Landscape of interventional clinical trials involving gastrectomy for gastric cancer

Background Gastric cancer (GC) is the third most common cause of malignancy associated mortality globally. The cornerstone of curative treatment involves surgical gastrectomy. In this study, we explore clinical trials involving gastrectomy for GC, highlighting inadequacies and underlining promising surgical interventions and strategies. Materials and methods On 1 May 2020, ClinicalTrials.gov was explored for interventional trials related to gastrectomy for GC, without adding limitations for location or date. All data pertaining to the trials were collected. Characteristics such as phase, duration, enrolment size, location, treatment allocation, masking and primary endpoint were analysed. Results One hundred thirty-eight clinical trials met the search criteria. Clinical trials were performed in only 14 countries; most of them occurring in China. Most trials (33%) were still in the recruiting phase. On average, the length of trials was 3.9 years. Most trials had parallel assignment, were randomised and masked. The primary endpoint which was mostly commonly studied was overall survival (33%). The most common intervention studied is laparoscopic gastrectomy in 43 (31%) trials. Conclusions Our study exposed a small number of trials, publication rate, absence of geographic variety in clinical trials involving gastrectomy for GC. Adequate management of trial design can help decrease duration and increase validity of results. More trials comparing different surgical techniques are needed to update the surgical practice of gastrectomy for GC.


Introduction
Among all causes of death worldwide, cancers and malignancies are the second most common cause [1]. The six most prevalent cancer globally and third common cause of malignancy related mortality is gastric cancer (GC) [2]. The majority (almost 90%) of GCs are adenocarcinomas, which arise from mucosal glands of the stomach [3]. GC incidence rates vary extensively between both sexes, and across different countries and geographic locations [3]. The areas with the highest rates of GC worldwide are East Asia and Eastern Europe, while North America is home to the lowest rates in the world [4].
The cornerstone of curative treatment for this disease is surgical resection with lymphadenectomy [5]. Yet, only about 50% of all GC patients may undergo resection with curative intent. Curative resection has a '5-year survival rate' of around 45%, with perioperative chemotherapy improving that rate by around 10% [6,7].
In the past couple of decades, surgeries for GC have witnessed a shift from the traditional open approach to more minimally invasive operations [8]. These new techniques include laparoscopic assisted, total laparoscopic, robot assisted and total robotic operations. Now, minimally invasive surgical approaches have become the new standard for GC. These surgeries provide briefer hospital stay, faster recovery and general enhancement in patient quality of life [9,10]. This shift to minimally invasive surgeries came about because of reasons such as advancements in surgical instruments, increased experience among surgeons and the better outcomes associated with them [11].
In recent years, surgery has become the primary intervention employed in the management of GC. Beneficial clinical outcomes depend heavily on finding of new surgical techniques and treatment plans. Therefore, it is vital to evaluate surgical interventions that are currently in trial or new ones that have arose. Here, we give an overview of gastrectomy clinical trials for GC, study the characteristics, discuss the inadequacies associated with them, highlight the effective interventions present and suggest potential rooms for enhancement.

Search strategy and selection criteria
On 1 May 2020, we retrieved all information on clinical trials involving gastrectomy for GC from ClinicalTrials.gov. This search was conducted without adding limitations for date or location. 'Clinicaltrials.gov' is a registry that archives new data on various clinical trials weekly. For investigators to submit entries into the registry, they are required to provide extensive specifics on their particular trial. These details include trial profile, a report of protocol used in their study and any history that may be relevant. Analysis and extrapolation of conclusions on the basis of information present in this wide-ranging registry has previously been described in a number studies [12,13].
Of the 326 total trails collected, 177 were eliminated as they were either 'non-interventional' or did not involve gastrectomy as an intervention. This exclusion was done using a parallel elimination plan as Nasrallah et al [14] where withdrawn/terminated and 'non-interventional' trials were not included in the final set of studies analysed (Figure 1).

