Description of the condition
Carcinoid syndrome, which was first described by Ransom in a patient with an ileal carcinoid tumour and hepatic metastasis in 1890, presents a typical set of symptoms as a result of functioning neuroendocrine tumours (Ransom 1890). It is generally accepted that carcinoid is a consequence of the bioactive substances (serotonin, histamine, tachykinins and prostaglandins among others) released by different types of neuroendocrine cells into the systemic circulation, however the exact pathogenesis for each symptom remains to be determined. The occurrence of carcinoid syndrome was about 8.4% according to a literature analysis of 8876 carcinoid patients (Soga 1999). Only few epidemiological studies have reported the incidence of carcinoid syndrome. The syndrome was observed in 49% of midgut (distal duodenum, jejunum, ileum, proximal colon) neuroendocrine tumours in France (Lombard-Bohas 2009). In contrast, the incidence of carcinoid syndrome was only 3.4% in gastrointestinal neuroendocrine tumours within the Japanese population (Ito 2010). Carcinoid syndrome is most frequently observed in small intestine neuroendocrine tumours with liver metastasis, which causes the direct release of hormones into the systemic circulation. Less frequent primary sites are lung and stomach (Kvols 1986; Modlin 2005).
The common manifestations of carcinoid syndrome include flushing (63% to 94% of patients), diarrhoea (68% to 85%), abdominal pain (10% to 55%), heart valvular lesions (10% to 53%), telangiectasia (25%), bronchoconstriction (3% to 19%), oedema (19%), arthropathy (7%), and pellagra (5% to 7%) (Creutzfeldt 1996; Massironi 2008; Schnirer 2003). Flushing as a characteristic marker of carcinoid syndrome occurs most frequently occurs on the face and sometimes the torso, and may last for 10 to 30 minutes (Bell 2005). Carcinoid cardiac disease characterized by plaque-like, ﬁbrous endocardial thickening is a major complication causing morbidity and mortality among patients with carcinoid syndrome. Nearly two-thirds of the patients manifest cardiac valvular lesions especially on the right side of the heart. Fibrosis at other locations besides the heart has also been reported in association with carcinoid syndrome, such as in the small intestine mesentery and lung (Modlin 2004).
The clinical manifestation and course of carcinoid syndrome may be highly variable and unpredictable. Depending on the composition of released vasoactive peptides, the syndrome may have a diverse constellation of presentations. The varied intensity and onset pattern of symptoms greatly increases the difficulty of clinical interpretation (Modlin 2005). The diagnosis of carcinoid syndrome requires a comprehensive analysis of clinical symptoms and biochemical testing. Two main biochemical tests are used nowadays. Urinary 5-hydroxyindoleacetic acid (5-HIAA), a metabolite of 5-hydroxytryptamine (5-HT), is not only associated with the pathogenesis of classic carcinoid symptoms (flushing, diarrhoea, carcinoid cardiac disease, etc) but is also related to the survival of carcinoid patients (Rorstad 2005). Chromogranin-A (Cg A) is a general biomarker of neuroendocrine tumours and is often used as a supportive indicator for carcinoid syndrome. Once the diagnosis of carcinoid syndrome is confirmed, topographic localization of neuroendocrine tumours (NETs) is needed using various imaging examinations including computed tomography (CT), magnetic resonance imaging (MRI), octreoscan, or positron emission tomography (PET).
Description of the intervention
Somatostatin is a native peptide hormone and plays a major inhibitory role in the digestive system and other organs. Since native somatostatin is short-acting and of limited usefulness in clinical treatment, synthetic somatostatin analogues were developed. Somatostatin analogues that are applied in clinical use includes octreotide (immediate release (IR) form), octreotide (long-acting release (LAR) formulations), lanreotide (fortnightly injection form), lanreotide autogel (monthly injection form), vapreotide, seglitide and pasireotide (Kiran 2011). The analogues form the mainstay of drug therapy for symptomatic control of carcinoid syndrome if the curative resection of a neuroendocrine tumour is beyond consideration (Oberg 2004). The optimal initial dosages of somatostatin analogues are flexible depending on the individual (Oberg 2004; Ramage 2005). Escalation is needed if symptom control remains unsatisfactory. IR octreotide should be used as a rescue procedure if there are re-emerging symptoms during treatment with LAR somatostatin (Oberg 2004). Additional treatment (cytoreduction surgery, embolization, radiofrequency ablation, etc) should be considered if the tumour progresses or symptoms worsen despite a dosage increase of somatostatin. Common adverse effects associated with somatostatin analogues include gastrointestinal disorders (nausea, abdominal pain, mild steatorrhoea and flatulence) and biliary disorders (gallstones or gallbladder sludge). Nearly half of the patients may present gastrointestinal complaints, which are usually modest and resolve with time (Jallad 2005). The risk of cholelithiasis reaches 10% to 50% (Norlen 2010; Trendle 1997). Fortunately most cases are asymptomatic and cholecystectomy is not necessary. Other rare adverse effects include bradycardia, abnormal metabolism of glucose and malabsorption of vitamins A, B12 and D. In general the side effects are mild and well tolerated (Gurusamy 2009; Lamberts 2002; Modlin 2008; Tomassetti 1998).
