Endocrine Reviews 18 (2): 259-280
Copyright © 1997 by The Endocrine Society
Anatomical and Functional Aspects of Testicular Descent and Cryptorchidism1
John M. Hutson,
Suzanne Hasthorpe and
Chris F. Heyns
F. Douglas Stephens Surgical Laboratory, Royal Children’s Hospital
Research Foundation, and Department of Paediatrics, University of
Melbourne Urology Department, Parkville, Victoria, Australia (J.M.H.,
S.H.); and Tygerberg Hospital and Faculty of Medicine, University of
Stellenbosch, South Africa (C.F.H.)
 |
Abstract
|
|---|
- I. Normal Development
- A. Anatomical aspects
- 1. Sexual development
- 2. The gubernaculum
- 3. Cranial suspensory ligament
- 4. Abdominal pressure
- B. Hormonal control and functional aspects of testicular descent
- 1. Müllerian inhibiting substance
- 2. Androgen
- 3. The genitofemoral nerve (GFN)
- 4. Calcitonin gene-related peptide (CGRP)
-
- II. Cryptorchidism
- A. Etiology
- B. Frequency
- C. Are some UDT acquired?
- D. Risks of infertility/malignancy
- E. Role of hormone therapy
- F. Timing of surgery
|
I. Normal Development
|
|---|
A. Anatomical aspects
TESTICULAR descent to the scrotum is a profound example of
sexual dimorphism that cannot be fully explained as yet. It is not a
simple process, but appears to be multistaged, with various anatomical
factors and hormonal influences. During mammalian evolution the male
gonad has assumed a progressively lower position relative to that of
the ovary, eventually taking up an extraabdominal location within a
scrotum in most modern mammals (1, 2). The exact site and structure of
the scrotum varies among species with, for example, some modern
marsupials having a prepenile scrotum (3). The essential physiological
feature, however, is that the scrotum is a specialized, low-temperature
environment that allows the extraabdominal testis to be maintained at a
temperature below that of the rest of the body (4).
1. Sexual development. The urogenital ridge in humans is
identical morphologically in males and females up to 7–8 weeks of
gestation (5). Sexual differentiation begins with the
testis-determining gene (SRY) on the Y chromosome triggering testicular
differentiation by unknown mechanisms (6). Recent studies suggest a
role for steroidogenic factor 1 (the Ad4-binding protein) in gonadal
development (7, 8) and regulation of the Müllerian inhibiting
substance (MIS) or anti-Müllerian hormone gene (9). MIS itself
also has been proposed to have a role in gonadal differentiation (10).
Once a testis is formed, both MIS and testosterone are involved in
altering the anatomy of the male embryo (11, 12). MIS causes regression
of the Müllerian ducts whereas testosterone secretion directly
into the Wolffian duct permits its continuing development into
epididymis, vas deferens, and seminal vesicles (13, 14).
In human embryos between 10 to 15 weeks of gestation, the testis
remains close to the future inguinal region during enlargement of the
abdominal cavity while the ovary moves relatively more cranially (5).
Similar relative movement is observed in fetal mice between 14 and 18
days of gestation on scanning electron microscopy (EM) (15) (Fig. 1
) and in fetal rats (16). There have been numerous
suggestions that no actual movement of the testis occurs (17, 18),
although quantitative assessment of fetal mice confirms relative
movement (15) (Fig. 2).
View larger version (166K):
[in this window]
[in a new window]
|
Figure 1. Scanning EM of male and female fetal mice. A, Day
14 male. Primitive testis (PT) and kidney (K) are situated near bladder
neck (BN). Mesonephros (ME) is connected to inguinal region by
gubernaculum (G). B, Day 14 female. Primitive ovary (PO) and kidney (K)
are located near bladder neck (BN) as in male. Gubernaculum (G) is also
similar to the male. C, Day 18 male. The testis (T) is near bladder
neck (BN) while the kidney (K) is now higher. The gubernacular bulb
(GB) is enlarged. D, Day 18 female. The ovary (O) and kidney (K) are
both high in the abdomen. The Müllerian ducts (M) are developing
into the uterus while the gubernaculum (G) remains long and thin.
