Laparoscopic Partial Splenectomy: A Critical Appraisal of an Emerging Technique. A Review of the First 457 Published Cases

Introduction

Since the late 1960s, partial splenectomy (PS) has been introduced^1–5 in selected patients to avoid the nonnegligible drawbacks of total splenectomy (TS), including 3.2% rate of severe infections, ⁶ overwhelming postsplenectomy sepsis, and 0.2%–1.4% mortality.^6–8 Since, PS has been proposed for various conditions whenever splenectomy was nonmandatory for oncological or hemostatic reasons, including hematological disease,^2,3,9–11 portal hypertension, ¹² trauma,^13–15 benign,^16–18 and malignant^19,20 focal lesions.

Although there is the obvious advantage of spleen partial preservation, PS has only slowly spread through the 1990s, and it is still not recognized as an option by most general surgeons.^1,21–26 Several factors might have played a role in such a slow spreading of this procedure, including the rarity of conditions suitable of PS, the suboptimal effectiveness of surgical technique/tools used for splenic parenchyma transection/hemostasis, and other technical issues of surgery, including the need of a large laparotomy to obtain an optimal vision of the operating field.

The introduction of laparoscopy and the development of new devices allowing for enhanced visibility and effective splenic parenchyma transection/hemostasis has had a pivotal role in the diffusion of PS during the 1990s and 2000s, as witnessed by the increasing number of performed procedures and published articles on the subject ²⁷ First introduced by Poulin in 1995, who reported the laparoscopic resection of spleen superior pole for trauma after preoperative embolization. ¹⁵ Since the early 2000's, laparoscopic partial splenectomy (LPS) is widely spreading and extending its indications.^28–30 In 2010, the first robotic PSs were reported by Vasilescu et al. ³¹

Present literature about LPS procedures is extremely heterogeneous, considering patients, indications, surgical technique, and used devices/tools. Unfortunately, the scientific quality of articles reporting LPS, mostly consisting of case reports and small retrospective case series, is low, with IIIb to V grade of evidence. With the primary aim to allow for a first assessment of technique's feasibility and safety and to give the reader any possible useful information of a spreading but heterogeneously performed technique, in this study, we propose a comprehensive review of the 457 published cases. Secondary aim of the article is to report the comparison between LPS and alternative techniques for splenic resection, namely laparoscopic TS and open PS.

Materials and Methods

Ethics approval

As a narrative review, according to Institutional Review Board approval guidelines, ethics approval is not required.

An extensive bibliographic research of literature according to PRISMA criteria was performed (Fig. 1). Medline, Embase, and PubMed were consulted to identify articles reporting the following items: hemisplenectomy, hemi-splenectomy, polar splenectomy, PS, subtotal splenectomy, and near TS up to December 2018. Those items were then meshed with the Boolean operator “AND” and “OR” with the following mesh terms: laparoscopic, minimally invasive, video-laparoscopic, and mini-invasive. Additional articles were searched by manual identification from the key articles.

OPEN IN VIEWER

FIG. 1.

Prisma 2009 flow diagram. ¹⁰⁶

Results

All gathered information about 457 patients undergoing LPS is reported in Table 1.

Technique

Details regarding the type of performed resection, spleen devascularization techniques, and used devices for parenchymal section/hemostasis are reported in Table 2.

Placement

A variable number of trocars is reported to perform LPS. Four trocars are reported in 170 procedures,^31,33–51 3 in 53,^7,52–56 and 5 in 47.^{15,41,42,57–61} Two-port-LPS (4 patients^44,62), single port (2 cases^63,64), and hand-assisted LPS (1 case ⁶⁵ ) are also reported.

Two types of settings are mostly reported: the so-called French position and American position (Fig. 2). According to the French position, the patient is placed in supine position (0° lateral, 20° reversed Trendelemburg position) and the leading surgeon is placed between patient's legs.^{44,56,59,66–68} According to the American position, the patient is placed in a left side-up position (30°–70° right-lateral semidecubitus position, 20° reversed Trendelemburg position) ⁶⁶ with the operating surgeon on his right side^{7,43,62,63,66,69–72}

OPEN IN VIEWER

FIG. 2.

Trocars' position. (A) Left-subcostal trocar placement in “French” position (0° lateral, 20° reversed Trendelemburg position) ⁶⁶ ; (B) Semicircular trocar placement in “American” position (30°–70° right-lateral semidecubitus position, 20° reversed Trendelemburg position)^43,71; (a) optic trocar (assistant surgeon); (b) trocar for operating instruments; (c) trocar for operating instruments; (d): trocar for assistant surgeon retractor; (*) Trocar for the assistant surgeon may be placed at the xiphoid ⁴³ or in the left flank. ⁴³

Two types of trocar placement are mainly reported: the “left-subcostal” one and “semicircular” one (Fig. 2).

