Treatment of peritoneal metastases from small bowel adenocarcinoma

Abstract

Background/purpose: Peritoneal metastases (PM) affect approximately one third of patients with metastatic small bowel adenocarcinoma (SBA). Treatment options are (1) systemic therapy ± palliative surgery and (2) cytoreductive surgery with intraperitoneal chemotherapy (CRS + IPC). Due to scarce evidence, PM from SBA represents a therapeutic challenge. This narrative review summarised and discussed the evidence that investigated available treatment options.

Methods: Studies were discussed if they investigated first line systemic therapy for advanced SBA or CRS + IPC for PM from SBA. Extracted outcomes were objective response rate (ORR), disease control rate (DCR), progression-free survival (PFS), disease-free survival (DFS), overall survival (OS), and grade III–V toxicity/morbidity.

Results: Eighteen studies (15 observational, 3 phase II) that investigated systemic therapy and six observational studies that investigated CRS + IPC were reviewed. In studies that investigated systemic therapy, ORR, DCR, median PFS, median OS, and grade III–V toxicity ranged from 6% to 50%, 50% to 90%, 3 to 11 months, 8 to 20 months, and 10% to 68%, respectively. Fluoropyrimidine–oxaliplatin revealed favourable survival outcomes compared to fluoropyrimidine–irinotecan, fluoropyrimidine–cisplatin, fluoropyrimidine monotherapy, and other regimens. In studies that investigated CRS + IPC, median DFS, median OS, and grade III–V morbidity ranged from 10 to 12 months, 16 to 47 months, and 12% to 35%, respectively.

Conclusion: Based on available evidence, fluoropyrimidine–oxaliplatin should be regarded as optimal first line systemic treatment. In selected patients, CRS + IPC appears safe and may be more effective than systemic therapy as single treatment. Future studies should evaluate survival and morbidity of CRS + IPC in larger cohorts, as well as the value of chemotherapy with targeted agents in metastatic SBA with subgroup analysis for PM from SBA.

Keywords:

• Small bowel adenocarcinoma
• peritoneal carcinomatosis
• peritoneal metastases
• surgery
• HIPEC
• chemotherapy

Introduction

Approximately 5% of all gastrointestinal malignancies arises from the small bowel [1]. Small bowel adenocarcinoma (SBA) accounts for 30%–45% of all small bowel malignancies [2–6], with estimated age-standardised incidences of 0.5 per 100 000 in 1999 and 0.7 per 100 000 in 2013 [7]. SBA is mostly diagnosed in an advanced stage due to a low index of suspicion, vagueness of symptoms, difficult endoscopic access, and poor detection by radiologic imaging. As a result, approximately one third of patients with SBA present with stage IV disease at the time of diagnosis [2,3,7–9]. Peritoneal metastases (PM) affect 25%–50% with stage IV SBA and are more common in SBA arising from the jejunum or ileum than SBA arising from the duodenum [7,8,10–18] (Table 1).

Table 1. Percentage of patients with PM from SBA.

	mSBA	PM	PM
Study	N	n	%
Legué et al. [7]	581	168	29
Debaja et al. [8]	75	19	25
Locher et al. [10]	16	6	38
Fishman et al. [11]	64	31	48
Overman et al. [12]	80	37	46
Overman et al. [13]	25	9	36
Zaanan et al. [14]	80	40	50
Zaanan et al. [15]	27	12	44
Koo et al. [16]	87	42	48
Tshushima et al. [17]	122	37	30
Xiang et al. [18]	29	9	31
Total	1186	410	35

mSBA: metastatic small bowel adenocarcinoma; PM: peritoneal metastases.

Systemic therapy has been increasingly used for stage IV SBA [7]. This is based on scarce evidence obtained from several observational studies and small phase II trials in patients with advanced SBA [8–27]. The extrapolation of this evidence to PM from SBA may be questioned, since clinical studies in colorectal cancer (CRC) demonstrated that PM are relatively resistant to systemic therapy compared to other isolated metastatic sites [28–32].