[Reproduced with permission from T. Shono et al.:
J Urol 152:781–784, 1994 (15). © Williams and
Wilkins.]
|
|
View larger version (21K):
[in this window]
[in a new window]
|
Figure 2. Distance from lower pole of gonad to bladder neck
(mm) vs. age of gestation (days). Positive
numbers reflect gonads cranial to bladder neck (n = number
of gonads measured). [Reproduced with permission from T. Shono
et al.: J Urol 152:781–784, 1994
(15) © Williams and Wilkins.]
|
|
The testis is anchored near the future inguinal canal by enlargement of
the caudal ligament of the testis, known as the gubernaculum, and
regression of the cranial suspensory ligament. The gubernaculum was
described first in 1762 as the genitoinguinal ligament or
’gubernaculum,’ because it appeared to direct the course of the
testis to the scrotum (19). In this century, enlargement of the
gubernaculum in males was observed to tether the testis near the groin
while the kidney migrated cranially (20, 21). Simultaneously,
regression of the cranial ligament holding the urogential tract near
the developing diaphragm allows gonadal descent (22) (see Fig. 1
).
The gonadal positions deviate further after 25 weeks of gestation, when
the gubernaculum bulges beyond the external inguinal ring and descends
to the scrotum, while simultaneously it is hollowed out by a peritoneal
diverticulum called the processus vaginalis (23, 24) (Fig. 3). The processus vaginalis allows the previously
intraabdominal testis to exit from the abdominal cavity (18). The bulky
caudal end of the gubernaculum, which is known as the bulb, is resorbed
after completion of migration in humans (24) and pigs and before
migration in rodents (25). The differences in timing of matrix
resorption among species are unexplained but do not change the
fundamental anatomical processes.
View larger version (134K):
[in this window]
[in a new window]
|
Figure 3. Dissection of 32-week human fetus showing the
testis (T) and gubernaculum (G) migrating across the pubic region
toward the scrotum (S). A pair of forceps holds the caudal end of the
gubernaculum.
|
|
Multistaged testicular descent was proposed first by Gier and Marion
(26) as initial downward displacement of the gonads by the developing
metanephros, transabdominal movement to the groin, and finally descent
through the inguinal canal and down to the scrotum. By 7 weeks of
gestation the first step is complete, thereby precluding it having a
major role in sexual dimorphism. Separate phases of testicular descent
have been proposed by other authors (27, 28, 30); however, a two-stage
model with separate morphological steps and hormonal control has been
suggested more recently (31, 32) (Fig. 4).
View larger version (19K):
[in this window]
[in a new window]
|
Figure 4. Schema proposing two main steps of testicular (T)
descent in the human. Between 8 and 15 weeks the gubernaculum (G)
enlarges in the male, holding the testis near the groin. The cranial
suspensory ligament (CSL) regresses. At 28–35 weeks, the gubernaculum
migrates across the pubic region to the scrotum. [Reproduced with
permission from J. M. Hutson and S. W. Beasley: Descent of the
Testis, 1992 (149).]
|
|
2. The gubernaculum. Caudal enlargement of the gubernaculum
during relative transabdominal movement of the testis is known as the
"swelling reaction" or "gubernacular outgrowth" and is caused
by cell division and an increase in glycosaminoglycans and hyaluronic
acid (33). The hydrophilic nature of hyaluronic acid makes the end of
the gubernaculum bulky and gelatinous. Subsequently, the gubernaculum
involutes, presumably by removal of the extracellular matrix, leaving a
fibrous remnant that attaches the testis and caudal epididymis to the
scrotum after descent.
The proximal gubernacular cord appears to shorten during testicular
descent, as it becomes incorporated into the enlarging bulb (34, 35, 36).
Shortening of the cord may be an important part of the mechanism of
positioning the testis over the inguinal ring to permit abdominal
pressure to push the testis out of the abdomen (18, 37, 38, 39).
Transection of the cord frequently leads to either accidental gonadal
descent into the contralateral inguinal canal or to aberrant
intraabdominal sites (38, 40, 41, 42). For example, abnormally long
gubernacular cords associated with intraabdominal testes have been
found recently in transgenic mice with a mutant Hoxa-10 gene
(43, 44). Both the cord and the bulb are sites of strong
Hoxa-10 expression, both during sexual differentiation and
postnatally in the male.
