Endoscopic resection modalities such as endoscopic mucosal resection (EMR) and endoscopic submucosal dissection (ESD) are standard procedures for treating noninvasive mucosal tumors and adenomas, which are precursors to cancer in the gastrointestinal (GI) tract.¹–⁴ To increase their efficacy and safety, EMR and ESD techniques require the injection of a fluid agent underneath the mucosa into the submucosal layer.⁵ Submucosal injection solutions are used to separate the lesion from the muscularis propria to allow the complete resection of the lesion and to prevent perforation and thermal injury to the GI wall.⁶ Normal saline (NS) solution is most commonly used in clinical practice because of its low cost and ease of use. However, its use is hampered by the quick absorption of the NS solution into the surrounding tissue, thereby resulting in the need for repeated injections.⁷ In recent years, various submucosal injection solutions such as hyaluronic acid (HA), glycerol, dextrose water (DW), fibrinogen mixture (FM), and hydroxypropyl methyl-cellulose (HPMC) have been developed and studied for safety and efficacy in EMR and ESD. In addition, there have been encouraging results with using autologous blood as a submucosal injection solution. Table 1 summarizes the characteristics of the various submucosal injection solutions. The appropriate selection of sub-mucosal injection solutions is important for successful EMR and ESD. This review addresses recent advances regarding the development of efficient submucosal injection solutions for use during EMR or ESD. The characteristics of various submucosal injection solutions are also described.

Ideal submucosal injection solutions should provide a long-lasting submucosal cushion and should be safe, inexpensive, readily available, nontoxic, easy to inject, and avoid histopathological tissue damage. They should also provide a sufficiently high submucosal elevation to facilitate an en bloc resection and reduce the risk of perforation. En bloc resections can reduce the risk of local recurrence after the endoscopic resection and can provide more accurate histopathological assessments.^1,8

Normal saline solution

Much research has been devoted to determining the ideal sub-mucosal injection for endoscopic resection.⁷ Except for NS solution, most solutions are difficult to administer or prepare, are associated with toxicity, or are expensive. Some solutions have limited human data on their safety and efficacy. Therefore, NS solution is widely used for EMR in most institutions. However, the greatest problem with NS solution is that it rapidly dissipates into the surrounding tissue. As a result, NS solution requires repeated injections, which makes it difficult to produce an adequate submucosal fluid cushion and maintain a sufficient height, especially for flat elevated lesions or large lesions. Normal saline solution appears to be adequate for EMR of non-flat elevated lesions and small lesions.

Hypertonic saline solution

Hypertonic saline solution is an inexpensive, readily available, easy-to-inject NS-like solution that creates a higher mucosal elevation than NS solution.⁵ However, tissue damage and local inflammatory reactions have been reported at the injection site of hypertonic sodium chloride (3.75%).⁹

Hyaluronic acid

Hyaluronic acid (HA) is a type of glycosaminoglycan found in connective tissue.⁶ The current indications for HA in clinical practice in many countries are for intra-articular injections for osteoarthritis and for use in eye surgery.¹⁰ In 2007 in Japan, the Japanese National Health Insurance approved HA as an injection solution for EMR.¹⁰

The efficacy of HA in mucosal elevation has been previously reported in an animal model.^5,10,11 Yoshida et al¹⁰ showed that mucosal elevation was greater with HA than with NS in resected and living animal models. Hyun et al¹¹ demonstrated that the mucosal elevation induced by 0.1% HA lasted longer than the elevation by NS or by mannitol in a resected mongrel colon. The efficacy of HA in EMR and ESD has also been reported in clinical practice. Several studies show that HA provides the longest lasting fluid cushion.¹²–¹⁴ Using a 0.4% HA solution enabled sufficient lifting of a colorectal and gastric intramucosal lesion during endoscopic resection, thereby reducing the need for additional injections and the risk of perforation.^15,16 Using HA moreover resulted in a greater number of successful en bloc resections and lower perforation complication rates, especially for colorectal ESD.^14,17,18

We have recently conducted a randomized, open-label, controlled, multicenter safety and efficacy trial of HA (Endo-Ease Unimed Pharm Inc., Seoul, South Korea). It was used as a submucosal injection solution in the endoscopic resection of gastric neoplasms. Similar to the aforementioned studies, we were able to confirm that HA is superior to NS solution with regard to its higher usefulness rate [33/37 (HA) vs. 21/36 (NA); P = 0.0027], higher subjective convenience scoring by the endoscopist (72.97% vs. 17.14%; P < 0.0001), lower percentage of patients needing an injection of 30 mL or more (37.84% vs. 63.89%; P = 0.0416), and a lower percentage of patients needing 3 or more injections (64.87% vs. 88.89%); there was no significant difference between HA and NA in the complication rate (10.81% vs. 5.71%; P = 0.6745) or serious adverse events (3.53% vs. 0%; P = 0.0951; unpublished data).