Data collection
All information relating to the clinical trials were gathered. This included: trial status ('active not recruiting', 'completed', 'enrolling by invitation', 'not yet recruiting', 'suspended', etc.), phase of trial ('I', 'I/II', 'II', 'II/III', 'IV'), the official start and end/completion dates, location (city/ country), selection criteria (inclusion and exclusion), primary endpoints, sample size, outcomes, interventions used and where any publications were produced. Trial duration was calculated from the official start date until the primary end/completion date. This was done to be in accordance to the Food and Drug Administration Amendments Act which was announced in 2007 (Section 801) [15]. Primary endpoints were defined as '30-day reoperation', 'number of lymph nodes harvested/collected', 'operation time', 'percentage body weight ratio', 'overall survival', 'postoperative length of stay', 'postoperative morbidity', 'postoperative outcomes', 'progression-free survival', 'rate of conversion', 'quality of life', 'time till drain removal' and 'tumour recurrence rate'.

Trial characteristics
One hundred and thirty eight trials had the criteria needed of our study. The distribution of these clinical trials was done according to characteristic details such as the number of participants, phase, status, location and duration in Table 1. This study showed that 39,954 participants were registered altogether in clinical trials involving gastrectomy for GC (Table 1). More than 70% of trials had >100 patients enrolled (Table  1). Clinical trials were performed across only 14 different countries, with the majority taking place in Asia/Australia (Table 1, Figure 2). The duration/length was stated in all trials, with an average length of 3.9 years (Table 1). Most trials (33%) were in the recruiting phase. Almost all (95%) of the clinical trials were for adults only, and all of them (100%) were for both genders.
Trials were further assorted by interventional model, treatment allocation, masking and primary end point as in Table 2. In terms of interventional models, 82% of trials had parallel assignment ( Table 2). 78% of trials were randomised and were not masked ( Table 2). The most common primary endpoint was progression-free survival, where it was present in 28% of all trials in our study.

Publications linked to trials
Of the 138 total clinical trials in our study, only 39 had linked publications related to gastrectomy surgeries for GC. A total number of 57 publications were retrieved (Table 1). Of the 33 completed trials, 28 publications were produced.

Not applicable (NA)
Sixty-two trials (45%) were in this phase, with only 14 trials being completed (Table 1). Patients enrolled in non-applicable phase trials were 17,444. All four trials which included paediatric cases were found in this category (Table 1). These trials were spread across 12 different countries, with China conducting the highest number at 39 (63%). Average trial duration in this phase was 3.3 years. Sixteen publications were linked to trials in this status (Table 1). The overwhelming majority of trials had parallel assignment (86%), were randomised (79%) and were not masked (67%) ( Table 2). Progression-free survival was the most common primary endpoint in these trials with a total number of 17 (28%).

Phase I trials
Only 2 (1%) trials were in phase I, both of which were of unknown status (Table 1). Phase I trials had 204 patients enrolled, both of which were only for adults. These trials were conducted in China and Republic of Korea (Table 1, Figure 2). Average trial length was around 3.5 years. A single publication was linked to these trials (Table 1). Both clinical trials were randomised, had parallel assignment and masked ( Table 2). One trial focused on progression-free survival, while the other had a primary endpoint of quality of life (Table 2).

Phase I/II trials
Only 3 (2%) trials were in Phase I/II, all of which were completed (Table 1). Phase I/II trials had 117 patients enrolled, all of which were only for adults. Two trials were conducted in Republic of Korea and one was done in China (Table 1, Figure 2). Average trial length was 1.8 years. Two published works were retrieved from phase I/II trials (Table 1). Single group assignment was found in two trials, while one trial had parallel assignment ( Table 2). Two trials had no specified treatment allocation, while one trial was randomised. Two trials were not masked, while one trial did not specify any details on masking. The primary endpoints studied were: 'number of harvested/collected lymph nodes', 'postoperative morbidity' and 'rate of conversion' (Table 2).