A locoregional treatment option for PM from SBA is cytoreductive surgery with hyperthermic intraperitoneal chemotherapy (CRS + HIPEC). This treatment modality is gaining acceptance as standard of care for pseudomyxoma peritonei, peritoneal mesothelioma, and selected patients with PM from CRC [33,34]. CRS + HIPEC may also be beneficial in PM from SBA, since molecular characteristics and carcinogenesis of SBA and colorectal adenocarcinoma are roughly similar [35,36]. Evidence to support this hypothetical benefit is based on several observational studies [37–43].

Due to the rarity of PM from SBA and the paucity of data regarding systemic and surgical treatment, PM from SBA represent a major therapeutic challenge for surgeons and medical oncologists. Therefore, this narrative review aims to summarise and discuss studies that investigated available treatment options in order to provide treatment recommendations and future research directions.

Methods

Study eligibility

All studies that included patients who received first line systemic therapy for advanced SBA or CRS + IPC for PM from SBA were discussed in this review. Studies were considered ineligible if they were published >20 years ago or included ≤3 patients.

Data collection and outcome parameters

If reported, the following data were collected for each study by using a standardised form: year of publication, years of patient inclusion, study design, number of patients, primary tumour location, locally advanced or metastatic disease, percentage of patients who received first line systemic therapy, systemic therapy regimens, percentage of patients who received (neo)adjuvant systemic therapy, procedure related characteristics of CRS + IPC (e.g. IPC technique, IPC drugs, completeness of cytoreduction score, peritoneal cancer index).

If reported, the following outcome parameters were extracted by using a standardised form: objective response rate (ORR) in %, disease control rate (DCR) in %, median progression-free survival (PFS) in months, median disease-free survival (DFS) in months, median overall survival (OS) in months, and grade III–V morbidity/toxicity in %. Results from chi-square test (X²), log-rank test, and multivariable regression analyses were extracted if studies compared predefined outcome parameters between different interventions within the study.

ORR and DCR were based on either the Response Criteria in Solid Tumours (RECIST) or the World Health Organisation (WHO) response measurement criteria [44–46]. Toxicity and morbidity were graded according to either the National Cancer Institute Common Toxicity Criteria, National Cancer Institute Common Terminology Criteria for Adverse Events, or Clavien-Dindo [47–49].

Results

Twenty-four studies were discussed. Supplementary Appendix 1 demonstrates the data collection form for all studies.

Systemic therapy

Three phase II studies (Table 2a) and 15 observational studies reported predefined outcome parameters in patients who received first line systemic for advanced SBA (Table 2b–d) [8–14,16–19,21–27]. Five observational studies were non-comparative and reported predefined outcomes parameters of their total cohorts (Table 2b) [11,19,23,24,26]. Ten observational studies were comparative. Of these studies, four compared predefined outcome parameters between different systemic therapy regimens (Table 2c) [10,12,14,17], and six compared survival outcomes in patients who received systemic therapy vs. no systemic therapy (Table 2d) [8,9,16,22,25,27].

Table 2. Studies that investigate ORR, DCR, PFS, OS, or grade III–V toxicity (NCICTC) in patients who receive first line systemic therapy for advanced SBA.

