As evidenced by a growing number of publications during the past 15 years, laparoscopic splenectomy (LS) has supplanted open splenectomy for the treatment of immune thrombocytopenic purpura (ITP). The transition in treatment was made to a minimal access approach relatively quickly, and it became clear from several large series that the procedure was safe, resulted in shortened hospital stay, and was effective in resolving thrombocytopenia [8, 12]. However as follow-up of ITP patients matured, several drawbacks to the laparoscopic procedure arose. The first concern is the incidence of missed accessory splenic tissue at the time of laparoscopy [5]. The other concern, which has been supported in an animal model [3], regards splenic injury or rupture at the time of LS, with resultant splenosis. Splenosis and missed accessory spleens are concerning because both have been postulated as causes for recurrent thrombocytopenia. Recommendations to avoid these two pitfalls include exploratory laparoscopy as an initial step to search for accessory spleens as well as exquisite care to avoid rupture and cell spillage [15]. In the case of recurrent ITP following LS, the only surgical option is reexploration and removal of accessory tissue, which has been performed laparoscopically and reported several times in the literature [2, 14, 16]. We describe the first published case of successful accessory splenectomy using a hand-assist device.

Case report

A 60-year-old female presented with recurrent thrombocytopenia secondary to ITP. At her initial presentation with ITP, she was treated medically and subsequently underwent splenectomy, which was performed at the age of 55. Her open operation was unremarkable and she had a recovery of her platelet counts to within normal limits. She presented 5 years later to her hematologist after noticing easy bruisability for several weeks, particularly on her legs. She was found to be thrombocytopenic and was given a course of steroids and intravenous gamma globulin. She had an initial response, with her platelet count rising to above 100,000/μl transiently and then dropping to less than 10,000/μl. A computed tomography (CT) scan of her abdomen (Fig. 1) and a liver/spleen scan were performed, which demonstrated the presence of four accessory spleens within the left upper quadrant. The patient was referred to the minimally invasive surgery service and underwent attempted laparoscopic accessory splenectomy.

Figure 1
figure 1

CT scans of the abdomen showing lesions in the left upper quadrant consistent with accessory splenic tissue (arrows).

Operative technique

On the morning of the operation, radiolabeled technetium was given intravenously to aid the intraoperative identification of splenic tissue. The patient was positioned supine under general anesthesia. Pneumoperitoneum was obtained with a Hasson trocar, and a 5-mm left lateral trocar was placed. After multiple adhesions were taken down, 12-mm ports were placed epigastrically and in the high left upper quadrant, with a 5-mm port laterally. Two nodules were found in a retrogastric location within the lesser sac. After dissection with an ultrasonic dissector, they were removed through a 12-mm port in an endoscopic bag. Despite an exhaustive search laparoscopically no additional nodules could be identified. The gamma camera could not localize the remaining splenic tissue seen on the preoperative imaging. A HandPort (Smith & Nephew Endoscopy, Andover, MA, USA) was placed, incorporating two trocar sites in the left upper quadrant, approximately 7 cm in length. Using this hand-assisted laparoscopy technique, three more nodules were palpable, and they were identified and removed in a similar fashion. Her postoperative course was uneventful, and she was discharged with a platelet count of 180,000/μl. Follow-up at 1 month showed the platelet level to be 350,000/μl. All of the nodules removed were subsequently confirmed to be splenic tissue on histology.

Discussion

As demonstrated by a number of published series, LS for ITP has been shown to be safe and associated with a low morbidity and good short-term outcomes [8, 12]. However, as follow-up has matured, it has become apparent that, as for open splenectomy, a subset of patients will develop recurrent disease [11]. The immediate platelet response (platelets >150,000/μl) following LS in ITP is expected to occur within the first 36 h postoperatively. In a series of 67 patients with ITP, Katkhouda et al. [9], found an immediate platelet response rate in 83% of patients, but with an average of 38 months of follow-up, the response rate decreased to 77%. These results are similar to those of other series [11]. Patients with recurrent thrombocytopenia who undergo reexploration can expect a variable platelet response rate of 20–73% [1, 15].

Several methods are available to the surgeon to help preoperatively localize splenic remnants. Radiolabeled, denatured, RBC spleen scanning has been promoted as the first investigation for recurrent thrombocytopenia to identify residual splenic tissue [10]. If splenic tissue is identified, consideration is then given for surgical excision. This test has been able to successfully identify and localize accessory splenic tissue in several published cases [2, 7, 16]. Furthermore, intraoperative use of nuclear imaging with a gamma probe provides real-time localization and allows confirmation of removal of the labeled splenic tissue [2, 7]. A hand-held gamma probe was utilized intraoperatively in our case, although it failed to localize the remaining splenic tissue subsequently found after insertion of the hand-assist device. CT scanning, used in conjunction with nuclear imaging, can similarly identify nodules consistent with splenic tissue, although this is less sensitive for smaller foci of splenic tissue.

In the operating room, laparoscopy affords good visualization but is hindered by the lack of sensitive tactile feedback that would allow detection of otherwise hidden splenic nodules. Other techniques, such as intraoperative ultrasound, may provide some benefit for identification of splenic nodules during laparoscopy but require surgeon experience and equipment availability. If laparoscopic exploration fails to detect the accessory spleen (AS) identified on preoperative imaging, few alternatives are available for the surgeon aside from laparotomy.

The rate of identification of accessory spleens at the time of open splenectomy for ITP has ranged from 18 to 40%. A wide range of ASs are similarly identified at the time of LS, ranging from 0 to 30% [15]. The inference that laparoscopy is inferior to open surgery in the detection of ASs is one possible explanation. The other, as suggested by Friederici [4], is that the difference in detection is related to the diligence of the surgeon. Regardless, the higher rate of detection found in some open splenectomy series suggests that manual palpation at open splenectomy is more sensitive than the indirect “palpation” with instruments at laparoscopy.

Although the best method for treatment of recurrent ITP is controversial, the feasibility of laparoscopic reexploration is clear [2, 16]. However, the disadvantage is that the same failed diagnostic technique, laparoscopy, is being used to identify the AS. The use of a hand-assist device increases the sensitivity of laparoscopy by allowing palpation for the nodule by the surgeon’s hand. As demonstrated by this case, only two nodules were identified by exhaustive laparoscopy, with three more splenic nodules identified only after placement of the hand-assist device.

We believe that the small incision (6 or 7 cm) for the hand-assist device, although larger than the incisions for laparoscopic access, results in less pain and earlier return to activity following surgery. There is evidence that the benefits of laparoscopic surgery are similar to those for hand-assisted surgery in other surgical procedures, such as colectomy and nephrectomy [6, 17]. Cases in which hand-assisted surgery has been utilized for splenectomy also show equitable postoperative outcomes with respect to hospital stay and pain medication use [13].

Although it may be argued that a hand-assist device may be used from the onset of the procedure in cases of laproscopic accessory splenectomy, we believe that ASs that are identified on CT or by other imaging may be amendable to a purely laproscopic excision. In cases in which the AS is not easily located at laparoscopy, as in our case, conversion to a hand-assisted approach would be the most appropriate step.

Conclusion

We propose that in the case of recurrent ITP following LS, repeat laparoscopy is the first step after Ass are identified by nuclear imaging or CT scan. At laparoscopy, a dedicated exploration for ASs should be made. If the ASs are not identified at laparoscopy, the insertion of a hand-assist device incorporating a port site is an alternative to a full laparotomy.