Hyaluronic acid is the one of the most ideal agents for submucosal injections, although it has several shortcomings. First, HA is expensive and has limited availability. Second, HA is very viscous and must be diluted to facilitate injection. Viscosity because of high concentrations of HA may make snaring difficult.¹⁰ Third, HA used for EMR and ESD may stimulate the growth of residual tumor cells. Matsui et al¹⁹ demonstrated that HA enhances tumor growth and CD44 expression of cancer cells at the wound sites. Therefore, HA should not be used for endoscopic piecemeal resection procedures because it increases the risk of residual tumors. Hyaluronic acid appears to be an adequate agent for ESD [particularly for the curative treatment of early gastric cancer (EGC) or early colorectal cancer (ECC)] because of its higher en bloc resection rate, even for large lesions.

Fujishiro and colleagues²⁰ evaluated the possibility of developing an appropriate low-cost HA solution by diversifying its molecular weight and mixing it with various solutions such as glycerol or DW. They report that a mixture of high-molecular-weight HA and glycerol resulted in excellent outcomes for ESD of gastrointestinal tumors.²¹ Glycerol (which contains glycerin and fructose) was the better mixing solution for HA, compared to pure NS solution, because glycerin produces hypertonic potency over extracellular fluid without tissue damage and fructose increases the viscoelasticity of the HA solution because of its cross-linking of HA molecules.²¹

Glycerol

Glycerol (glycerol, Chugai Pharmaceutical Co., Tokyo, Japan) is a hypertonic solution consisting of 10% glycerin and 5% fructose in a NS solution.⁶ It has been widely used as an osmotic agent against cerebral edema and has no toxic systemic effects.²² In 1995, glycerol was first introduced as a submucosal injection solution in a report of 24 patients with EGC who were treated by the endoscopic aspiration mucosectomy method.²³ Since then, a few studies have evaluated its effect as a submucosal injection solution. Uraoka et al²⁴ compared the efficacy of using glycerol or using NS as the injecting solution for EMR of colorectal laterally spreading tumors (LSTs). They showed that the en bloc resection rate and the complete resection rate were significantly higher in the glycerol group than in the NS group, whereas the rates of associated complications such as perforation and bleeding were similar in both groups. According to the Uraoka study, glycerol seems to safely increase en bloc resection rates of colorectal LST. Another study demonstrates that glycerol does not induce tissue damage.⁹ Moreover, glycerol is relatively cheap and readily available. However, glycerol produces smoke when electrocautery is performed and so it interferes with visualization during the procedure.

Dextrose water

Dextrose water (DW) is a hypertonic solution that produces and maintains a higher and more prolonged mucosal elevation, compared to NS solution.^25,26 Dextrose water is furthermore an inexpensive and easily available solution. Varadarajulu et al²⁵ showed that 50% DW is superior to NS as a submucosal fluid cushion because it allows a better en bloc resection during injection-associated polypectomy. Katsinelos et al²⁷ report that EMR with submucosal injections of hypertonic 50% DW plus epinephrine solution seems to be a safe and effective treatment for large sessile colorectal polyps. Katsinelos subsequently compared 50% DW plus epinephrine versus NS plus epinephrine during EMRs of sessile rectosigmoid polyps (>10 mm) and demonstrated that 50% DW plus epinephrine was superior to NS plus epinephrine for EMR, particularly for large and giant sessile polyps.²⁶ However, they also showed that the risk of thermal tissue injury (e.g., postpolypectomy syndrome) was significantly higher in the 50% DW plus epinephrine group than in the NS plus epinephrine group.²⁶ Furthermore, in a comparison study using mini-pig stomachs and injection solutions consisting of just NS, DW (at five different concentrations: 5%, 10%, 15%, 30%, and 40%), glycerol, or HA, researchers have observed considerable tissue damage with DW at concentrations of ≥20%.⁹ An accurate histopathological assessment of the margin and depth of resected EMR or ESD specimens is important for determining the appropriate therapeutic strategy, especially for EGC or ECC. The disadvantage of DW is that it may cause histopathological tissue damage and delay ulcer healing and may cause postpolypectomy syndrome. Therefore, DW at concentrations of ≥20% is not recommended as a submucosal injection solution.

Fibrinogen mixture

In 2004, Lee et al²⁸ investigated the clinical outcome of EMR with a submucosal injection of fibrinogen mixture (FM). Lee et al²⁸ confirmed two significant characteristics of FM: (1) its viscosity produces a long-lasting submucosal elevation that allows EMR to be performed more easily without the need of additional injections, and (2) it has a microvascular hemostatic effect, thereby keeping the visual field clear. The authors subsequently conducted a prospective randomized study comparing the efficacy of FM with that of the NS solution.²⁹ A total of 72 patients with EGC were randomly assigned to receive EMR with submucosal injections of either NS or FM. The authors reported that the mean procedure time, the mean submucosal injection volume, and the need for additional submucosal injections were lower in the FM group than in the NS group, although no significant differences were observed between the two groups in the rates of en bloc resection, complete resection, or recurrence.²⁹ For easier and safer EMR, FM may be a suitable submucosal injection solution and may be a good alternative to HA because of the advantages of FM such as its shorter procedure time, a clearer visual field, fewer required injections, and acceptable cost [US $0.20/mL in the United States (approximately ₩214.47/mL in Korea)]. However, FM has the risk of transmitting hepatitis or other viruses: fibrinogen is produced by the fractionalization of coagulation proteins in human serum, which may be contaminated with these viruses.⁶

