Ultrasonography

Ultrasonography was used in nine of 18 studies to evaluate a total of 339 boys (391 testicles, 15.3 percent of boys with bilateral, nonpalpable testicles). One study was of good,⁶² two of fair,^{63, 64} and six were of poor quality.^{68, 72, 73, 75, 77, 79} The participants ranged in age from 1 month to 17 years. Surgery confirmed the presence of 358 testicles (91.6 percent) with 33 (8.4 percent) absent or vanishing. Eighty-six (24.0 percent) of the testicles were found in the abdomen, 188 (52.5 percent) in the inguinal or scrotal region and 84 (23.5 percent) were found to be atrophied. Prior to surgery, 222 (56.8 percent) had been identified as present with US and 169 were indicated to be absent. The one good quality study ⁶² described selection criteria of participants, reported patient characteristics including previous surgeries, provided the time interval between imaging and surgery, and provided details on how imaging and surgery were conducted. The study reported only one intra-abdominal testis as identified by surgery.⁶²

Of the 358 testicles found during surgery, US had identified 219 (61.2 percent), and correctly predicted that 30 testicles would be absent or vanishing (90.9 percent). Ultrasonography falsely identified three testicles as present when they were absent, and failed to identify 139 testicles that were found surgically. The sensitivity with which US identified the presence of surgically confirmed testicles ranged from 15 percent⁷³ to 80 percent across studies,⁶² and the specificity was generally close to 100 percent except in two studies that reported substantially lower specificities of 67 percent⁶² and 80 percent.⁷²

The study-specific accuracy rate (using surgery as a reference standard) for US ranged from 21.4 percent⁷³ to 76 percent (Table 6).⁷² The one good quality study had an accuracy rate of 73 percent ⁶². In the three cases where US suggested the presence of testicles that turned out to be false positives, two were scar tissue⁶² and one was a lymph node structure.⁷²

Table 6

Studies assessing US for locating nonpalpable undescended testicles.

The degree to which US correctly identified testicles differed by location of the testicles (Table 7). For example, although there were 86 normal testicles found by surgery in the abdomen, US identified only 29 of these, missing more than 65 percent. On the other hand, US correctly predicted the presence of more than 90 percent of testicles found by surgery to be in the inguino-scrotal area. When testicles were atrophied, US was more likely to miss them altogether, failing to identify more than 80 percent of the 84 atrophied testicles across studies. Ultrasonography also “misplaced” three testicles, indicating that they were inguinally located when they were in fact in the abdomen.⁷⁵

Table 7

Position and type of undescended testicles identified by surgery and US.

Computed Tomography Scan

Computed tomography scan was used in one poor quality prospective case series of 26 boys, ranging in age from 2 to 18 years.⁶⁷ A total of 30 possible testicles were evaluated. Four boys had bilateral, nonpalpable testicles. Those boys in whom CT scan failed to locate the testicles were also assessed with spermatic venography. Due to technical difficulties or family refusals, relevant data for spermatic venography could be extracted only for nine patients and are not discussed here.

Surgery determined that 28 testicles (93.3 percent) were present and 2 (6.7 percent) were absent. Nine (32.1 percent of 28) were found in the abdominal region and 19 (67.9 percent) in the inguino-scrotal region, including 17 at the internal ring, one each in the inguinal canal and in the canalicular region. Computed tomography scan identified 16 testicles as present (53.3 percent) and the proportion of absent testicles was 0.47 (46.7 percent). Computed tomography scan did not falsely predict any testicles to be present when they were not, but it did miss 12 testicles. Thus CT scan was able to identify testicles with a sensitivity of 57 percent, and specificity of 100 percent. Negative predictive value was 14.3 percent, with a 100 percent positive predictive value. Overall accuracy was 60 percent but the study was of poor quality (Table 8).

Table 8

Studies assessing CT scan for locating nonpalpable undescended testicles.

CT scan correctly located 11 testicles in the inguino-scrotal region with 10 testicles at internal ring, 1 at inguinal canal and 5 in the abdominal region in 1 poor quality study (Table 9). However, it missed four intra-abdominal, one canalicular testicles and seven testicles at internal ring. Overall, CT scan located similar proportion of testicles at inguinal (57.9 percent) and in the abdominal regions (55.6 percent).

Table 9

Position and type of undescended testicles identified by surgery and CT scan.

Magnetic Resonance Imaging

Ten studies (3 of fair^{63, 65, 78} and 7 of poor quality^{68–70, 72, 74, 76, 77}) evaluated the accuracy of MRI for identifying nonpalpable testicles in a total of 263 boys with 294 testicles (31 boys with bilateral testicles). Seven of 10 studies were conducted in Asia. The boys were 8 months to 14.6 years in age. Both conventional and diffusion-weighted MRI (DWI) techniques were used (Table 10).

Table 10

Studies assessing MRI for locating nonpalpable undescended testicles.