The inguinoscrotal phase of descent requires significant movement of
the gubernaculum. Regression of the bulky gubernacular bulb, which is
particularly prominent in the pig, was thought to be sufficient to
allow the testis to reach the scrotum (34, 35, 36). A significant migratory
phase of the gubernaculum in humans (23) and rodents (25) has been
demonstrated recently, although this is disputed by some authors
in favor of inversion/eversion of the gubernacular cone, as seen in
rodents (45, 46, 47). The caudal end of the gubernaculum extends
progressively from the inguinal ring across the pubic bone and into the
scrotum. This migration cannot be explained by eversion of the
gubernacular cone, as its length is significantly less than the
distance to the scrotum (S. Lam, T. Clarnette, and J. M. Hutson,
unpublished results).
3. Cranial suspensory ligament. The cranial suspensory
ligament regresses in male embryos to allow normal gonadal descent
(22). Whether or not it is the key factor in descent, however, is
disputed. Most authors agree that the cranial ligament, and probably
also the gubernacular cord, has a limited role in at least some species
(37, 48, 49).
4. Abdominal pressure. Abdominal pressure has an ancillary
role in facilitating gonadal exit from the abdomen (18, 39, 41). While
intraabdominal pressure is not a factor during transabdominal descent,
it assumes much greater importance in transit through the inguinal
canal and subsequent migration to the scrotum (26, 40, 41, 45, 46).
Pressure to push the gonad down the processus vaginalis is likely to be
important in males but nevertheless does not occur in females because
of their different gubernacular and cranial suspensory ligament anatomy
(37). The force of intraabdominal pressure not only may be applied
directly to the testis, but also may act indirectly by creating the tip
of the processus vaginalis and thereby stabilizing it so that the
gubernaculum can exert some traction on the testis (49).
In summary, although controversial, current knowledge suggests
that transabdominal testicular descent is associated with regression of
the cranial suspensory ligament and enlargement of the gubernacular
bulb with traction applied by the bulb through the gubernacular cord to
the urogenital ridge. The net effect of these processes appears to be
to anchor the developing testis near the inguinal region during growth
of the fetal abdomen (Fig. 5). Passage through the
inguinal canal needs development of the processus vaginalis, prior
dilation of the canal by the gubernacular bulb, and some intraabdominal
pressure to force the testis through the canal. Inguinoscrotal descent
requires migration of the gubernaculum over a considerable distance
compared with its size, along with an increase in length of the
processus vaginalis. The force for movement may come from the
intraabdominal pressure, transmitted directly and indirectly to the
testis via the lumen of the processus vaginalis and the gubernacular
cord, respectively. The factor(s) controlling the direction of
migration remain unknown.
View larger version (25K):
[in this window]
[in a new window]
|
Figure 5. Schema showing cranial suspensory ligament (CSL)
and gubernaculum (G) in sexual differentiation of rodents. Both
ligaments are present in the indifferent stage. In males, the CSL
regresses while the gubernaculum enlarges. By contrast, in females the
CSL persists and the gubernaculum also remains, but as thin and
elongated.
|
|
B. Hormonal control and functional aspects of testicular descent
Regulation of transabdominal descent centers around the control of
gubernacular enlargement and regression of the cranial suspensory
ligament. A detailed historical review of the endocrine factors that
may be involved is available (18); therefore the emphasis here is on
recent developments. In 1985 it was suggested that the swelling
reaction in the gubernaculum was not under androgenic control because
it occurred normally in both mice and humans with complete androgen
resistance (31) and also in rats exposed prenatally to flutamide (50).
Because estrogen-treated fetal male mice and rats had retained
Müllerian ducts, atrophy of the gubernaculum, and high
intraabdominal testes (50, 51), a causal link between MIS and the
gubernacular swelling reaction was suggested (32, 52, 53) (Fig. 6).
View larger version (148K):
[in this window]
[in a new window]
|
Figure 6. Scanning EM of day 20 rat fetuses exposed to
flutamide (100 mg/kg per day 16–19) or estradiol benzoate (15 mg/day
16). A, Control male. Testes (T) have descended from kidney (K) to
lower abdomen near bladder neck (BN). The gubernacular bulb (GB) and
cord (GC) are visible. B, Flutamide-treated male, with testes in a
similar position to that in control, despite persistence of cranial
suspensory ligament (white arrow). Note well-developed
gubernacular bulb (GB). C, Estrogen-treated male with high testes near
kidney. The retained Müllerian ducts (M) are adjacent to the
epididymis (E). The gubernacular bulb (GB) and cord (GC) are longer and
thinner than in controls. D, Control female with ovary (O) just under
lower pole of kidney (K). Colon (C) is seen between Müllerian
ducts and elongated gubernacula (G). [Reproduced with permission from
T. Shono et al.: Int J Androl
19:263–270, 1996. © Blackwell Science Ltd.]