Author		Systemic therapy		ORR	DCR	X^2	Median PFS		Median OS		III–V toxicity
Year (inclusion)	n	Regimen	%	%	%	p	Months	Log rank	Months	Log rank	%	X^2 p
(a) Phase II studies that investigate predefined outcome parameters
Gibson et al. [21] 2005 (1983–1985)	38	5-FU + mitomycin C + doxorubicin	100	19	31	–	5	–	8	–	68	–
Overman et al. [13] 2009 (2004–2007)	30	CAPOX	100	50	90	–	11	–	20	–	–	–
Xiang et al. [18] 2012 (2007–2011)	33	FOLFOX	100	49	85	–	8	–	15	–	39	–
(b) Non-comparative observational studies that investigate predefined outcome parameters
Crawley et al. [19] 1998 (1990–1995)	8	Yes 5-FU 5-FU + cisplatin	100 38 62	37	75	–	8	–	13	–	25	–
Fishman et al.[11] 2006 (1986–2004)	44	Yes Various regimens	100	30	50	–	–	–	11	–	–	–
Ono et al. [23] 2008 (1999–2007)	8	Yes Irinotecan + cisplatin	100 100	13	75	–	5	–	17	–	–	–
Suenaga et al. [24] 2009 (2001–2006)	10	Yes 5-FU	100 100	11	56	–	3	–	12	–	10	–
Zhang et al. [26] 2011 (2004–2010)	34	Yes FOLFOX CAPOX	100 82 18	32	62	–	6	–	14	–	–	–
(c) Comparative observational studies that compare predefined outcome parameters between different systemic therapy regimens
Locher et al. [10] 2005 (1992–2000)	20	Yes 5-FU + cisplatin 5-FU + carboplatin FOLFOX	100 75 10 15	21 20 0 33	79	–	8	–	14	–	45 53 0 33	–
Overman et al. [12] 2008 (1978–2005)	80	Yes 5-FU + platinum 5-FU or other	100 36 64	25 46 16	–	0.01	5 9 4	<0.01	13 15 12	0.10	–	–
Zaanan et al. [14] 2010 (1996–2008)	93	Yes 5-FU 5-FU + cisplatin FOLFIRI FOLFOX	100 11 17 20 52	26 0 31 9 34	74 50 69 73 79	0.18	7 8 5 6 7	0.16  0.02^a	15 14 9 11 18	0.25  0.04^a	33 0 75 39 46	<0.01^c
Tsushima et al. [17] 2012 (1999–2009)	132	Yes 5-FU 5-FU + cisplatin FOLFIRI FOLFOX Other	100 45 13 8 17 17	– 20 38 25 42 21	–	–	– 5 4 6 8 3	0.03^b	– 14 13 9 22 8	0.16^b	–	–
(d) Comparative observational studies that compare survival outcomes between systemic therapy vs. no systemic therapy
Debaja et al. [8] 2004 (1978–1998)	71	Yes No	68 32	–	–	–	–	–	12 2	0.02	–	–
Czaykowski et al. [22] 2007 (1990–2000)	37	Yes 5-FU based No	43 100 57	6	–	–	–	–	16 8	–	–	–
Moon et al. [25] 2009 (1986–2007)	87	Yes 5-FU 5-FU + cisplatin Gemcitabine based Other No	39 56 32 6 6 61	28	59	–	4	–	9 4	0.01	–	–
Halfdanarson et al. [9] 2010 (1970–2005)	165	Yes No	– –	–	–	–	–	–	16 3	<0.01	–	–
Koo et al. [16] 2011 (1989–2009)	91	Yes 5-FU 5-FU + platinum 5-FU + irinotecan No	44 30 63 8 56	11	37	–	6 1	<0.01	12 4	<0.01	–	–
Khan et al. [27] 2015 (1996–2011)	59	Yes Triplet Doublet Other No	78 50 39 11 22	50	67	–	–	–	61%^d 27%^d	0.04	–	–

5-FU: 5-fluorouracil; CAPOX: capecitabine, oxaliplatin; DCR: disease control rate; FOLFOX: 5-fluorouracil, leucovorin, oxaliplatin; FOLFIRI: 5-fluorouracil, irinotecan; ORR: objective response rate; OS: overall survival; PFS: progression-free survival, X²: chi-square test.

^a FOLFOX vs. 5-FU with cisplatin.

^b FOLFOX vs. 5-FU monotherapy.

^c 5-FU and cisplatin vs. other regimens.

^d One-year overall survival.

Tumour response

Twelve studies reported ORR or DCR. ORR ranged from 6% to 50% and DCR ranged from 50% to 90% between studies [10,11,13,14,18,19,22–27].

In phase II studies, ORR was 49% after 5-fluorouracil (5-FU) with leucovorin and oxaliplatin (FOLFOX) and 50% after capecitabine and oxaliplatin (CAPOX) [13,18]. DCR was 85% after FOLFOX and 90% after CAPOX [13,18].

In non-comparative observational studies, ORR and DCR were 6%–11% and 56% after 5-FU monotherapy [22,24], 13% and 75% after irinotecan and cisplatin [23], 32% and 62% after a fluoropyrimidine and oxaliplatin [26], and 50% and 67% after modern doublet and triplet systemic therapy regimens, respectively [27]. Other studies that used various regimens reported an ORR and DCR ranging from 28% to 37% and 50% to 79%, respectively [10,11,19,25].