One fair quality case series used three types of MRI (conventional, DWI, and a combination of MRI and DWI) and two observers to read the pre-operative MRI images of 36 boys with 38 testicles.⁷⁸ Thirty-four testicles were found laparoscopically (89.5 percent) while 4 were not identified (10.5 percent). Using conventional MRI, observer 1 and observer 2, identified 29 and 30 testicles, respectively, as present. The overall accuracy rate (using surgery as a reference standard) for conventional MRI was 86 and 84 percent. Using DWI, observer 1 and observer 2 correctly identified the presence and absence of testicles with an accuracy rate of 86 percent and 81 percent, respectively. The diagnostic performance was superior when DWI and conventional MRI were used together. Both observers identified 31 testicles with a sensitivity of 0.88 and 0.91, specificity and PPV of 100 percent and an overall accuracy rates of 92 and 86 percent. The one false positive case identified with all the methods was confirmed to be an infected lymph node. Of the 34 testicles identified by laparoscopy, 19 (50 percent) were intracanalicular in position, while 11 were low intra-abdominal (29 percent) and 4 (10.5 percent) were high-intra-abdominal. Among the intra-abdominal testicles, one was reported to be atrophic (2.6 percent) which both observers failed to locate. Though both observers correctly located all 29 testicles by conventional MRI, details of the actual locations of these testicles were not reported.

Of the remaining nine studies, most of which were poor quality,^{68–70, 72, 74, 76, 77} 256 testicles were evaluated with MRI (227 boys, 12.8 percent with bilateral, nonpalpable testicles). Surgery identified 214 testicles (83.6 percent) as present and 42 as absent, while MRI found 146 testicles (57 percent) as present and 110 (43 percent) as absent. MRI falsely identified 5 testicles as present including 4 lymph node structures⁶⁸ and failed to identify 73 testicles that were found surgically. Overall, MRI was able to correctly identify testicles with a sensitivity ranging from 0.33 to 0.82, specificity from 0.56 to 1.0, PPV of 0.83 to 1.0, and an overall accuracy rate of 42 to 83 percent. Overall accuracy ranged from 41.7 to 92 percent among fair studies and 43.5 to 88 percent in poor studies.

Table 11 lists the location of testicles as identified by surgery and MRI. One study did not provide MRI results for the position or type of testicles and therefore does not contribute data to this table.⁷⁸ Surgery located 52 testicles in the intra-abdominal region and 100 in the inguino-scrotal region, while MRI located 38 (26 percent) testicles intra-abdominally and 88 (60.3 percent) in the inguino-scrotal region. Surgery and MRI also located 62 and 20 atrophic testicles respectively. MRI wrongly classified one testicular nubbin⁶⁹ as an intra-abdominal testicle. Overall, MRI was able to correctly locate 71.2 percent of the intra-abdominal testicles, 83 percent of the testicles in inguinal or scrotal region and only 32.3 percent of the atrophied testicles.

Table 11

Position and type of undescended testicles identified by surgery and MRI.

Magnetic Resonance Arteriography/Venography

Fourteen boys with a mean age of 28 months (range: 3–144) underwent MRI and MRAr/V as part of a larger, poor quality, prospective case series.⁶⁶ Imaging predicted that eight of the 14 testicles (57 percent) would be present, and that six (42.9 percent) were absent. No testicles were falsely predicted to be present by imaging, resulting in an overall accuracy rate for identifying testicles of 57 percent, and a positive predictive value of 100 percent (Table 12).

Table 12

Studies assessing other magnetic resonance techniques for locating nonpalpable undescended testicles.

Of the 14 testicles found surgically, five were located intra-abdominally, five were testicular nubbins, three were normal testicles in the inguinal canal and one was in the scrotum. MRI with MRAr/V correctly located 80 percent (4 of 5) of intra-abdominal testicles, 40 percent (2 of 5) of testicular nubbins, and 50 percent (2 of 4) of testicles in the inguino-scrotal region (Table 13). However it failed to correctly locate one of each scrotal, intra-abdominal, intracanalicular and three testicular nubbins.

Table 13

Position and type of undescended testicles identified by surgery and other magnetic resonance related techniques.

Magnetic Resonance Venography

Magnetic resonance venography was assessed in 34 boys with a mean age of 6.4 years (range: 1–16 years), for a total of 44 testicles in one poor quality case series.⁷¹ Ten of the boys had bilateral nonpalpable testicles. Patients younger than 5 years old received oral sedation while others received IV sedative. Magnetic resonance venography identified 37 of 44 testicles as present (84.1 percent) while identifying seven of 44 testicles as vanishing testicles (15.9 percent), all of which were confirmed surgically for 100 percent accuracy (Table 12). Magnetic resonance venography correctly located all 28 inguino-scrotal testicles (26 canalicular, and two testicles at pelvic skinfold), five intra-abdominal and four atrophic testicles, yielding a 100 percent accuracy rate in locating nonpalpable testicles (Table 13).

Magnetic Resonance Angiography

Two studies assessed gadolinium-infusion MRA in the identification of bilateral, nonpalpable undescended testicles in 35 boys (40 testicles), ages 1 to 16.^{65, 77} Both studies were conducted in China and were prospective. One had fair quality,⁶⁵ and one was poor quality.⁷⁷ No adverse effects were reported with gadolinium infusions.