|
|
1. Müllerian inhibiting substance. MIS (or
anti-Müllerian hormone) is a 140-kDa glycoprotein produced by
Sertoli cells and is responsible for regression of the embryonic
Müllerian ducts (10, 11, 12). The gene for MIS has been cloned for
the human (54) cow (54), rat (55), and mouse (56). It contains five
exons and is localized to chromosome 19p13.3 (57). Cleavage of
the glycosylated dimer by a protease produces a carboxy terminus dimer
that is still biologically active (58). Although once thought to be a
paracrine factor in the male fetus (59), it is now known to be produced
postnatally by both testis and ovary (60). A number of functions have
been suggested for MIS in addition to regression of embryonic
Müllerian ducts, including early differentiation of the testis
(11, 61), prenatal lung maturation (62), and postnatal germ cell
maturation (63, 64, 65).
A number of observations support a role for MIS in the first phase of
testicular descent (Table 1A). First, animal models with
intraabdominal cryptorchidism also have retained Müllerian ducts
(52, 53, 66) (Fig. 6C). Second, gonadal maldescent is proportional to
Müllerian duct retention in humans with intersex (67, 68) as well
as in estrogen-treated fetal mice (53). Third, in humans with genetic
defects in the MIS gene or its receptor with so-called persisting
Müllerian duct syndrome (PMDS) (69), the testes are undescended
and the gubernaculum is thin and enlongated (70) (Fig. 7). The latter defect suggests that the gubernacular
swelling reaction fails to occur in PMDS, leading to cryptorchidism
(70, 71). Transgenic mice with MIS deficiency have retained
Müllerian ducts and variable gonadal position depending on their
androgenic status: those with normal androgen receptors have almost
normally descended testes while those with combined androgen resistance
have completely undescended testes (UDT) (72).
View larger version (156K):
[in this window]
[in a new window]
|
Figure 7. Surgical photograph of a patient with PMDS
presenting with bilateral UDT and an inguinal hernia containing both
testes (T) and Müllerian ducts (MD). Note the fact that the
gonads and uterus are free inside the hernial sac, and the absence of a
normal gubernacular bulb or any attachment of either testis to the
scrotum. [Reproduced with permission from J. M. Hutson et
al.: Pediatr Surg Int 2:191–194,1987 (71).]
|
|
Arguments against a role for MIS in the gubernacular swelling reaction
include the failure of fetal rabbits immunized against bovine MIS to
have UDT, despite persistence or partial persistence of the
Müllerian ducts (73) (Table 1B). In addition, semipurified bovine
MIS failed to cause cell division of cultured fibroblasts from the
fetal pig gubernaculum, suggesting that MIS does not stimulate the
swelling reaction (74). A further argument against a role for MIS is
the view that the intraabdominal testes in patients with PMDS are
caused, not by absence of the gubernacular swelling reaction, but by
anatomic connection of the testis with the persistent Müllerian
ducts (12, 75). Cryptorchidism was believed to be caused by anatomic
blockade by the retained Müllerian ducts, a view supported by
Husmann and Levy (37), who also cite a final argument against MIS,
which is the finding that nearly all patients with intraabdominal
cryptorchidism do not have persistent Müllerian ducts (76).
Studies of the gubernaculum in intersex and normal pigs have indicated
that growth of the gubernaculum may be stimulated by a nonandrogenic
hormone from the testis (27, 77, 78, 79, 80, 81). Proliferation of pig fetal
gubernaculum in culture appears to be stimulated by a low molecular
weight testicular factor that is distinct from known polypeptide growth
factors and MIS (74, 82). Fentener van Vlissingen et al.
(74) proposed the name "descendin" for this hormonal activity,
whereas Visser and Heyns (82) suggested that, in line with the names of
other trophic hormones (e.g. gonadotropin, thyrotropin),
"gubernaculotropin" would be a more appropriate name for the
putative hormone responsible for the proliferation of gubernaculum
cells.