In a French single-centre comparative observational study (n = 20), ORR after 5-FU and carboplatin, 5-FU and cisplatin, and FOLFOX was 0%, 20%, and 33%, respectively (no statistical analysis) [10]. In a larger American single-centre comparative observational study (n = 80), a higher ORR was found after 5-FU with a platinum compound vs. 5-FU monotherapy/other regimens (46% vs. 16%; X²: p = 0.01) [12]. In a French multicentre comparative observational study (n = 93), ORR and DCR after 5-FU monotherapy, 5-FU with leucovorin and irinotecan (FOLFIRI), 5-FU and cisplatin, and FOLFOX were 0% and 50%, 9% and 73%, 31% and 69%, and 34% and 79%, respectively (X²: p = 0.18) [14]. In a Japanese multicentre comparative observational study (n = 132), ORR after 5-FU monotherapy, FOLFIRI, 5-FU and cisplatin, FOLFOX, and other regimens was 20%, 25%, 38%, 42% and 21%, respectively (no statistical analysis) [17].

Progression-free survival

Twelve studies reported median PFS, which ranged from 3 to 11 months between studies [10,12–14,16,18,19,21,23–26].

In phase II studies, median PFS was 8 months after FOLFOX [18], 11 months after CAPOX [13], and 5 months after 5-FU, mitomycin C, and doxorubicin [21].

In non-comparative observational studies, median PFS was 3 months after 5-FU monotherapy [24], 5 months after irinotecan and cisplatin [23], 6–8 months after fluoropyrimidines and a platinum compound [10,26], and 4–8 months in studies that used various regimens [19,25].

In a Korean single-centre comparative observational study, median PFS was 6 months in patients who received chemotherapy compared to 1 month in patients who did not receive chemotherapy (log-rank: p < 0.01) [16]. In the American single-centre study, median PFS after 5-FU with and without a platinum compound was 9 months vs. 4 months, respectively (log-rank: p < 0.01) [12]. In the French multicentre study, median PFS after 5-FU monotherapy, FOLFIRI, 5-FU and cisplatin, and FOLFOX was 8 months, 6 months, 5 months, and 7 months, respectively (log-rank: p = 0.16) [14]. In this study, median PFS was shorter after 5-FU and cisplatin than after FOLFOX on univariable (log-rank: p = 0.02) and multivariable analysis (HR: 2.25 [95%CI 1.12–4.50]) [14]. In the Japanese multicentre study, median PFS after 5-FU monotherapy, FOLFIRI, 5-FU and cisplatin, FOLFOX, and other regimens were 5 months, 6 months, 4 months, 8 months, and 3 months, respectively [17]. In this study, median PFS was higher after FOLFOX vs. 5-FU monotherapy on univariable (log-rank p = 0.03) and multivariable analysis (HR 0.48 [95%CI 0.23–0.84]), whereas no statistically significant difference in PFS was found between 5-FU monotherapy and FOLFIRI, 5-FU and cisplatin, or other regimens [17].

Overall survival

Seventeen studies reported median OS, which ranged from 8 to 20 months between studies [8–14,16,18,19,21,22–27].

In phase II studies, median OS was 15 months after FOLFOX [18], 20 months after CAPOX [13], and 8 months after 5-FU, mitomycin C, and doxorubicin [21].

In non-comparative observational studies, median OS was 12 months after 5-FU monotherapy [24], 17 months after irinotecan and cisplatin [23], 14 months after fluoropyrimidines and a platinum compound [10,26], and 11–13 months in studies that used various regimens [11,19].

Six comparative studies revealed higher median OS in patients who received chemotherapy (12–16 months) vs. patients who did not receive chemotherapy (2–8 months) (log-rank: p < 0.05 in all studies) [8,9,16,22,25,27]. In the American single-centre study, median OS after 5-FU with and without a platinum compound was 15 vs. 12 months, respectively (log-rank: p = 0.10) [12]. In the French multicentre study, median OS after 5-FU monotherapy, FOLFIRI, 5-FU and cisplatin, and FOLFOX was 14 months, 11 months, 9 months and 18 months, respectively (log-rank p = 0.25) [14]. In this study, median OS was shorter after 5-FU and cisplatin compared to FOLFOX on univariable (log-rank: p = 0.04) and multivariable analysis (HR: 2.75 [95%CI 1.18–6.41]) [14]. In the Japanese multicentre study, median OS after 5-FU monotherapy, FOLFIRI, 5-FU and cisplatin, FOLFOX, and other regimens was 14 months, 9 months, 13 months, 22 months, and 8 months, respectively [17]. In this study, median OS did not differ between FOLFOX and 5-FU monotherapy on univariable analysis (log-rank: p = 0.16). However, median OS was higher after FOLFOX vs. 5-FU monotherapy on multivariable analysis (HR: 0.45 [95%CI 0.23–0.88]), whereas no difference in OS was found between 5-FU monotherapy and FOLFIRI, 5-FU and cisplatin, or other regimens [17].