Left-subcostal trocar placement ⁶⁶ is reported in Figure 2A. A periumbilical optic trocar is placed, and other ports can be inserted in the left upper quadrant or in the epigastric region. Trocars' number varies according to surgeon preference (see below), including an additional trocar placed in right hypochondrium if cholecystectomy is performed contextually. ⁷³

Semicircular trocar placement^43,71 is reported in Figure 2B. The optic trocar is placed in the umbilical or periumbilical region, while remaining ports can be placed on the median line in the upper umbilical and epigastric region, for spleen elevation and retraction, or along the left rib cage. Some authors report in their series procedures performed with both French and American positions. ⁶⁶

Dissection, resection, and hemostasis

After trocar placement, spleno-gastro-colic ligament is severed to expose the splenic hilum. Two approaches regarding vessels' dissection/devascularization are mainly reported: (i). division of splenic artery and vein^{42,48,70,74–77} and (ii) division of splenic vessels' collateral branches at the hilum of the spleen.^{7,15,33,34,36–38,40,41,43–47,49,50,53–56,58–62,65–68,71–73,78–88}

To limit bleeding during splenic parenchyma transection, some authors propose the preoperative embolization of the spleen,^15,44 or the intraoperative temporary clamping, both by clamping of main splenic vessels,^{7,44,62,66,86} or by compression of splenic parenchyma at the resection margin, by Lin's clamps. ⁷

In most cases, splenic resection consists in a polar resection, either the upper or the lower pole. The upper pole is resected in 39 patients (8.5%),^{15,46,48,50,54,58–60,62,65,66,68,72,74,78–80,82,85,88–90} whereas lower pole resection is performed in 81 cases (17.7%).^{34,40–43,49,52,53,55,56,73,75–77,81} In 337 cases (73.7%), the information concerning the resected pole is not reported or not possible to retrieve. Upper spleen pole resection is never performed for hematological disease, while 69/81 cases undergoing lower spleen pole resection are affected by hematological disease.

As reported in Table 1, various devices, including mechanical stapler, harmonic scalpel, radiofrequency, unipolar/bipolar electrocautery, and argon are used for the dissection of splenic vessel and the transection/hemostasis of splenic parenchyma.^{7,15,33,34,36–38,40–43,45–50,52–56,58–61,66,67,70–74,76–86,88,89} Mechanical stapler is the most frequently used device for parenchymal transection (Table 1). Some authors report the successful accomplishment of spleen section by means of unipolar and bipolar electrocautery only.^15,49,55,82

Several articles report the use of various hemostatic means/agents with the purpose of improving the hemostasis of splenic raw surface after section, including fibrin glue, fibrillar hemostatic products, and so on.^{7,15,33,34,36–38,40–45,47–50,52–54,56,58–62,65–68,70–74,76–89,91}

Discussion

LPS is seemingly feasible and safe, as it is reported to be associated with no mortality, low morbidity, and low rate of conversion of the access from laparoscopy to laparotomy as well as low rate of conversion from partial to TS (Tables 2–4).

Literature on LPS is extremely heterogeneous considering patients, indications, and surgical procedure, which is reportedly accomplished with evolving techniques and tools through last quarter of a century. Such a heterogeneous variety of indications for LPS^{7,15,36,39,41,47,54,67} eventually led to different approaches to LPS, since the purpose of surgery changes with the affection needing spleen resection. LPS, in fact, consists in a massive, 85%–95% spleen resection^70,95,96 aimed to decrease erythrophagocytosis and symptoms such as fatigue and anemia of patients affected by hematological disease; on the contrary, LPS is a limited, parenchyma-preserving resection of focal lesions, ³⁶ as well as in the case of splattered spleen, where tissue debriding is aimed to achieve the hemostasis after traumatic injury. ⁴⁷ Moreover, the indication for LPS intrinsically affects patient outcome too, as it is in the case of LPS performed for trauma, where results of surgery are related to multivisceral trauma other than LPS.^47,66 Hematological disease, also present peculiar outcome, as it is mostly performed in children with a sistemic disease undergoing surgery with all the technical implications of performing a massive resection of a large spleen in a small-sized body, and where a poor perioperative outcome may be also due to a nonvital, small splenic remnant or an early recurrence of hematological disease.

Although the intrinsic bias is associated with such a heterogeneous literature, we finally included most articles describing LPS, regardless of surgical indication, patients' population size, or article intrinsic scientific quality, to give the reader all the possible information regarding this procedure.