Several factors may account for the above contradictory opinions, not
the least of which is that most studies focus on gonadal position while
few examine the swelling reaction in the gubernaculum, which is the
likely target organ for hormonal action. Also, despite strong evidence
that it is an important accessory factor, the role of the cranial
suspensory ligament has been ignored (22). The cranial ligament is
responsive to androgens, which induce its regression (83, 84). Because
gonadal position appears to be the net result of the opposing actions
of the cranial suspensory ligament and the gubernaculum, it cannot be
assessed without considering these opposing actions (Fig. 8).
View larger version (15K):
[in this window]
[in a new window]
|
Figure 8. Schema showing the gonadal position as a vector
summation of the opposing forces of the cranial suspensory ligament and
gubernaculum. Evidence would suggest that testosterone causes
regression of the cranial ligament and MIS is proposed as the possible
cause of gubernacular enlargement. Without gubernacular swelling, the
ovary is held high in the abdomen by the cranial ligament. In the male,
the enlarged gubernaculum holds the testis low near the groin.
|
|
In the normal male, the gubernacular outgrowth and regression of the
cranial ligament act in concert to permit the gubernaculum to hold the
testis near the inguinal region. By contrast, in the female, the
cranial ligament holds the ovary higher in the abdomen, and the
gubernaculum remains long and thin as the round ligament and ligament
of the ovary. Complete androgen resistance in the mouse (or human) does
not prevent gubernacular swelling but does prevent regression of the
suspensory ligament; as the testes descend to the bladder neck, it can
be inferred that the suspensory ligament cannot overcome the traction
of the gubernaculum (31, 85).
In the estrogen-treated, androgen-resistant mouse the testes are high
in the abdomen at birth, similar in position to ovaries (52) (Fig. 9). The gubernacular swelling reaction is absent, and
the unopposed cranial suspensory ligament is now able to keep the
testis and ducts high in the abdomen. This gonadal position is similar
in the combined MIS/TFM mutant, in which the testis, with retained
Müllerian duct, is near the kidney (72). The cranial suspensory
ligament alone apparently anchors the gonad in this model, although the
status of the gubernacular swelling reaction was not described.
View larger version (70K):
[in this window]
[in a new window]
|
Figure 9. A, Untreated neonatal male mouse with complete
androgen resistance (TFM). The testes are in the normal position near
the bladder neck despite absence of Wolffian duct structures. The
gubernacular swelling has occurred (big arrow). B,
Neonatal male TFM mouse treated with estradiol benzoate in
utero. Testes are located high in the abdomen near the kidneys.
The Müllerian ducts (M) are retained and the gubernaculum is long
and thin (arrow) as in normal females. [Reproduced with
permission from J. M. Hutson: Pediatr Surg Int
2:242–246, 1987 (52).]
|
|
In the MIS-deficient mouse, the testes appear to be normally descended
(72). However, here the cranial suspensory ligament is absent (as
androgen function is normal), and whether the swelling reaction occurs
in the gubernaculum is not yet known, but hypermobility of the gonads
could allow accidental descent, which is observed in the equivalent
human mutant with PMDS.
Recent studies of the anatomy of an MIS receptor-deficient transgenic
mouse show normal transabdominal testicular descent but lack the
swelling reaction with minimal deposition of extracellular matrix in
the mutant gubernaculum. (R. R. Behringer and J. M. Hutson,
unpublished). In addition, the MIS receptor deficiency fails to prevent
cell division within the gubernaculum, consistent with the view that
this is not dependent on MIS (74, 82). Further detailed analysis of the
MIS-deficient and receptor-deficient transgenic mouse models is needed,
but these preliminary observations suggest that MIS may indeed control
deposition of matrix in the gubernacular swelling reaction.
Mutations of both the MIS gene and its receptor lead to PMDS in the
human (75, 86). The gonadal position in this syndrome is variable, but
all have retention of the Müllerian ducts and normal
masculinization of androgens (71, 87). A review of the literature
reveals that most testes are intraabdominal, while in some patients one
or both testes may be found in an inguinal hernia (70, 71). A recent
surgical patient was observed to have a very long, thin gubernacular
cord similar to the round ligament in females (70). This led to the
suggestion that the transverse testicular ectopia commonly seen in this
syndrome is caused by prolapse of one or both testes into the patent
processus vaginalis (88) (Fig. 10). Hypermobility of
intraabdominal testes in PMDS is supported by the high frequency of
bilateral torsion and involution of the testes (89). Rather than the
retained Müllerian duct blocking descent (12, 75), the
hypoplastic uterus and tubes appear to play no role; meanwhile, the
absence of tight anchoring by either cranial or caudal gonadal
ligaments allows accidental descent, transverse ectopia, or gonadal
torsion. Nearly all surgical photographs of PMDS show extremely mobile
gonads and elongated ducts (90), suggesting that blockade of the MIS
gene increases, rather than restricts, gonadal mobility. A telling
factor in the argument in favor of MIS causing the gubernacular
swelling reaction is its apparent absence in PMDS.