Phase II trials
Twenty-six (19%) trials were in phase II, only five of which were completed (Table 1). Trials of this status had 2,967 patients enrolled, all of which were adults only. Trials were distributed to six countries, with China conducting the highest number at 16 (Table 1, Figure 2). Average trial length was 2.9 years. Ten published works were retrieved from trials in this phase (Table 1). The majority of trials (54%) had parallel assignment and were randomised ( Table 2). Almost all (92%) of trials had no masking (Table 2). Progression-free survival was the most common primary endpoint, which was found in seven trials ( Table 2).
Main primary endpoint Number of harvested lymph nodes 5 - Phase II/III trials Three (2%) trials were in phase II/III, only one of which was completed (Table 1). Trials of this status had 440 patients enrolled, all of which were adults only. All trials were conducted in China (Table 1, Figure 2). Average trial length was 4.6 years. No published works were retrieved from trials in this phase. All clinical trials had parallel assignment and were masked ( Table 2). Two trials were randomised, while one was nonrandomised (Table 2). Two studies had a primary endpoint of overall survival, while one was interested in progression-free survival ( Table 2).

Phase III trials
Forty-one trials (30%) were in phase III, where 11 were completed (  Figure 2). Average trial length was 5.4 years. Twenty-six publications were linked to trials in this phase (Table 1). Almost all (98%) trials had parallel assignment ( Table 2), and the majority were randomised (95%) and had no masking (88%) ( Table 2). The most common primary endpoint was overall survival (37%) ( Table 2).

Phase IV trials
A single (1%) trial was in phase IV, and it has an unknown status (Table 1). The trial in this phase had 150 patients enrolled, where it was for adults only. The trial was conducted in Italy (Table 1, Figure 2). Trial length was 2 years. Two publications were linked to this trial (Table 1). The trial had parallel assignment, was randomised and non-masked ( Table 2). The primary endpoint was postoperative outcomes ( Table 2).

Inadequacies of clinical trials
As of May 2020, there have been 138 interventional clinical trials involving gastrectomy for GC. Surgical clinical trials for GC are rare. Of the 2,028 listed trials for GC, only 138 had experimental interventions involving gastrectomy, constituting 6.8% of the total number of trials. This low percentage and number may be accredited to several reasons.
Results from our study show that clinical trials involving gastrectomy for GC lack diversity, and that there is somewhat of a proportionality between disease burden and number of trials conducted. The overwhelming majority of trials were found in eastern Asia, which is also the most heavily burdened region in the world by the malignancy [4].
Our data shows that clinicals trials were conducted across only 14 different countries worldwide. Brazil was the only country in South America to have conducted a clinical trial on gastrectomy for GC and only Egypt had conducted a couple trials in the whole of Africa. This observation can be explained by several reasons. First, states present in less developed areas like Africa perhaps lack fiscal means and the set-up needed for research studies [16]. Globally, a great discrepancy is present in the prevalence and incidence of GC. For example, yearly age standardised incidence rates of GC in the Republic of Korea is 65.9 cases per 100,000 versus a mere 7.8 cases per 100,000 in the United States [17]. This can be explained by the notion that research on specific diseases is dependent on the burden of the disease itself. Furthermore, in a large number of countries, trials do not need to be registered in 'ClinicalTrials.gov'. This could explain why no studies were found in many nations around the world.    On average, the length of trials in this work was 3.9 years. It may be hypothesised that the long lengths of trials are owed to schemas of the respective studies. In addition, obligatory approvals and financial/logistical backing often taken a lot of time. With better planning and optimised time managing, the lengths of studies may become shorter. This would result in faster development of new gastrectomy surgical management plans for GC. Introducing 'new master protocols' for the screening of patients in regard to numerous features such as race and ethnicity, genetic profile and sex could help in restructuring selection processes and in the assignment of volunteers into trials in a well-matched manner according to their profile [18]. The termination of most trials was in fact due to the slow or poor participation of patients, and discontinuation of funding. The recruitment for clinical trials has been inept to increase the number of participants [19]. Health communication strategies using advertisement and media outlets did not succeed in expanding the overall volunteer number until now [20]. However, the optimal use of electronic health records (EHR) in screening for potential candidates has proved to improve volunteer requirement into trials [21]. The recruitment of participants to gastrectomy clinical trials for GC may become more effective and quicker now that a growing number of medical centres and institutions have started to adopt EHR systems. Also, surgeons that treat patients with GC are required to stay up to date on the current gastrectomy clinical trials in order to advise said patients to volunteer in them. Previous reports shown that people are more prone to volunteer in clinical trials if their primary care physician recommends it [22]. Decreased funding has led to a substantial decrease in the number of new clinical trials and increase in the number of terminated ones [23].
Reasons such as increase in trial cost and presence of constant budgets with price inflations have caused the decreased funding of interventional clinical trials [23]. These factors can help in decreasing design and recruitment duration, prevent termination of trials and generate more viable results with larger sample sizes.
Of the 138 trials included in our study, 57 publications attributed to 39 clinical trials have been produced. This amounts to a 28% publishing rate from clinical trials involving gastrectomy for GC. Several reasons may explain this low figure. Since significant time and efforts are put into designing, setting up, conducting and analysing results of clinical trials, it may be thought that taking the decision of not publishing one made by the investigator(s) or sponsor of study. Such decisions might be taken due to discrepancies in desired versus observed results [24]. Another factor would be the decision of non-publishing bias of negative results, a phenomena that has already made its way into many clinical trials [24]. However, publishing of negative results may help other researchers in focusing future research efforts by informing others of the potential difficulties and obstacles faced during any respective trial. Publishing of said negative results could help surgeons in avoiding repetition of failing gastrectomy/gastrectomy related plans and thus divert efforts to other possible new interventions.