Surgery and MRA identified all 39 testicles as present (100 percent) and 1 as absent (Table 12).⁷⁷ MRA correctly located all the 7 intra-abdominal testicles, 21 inguino-scrotal testicles and 11 atrophied testicles confirmed by surgery, yielding an overall accuracy rate of 100 percent along with all the diagnostic performance indices at 100 percent (Table 13).

Studies of LHRH and/or Its Analogs

Six studies used a placebo-controlled design to assess the effectiveness of LHRH in the treatment of cryptorchidism.^{18, 19, 23–25, 27, 29, 30} All were double-blind RCTs, with the most common outcome testicular descent (Table 15). One fair-quality, three-arm study compared hCG, LHRH and placebo and found hCG to be more effective than placebo at inducing successful descent of the undescended testicle into the scrotum.¹⁸ This study is discussed in greater detail in the section on studies comparing different hormonal regimens.

Table 15

Short-term testicular descent in two-arm, randomized, placebo-controlled studies.

One fair-quality RCT compared LHRH to placebo.¹⁹ This study randomized 141 boys with cryptorchidism age 2 to 12 years to receive either LHRH 0.4mg or placebo intranasal three times a day for 4 weeks. One hundred twenty-three (87 percent) participants had complete followup, including 62 (97 testicles) in the LHRH group and 61 (90 testicles) in the placebo group. Success was assessed immediately following treatment and was defined as having both testicles fully descended into the scrotum. In the LHRH arm, 9.7 percent of participants were considered successfully treated compared to 1.6 percent in the placebo arm, but the difference was not statistically significant.

Three papers were published from one poor-quality RCT of the effectiveness of LHRH in achieving complete testicular descent in children from 1 to 12 years old.^23–25 Of 252 participants, 237 (with 281 cryptorchid testicles) had complete followup. Neither the approach to random sequence generation, nor the allocation concealment was adequate or clear in this study. Participants randomized to LHRH received a total dose of 1.2mg/day (200μg inhaled in each nostril three times a day) for 4 weeks. Treatment was unblinded 8 weeks after randomization (4 weeks of study drug followed by 4 weeks off of treatment) and outcomes assessed. Nine percent of participants randomized to LHRH achieved complete testicular descent compared to eight percent in the placebo group. Although no differences were seen in the primary comparison, initial testicular location may have been associated with success in that no testicle initially located above the external inguinal ring descended.

The remainder of the RCTs comparing LHRH to placebo had slightly higher rates of descent in the treatment arms relative to placebo, but all were of poor quality. One assessed clinically acceptable descent into the scrotum following LHRH 100ug three times a day for 28 days in each nostril.²⁷ Fifty cryptorchid boys aged 1.5 to 10.5 years enrolled in the study. Outcomes were assessed immediately following the completion of treatment. Three participants dropped out of the study, leaving 23 (with 29 cryptorchid testicles) in the LHRH arm and 24 (with 32 cryptorchid testicles) in the placebo arm. Although 18 testicles (62 percent) had a “therapeutic effect” (defined as “significant move from the pretreatment location towards the bottom of the scrotum”) with LHRH compared to one (3 percent) in the placebo arm, only eight (28 percent) testicles treated with LHRH had complete descent into the scrotum (number with complete descent not reported for placebo arm). Six of these successfully treated testicles were located at the scrotal neck at the start of treatment while two were located in the inguinal canal. Six to 12 months later, these last two had re-ascended out of the scrotum and required surgical repair.

Another poor-quality double-blinded RCT enrolled 50 boys aged 3 to 8 years with unilateral undescended testicles to receive either LHRH 100ug in each nostril or placebo six times a day for 28 days.²⁹ The primary outcome of interest was complete descent into the scrotum, which was assessed immediately following the completion of treatment and 6 months after randomization. Immediately following completion of treatment, 20 percent (5 of 25) of the participants in the LHRH arm had responded to treatment. Of these, three were considered complete responses and two were “borderline.” Twelve percent (3 of 25) of patients in the placebo group experienced testicular descent immediately following treatment. Response to treatment was not durable, with only eight percent (2 of 25) in the LHRH arm and four percent (1 of 25) in the placebo arm still descended after 6 months. Neither this study nor the one described above identified factors associated with testicular re-ascent.

A similar RCT of poor quality randomized 49 participants aged 1.2 to 11.9 years to either LHRH 400ug in each nostril or placebo three times a day for a period of 28 days.³⁰ Outcomes were assessed 8 weeks after randomization. The authors note that 37 percent of 35 testicles in the LHRH experience some degree of descent at 8 weeks compared to 18 percent of 34 in the placebo arm. However, no patients in the placebo group and only three (9 percent) in the LHRH group experienced complete descent.

Studies Focused on hCG and/or Its Analogs

One fair-quality, three-arm study compared hCG, LHRH and placebo and found hCG to be more effective than placebo at inducing successful descent of the undescended testicle into the scrotum.¹⁸ This study is discussed in greater detail below in the section on studies comparing different hormonal regimens.