View larger version (33K):
[in this window]
[in a new window]
|
Figure 10. Schema of gonadal descent showing two phases in
normal males compared with those with PMDS and normal females. Normal
development of the processus vaginalis, which is part of the second
phase of descent, along with hypermobility of the testis because the
gubernaculum is long allows the testis (± ducts) to prolapse into the
groin. [Reproduced with permission from J. M. Hutson and M. L. Baker:
Pediatr Surg Int 9:542–543, 1994 (88).]
|
|
The failure of most patients with intraabdominal testes to have
Müllerian duct remnants does not preclude a role for MIS, as
suggested by Husmann and Levy (37). There are numerous potential
anatomical anomalies, such as the atrophic gubernaculum found in the
recent transgenic Hoxa-10 mutant mouse, that could be present in the
gubernaculum or developing inguinal canal that could prevent the gonad
leaving the abdominal cavity (43, 44).
2. Androgen. It is generally agreed that the inguinoscrotal
phase of testicular descent requires androgen (18, 37, 76, 91).
Migration of the gubernaculum beyond the inguinal region is absent in
gonadotropin-deficient animals (92) and those with complete androgen
resistance (85). About 50% of animals treated prenatally with the
antiandrogen flutamide also have deranged gubernacular migration and
delayed regression (50, 93, 94, 95, 96, 97) (Fig. 11). Regression
of gubernacular bulk also appears androgen-dependent since in the human
with complete androgen resistance the gubernaculum remains enlarged,
with failure of extracellular matrix resorption (32, 47). Recent
studies suggest that regression of the gubernacular cord may also be
androgen dependent, as the prenatal flutamide treatment prevented its
regression (98).
View larger version (113K):
[in this window]
[in a new window]
|
Figure 11. A 30-day-old flutamide-treated rat after 100
mg/kg/day on days 16–19 of fetal development. The right testis (T) is
undescended in the suprainguinal position. (Unilateral cryptorchidism
is common in this model and is unexplained.)
|
|
The site of androgenic action on the gubernaculum remains
controversial. High local concentrations are thought to be required
(17, 46). The gubernaculum of the newborn rat, however, contains only
20% of the level of androgen binding
([3H]dihydrotestosterone) seen in the urogenital sinus,
which is a known target organ (99). Fibroblast cultures from fetal pig
gubernacula also are reported to contain binding for androgen analogs
(3H-R1881) (100), although other studies have found that
the total androgen receptor concentration in the fetal pig gubernaculum
was much less than in a recognized androgen target organ such as the
prostate (101). In a more detailed study of androgen binding in the
gubernaculum of fetal pigs, Heyns and Pape (102) found
[3H]methyltrienolone (3H-R1881) binding in
the gubernaculum was of lower affinity and capacity than in prostate,
and on a par with that in striated muscle (male or female). They
concluded that direct androgen stimulation may not account for growth
of the gubernaculum during descent. In addition, 5
-reductase
activity in the gubernaculum remained constant throughout gestation in
the pig, at levels well below that seen in the prostate and urethra
(103).
The possibility that androgens may act by indirect means on the
gubernaculum was considered after a literature review identified a 1948
study by Lewis (21), who transected the genitofemoral nerve (GFN) while
testing the traction theory. At that time the cremaster was being
considered as a source of muscular traction to pull the testis into the
scrotum. Neonatal nerve transection would theoretically prevent
traction by denervation of the muscle: cryptorchidism was produced, but
the study was not followed up until many years later (104). It was
intriguing that denervation should block a process that is under
apparent androgenic control, leading to the speculation that androgens
may act via the nerve (105). This hypothesis was consistent with the
results of distal transection of the gubernaculum in neonatal rats,
which prevents postnatal migration of the gubernaculum and testicular
descent (42, 106). The GFN supplies the gubernaculum from its posterior
and caudal surface, so that distal transection of it would also cause
denervation (107, 108). Proximal transection of the neonatal
gubernaculum, which preserves its inguinal attachment and nerve supply,
failed to block inguinoscrotal descent (42, 106).
