chemotherapy regimens, and prognosis were analyzed.

Fig. 1.

Consort diagram. Cases with history of previous surgery and those with missing data were excluded

Of the 24 patients with cT4b LAGC, 5 received NAC (Table 2). In Cases 1–4, pancreatic head invasion by gastric cancer was identified through preoperative imaging, and in Case 5, this was confirmed through laparotomy findings. Although curative resection would have been achieved through PD, we opted for NAC to minimize excessive surgical invasion and reduce the risk of microscopic residual cancer. The patients were aged between 64 and 78 years, with a majority being male (four male, one female). Regarding medical history, four patients had comorbidities, and one underwent bypass prior to chemotherapy. The NAC regimens were as follows: three patients received Tegafur Gimeracil Oteracil Potassium (S1) + Oxaliplatin (SOX), one patient received Trastuzumab + S1 + Oxaliplatin (HERSOX), and one patient received S1 + Cisplatin. Treatment duration ranged from 2 to 4.5 months (three-to-eight cycles). Due to insufficient efficacy in Cases 3 and 5, additional doses were administered.

Table 2.

Our case reports of neoadjuvant chemotherapy for gastric cancer with pancreatic head invasion

Case	Age, gender	BMIa (kg/m2)	Type	Tumor locationb	Histological type	NACc	Therapy before NAC	Course of NAC	ASA-PSd	Clinical stage at the time of surgerye	Surgical methodf	Rg
1	68, Male	17.3	2	LMD, Circ	tub1	T-mab +  S1 +  L-OHP	–	3	2	T4a	DG	R0
2	78, Male	29.8	3	LMDU, Less	por2	S-1 +  L-OHP	–	3	3	T4b (panc), N: positive	TG (+ local pancreatic resection)	R1 (Cy1)
3	64, Male	22.2	4	MLDU, Circ	por2	S-1 +  CDDP	–	5	2	T3	TG	R0
4	69, Male	21.1	4	MLDU, Circ	por2	S-1 +  L-OHP	–	3	2	T4b (mesocolon), N: positive	TG	R0
5	76, Female	23.4	1	LD, Less	tub2	S-1 +  L-OHP	Bypass	8	2	T3	DG	R0

Surgical time (min)	Blood loss (ml)	Depth of invasione (pT)	Lymph-node metastasise (pN)	Pathological stagee	Pathological effective grade	Mobidityh	ACi	Recurrence	RFSj (months)	Chemo-therapy after recurrence	OSk (months)
235	280	T2	N0	IB	1b	–	T-mab +  S1	-	15.8	–	alive, 15.8
372	1010	T4b (panc)	N3a	IV	1b	Abscess G3a	S1  + DTX	Liver, local	7.6	Done	dead, 26.1
260	66	T3	N0	IIB	2a	–	S1	Peritoneum	13.7	Done	dead, 18.9
229	220	T4b (mesocolon)	N1	IIIB	2b	–	S1 +  L-OHP	Peritoneum	8.5	Done	dead, 12.0
213	70	T3	N0	IIA	1b	–	–	–	56.7	–	alive, 56.7

^aBMI: body mass index

^bTumor location: L (antrum and pylorus), M (corpus), D (duodenum), U (fundus), Circ (circle), Less (lesser curve)

^cNAC: neoadjuvant chemotherapy; S-1 (Tegafur Gimeracil Oteracil Potassium), L-OHP (Oxaliplatin), T-mab (Trastuzumab)

^dASA-PS: American Society of. Anesthesiologists physical status

^eStaging: according to Japanese Classification of the15th edition

^fSurgical method: DG (distal gastrectomy), TG (total gastrectomy)

^gR: residual tumor

^hMobidity: according to Clavien–Dindo classification

ⁱAC: postoperative adjuvant chemotherapy; DTX (Docetaxel)