Several technical details of LPS procedure are worth a commentary. Similarly to the approach described for other upper-abdomen, one-side-directed procedures (cholecistectomy, surrenalectomy), two main options for surgeons' placement are reported, the French,^{44,56,59,66–68} and the American position,^{7,43,62,63,66,69–72,86} according to surgeons' background and preference. A variable number of trocars is reported to carry out LPS, ranging from one to five, four being the preferred option overall.^{44,56,59,66–68} Also trocars' position varied consistently through the series, being the “semicircular” and “lef-subcostal” the most popular (Fig. 2).

Similarly to the open PS procedure, two main approaches to spleen devascularization, aimed to identify the transection line (the line of demarcation between the vascularized parenchyma and the devascularized remainder) and to reduce bleeding, are reported. The main splenic vessels ligation^{42,48,70,74–77} relies on spleen blood supply from gastric short vessels/phrenic collaterals or left gastroepiploic artery, depending on the splenic pole, which is intended to be resected, while the selective segmental splenic devascularization implies the collateral branches division at the hilum.^{7,15,33,34,36–38,40,41,43–47,49,50,53–56,58–62,65–68,71–73,78–88} The first approach is the laparoscopic version of the original procedure described by Morgenstern in 1966 by ligating the splenic artery proximal to the origin of the superior polar artery, ¹ thus preserving the blood supply to the superior pole by the gastric vasa brevia and the phrenic collaterals. Although few authors still prefere such an approach,^{34,40–43,49,52,53,55,56,73,75–77,81} the segmental vessels' dissection and division at the hilum has become the most diffused approach by laparoscopy,^{7,15,33,34,36–38,40,41,43–46,49,50,53–56,58–62,65–68,71–73,78–88} as it is by open approach. To reduce bleeding when transecting the spleen, splenic vessels' temporary clamping is reported.^{7,44,62,66,86} Alternatively, preoperative arterial embolization^15,44 and spleen parenchyma compression by clamps ⁷ have been reported for a safer parenchymal section. Although many biases may definitely hamper our analysis' results, less AOs are associated with the periferal vessel dissection (Table 2) and may suggest this latter as the technique to be preferred.

Although in the majority of cases it was not possible to identify which portion of the spleen was resected, the lower pole of the spleen resulted as more often resected^{34,40–43,49,52,53,55,56,73,75–77,81} than the upper one.^{15,46,48,50,54,58–60,62,65,66,68,72,74,78–80,82,85,88–90} Since focal masses may be thought to harbor (and traumatic lesions to occurr) randomly in splenic parenchyma, thus leading to the resection of the affected pole with the same frequency, such a trend toward a more frequent resection of the inferior pole of the spleen is due to surgeon's preference when performing LPS for hematological disease. This is in fact the only situation where the operating surgeon actually has the choice between resecting the upper or the lower pole, as the purpose of surgery in this setting is just to reduce spleen size. The reason why lower pole resection is preferred is probably that the lower pole is nearer to laparoscope and instruments, thus it is considered easier to reach. In the case of selective segmental splenic devascularization, the mobilization of colonic splenic flexure and the dissection of the lower pole is seemingly perceived as an easier task than dissecting the upper pole of the spleen from gastric fundus and diaphragm. When the main vessels' dissection is preferred, the collateral vascularization by gastric short vessels is probably considered as more effective than gastroepiploic vascular anastomosis in preventing splenic ischaemia/infarction. In our opinion, selection bias plays a major role in the reported higher rate of AOs after lower pole resection, as it is frequently performed for hematological disease, where reoperations and completation TS are more often needed and follow-up is usually longer.

Although retrospective nature, heterogeneity of procedures, and small numbers do not allow for any statistical analysis, the present series may allow for an overview of means of parenchymal section and hemostasis. Since 1995, a variety of techniques/devices has been proposed, including mechanical staplers,^{7,38,41,42,47,48,53,56,61,73,77,80,83,84,86,87,91} various types of electrocoagulation,^{15,33,36–38,43,44,47–49,53–56,59–62,65,68,70–73,78,79,81–83,85–87} and ultrasonic dissectors.^{34,36–38,41–43,45–47,50,56,58–62,66–68,71–73,78,80,81,84–87,89}

Since the mid-2000's, new energies' cutting/coagulating devices (harmonic scalpel),^{34,36–38,41–43,45–47,50,56,58–63,66–68,71–73,78,80,81,84–89,97–99} radiofrequency,^{7,33,40,44–46,53,59,71,72,74,76,79,86,90,100} and hemostatic agents (TachoComb^®70 Tachosil^®52 and FloSeal^®)^62,80,101 have been proposed for safer LPS. In general, during the studied period, the choice of section/hemostasis tools seems more to follow technology progress, surgeon's curiosity, and instrument availability in a specific environment than proven effectiveness.