Three different studies (one of good³⁵ and two of poor^{26, 28} quality) compared different dosing regimens of hCG (Table 16). These studies primarily focused on comparing higher vs. lower doses of the drug, with the intent to make the administration schedule easier and minimize side-effects. One poor-quality study randomized 183 cryptorchid boys to receive either hCG 1500 IU by intramuscular (IM) injection every other day for 14 days (88 participants) or four IM injections in a dose related to body weight (100 IU/kg) every 4–5 days to a maximal dose of 3000 IU (95 participants).²⁶ Successful outcome was defined as complete testicular descent into the scrotum and was assessed immediately following the conclusion of treatment. In children with unilateral cryptorchidism, 50.8 percent and 50.9 percent of participants in each groups had successful outcomes. In children with bilateral cryptorchidism, 48.3 percent of participants in the lower dose group and 50 percent of those in the higher dose group had successful outcomes. Not surprisingly, children in whom the undescended testicle(s) was/were initially in the mid-inguinal canal or lower had higher success rates than those who had testicles located above that level.

Table 16

Testicular descent in studies comparing dosages of hCG.

The second study examining different dosing strategies of hCG primarily focused on schedule of administration as opposed to total dose received.²⁸ This poor-quality RCT randomized 332 boys with 435 cryptorchid testicles aged 1 to 13 years to receive either 2 lower dose hCG injections per week for 5 weeks or 1 higher dose hCG injection every 7–10 days for 3 weeks. The total dose received was effectively the same between the two arms and was determined by the age of the participant at the time of randomization. Complete success was defined as full descent of the testicle into the scrotum and was assessed between 8 weeks and 6 months after the conclusion of treatment. Restricting the analysis to cryptorchid testicles located above the high scrotum, the dosing schedule with more injections had a significantly higher complete success rate than the schedule with fewer injections (39.4 vs. 29.9 percent, p<0.05). However, children with abdominal testicles were significantly more likely to report complete success with more intensive injection regimen (31.5 vs. 13.5 percent, p<0.05). In addition, older children randomized to the more intensive dosing schedule (age 6–13) were more likely to report complete success than those assigned to the strategy with fewer injections (46.4 vs. 21.6 percent, p<0.05).

Similar to the study above that failed to show a difference between higher and lower doses of hCG, one good-quality prospective observational study comparing low-dose hCG (500 IU/week for 3 weeks) to a higher-dosing regimen (1,500 IU/m2 three times a week for 3 weeks) and found similar results.³⁵ Again, success was defined as complete descent into the scrotum and was assessed immediately following the conclusion of treatment. Twenty-one children were treated with the lower dose, which had a success rate of 66.7 percent, compared with 14 children treated with the higher dose, which had a success rate of 57.1 percent (p=NS)

Studies Comparing Hormonal Regimens

Four studies compared the effectiveness of hCG and LHRH therapy (Table 17).^{16, 18, 20, 38} A good-quality randomized double-blind study compared the effectiveness of intranasal LHRH 100μg inhaled in each nostril six times a day for 4 weeks to parenteral hCG 1.2mg daily for 4 weeks.¹⁶ In each arm, placebo injections or inhalations using the same administration schedule were also given to ensure blinding. Thirty-three boys with unilateral (29) or bilateral (4) cryptorchidism were enrolled in the study. The primary outcome was complete scrotal descent and was assessed immediately following treatment and then monthly for up to three months after the conclusion of treatment. One participant in the hCG (5.9 percent) had complete testicular descent into the scrotum as opposed to three in the LHRH arm (18.8 percent). The difference between the two arms was nonsignificant (p=0.23). The investigators conducted a parallel uncontrolled study of 13 boys with retractile testicle treated with the same hCG regimen which is presented in the same report. In this study, five of the 13 children (38.4 percent) had complete descent of the testicle into the scrotum¹⁶

Table 17

Testicular descent in studies comparing LHRH with hCG.

One fair-quality RCT¹⁸ that stratified by sidedness is not included in the table because it lacked pooled results, but is described here. The study describes a three-arm double-blind placebo controlled RCT assessing the effectiveness of hCG or LHRH. Children aged 1.8 to 13 years with unilateral or bilateral cryptorchidism were randomized to either hCG 100 IU/kg IM (maximum 1500 IU) twice weekly for 3 weeks, LHRH 200 μg intranasal in each nostril three times a day for 4 weeks, or inhaled placebo three times a day for 4 weeks. Although 330 participants were randomized, complete followup was available in only 243, of whom 155 had bilateral and 88 had unilateral cryptorchidism. After excluding 23 participants who were felt to have retractile testicles, the authors reported their results stratified by bilateral or unilateral disease. For participants with bilateral cryptorchidism, 23 percent treated with hCG reported complete descent of both testicles into the scrotum, as opposed to 9 percent in the LHRH group and 0 percent in the placebo group (Fisher’s exact p=0.001). Among those with unilateral cryptorchidism, hCG was effective in inducing complete testicular descent in 15 percent of those treated with hCG, as opposed to 0 percent in either the LHRH or placebo groups (Fisher’s exact p=0.02).