^jRFS: relapse free survival

^kOS: overall survival

Preoperative American Society of Anesthesiologists performance status (ASA-PS) scores ranged from 2 to 3. Radical resection was performed for all five patients, including total gastrectomy in three patients and distal gastrectomy in two patients. Blood loss ranged from 66 to 1010 ml, and operative times ranged from 213 to 372 min. Only one patient experienced postoperative complications (Grade 3a: intra-abdominal abscess). Pathological examination revealed residual pancreatic invasion in one – (Case 2) (Figs. 2 and 3). This patient was ultimately diagnosed with Stage IV disease due to positive peritoneal lavage cytology (Cy-positive). In contrast, Case 5 demonstrated tumor shrinkage following NAC (Fig. 4a/b). Histopathological examination of the resected specimens showed residual cancer cells on hematoxylin and eosin staining, although infiltration beyond the serosa was not observed (Fig. 5). Notably, all cases achieved a chemotherapy response of Grade 1b or higher. Postoperative S1-based adjuvant chemotherapy was initiated for four patients, and three patients with recurrence received the second-line or later chemotherapy. Relapse-free survival and overall survival ranged from 7.6 to 56.7 months and 12.0 to 56.7 months, respectively. Among them, three patients succumbed to their disease.

Fig. 2.

Computed tomography (CT) following chemotherapy in Case 2. CT scans show a considerably reduced tumor mass after chemotherapy. However, the border between the tumor and the pancreatic head is unclear (yellow mark)

Fig. 3.

Histopathological findings under hematoxylin and eosin staining (× 10) in Case 2. Pancreatic infiltration of cancer cells is observed within the resected specimens

Fig. 4.

a Computed tomography scan obtained before neoadjuvant chemotherapy in Case 5. b Computed tomography scan obtained after neoadjuvant chemotherapy in Case 5. Tumor shrinkage is confirmed after undergoing chemotherapy

Fig. 5.

Histopathological examination of the resected specimens of Case 5. Histopathological examination of the resected specimens reveals residual cancer cells on hematoxylin and eosin staining. Infiltration beyond the serosa is not observed

Discussion

LAGC with deep invasion (T3 or higher) has always been associated with poor prognosis, even with radical gastrectomy and D2 lymphadenectomy followed by adjuvant chemotherapy [1]. In fact, studies report a 5-year survival rate of approximately 21.5% for cases of LAGC with organ involvement [10].

PD has been explored as a radical resection approach for LAGC with pancreatic invasion [2–5, 11, 12]. A recent systematic review by Li et al. [12] suggested PD as a viable option for select cases of LAGC with pancreatic invasion, offering acceptable surgical risk and potential survival benefits. However, studies have also indicated significant mortality and morbidity rates of 0–12.9% and 45.1–73.9%, respectively, in patients who underwent PD for gastric cancer [4, 11, 12]. One study identified pancreatic fistula as the most common and severe inflammatory complication of PD (43.5%), contributing to the adverse outcomes associated with PD in gastric cancer [11, 13]. Moreover, the highly invasive nature of PD may compromise postoperative quality of life, possibly hindering optimal adjuvant chemotherapy [14]. These factors limit the widespread adoption of this procedure in general hospitals, resulting in scarce data needed to determine the indications of PD for LAGC. Studies estimate that PD is required in about 1 in 500–1000 gastric cancer cases, with approximately one-third of cases harboring incurable factors that may affect survival [11]. Furthermore, accurate intraoperative assessment of invasion depth, lymph-node involvement, and peritoneal dissemination remains challenging [15].