How the intervention might work
Somatostatin plays a general inhibitory role in the physiological function of the gastrointestinal tract including motility and hormone secretion. This is through interaction with five different G-protein coupled receptors (somatostatin receptor subtypes (SST) 1 to 5) (Oberg 2004). Among the five receptors, SST2 and SST5 have the most broad ranging inhibitory effects and present the main targets of octreotide and lanreotide (Susini 2006).
Why it is important to do this review
Neuroendocrine tumours have caused much concern due to their rapidly increasing incidence and relatively benign prognosis. Carcinoid syndrome is associated with functional neuroendocrine tumour and poses an important issue in clinical practice. Somatostatin analogues have long been accepted as the major treatment for carcinoid syndrome. However, no systematic reviews relating to the efficiency of somatostatin analogues for carcinoid syndrome, nor any comparisons regarding different kinds and forms of somatostatin analogues, have been published so far. Summarising the evidence and identifying the best form of somatostatin for the treatment of carcinoid syndrome is important for both clinicians and patients.
To assess the therapeutic and adverse effects of somatostatin analogues in carcinoid syndrome, and the differences between various somatostatin analogues and formulations.
Criteria for considering studies for this review
Types of studies
We will include all randomised controlled trials (RCTs) and quasi-randomised trials concerning somatostatin analogue treatment for carcinoid syndrome as a result of neuroendocrine tumours.
Types of participants
Patients with a clinical diagnosis of carcinoid syndrome resulting from neuroendocrine tumours. We will include all ages of patients.
Types of interventions
- Treatment with somatostatin analogues (IR octreotide, LAR octreotide, fortnightly injection of lanreotide form, lanreotide autogel, vapreotide, seglitide and pasireotide) versus placebo
- Comparisons between different somatostatin analogues
Types of outcome measures
- Symptom response, including diarrhoea and flushing: defined by improvement in the number of daily episodes of the symptom, or by authors.
- Biochemical response: defined by reduction in 24-hour urinary level of 5-hydroxyindoleacetic acid (5-HIAA).
- Adverse outcomes including gastrointestinal disorders (nausea, abdominal pain, mild steatorrhoea and flatulence) and biliary disorders (gallstones or gallbladder sludge).
Search methods for identification of studies
We will follow the search strategy of the Cochrane Upper Gastrointestinal and Pancreatic Diseases Group. We will combine carcinoid syndrome with somatostatin and its different analogues (IR octreotide, LAR octreotide, the fortnightly injection form of lanreotide, lanreotide autogel, vapreotide, seglitide and pasireotide) for both text word and MeSH term searching.
- The Cochrane Upper Gastrointestinal and Pancreatic Diseases Group Specialized Register, and the Cochrane Central Register of Controlled Trials (CENTRAL) in The Cochrane Library (latest issue)
- MEDLINE (from 1950 to present with Daily Update)
- EMBASE (from 1966 to present)
- PubMed online (latest update)
- Chinese Biomedical Database (from 1976 to present)
- China National Knowledge Infrastructure (CNKI) (from 1979 to present)
Searching other resources
We will conduct a handsearch of journals, if necessary.
Data collection and analysis
Selection of studies
Two review authors will independently assess the studies according to the inclusion and exclusion criteria. Reasons for exclusion will be documented. Any disagreements about inclusion will be discussed to reach consensus.
Data extraction and management
Data extraction will be performed by two review author independently using a standard data extraction form. The authors will assess the study quality by determining the clarity of items such as the inclusion and exclusion criteria, randomisation, adequacy of allocation concealment, blinding, management of withdrawals or drop-outs. If any disagreement occurs, the whole group of authors will discuss the issues to reach a consensus by following the instructions of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011). The following data will also be recorded: year of study and publication, country, language of publication, intervention (dosage, duration and regimen), number of patients, outcome (as mentioned above). If there is any doubt about overlap of patients in separate papers, we plan to acquire relevant information from the authors of the trials.