The remaining studies focused on combinations of hormonal therapy to determine if multidrug regimens worked better than single agents. One poor-quality RCT compared the effectiveness of four different hormonal regimens in 155 boys with unilateral palpable cryptorchidism age 10 to 48 months.²⁰ The four treatment regimens included: hCG 500–1000 IU intramuscular (IM) per week (depending upon age) for 6 weeks (n=37); hCG 500–1000 IU IM per week plus hMG 75 IU IM per week for 6 weeks; LHRH 1.2mg inhaled daily for 4 weeks and; LHRH 1.2 mg inhaled dialed for 4 weeks plus hCG 1500 IU IM per week for 3 weeks. Short-term testicular descent rates were similar among the four groups, with 22.5 percent of those in the LHRH-hCG group reporting complete testicular decent compared to 21.6 percent in the hCG alone group, 17.9 percent in the hCG-hMG group and 15.4 percent in the LHRH alone group. Roughly 5 percent of descended testicles had relapsed at 6 months with no differences noted among the groups. The overall long-term success rate in the study was 15 percent, with success rates within the four groups varying from 13 to 19 percent. The authors note that 74 percent of boys on hCG and 5 percent of those on LHRH initially reported signs of androgenization, such as erections and penis growth that receded at long-term followup.

Two papers from a single poor-quality RCT compared buserelin acetate (a gonadotropin-releasing hormone agonist which would function much like LHRH) 20 μg per day delivered in three inhaled doses for 4 weeks followed by hCG 1,500 IU IM per week for 3 weeks to inhaled placebo followed by hCG 1,500 IU IM per week for 3 weeks.^{21, 22} Twenty-two participants with cryptorchidism were randomized to buserelin/hCG arm while 19 were randomized to the placebo/hCG arm. The primary outcome testicular descent assessed 3 months after the conclusion of hCG therapy. Thirty-six percent (8 of 22) of patients in the buserelin/hCG group were noted to have complete testicular descent at 3 months, as opposed to 11 percent (2 of 19) in the placebo/hCG group (p<0.01). It is worth noting that in both groups all but one of the testicles that responded to treatment was located in the prescrotal position at the time of randomization, indicating that cryptorchid testicle located closer to the scrotum are more likely to respond to hormonal treatment.

Finally, one poor-quality prospective cohort study examined 324 children with palpable unilateral or bilateral cryptorchidism aged 6 months to 13 years who were treated with one of five hormonal regimens: hCG 500 IU IM twice a week for 6 weeks (n=113); human menopausal gonadotropin (hMG) 150 IU IM twice a week for 4 weeks (n=35); LHRH 1.2mg inhaled daily for 4 weeks (n=85); hMG 150 IU IM twice a week for 4 weeks followed by hCG 500 IU IM twice a week for 6 weeks (n=27) and; LHRH 1.2mg inhaled daily for 4 weeks followed by hCG 500 IU IM twice a week for 6 weeks (n=64).³⁸ The authors reported that treatment was randomly assigned by 6-month period, whereby every six months a new treatment was randomly assigned and all participants enrolled during that time would receive that therapy. The primary outcome was complete descent into the scrotum (mid- or low scrotal position) after the completion of therapy although the exact timing of the assessment of the outcome is not reported, hCG alone was the most effective treatment, resulting in complete testicular descent in 34.5 percent of participants. The two next most effective therapies were LHRH alone and LHRH and hCG treatment resulting in testicular descent 29.4 and 29.6 percent of the time, respectively. Combined hCG and hMG resulted in testicular descent 25.9 percent of the time, while hMG alone was completely ineffective with no participants reporting complete testicular descent. There were no significant differences in effectiveness between the four effective therapies. Successful outcomes were defined as complete descent into the scrotum and were assessed 1 week and 6 months after the conclusion of therapy.

Long-Term Fertility Outcomes Following Hormonal Therapy

No reports on long-term fertility outcomes following isolated hormonal therapy were found in our literature search. One good-quality study explored long-term fertility outcomes of hormonal therapy as an adjunct to surgical treatment in cryptorchidism.^{43, 44} This study compared 15 young men who had cryptorchidism and were treated as children with orchiopexy followed by Buserelin 10ug intranasal every other day for 6 months to 15 age-matched controls who were treated by orchiopexy alone. The primary outcomes were semen analysis parameters measured in early adulthood (mean age=19 years). Those participants who received surgery and buserelin had significantly higher sperm counts (90 ×10⁶ sperm per ejaculate compared to 1×10⁶ in the surgery only group, p<0.001). In addition, 11 percent of those who received surgery and hormone therapy had normal morphology as opposed to none in the surgery alone group.

Comparisons of Single-Stage, Two-Stage Fowler-Stephens and Primary Orchiopexy Procedures

The studies reporting results of surgical techniques were all retrospective, observational studies. Decisions about which method of surgical repair to use are made clinically, and not with the intent of comparing the effectiveness of the procedures in comparable groups of patients. The groups were therefore essentially different, and specific treatment choices were most likely made on the basis of where the affected testicle was located. Because these studies did not control for initial testicular location, the results can only be interpreted as providing noncomparative data on outcomes in groups with differing clinical presentations treated surgically. We identified 11 such studies, one of which was assessed as having good quality,⁴⁰ and the remainder poor quality.^{47, 48, 50, 52, 54, 56, 57, 59–61} In all but one of these reports,⁴⁸ the authors failed to control for the location of the testicle in the abdomen. One study was of good quality and reported lower success rates across intervention types, but this study also did not control for testicular location.⁴⁰