Hydroxypropyl methylcellulose

Hydroxypropyl methylcellulose (HPMC), a cellulose derivative with viscoelastic properties, is primarily used by ophthalmologists in Western countries for creating artificial tears.^6,30 Feitoza et al³¹ conducted a preliminary study regarding the efficacy of HPMC. They demonstrated that HPMC creates a long-lasting submucosal fluid cushion with minimal tissue reaction.³¹ Feitoza in a subsequent study successfully performed a novel EMR technique by using HPMC for removing large areas of the mucosa in a porcine model.³² A Korean study reveals that the mucosal elevation lasted longer with 0.1% HA, 0.3% HPMC, and 2% FM than with 0.9% NaCl or 20% mannitol, and that the duration of the mucosal elevation was correlated with viscosity, but not with osmolarity.¹¹ Lenz et al³³ in a recent study showed that in the porcine stomach, the length of the HPMC submucosal fluid cushion is longer than what can be achieved with NS solution.

Hydroxypropyl methylcellulose and HA are very viscous and must be diluted to facilitate injection.⁷ Hydroxypropyl methylcellulose is less expensive than HA [0.83% HPMC as a generic product costs US $0.15/mL in the United States (approximately ₩160.85/mL in Korea)].⁶ However, HPMC is a synthetic product that may give rise to antigenic reactions, whereas HA exists in the connective tissue of mammals and is not antigenic.⁶ Furthermore HPMC toxicity testing in animal models is required for clinical applications in patients.

Autologous blood

Several studies have recently reported encouraging results when autologous blood is used for submucosal injection.³⁴–³⁶ Giday et al³⁴ evaluated six solutions as cushioning agents in live pigs: 5 mL of NS; NS plus epinephrine; albumin 12.5%; albumin 25%; HPMC; and the pig’s own whole blood. The injection of blood resulted in significantly longer mucosal elevations than any other solution. The authors showed that the cushion created with autologous blood does not interfere with visualization during EMR and lasts up to seven times longer than with NS solution. Shastri et al³⁵ also showed that the mucosal elevation persists significantly longer after injecting autologous blood, compared to NS solution, in a resected porcine stomach and autologous blood generated adequate mucosal elevation for the resection of high-quality specimens. Another study investigated the efficacy of different blood components (serum, plasma, and whole blood) in comparison with NS, HA, glycerol, and hydroxyethyl starch for use as a submucosal fluid cushion.³⁷ The study demonstrated that whole blood generates a mucosal elevation that lasts longer than that by all other agents. A Japanese study successfully performed a novel EMR method assisted by submucosal injection of autologous blood in 28 patients with 35 colorectal polyps.³⁶ According to the study’s findings, the injection of autologous blood does not result in complications such as bleeding or perforation and does not negatively affect pathologic examinations.³⁶

Submucosal injection of autologous blood offers several advantages because of its specific properties: its corpuscular components ensure prolonged submucosal elevation and its procoagulatory constituents promote local hemostasis and prevent postinterventional bleeding. Autologous blood appears to be an ideal submucosal injection solution that provides a long-lasting submucosal cushion, and it is safe, economic, readily available, and easy to inject. Autologous blood may be a promising medium for submucosal injection during EMR and ESD. However, procoagulatory constituents in blood can cause clotting in the syringe if an injection is delayed. In addition, human data are lacking regarding autologous blood as a submucosal injection solution. More in vivo studies in humans are necessary to substantiate its use for long-lasting EMR or ESD.

Other solutions

Besides the aforementioned solutions, various submucosal injection solutions have been developed and studied for EMR or ESD. These solutions include carboxymethylcellulose,^33,38 photo-crosslinkable chitosan hydrogel,³⁹–⁴² polyethylene glycol maleate citrate-based injectable drug-eluting elastomeric polymer,⁴³ and sodium alginate.⁴⁴–⁴⁶

The addition of a staining dye (e.g., 0.004% indigo carmine or methylene blue) to the injection solution can assist in identifying the area of submucosal injection and clearly distinguish between the muscle layer and the submucosal layer.^6,47

The adequate selection of a submucosal injection solution is required for successful EMR and ESD. Hydroxypropyl methylcellulose is a very effective submucosal injection solution, but its use is limited by its high cost. Because of the cost-effectiveness, a mixture of HA and glycerol appears to be a good candidate for use as a submucosal injection solution. Autologous blood has recently shown promising preliminary results, but human data are lacking. Further research is needed on the longterm outcomes of submucosal injection solutions to enhance the safety and efficacy of EMR and ESD. In addition, the search should be continued for an ideal injection solution for endoscopic procedures.