Grade III–V toxicity

Six studies reported grade III–V toxicity, which ranged from 10% to 68% between studies [10,14,18,19,21,24].

In phase II studies, grade III–V toxicity was 39% with FOLFOX [18], and 68% with 5-FU, mitomycin C, and doxorubicin [21].

In non-comparative observational studies, grade III–V toxicity was 10% with 5-FU monotherapy [24], and 25% in a study that used various regimens [19].

In the French single-centre study, grade III–V toxicity of 5-FU and carboplatin, 5-FU and cisplatin, and FOLFOX were 0%, 53%, and 33%, respectively (no statistical analysis) [10]. In the French multicentre study, grade III–V toxicity of 5-FU monotherapy, FOLFIRI, 5-FU and cisplatin, and FOLFOX was 0%, 39%, 75%, and 46%, respectively [14]. In this study, grade III–V toxicity was higher with 5-FU and cisplatin compared to other regimens (X²: p < 0.01) [14].

Cytoreductive surgery with intraperitoneal chemotherapy

Six observational studies reported predefined outcome parameters in patients who underwent CRS + IPC for PM from SBA (Table 3) [37,39–43]. The percentage of patients who received neoadjuvant and adjuvant systemic therapy ranged from 14% to 90% and 29% to 57% between studies, respectively [39–43].

Disease-free and overall survival

Two studies reported median DFS, which was 12 months after CRS + HIPEC with mitomycin C [42] and 10 months in a study that used various IPC techniques and drugs [39].

Table 3. Observational studies that investigate DFS, OS, and grade III–V morbidity in patients who undergo cytoreductive surgery with intraperitoneal chemotherapy for peritoneal metastases arising from SBA.

Author		IPC technique		Median DFS	Median OS^a	Median OS^b	III–V morbidity
Year (inclusion)	n	CRS with	%	Months	Months	Months	%
Marchettini et al. [37] 2002 (–)	6	HIPEC + EPIC	100	–	16	–	–
Chua et al. [39] 2009 (1997–2009)	7	HIPEC + EPIC EPIC HIPEC	57  29  14	12	25	–	29
Elias et al. [40] 2010 (1989–2007)	31	HIPEC EPIC	68  32	–	47	–	35
Sun et al. [41] 2013 (1995–2011)	17	HIPEC	100	–	18	37	12
Van Oudheusden et al. [42] 2015 (2005–2014)	16	HIPEC	100	10	31	–	25
Liu et al. [43] 2015 (2006–2014)	31	HIPEC	100	–	36	51	26

CRS: cytoreductive surgery; DFS: disease-free survival; EPIC: early postoperative intraperitoneal chemotherapy; HIPEC: hyperthermic intraperitoneal chemotherapy; OS: overall survival; SBA: small bowel adenocarcinoma.

^a Measured from CRS + IPC.

^b Measured from diagnosis with PM from SBA.

Six studies reported median OS measured from CRS + IPC, which ranged from 16 to 47 months between studies [37,39–43]. Median OS was 16 months after CRS + HIPEC with mitomycin C and early postoperative intraperitoneal chemotherapy (EPIC) with 5-FU [37], 18–36 months after CRS + HIPEC with mitomycin C [41–43], and 25–47 months in studies that used various IPC techniques and drugs [39,40]. Median OS measured from diagnosis was reported in two studies that used CRS + HIPEC with mitomycin C and ranged from 37 to 51 months [41,43].

Grade III–V morbidity

Five studies reported grade III–V morbidity, which ranged from 12% to 35% between studies [39–43]. Grade III–V morbidity was 12%–26% after CRS + HIPEC with mitomycin C [41–43], and 29%–35% in studies that used various IPC techniques and drugs [39,40].