Treated topics
The most commonly treated topics in our evaluated studies were on nonspecific GC, locally advanced gastric cancer (LAGC) and early-stage GC. This is because patients with these tumours are free from any contraindications of gastric surgery. The commonly used procedures are shown in Figure 3. The most common contraindications for gastric resection include patients being unfit for general anaesthesia and those who have extremely poor prognosis (distant metastasis) [25]. As such, clinical trials involving gastrectomy can be for these malignancies. The most commonly studied primary outcomes involved overall survival and 3-year disease-free survival (DFS) in published works from the clinical trials. Studies showed for the most commonly treated topics that different surgical approaches show no significant difference in terms of long-term survival but do help in immediate postoperative course [26,27]. Current research lines are mainly focused on procedures including laparoscopy, HIPEC and robotic-assisted operations. This is because of the trend of utilising more and more minimally invasive procedures such as laparoscopy and robotic techniques for their better post-operative outcomes as compared to open approaches [28]. The increased use of HIPEC has also been reported in other cancers such as ovarian cancer, explaining the trend for increased trials in GC [29]. More clinical trials and publications must be produced on the common treated topics of GC (early stage and local advancement with no metastasis) to investigate ways which improve long-term outcomes. Better publishing of the current research lines and techniques being utilised is important to know their clinical outcomes.

Limitations
The focus of this study is primarily on trials involving gastrectomy for patients with GC only. The strength of the results generated depends on the accuracy of the information from the source database ('ClinicalTrials.gov'). Our study utilises the ClinicalTrials.gov database alone. The use of other databases such as the WHO International Clinical Trials Registry Platform or Cochrane database for clinical trials could have generated possibly more clinical trials. Inaccuracies might have been present in the trial data such as whether the data is up-to-date or not. Other issues could have been that data might be missing altogether from the registry.

Conclusion
Clinical trials involving gastrectomy for GC have a small number, minimal publishing rate and lack of geographic variety. Laparoscopy is the most common intervention being studied in clinical trials involving gastrectomy. Increased research efforts, funding and proper management are needed to improve and expand clinical trials, which in turn will improve patient outcomes.

Funding
The authors declare that this study did not receive any funding.

Conflicts of interest
The authors declare no conflict of interest.