This "GFN hypothesis" stimulated a number of new studies to test
its predictions. Androgens should lead to modification of the GFN
prenatally, as even in species with postnatal gubernacular migration,
the nervous system would require earlier modification. Studies with
antiandrogens confirm that inguinoscrotal migration, although occurring
postnatally in rodents, can only be blocked by prenatal treatment (50, 94, 95, 96, 97, 98). This coincides with the time of development of other sexually
dimorphic nuclei (109).
3. The genitofemoral nerve (GFN). The spinal nucleus of the
GFN is located at L1–2 in the spinal cord and is sexually dimorphic in
rodents (97, 110). Prenatal blockade of androgens with flutamide does
inhibit the sexual dimorphism (97) (Table 2). Rather
than induce feminization of the nucleus in the male, flutamide appears
to partially masculinize the nucleus in the female (97). Other authors
have claimed that androgens had no effect on the GFN nucleus, as the
nucleus in TFM rats was found to be of similar size as in normal male
rats (48). This is similar to our own results in the TFM mouse (110)
(Table 2).
The location of androgen receptors that might mediate androgenic
modification of the genitofemoral nerve remains controversial (37),
since androgen binding was not reported in the rat spinal cord until 1
week after birth.
A specific search for androgen receptors in the fetal rat spinal cord
has recently shown androgenic binding as early as 15 days of gestation,
which would be consistent with androgens stimulating differentiation of
the GFN and its nucleus (111). Alternatively, the androgen receptors
affecting the gubernaculum may be in the cremaster muscle itself,
similar to the situation in the bulbocavernosus muscle (37, 112),
although this has not been supported by studies in the pig fetus (103).
The fetal rat gubernaculum, however, does show staining with antiserum
raised against the androgen receptor at 18 and 20 days post coitum and
1 day after birth but declines rapidly thereafter (96). During the
migration phase of the rat gubernaculum, between 3–10 days postnatally
(25), the developing cremaster muscle shows minimal staining for
androgen receptors (96).
The GFN hypothesis predicts that neuronal anomalies affecting the GFN
nucleus should be associated with cryptorchidism. In 345 boys with
spina bifida, in whom the position of the testes and the neurological
defect were recorded, 23% had UDT. Cryptorchidism was present in 19%
of 186 boys with lesions below L4 compared with 36% of 59 boys with
high lumbar defects (113). In rats undergoing transection of the spinal
cord at birth, 39% had cryptorchidism (7/18) with midlumbar
transections (113, 114).
4. Calcitonin gene-related peptide (CGRP). Anatomical studies
of the GFN in neonatal rodents identified CGRP in the nerve (97, 110).
CGRP is a neuropeptide with close homology with calcitonin and is
produced by alternative splicing of the calcitonin RNA transcript
(115). Numerous studies in recent years have demonstrated multiple
sites and functions for this peptide in the nervous system (116). In
motor nerves it is involved in regulation of acetylcholine receptor
synthesis in skeletal muscle (117). It has an important role in pain
perception in sensory nerves (118) while in the autonomic system it has
different effects on smooth muscle, including peripheral vasodilation
(119, 120, 121).
Sexually dimorphic nuclei, such as the bulbocavernosus spinal nucleus
which contains CGRP, may be regulated by retrograde transport of a
soluble factor(s) from the bulbocavernosus muscle (112). Castration of
adult male rats stimulates the expression of CGRP mRNA and increases
the amount of CGRP in the cell bodies of the bulbocavernosus spinal
nucleus (112). The mature GFN or cremaster spinal nucleus in adult rats
contains only a few CGRP-positive cell bodies (122), and neonatal
castration leads to a decrease in the size of the GFN nucleus in adult
rats (123), which is in contrast to the bulbocavernosus nucleus. The
adult GFN nucleus contains neuropeptide Y immunoreactivity (124) and
progressive postnatal connections with nerve endings containing
substance P and 5-hydroxytryptamine (125). The mature male rat has a
GFN nucleus with 3-fold more cell bodies than the female (122), but the
androgen-resistant (TFM) rat (King-Holzman) has similar nuclear size
and connections as a normal male. By contrast, the bulbocavernosus
nucleus is significantly feminized (126).