Primary Orchiopexy Versus One-Stage Fowler Stephens Versus Two-Stage Fowler Stephens

A single study controlled for location of the testicle by stratifying the analysis (“peeping” testicle vs. within 3 cm of the internal inguinal ring vs. beyond 3 cm of the internal inguinal ring).⁴⁸ The 3 cm cutoff was arbitrarily chosen by the study authors. The study compared laparoscopic primary orchiopexy to laparoscopy single and two-stage FS procedures. Eighty-six testicles were studied, but only 63 had more than 3 months of followup. One-stage FS orchiopexy was used to repair 11 testicles, of which three were located within 3 cm of the inguinal ring and eight were beyond 3cm from the inguinal ring. Nine (82 percent) of the testicles treated with a one-stage FS were viable 3-months post-operatively and both of the nonviable testicles were initially located beyond 3 cm from the inguinal ring.

The remaining 49 testicles were all treated with primary orchiopexy. Only one of these was initially located beyond 3 cm from the inguinal ring testicle and this one was viable 3 months after surgery. The remainder were all either within 3 cm of the inguinal ring or “peeping” and of these, 98 percent (47 of 48) were viable three months after surgery. Any direct comparison of surgical approaches in this study is limited by small sample size in the FS groups and the fact that location in the testicle influences both choice of approach and the underlying probability of post-operative viability.

Two studies comparing the three approaches reported intra-abdominal location of the nonpalpable testicle at the time of surgery, but neither stratified results by location. In one series of 101 nonpalpable cryptorchid testicles, 46 were “intra-abdominal,” 14 were located at the iliac vessels, 22 were located at the inguinal ring, 12 were “peeping” through the ring, 3 were behind the bladder, and 4 were intracanicular.⁵⁶ Success rates (no specific definition given) were 92 percent for primary orchiopexy, 84 percent for one-stage FS, and 96 percent for two-stage FS orchiopexy. In a pooled series of 281 testicles, 65.2 percent of testicles treated by primary orchiopexy were either within 2 cm of the inguinal ring or peeping through the ring, in contrast to 6.9 percent of those repaired with one-stage FS orchiopexy and 21.1 percent with two-stage FS orchiopexy.⁵⁴ Among 19 “surgical failures,” 13 were more than 2 cm from the internal ring, while only three were ectopic, two within 2 cm of the ring and two were “peeping.”

Four studies^{40, 47, 52, 60} did not report the distribution of intra-abdominal testicular location at the time of surgery, although three^{40, 47, 60} specifically stated that when a testicle was located higher in the abdomen, one or two-stage FS procedure was performed, while when it was lower in the abdomen one-stage FS or primary orchiopexy was performed. Although post-operative atrophy and/or success rates for these studies are presented in Tables 19–24, it is impossible to determine if superior outcomes noted with one approach or another are due to the operation itself or to confounding by indication. Nine of the 11 studies provided success rates for primary orchiopexy, one-stage FS, or two-stage FS procedures. When the results of these nine studies are pooled, the overall success rate for primary orchiopexy is 96.4 percent (Table 19).

Table 19

Success rates after primary orchiopexy.

The overall success rate for one-stage FS is 78.7 percent (Table 20).

Table 20

Success rates after one-stage Fowler-Stephens.

The overall success rate for two-stage FS is 86.0 percent (Table 21).

Table 21

Success rates after two-stage Fowler-Stephens.

Five of the 11 studies provided atrophy rates for primary orchiopexy, one-stage FS, and two-stage FS procedures. When the results of these five studies are pooled, the overall atrophy rate for primary orchiopexy is 1.83 percent (Table 22).

Table 22

Atrophy rates after primary orchiopexy.

The overall atrophy rate for one-stage FS is 28.1 percent (Table 23).

Table 23

Atrophy rates after one-stage Fowler-Stephens.

The overall atrophy rate for two-stage FS is 8.2 percent (Table 24).

Table 24

Atrophy rates after two-stage Fowler-Stephens.

While, on the surface, this may appear to favor primary orchiopexy over the two other approaches, all of these studies are observational cohort designs in which confounding by indication limits our ability to draw any conclusions regarding the superiority of primary orchiopexy to either the one- or two-stage FS approach. Nonetheless, these studies may provide some insight into surgical outcomes among these three clinically distinct groups.

Comparisons of the Same Procedure Performed Through Either an Open or Laparoscopic Approach

Given that the comparative effectiveness of the possible surgeries is unknown due to the confounding by indication noted above, the question of whether it might be possible to use a less invasive approach (laparoscopy rather than open surgery) becomes eminent. Five studies (one of good,¹⁷ one of fair,³⁶ and three of poor quality^{31, 49, 55}) compared laparoscopic versus open approaches of essentially the same procedure (Table 25). Two focused on approaches for exploration; three focused on surgical repair.

Table 25

Laparoscopic versus open exploratory approaches.