Discussion

Based on results of phase II and comparative observational studies, first line systemic treatment with a fluoropyrimidine and oxaliplatin seems to achieve favourable survival outcomes and at least equal toxicity and tumour response outcomes compared to regimens with a fluoropyrimidine and irinotecan or cisplatin. In selected patients with PM from SBA, CRS + IPC with or without (neo)adjuvant systemic therapy seems safe and may be more effective than systemic therapy as single treatment, with promising survival and morbidity outcomes.

Systemic therapy

In daily practice, systemic therapy is the most common treatment option for PM from SBA. In the absence of prospective randomised studies, there is no consensus on the preferred first line regimen in metastatic SBA, even though observational studies in this review point towards superior results of a fluoropyrimidine with oxaliplatin. Therefore, information obtained from observational and phase II studies in metastatic SBA may be complemented with evidence derived from metastatic CRC, which has similar molecular characteristics and carcinogenesis pathways [35,36].

In metastatic CRC, a fluoropyrimidine with oxaliplatin has been standard of care for many years [50]. In 2011, a phase II study revealed a promising median OS of 21.5 months in patients who received FOLFOX for metastatic CRC with PM [51]. More recently, addition of targeted agents to combination chemotherapy significantly improved survival of metastatic CRC [52]. The role of targeted agents in metastatic SBA is less clear. However, the high expression of vascular endothelial growth factor (VEGF) A (96%) and epidermal growth factor receptor (EGFR) (71%) in SBA may suggest a role for treatment with targeted agents such as bevacizumab or cetuximab [53].

Interestingly, addition of bevacizumab to chemotherapy improved survival of PM from CRC [54]. Specifically for peritoneal tumour depositions, VEGF is known to have a role in tumour related angiogenesis and formation of ascites [55]. Therefore, VEGF inhibitors may be of value as addition to chemotherapy for the treatment of PM from various gastrointestinal origins, including SBA. However, the role of VEGF inhibitors in this setting needs to be further clarified.

Limitations of available evidence

None of the studies that investigated systemic therapy for advanced SBA provided a subgroup analysis of patients with PM. In a recently published pooled subgroup analysis of 14 prospective randomised controlled trials in metastatic CRC, patients with PM demonstrated shorter OS compared to patients with other isolated sites of metastases [32]. However, presence of PM was no predictor of survival in studies in this review [12,14–16]. Nevertheless, the absence of this subgroup analysis impedes the extrapolation of results of systemic therapy for patients with any form of advanced SBA to patients with PM from SBA.

Furthermore, all studies in this review included patients with SBA arising from duodenum, ileum, and jejunum. However, duodenal adenocarcinomas arise mostly retroperitoneal and are known for a more advanced disease at presentation, worse prognosis, and a different metastatic pattern with less peritoneal seeding [7]. Therefore, it could be questioned whether these malignancies should be regarded as one disease entity in oncological studies.

Finally, given the large inclusion periods of studies in this review, several non-comparative studies only used conventional regimens [21,22,24], and other studies that used both conventional and modern regimens reported outcomes in their total cohort [10,11,16,19,25]. This interventional heterogeneity impedes extrapolation of results to patients who are treated in the era of modern systemic therapy regimens.

Future research directions

The value of targeted agents in metastatic SBA is currently under investigation in a phase IB (ClinicalTrials.gov: NCT00987766) and two phase II studies (ClinicalTrials.gov: NCT01202409, NCT01208103). Ideally, future studies that focus on systemic therapy for metastatic SBA should provide subgroup analyses for PM from SBA, in order to facilitate a more reliable comparison of survival outcomes between systemic therapy as single treatment and CRS + IPC. Based on the current review, patients with PM from SBA may be treated with a fluoropyrimidine with oxaliplatin until results of on-going trials become available. Since there is only one observational study that investigated second line chemotherapy (FOLFIRI) in advanced SBA [15], future studies are needed to identify the optimal second line regimen.