Assessment of risk of bias in included studies
We will assess the quality of the following methodological items in trials. Measures of quality for each item will be divided into three categories, as 'Adequate', 'Inadequate' or 'Unclear'.
- Adequate, if allocation sequence was generated by procedures such as random number table, computer software or other simple random methods.
- Unclear, if the trials were only mentioned as 'randomised', but the detailed methods were not described.
- Inadequate, if the generation of allocation sequence was not randomised or only quasi-randomised.
- Adequate, if specified personnel were responsible for the generation of allocation sequence and proper measures were taken to keep allocation information from both the patients and investigators.
- Unclear, if the method of allocation concealment was not mentioned.
- Inadequate, if the approach for allocation concealment was not performed or the allocation sequence was predictable for investigators and patients.
- Adequate, if both participants and investigators were blinded and the blinding methods were recorded.
- Unclear, if both participants and investigators were reported to be 'blinded' but the blinding methods were not recorded.
- inadequate, if only single blinding was achieved or there was no blinding at all.
Incomplete outcome data
- Adequate, if few drop-outs or withdrawals occurred and both the proportion and reasons for incomplete data were balanced between groups, or if the missing data were unlikely to affect the true outcome, or if there were no missing data at all.
- Unclear, if the proportion and reasons for drop-outs and withdrawals were not recorded.
- Inadequate, if wide differences in exclusions between groups were observed and were likely to affect the true outcome.
Selective outcome reporting
- Adequate, if all the major outcomes that were pre-specified in the trial's protocol were reported, or if all outcomes relating to this review were reported.
- Unclear, if available information is not adequate to assess selective outcome reporting.
- Inadequate, if outcomes relating to this review were incompletely reported.
- Adequate, if the trial didn't receive funding or donations from pharmaceutical companies or other facilities that were likely to influence the true outcome.
- Unclear, if the information about funding was not clear.
- Inadequate, if the trial received funding or donations from pharmaceutical companies or other facilities that were likely to influence the true outcome.
Measures of treatment effect
- For dichotomous data, we will calculate risk ratios of dichotomous outcomes and 95% confidence intervals (CI).
- For continuous data, we will calculate the mean difference (MD) or standardised mean difference (SMD) combined with the 95% CI.
Unit of analysis issues
For cross-over trials, we will incorporate data only from the first period to avoid unit of analysis error, if carry-over is problematic. Statistical advice will be sought to find the best solution.
Dealing with missing data
We will contact the original authors to retrieve missing data. If we do not receive a response, we will use intention-to-treat analysis for the missing data. All missing data will be assumed to be from unresponsive cases. Sensitivity analysis will be conducted to check the impact of imputation using 'best-case' and worst-case' scenarios.
Assessment of heterogeneity
We will assess clinical heterogeneity by the setting of the studies, the baseline characteristics of participants, and the diversity of interventions. We will using both the Chi² statistic and the I² statistic to analyse heterogeneity. Heterogeneity is considered statistically significant if the Chi² test P value is less than 0.10 or I² statistic is greater than 50% (Higgins 2011). Efforts will be made to find possible reasons if heterogeneity exists.
Assessment of reporting biases
we will assess reporting biases using the funnel plot model. However, if the number and sample size of included studies are very limited, reporting bias will be reported without applying the funnel plot model.
For continuous outcomes, we will calculate the MD or SMD combined with the 95% CI. A fixed-effect model meta-analysis will be used and a random-effects model will also be used to check the robustness of the results.
For dichotomous outcomes, we will calculate the risk ratios (RR) and 95% CIs. A fixed-effect model will be used for meta-analysis and a random-effects model analysis will be conducted to check the robustness of the results.
Subgroup analysis and investigation of heterogeneity
We will perform subgroup analysis according to the following aspects.
- Different forms of somatostatin analogues.
- Intervention duration and dosage.
We will perform sensitivity analyses by evaluating factors including study design, study quality and management of the missing data.
We thank the Cochrane Upper Gastrointestinal and Pancreatic Diseases Group for their professional support. We thank Sarah Rhodes and Marilyn Walsh for their advice, support and comments on the protocol. We thank Professor Li YouPing (Chinese Cochrane Center, China) for her kind advice and support of this review.
Contributions of authors
LinJie Guo: title registration, protocol design, data extraction, data analysis, draft of final review.
Pu Wang: data extraction, data analysis, draft of final review.
Bin Chen: data analysis, draft of final review.
Chen Li: data analysis, draft of final review.
CW Tang: protocol design, final review.
Declarations of interest
Sources of support
- Department of Gastroenterology, West China Hospital, Sichuan University, China.professional support
- No sources of support supplied