Two of the studies, one good¹⁷ and one poor⁵⁵ quality, compared initial laparoscopic evaluation of the abdomen to open exploration in children with nonpalpable testicles. The good quality RCT reported randomizing 64 patients with 68 nonpalpable testicles to laparoscopic evaluation followed by open surgical repair depending upon laparoscopic findings to initial open exploration followed by appropriate treatment.¹⁷ In 6 of 24 patients randomized to laparoscopy, intra-abdominal blind ending vessels and a closed inguinal ring was noted, and so open surgery was not performed. Twelve of 31 patients randomized to open exploration had a vanishing testicle. Mean operative time and cost were significantly higher in the laparoscopic group, but no differences were noted in recurrence rates. The second study was a poor-quality retrospective analysis to make a similar comparison of laparoscopic exploration followed by open repair to open exploration and repair.⁵⁵ Seven of 20 testicles in the laparoscopy arm were found to be nonviable, although the location of these vanishing testicles (intra-abdominal vs. canalicular) is not reported making it impossible to determine the number of patients that would have avoided open surgery with laparoscopic exploration. Rate of successful orchiopexy was similar between the two arms, although a specific definition of success was not given.

The remaining three studies (one fair quality³⁶ and two poor^{31, 49}) compared similar procedures using either a laparoscopic or an open approach (Table 26). One poor-quality RCT compared outcomes following various types of laparoscopic or open orchiopexies for nonpalpable testicles.³¹ This study is one of the few reports in the literature that controlled for location of the testicle within the abdomen, allowing for comparisons between procedures. Specifically, the study design required that children who were found to have high abdominal testicles at the time of laparoscopic exploration underwent a two-stage FS procedure (with the second stage randomized to a open vs. laparoscopic approach) while those with low abdominal testicles underwent either open or laparoscopic primary orchiopexy. This greatly minimized the possibility of confounding by indication. The study was judged to be of poor quality because no information was given regarding the randomization process and the study was at high risk for both performance and detection biases. All patients in this study initially underwent laparoscopic evaluation of the abdomen to confirm the presence and location of a viable testicle. If the testicle was noted to be high in the abdomen, the patient underwent a laparoscopic one-stage FS procedure (laparoscopic clipping of the testicular vessels). They were then randomized to receive either open or laparoscopic two-stage FS orchiopexy. Perioperative outcomes between participants undergoing laparoscopic or open second procedures were compared, with patients undergoing laparoscopic two-stage FS orchiopexy noted to have statistically significantly shorter operative times (p=0.000), time to oral feeding (p=0.004), hospital stays (p=0.008) and return to normal activities (p=0.000). While all testicles in both groups were noted to have satisfactory scrotal position after surgery, two of the 20 (10 percent) testicles in the laparoscopic arm and three of the 16 (19 percent) testicles in the open arm had atrophied after 1 year of follow-up. Patients in this study who had viable testicles located in the lower portion of the abdomen (closed to the inguinal ring) were randomized to undergo either laparoscopic or open primary orchiopexy. Like the high abdominal group, patients randomized to laparoscopic orchiopexy had statistically superior peri-operative outcomes. Of the 21 testicles randomized to laparoscopic orchiopexy and the 18 randomized to open orchiopexy, all were satisfactorily placed in the scrotum and no cases of atrophic testicles were noted after 1 year of follow-up.

Table 26

Laparoscopic versus open surgical repair approaches.

The other two studies in this category were observational.^{36, 49} One fair-quality cohort study included only palpable testicles and compared laparoscopic and open primary orchiopexy.³⁶ Despite having reasonable comparison groups, the results are difficult to interpret. The authors report that the participants undergoing the laparoscopic approach “had less pain when compared to the open technique in 80 percent of cases” using a visual analog scale, but how this comparison was made is unclear. Similarly, while the authors note that all patients had “satisfactory results in relation to size and location of testicle,” details regarding these outcomes are lacking. Another poor quality cohort study included both palpable and nonpalpable testicles, failed to control for the location of the testicle in the analysis, and grouped both primary and FS orchiopexies into two heterogeneous groups based upon whether an open or laparoscopic approach was used, making it difficult to draw meaningful conclusions in terms of post-operative testicular position or viability.⁴⁹

Comparisons of Various Surgical Modifications of Primary Orchiopexy

Six studies, one good⁴¹ and five poor quality,^{32–34, 45, 58} compared different modifications or techniques of orchiopexy. Three studies compared the type of incision used when performing primary orchiopexy.^{33, 41, 58} The first study,³³ a poor-quality RCT, assigned 212 participants with low lying cryptorchid testes to undergo primary orchiopexy through either a single incision versus two incision approach. The study did not show a difference in success rates between the two procedures, with the single incision approach having a 92.5 percent success rate and the two incision approach having a 96.5 percent success rate. The only significant difference noted between the two approaches was in operative time, with the single incision approach having shorter average operative times.

The good-quality retrospective cohort study compared 63 orchiopexies performed using a single prescrotal incision to 53 orchiopexies performed using the inguinal two-incision technique.⁴¹ The two groups appeared comparable in terms of location and age at surgery. No differences were noted in testicular atrophy or re-ascent rates between the groups. The final poor quality retrospective cohort study reviewed 286 orchiopexy procedures clustered into three groups: (1) 125 performed through a low-scrotal incision; (2) 60 performed through a high scrotal incision and; (3) 101 performed using the inguinal two-incision technique.⁵⁸ Success rates between the three groups ranged from 98–100 percent and were not significantly different. It is worth noting that the two scrotal approaches were not attempted in any testicle that could not be milked down into the upper third of the scrotum under anesthesia, which may affect the comparability of the groups.

