Three‐dimensional simulation analysis of microdissection testicular sperm extraction for patients with non‐obstructive azoospermia

Abstract Background Microdissection testicular sperm extraction (microTESE) is considered the gold standard method of sperm retrieval from patients with non‐obstructive azoospermia (NOA). For careful and thorough examination of seminiferous tubules during microTESE, maximizing surface area of the testicles which we are able to search is essential. Objectives To develop a systematic procedure for microTESE to maximize surface area and to achieve high sperm retrieval rate (SRR) in microTESE. Materials and methods We simulated microTESE using three‐dimensional (3D) simulation model and analyzed mathematically the sum of the surface area in various methods. The best method obtained from this simulation model was applied to 102 patients with NOA from 2014 to 2018. These new clinical results were compared with those of 56 patients who underwent a previous method of microTESE from 2011 to 2014. Results The mathematical 3D simulation model of microTESE indicated that a longitudinal incision on the tunica albuginea and following transverse slicing incisions of testicular parenchyma maximized the surface area coverage. Forty‐six (45%) out of 102 patients who underwent microTESE with the new method had successful retrieval of testicular spermatozoa compared with 16 (29%) of 56 patients with the previous method of microTESE (P = .04). Discussion Transverse resections of parenchyma in our method run parallel to the courses of intratesticular arteries and do not interfere with the blood supply. The small amount of extracted seminiferous tubules was equivalent to that of the previous method, and no patients exhibited post‐operative symptoms of androgen deficiency in our study. As for post‐operative pain, our new method was comparable with the previous method. Although our study needs a longer follow‐up, there will be limited effects on testicular functions. Conclusion Longitudinal incision on the tunica albuginea and following transverse slicing incisions in the testicular parenchyma maximized the surface area and improved the SRR of microTESE.


| INTRODUC TI ON
MicroTESE is considered the gold standard method of sperm retrieval from patients with NOA. It is well known that the dilated, white seminiferous tubules are more likely to have intact spermatogenesis and that the use of surgical microscope to search suitable tubules benefits the successful retrieval of spermatozoa in NOA patients. 1 However, despite careful and thorough examination of seminiferous tubules during microTESE, there is still a possibility that we may overlook suitable tubules and come to a wrong conclusion. Several previous papers concerning repeat microTESE have reported the possibility of sperm retrieval in cases of previous failed surgery. [2][3][4][5] For the patients to accept the final outcomes, regardless of whether it is hopeful or not, we need to maximize the area of the testicles which we are able to search.
Generally, microscopic surgery is technically demanding and needs advanced training. Ishikawa et al reported that the surgical outcomes and SRR of microTESE performed by one surgeon will improve with experience and showed substantial learning curves for microTESE. 6 However, there is no standardized idea to search for suitable tubules during microTESE. Furthermore, there is controversy about the types of initial incision for microTESE, transverse or longitudinal, with supporters on both sides. 7 Thus, we need to develop a standardized, systematic procedure for microTESE to minimize the possibility of overlooking suitable tubules and sperm retrieval.
Here, we used 3D simulation of microTESE to develop the best method to maximize the surface area of cut planes and systematically check all the seminiferous tubules of testicular parenchyma.
Also, we applied this new method to recent NOA patients, to demonstrate its safety and improvement of surgical outcomes for the rate of successful sperm retrieval.

| 3D simulation of microTESE
We assumed that the testis had an ellipsoid shape with the lengths of each axis denoted as A, B, and C, with A being the longest and C the shortest (A > B>C). We placed this testis model into the X-Y-Z axis, where the longitudinal incision was parallel to the X-axis and the coronal and sagittal planes were equivalent to the X-Y and X-Z planes, respectively. After a long incision on the tunica albuginea, testicular parenchyma was divided into several thin-sliced pieces in order to fully search the testicular contents. The width of a sliced piece of testicular parenchyma was assumed to be l, and we assumed A = al, B = bl, and C = cl ( Figure 1). This testis model could be described as follows: Using this testis model, we simulated microTESE using various methods. Regarding the initial incision on the tunica albuginea, two types of incision are possible: longitudinal or transverse. The initial incision on the tunica albuginea would produce a bivalved testis, and the tunica albuginea could be pushed upward turning the testicular parenchyma inside out. Then, the parenchyma could be divided into several thin-sliced pieces. Several directions are possible for the cutting planes through the testicular parenchyma: transverse, sagittal, and coronal planes. We excluded the example of a longitu-

K E Y W O R D S
male infertility, microdissection, non-obstructive azoospermia, simulation, TESE, testis F I G U R E 1 Mathematical model of the testis, which has an ellipsoid shape with the lengths of each axis denoted as Al, Bl, and Cl, with Al being the longest and Cl the shortest (Al > Bl>Cl). This testis model was placed into the X-Y-Z axis with the longitudinal incision parallel to the X-axis and the coronal and sagittal planes corresponding to the X-Y and X-Z planes, respectively sagittal incisions in the parenchyma ( Figure 2C), and transverse incision on the tunica albuginea and coronal incisions in the parenchyma ( Figure 2D). We calculated the estimated surface areas for each method and compared them to find the best method for maximizing the cut surface.

| NOA patient selection
Diagnoses of NOA were confirmed with several semen analyses by centrifugation as well as reduced testicular volume and increased serum follicle-stimulating hormone level. Patients with any sign of obstructive azoospermia were excluded from this study. Patients with Y chromosome microdeletions of AZFa and AZFb were also excluded. All patients were counselled pre-operatively, and all gave written informed consent to perform the surgery.