Cytoreductive surgery with intraperitoneal chemotherapy

In selected patients with PM from SBA, CRS + IPC seems to be an attractive treatment option. Selection criteria include absence of unresectable systemic disease, achievability of a complete CRS, and a physical status that allows a major surgical procedure. In the two most recent studies by van Oudheusden et al. [42] and Liu et al. [43], survival and morbidity outcomes were comparable with those after CRS + HIPEC for PM from CRC [34]. The median OS of 31–36 months in these studies is compatible with the pursuit of at least 30 months after CRS + HIPEC in PM from CRC as defined by the Peritoneal Surface Oncology Group International [56]. Given these promising survival results, efforts should be made to increase awareness regarding this treatment option. The low number of patients with PM from SBA who were treated with CRS + HIPEC demonstrates that this awareness is currently lacking. In the Netherlands, 16 patients were treated with CRS + HIPEC for PM from SBA over a 10 year time period [42], while the number of diagnosed patients was significantly higher. A Dutch epidemiological study revealed that 167 patients were diagnosed with synchronous PM from SBA between 1999 and 2013 [7]. Approximately 60% of these patients presented with isolated PM without concurrent systemic metastases (Dutch Cancer Registry), and are therefore potential candidates for CRS + HIPEC. These results underestimate the true number of patients with isolated PM from SBA, since data on metachronous PM from SBA are not available. The difference between the number of patients diagnosed with isolated PM from SBA and the number of patients who underwent CRS + HIPEC may indicate underuse of CRS + HIPEC for potentially eligible patients.

Limitations of available evidence

All studies that investigated CRS + HIPEC for PM from SBA have an observational design and no prospective studies on this topic have been conducted. Due to large inclusion periods within studies and different inclusion periods across studies, heterogeneity in interventions both within and across studies was observed. Several studies only used CRS + HIPEC [41–43], whereas other studies mainly used CRS + EPIC [37,39,40]. Moreover, (neo)adjuvant treatment varied across studies from mainly adjuvant [39,40,42], to mainly neoadjuvant [43], or both [41]. Since both HIPEC or EPIC and neoadjuvant or adjuvant systemic therapy appear to have a different influence on morbidity and survival in CRC [57–64], this heterogeneity may have biased reported outcomes.

Besides interventional heterogeneity, heterogeneity was also observed in included patients. Some studies only included patients with a complete cytoreduction [38,39] and a low peritoneal cancer index (PCI) [39,41], whereas other studies also included patients with an incomplete cytoreduction [40–42] and a high PCI [38,42]. Since completeness of cytoreduction and PCI are known to be the most important prognostic factors for survival in CRC [65], this heterogeneity may have influenced survival outcomes.

Finally, the small number of included patients impedes statistical analysis for prognostic variables that predict and influence survival and morbidity.

Future research directions

Limitations of available evidence stress the need for higher quality studies in order to confirm the promising results of CRS + HIPEC in selected patients with PM from SBA. Given the rarity of the disease, it is unlikely that this evidence will ever be provided by a prospective randomised study. In order to increase the number of patients in observational studies, a multi-institutional data registry on PM from SBA was established during the 9th International Congress on Peritoneal Surface Malignancies in 2014. Forthcoming results of this data registry will provide more insight in (1) survival and morbidity outcomes in a larger cohort and (2) prognostic patient, tumour and treatment related variables that predict and influence survival and morbidity.

If a patient is no candidate for CRS + HIPEC due to extensive peritoneal disease or a poor performance status, pressurised intraperitoneal aerosol chemotherapy (PIPAC) may be a future treatment option for symptomatic peritoneal disease [66]. Preliminary results of this innovative locoregional treatment modality, which may be combined with systemic therapy in case of concurrent systemic disease, revealed interesting peritoneal tumour regression in end-stage PM arising from gastric [67], colorectal [68], and ovarian cancer [69], without quality of life deterioration [70]. Several on-going studies are currently investigating this technique for PM from various primary malignancies (ClinicalTrials.gov: NCT02475772, NCT01809379, NCT01854255, NCT02320448, NCT02735928, NCT02604784). However, by our knowledge, no PIPAC has been performed in patients with PM from SBA thus far.

Conclusions

Based on available evidence, a fluoropyrimidine with oxaliplatin seems to be the optimal first line systemic therapy regimen for patients with PM from SBA. In selected patients, CRS + IPC appears safe and may be more effective than systemic therapy as single treatment. Future studies should evaluate survival and morbidity of CRS + IPC in larger cohorts, as well as the value of systemic chemotherapy with targeted agents in metastatic SBA with subgroup analysis for PM from SBA.

Disclosure statement

The authors report no declarations of interest.