Although LPS seems a safe procedure with no mortality, low morbidity/blood loss, and a 4.9-day-mean HS, nevertheless, major complications are also reported, including life-threatening situations, which may be challanging to manage laparoscopically, including three, namely colonic perforation, ⁵⁶ massive bleeding, ⁹² and diaphragm perforation, ³⁷ eventually needing emergency laparotomy. Although the rates of conversion to open surgery and to TS are low overall, being 3.9% (18 cases)^{7,37,41,67,77,86,92,94} and 3.7% (17),^{36,37,42,47,56,70,71,77,83,85,86} respectively, the risk of potentially life-threatening IAOs should not be underestimated. Our analysis of the oucomes of LPS in our series match the findings of a recent systematic review, ¹⁰² including 252 patients from 44 articles.

Considering TS, which may be considered one of the main concerns when performing LPS, it is worthwhile to point out that conversion from LPS to TS is reported to occur intraoperatively in only 5 cases^36,37,47,85 (in 1 case owing to intraoperatively suspected malignancy ³⁷ ), and in 5 cases perioperatively.^36,47,83,85 In fact, the most frequently reported reason for completion TS is LPS's failure in managing symptoms (anemia, etc.) of hematologic disease or the early recurrence of this latter (Table 4).

Considering PS mortality, although the only three reported fatalities actually occurred after PS performed by open approach,^103–105 those outcomes are worth being known by laparoscopic surgeons. Considering a 27-year-old patient dying from streptococcus meningitis 13 years after 2/5 PS for trauma, ¹⁰⁵ interestingly, the patient underwent technetium liver-spleen scan 3 months, and 2, 3, and 5 years postoperatively, showing a normal pattern and did not receive pneumococcal vaccination at any time. ¹⁰⁵ The record reported patient eventually died for fecal peritonitis by accidental colonic perforation. ¹⁰⁴ The last one ¹⁰³ was a 17-month-child with hematologic disease eventually dying of sepsis 8 months after PS was postoperatively converted to TS. Although this latter complication seems more an argument in favor than against PS, as the purpose of LPS is to avoid TS, these data should not be forgotten when planning LPS, as, in roughly 1/25 cases, LPS will end up in TS. The first reported fatality seems to suggest that pneumococcal vaccination is indicated even after successful PS with evidence of vital spleen remnant at long-term postoperative follow-up.

A comparison between the results of LPS and alternative mini-invasive surgical options, namely laparoscopic TS and open PS, is attempted by some groups,^{38,41,47,84,87} with variable results. Compared to laparoscopic TS, it is not clear whether longer OT or greater intraoperative bleeding are associated with LPS.^38,41,47 Results concerning postoperative course are also conflicting, as LPS' HS is reported as being shorter ³⁸ or longer ⁴¹ compared to laparoscopic TS. LPS and laparoscopic TS are reported to have similar morbidity overall,^38,41,47 but, significantly, no pleural effusion nor splenic vein thrombosis are reported after LPS (versus 40.9% and 45.5%, respectively, in the laparoscopic TS group, P < .01 in both cases). Since significantly lower mean platelets and leukocyte count are reported 6 months after LPS, it may be argued that LPS may be associated with lesser thrombocytosis and its related complications.

Results of the comparison between LPS and open PS are also heterogeneous. In detail, they are conflicting regarding intraoperative blood loss,^84,87 while, not surprisingly, LPS is associated with longer OT and shorter HS. ⁸⁴ Although postoperative morbidity is similar,^84,87 interestingly, most symptoms' recurrences and all completion TS are performed in patients undergoing open PS. ⁸⁷ Such results, at times confirming some of recognized advantages of mini-invasive approach, definitely need validation in larger series, since retrospective nature, patients' selection bias and small size of studied populations limit the interest of such analyses.

Analyzing 59 articles reporting such heterogeneous population of 457 patients with various affections undergoing various procedures with low rates of complications and AOs prevented us from an analysis of LPS results and effectiveness.

The main drawback of the present series is the heterogeneity and retrospective nature of small-sized series/case reports through a quarter of a century. In such a series, surgical complications are probably underestimated, as they are not always reported by articles focused on hematology or trauma issues, or not collected by short follow-up. Since successful surgery is more likely to be retrospectively reported than bad outcome, good results of LPS may be overestimated. Nevertheless, such a review may allow surgeons to acknowledge LPS indications and techniques and to suggest how to perform it safely with the means potentially available in their environments.

Conclusion

LPS seems feasible and safe, with no mortality, low morbidity, and low rate of conversion to laparotomy and to TS. Present literature on the subject is extremely heterogeneous considering patients, indications and surgical resection, which may be accomplished with various techniques and tools. Major surgical complications are sporadically reported, thus potential risks of LPS should not be underestimated.

Availability of Data and Material

The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.