| Application to new NOA patients
We performed microTESE under local anesthesia. The larger of two testes was delivered first through an incision in the median raphe.
Under the operating microscope, the tunica albuginea was incised, taking care to avoid equatorial testicular vessels. The testicular parenchyma was examined at a magnification of 10×-20× under the operating microscope. Each sample was taken from the dilated, white tubules and examined by one experienced embryologist in the operating room under a phase-contrast microscope at 200× magnification. If there were no spermatozoa identified, the contralateral testis was explored. The procedure was terminated when enough spermatozoa were obtained or all regions of both testes had been examined and tubules with the most suitable appearance had been excised. The number of biopsies examined ranged from 1 to 28 per testicle.
From 2011 to 2014, we used a previous method in which the tunica albuginea was incised transversely, and the parenchyma was opened and exposed as a bivalved testis. We searched initially on the cut surface and then in deeper regions of the testicular parenchyma but not systematically.
From 2014 to 2018, we adopted our new method of micro-TESE, which was developed using the findings obtained from the 3D simulation of microTESE. In this method, the tunica albuginea was incised longitudinally, and Mosquito clamps were placed on each respective side of the tunical incision. The testicle was incised and opened to form a bivalved testis, and the tunica albuginea was pushed upward with gentle pressure to turn the F I G U R E 2 Schematic illustrations of (A) a longitudinal incision on the tunica albuginea and transverse slicing incisions in the parenchyma, (B) a longitudinal incision on the tunica albuginea and sagittal incisions in the parenchyma, (C) a transverse incision on the tunica albuginea and sagittal incisions in the parenchyma, and (D) a transverse incision on the tunica albuginea and coronal incisions in the parenchyma. The width of the sliced piece of testicular parenchyma was assumed to be l parenchyma of the testicle inside out. Then, the parenchyma was divided transversely into several slices, the width of which was as thin as possible (2-3 mm).
All procedures were performed by one experienced surgeon (KI), who had performed more than 200 procedures of microTESE prior to 2011. Institutional review board approval (No.201336) was obtained for this study.

| Follow-up after microTESE
All patients were prescribed 10 tablets of loxoprofen (60 mg) and three diclofenac suppositories after the operation, and were informed that additional pain killer was available if necessary. Postoperative follow-up visits were scheduled at 1 week, 3, 6, 9, and 12 months after the operation. Post-operative complications, including wound infection, hematoma formation, and physical or mental symptoms related to androgen deficiency, were checked at every follow-up visit.

| Data analysis
We retrospectively collected data from the medical charts of patients who underwent the previous method between 2011 and 2014, and from 2014, we prospectively followed patients who would undergo the new method. Pre-operative basic characteristics of patients were compared between both groups, including serum luteinizing hormone, follicle-stimulating hormone, and testosterone. Sperm retrieval rates, occurrence rates of post-operative complications, and additional use of pain killer were also compared. Inter-group differences of categorial, continuous variables were analyzed with chi-square test and Mann-Whitney U test, respectively. Statistical significance was considered at P < .05, and analysis was performed using JMP ® 14 (SAS Institute Inc).

| 3D simulation of microTESE
Area of cut planes of bivalved testis is: Surface area of transverse sections at Z = lz is: Thus, the sum of the surface area of all coronal sections is: Total searching area in this example is:

| Application of 3D simulation findings to new NOA patients
Pre-operative basic characteristics of patients are shown in there was no significant difference between the two groups (Table 3). In the subgroup analysis, high SRRs for the new microTESE method were observed in Klinefelter syndrome and uniform maturation arrest subgroups. In these cases, suitable tubules containing Post-operative complications of microTESE have been examined in past papers. [12][13][14][15] While several comparative studies with conventional TESE and microTESE concluded that microTESE does not worsen the post-operative decrease in testosterone compared with conventional TESE, 16-20 a temporal decrease in testosterone after conventional and microTESE was reported. 21,22 Our new method of microTESE requires multiple cuts in testicular parenchyma, but the small amount of extracted seminiferous tubules was equivalent to that of the previous method of microTESE. As a result, no patients exhibited post-operative symptoms of androgen deficiency in our study. Although our study showed the short-term effects and needs a longer follow-up, we believe there will be limited effects on testosterone production, and testicular damage or loss will not be a significant risk in our new method.
There were no significant wound-related complications in our patients. To prevent post-operative complications, it is important to ensure complete hemostasis and meticulous closure of the wound.
We performed microTESE at our outpatient-based clinic, which had no admission facility, and paid considerable attention to hemostasis and to allow sufficient time for surgical wound closure. In this study, we used a 5-layer wound closure procedure after microTESE.
Sutures were run between the tunica albuginea, tunica vaginalis, Scarpa's fascia, Camper's fascia, and epidermal layers. We think that this procedure contributed to the low post-operative complication rate in our study.
As for post-operative pain, our new method was comparable with the previous method of microTESE. We completed the new method of microTESE using local anesthesia alone, and all patients could walk out of our clinic right after the surgery. The patients managed post-operative pain with NSAIDs prescribed after the surgery, and only a few required additional prescriptions.
In conclusion, our mathematical 3D simulation model highlighted a new way to maximize the surface area of testicular parenchyma in microTESE. Our new method achieved a better SRR and had similar post-operative complications and level of pain management to those of our previous method of microTESE.

CO N FLI C T O F I NTE R E S T
None.

AUTH O R S' CO NTR I B UTI O N S
All authors qualify for authorship by contributing substantially to this article. KI and OO developed the original concept of this study.
Surgeries were performed by KI, YM, NT, HN, and TG Data collection was performed by YM, NT, and HN Statistical analysis was performed by TG All authors have contributed to the discussion, reviewed the final version of this article, and approved it for publication.