One poor-quality RCT randomized 104 participants to undergo either traditional primary orchiopexy, where the external oblique fascia is opened, or a “closed” technique, where the procedure is performed without opening the fascia.³⁴ The study did not report final location of the testicle in the scrotum but did report atrophy rates and found no significant differences between the two arms (2 percent vs. 4 percent). The primary outcome of the study was post-operative complication rates, with no differences noted between the two arms in terms of post-operative hematoma, infection or medial thigh sensory loss.

Another poor-quality RCT randomized boys with a total of 150 unilateral palpable undescended testicles undergoing scrotal pouch orchiopexy in three groups: (1) testicular fixation to the scrotal wall; (2) narrowing of the neck of the Dartos pouch and; (3) both testicular fixation to the scrotal wall and narrowing of the neck of the Dartos pouch.³² Results were stratified by initial position of the testicle. No postoperative ascents were noted in the second or third group, but four of 50 testicles in the first group re-ascended out of the scrotum with a mean follow-up of 28 months. One testicle in the first group and one in the second group atrophied, while no atrophic testicles were noted in the third group. The authors conclude that only narrowing of the neck of the Dartos pouch is required to prevent re-ascent.

Finally, 1 poor-quality retrospective review compared 19 orchiopexies using a Jones approach (extensive retroperitoneal mobilization of the spermatic vessels without ligation through an incision medial to the superior anterior iliac spine) with 10 inguinal orchiopexies with testicular vessel ligation (effectively one-stage FS procedures).⁴⁵ Eighteen of the 19 orchiopexies performed using the Jones approach had a “satisfactory” outcome compared with 7 out of 10 in the inguinal orchiopexy group.

Fertility Indicators

Four retrospective cohort studies provided data on endocrine and fertility status and are discussed separately here.^{42, 46, 51, 53} One good-quality study studied testicular endocrine function in children.⁴² Three poor-quality studies focused on fertility and endocrine outcomes in adult men treated for cryptorchidism as children.^{46, 51, 53}

One study of good quality compared plasma testosterone levels after hCG injection in 52 children with cryptorchidism and 10 age-matched controls.⁴² Thirteen had untreated unilateral cryptorchidism, 10 had unilateral cryptorchidism and had previously undergone successful orchiopexy. The mean time between surgery and the study was 23 months. Seventeen had untreated bilateral cryptorchidism and 12 had bilateral cryptorchidism that had been successfully treated with orchiopexy, generally about 24 months prior to the study. No significant differences were noted in post-hCG plasma testosterone levels between treated and untreated children with unilateral cryptorchidism (1.59 and 1.53 ng/ml respectively) and controls after hCG injection (1.67 ng/ml). In the bilateral group, however, untreated children had significantly lower testosterone levels when compared to treated children (0.94 vs. 1.58 ng/ml, p<0.001). Children who were treated for bilateral cryptorchidism had post-hCG testosterone levels that were not significantly different from controls.

Three poor-quality studies examined long-term fertility outcomes in men who were treated for cryptorchidism in childhood (Table 27).^{46, 51, 53} One⁴⁶ reported actual paternity (ability to father children) rates, while the others^{51, 53} reported semen analysis parameters. Men with unilateral cryptorchidism had higher paternity rates regardless of treatment when compared to men with bilateral cryptorchidism⁴⁶. The addition of hCG treatment to orchiopexy was not associated with higher paternity rates than orchiopexy alone.

Table 27

Long-term endocrine and fertility outcomes in adult men with cryptorchidism treated in childhood.

In the first study of semen parameters, no man with bilateral cryptorchidism treated with orchiopexy in childhood had a normal sperm density and only 7 percent had normal sperm motility.⁵¹ These outcomes were considerably worse than those observed in the unilateral cryptorchid group, in which 72 percent of participants had normal sperm density and 70 percent had normal motility (p<0.001). Sperm density and motility were similar in men with unilateral cryptorchidism regardless of whether they were treated with orchiopexy or orchiectomy. Both groups had better semen parameters than men with untreated unilateral cryptorchidism, suggesting that surgical treatment of any type was associated with better semen parameters later in life, but that one surgical treatment was not more effective than another in this one poor quality study. The second study of semen parameters included four groups: patients with unilateral cryptorchidism who were successfully treated with hormonal therapy in childhood; patients with unilateral cryptorchidism treated with orchiopexy; patients with bilateral cryptorchidism treated with orchiopexy in childhood and; adult patients with untreated unilateral cryptorchidism.⁵³ Among the patients with unilateral cryptorchidism, the group treated with orchiopexy had similar sperm concentrations to untreated patients (17.0±2.9 vs. 16.6±4.6 millions ml⁻¹± standard deviation, respectively) or those who were treated with hormonal therapy (15.1±5.8) but had a higher proportion of patients with normal sperm motility (57.2 percent ± 21.4, statistical comparisons not reported).