Surgery remains the cornerstone for gastric cancer treatment; however, surgical excision alone often proves inadequate for patients with LAGC. NAC has emerged as a promising strategy for gastric cancer, with ongoing clinical trials investigating its efficacy [7]. Particularly, this approach can downstage tumors and improve the likelihood of achieving an R0 resection in LAGC [16]. In addition, NAC avoids the possibility of an over-invasive radical surgery, while ensuring a complete response after sufficient chemotherapy. It is for these reasons that NAC was considered in our study. Although a few studies have demonstrated the potential improvement in long-term prognosis with NAC for LAGC [17, 18], reports on the therapeutic effect of radical resection after NAC for LAGC with pancreatic invasion are also scarce [8, 9]. A recent single-institution analysis of surgical procedures for pathological T4b gastric cancer with pancreatic head invasion was published [19]. This study, which involved 54 of 76 cases (71%) treated with NAC, identified NAC, N stage, surgical margin, tumor type, and postoperative treatment as independent prognostic factors on multivariate analysis. Moreover, the overall 3-year survival rate of patients undergoing gastrectomy with pancreatectomy (GP) was significantly longer than that of those undergoing gastrectomy alone (GA) (47.6% vs 20.4%, p = 0.010). However, this study included only patients with pathological T4b diagnoses determined through postoperative pathological examinations. Cases where clinical T4b tumors identified in preoperative imaging were effectively downstaged to pathological T4a or below after NAC were excluded. This exclusion likely resulted in bias, as it omitted cases where NAC demonstrated efficacy. Consequently, the R0 resection rate was significantly higher in the PD group (87.5%) than in the GA group (0%). The low R0 resection rate in the GA group may have negatively affected the prognosis of those patients. Furthermore, in the multivariate analysis of prognostic factors presented in that study, the surgical procedure itself was not identified as an independent prognostic factor.

In Japan, the JCOG1509 trial is currently investigating the superiority of NAC with SOX therapy, with the aim of establishing this regimen as the standard treatment for LAGC [20].

Our case series, although retrospective in nature, demonstrates the feasibility of radical resection with NAC in select cases. Notably, only one of the five cases exhibited residual pancreatic invasion on pathology, indicating the effectiveness of NAC for tumor shrinkage and decreased invasion. Despite these advantages, Case 2, involving total gastrectomy with partial pancreatic resection, highlights the drawbacks of extensive surgery, where prolonged operative time and led to considerable blood loss, contributing to severe complications. This patient eventually succumbed to his disease following early recurrence of local and liver metastasis. This underscores the limitation of surgery alone in Cy-positive patients with a high risk of recurrence. In this study, PD was not performed in any of the five cases, and we instead opted for total and distal gastrectomy. While postoperative adjuvant therapy was performed in three cases, achieving sufficient treatment intensity was difficult. Even with radical resection, only a few cases would have achieved a sufficiently good prognosis due to the high risk of recurrence. Determining which cases can benefit from less-invasive treatments and providing a balanced multidisciplinary treatment approach become paramount in such cases.

Recent developments in targeted therapies, such as trastuzumab and immune checkpoint inhibitors, have been reported to improve treatment efficacy for non-curative gastric cancer [21]. The potential for targeted therapies to enhance radical conversion surgery for resectable lesions warrants further investigation in future studies.

Despite the insights offered in this study, certain limitations should be acknowledged. First, while this study involved multiple institutions, the sample size was relatively small, and PD cases were excluded. Second, the retrospective nature of this analysis likely excluded patients who received chemotherapy alone.

This case series analyzed cases of LAGC with pancreatic head invasion. Our findings demonstrated that NAC led to tumor downstaging and helped minimize the extent of surgery. In cases of Stage T4b LAGC with pancreatic head invasion, NAC may hold the potential to eliminate the necessity for PD.

Author contributions

M. Kano, N. Tokumoto, and K. Tanabe designed the study; M. Kano acquired, analyzed, interpreted the data, and drafted the initial manuscript; K. Toyota, R. Hotta, Y. Saeki, H. Tazawa, and N. Fujikuni acquired the data; K. Tanabe and J. Hihara revised the manuscript.

Funding

No funding was received for conducting this study.

Data availability

The data that support the findings of this study are not publicly available to preserve individuals’ privacy. Data requests can be made to Hiroshima Surgical Study Group of Clinical Oncology via this email: gnoritoku0424@gmail.com.

Declarations

Conflict of interests

The authors have no financial or proprietary interests in any material discussed in this article.

Ethical approval

The trial was conducted in accordance with the World Medical Association Declaration of Helsinki and Japanese Good Clinical Practice guidelines. The protocol was approved by the institutional review board of each participating hospital.

Informed consent

As an alternative to written informed consent for each patient, we used our institution's official website to provide an opt-out method.