CGRP has been localized in the soma of GFN neurons in the first and
second lumbar segments (97, 110, 127), using both immunofluorescence
and immunohistochemistry (Fig. 12). The total number of
neurons in the nucleus was determined by retrograde labeling of the GFN
with fluorescent dyes (diamidinophenyl indole and Fast Blue): neonatal
rats/mice underwent laparotomy, exposure of the GFN in the
retroperitoneum where it runs caudally on the anteromedial border of
the psoas muscle, transection of the nerve, and application of a few
dye crystals to the cut ends. The spinal cord was removed after a
further 48 h (Table 2). Vasoactive intestinal peptide, TRH,
neuropeptide Y, met-enkephalin, 5-hydroxytryptamine, substance P, and
somatostatin-8 were not present above background levels. The reason for
these differences compared with Newton’s work on the cremaster nucleus
is not known. Several possible explanations include alteration of the
GFN nucleus at sexual maturity, differences in different strains of
rats, or the fact that the GFN nucleus in the neonatal animal, as
identified by retrograde labeling, is different from the cremaster
nucleus localized by its synapses in adult rats. The presence of CGRP,
but not neuropeptide Y, in the neonatal nucleus and a difference in the
sex ratio of nerves suggests that either the neurons undergo major
changes at puberty or are not identical in location.
View larger version (94K):
[in this window]
[in a new window]
|
Figure 12. A, Retrograde Fast Blue fluorescent labeling of
motor neurons in the genitofemoral nerve (GFN) spinal nucleus of a
neonatal rat (x400). B, CGRP-like immunoreactivity of cell bodies in
GFN nucleus (different section from panel A) (x400) [Reproduced with
permission from D. W. Goh et al.: J Reprod
Fertil 102:195–199, 1994 (127).]
|
|
One argument against the GFN nucleus and CGRP having a role in descent
is the notion that the cremaster muscle, at least after sexual
maturity, controls neural morphology rather than the nerve regulating
the target organ (37, 123, 128). Regardless of whether the cremaster
controls or is controlled by the GFN, the neonatal gubernaculum
responds as if the GFN is the primary regulator (see below). Other
views include the fact that CGRP is known to be a neuromuscular
transmitter (129) and that CGRP receptors are localized to the
developing rat cremaster (130), whereas the primate gubernaculum is
mostly mesenchyme and extracellular matrix rather than muscle (23). In
addition, in some dimorphic nuclei CGRP is released by withdrawal of
androgens rather than stimulation (112, 131). These views are
reasonable conclusions by analogy with the bulbocavernosus nucleus, but
are contrary to what has been found in the GFN and rodent gubernaculum.
How they can be applied to the human remains speculative.
Rodent models with cryptorchidism have abnormalities in their GFN
nucleus that may support its putative role in testicular descent. As
mentioned previously, the TFM rat GFN is not feminized (48), which is
similar to what our own studies had shown for the TFM mouse (110)
(Table 2). However, the gubernaculum of TFM mice, which is
hypersensitive to CGRP in organ culture (see below), is consistent with
failure of CGRP to be released in the TFM. In the flutamide-treated rat
the male GFN had a small decrease in neuron number and CGRP-positive
cells, but the female rat had a similar number as the male (97) (Fig. 13A). Although the nucleus in male flutamide-treated
rats was not the same as normal females, the gubernaculum responded as
if the nerve was not releasing CGRP, as it became hypersensitive to the
exogenous CGRP added to the culture (see below). In a new rat model of
cryptorchidism, androgen function appears normal (132), but the scrotum
was initially thought to be ectopic, hence the name TS or
"trans-scrotal" rat. Dissections of this rat during development
have shown, however, that the scrotum is not ectopic but normally
located, hypoplastic, and empty (133). The GFN nucleus in TS rats has
similar numbers of neurons in males and females and increased numbers
of CGRP-immunoreactive neurons compared with normal controls (127)
(Fig. 13B). The target organ, which is the TS gubernaculum, behaves as
if it is desensitized and is not responsive to exogenous CGRP in
culture (see below).
View larger version (26K):
[in this window]
[in a new window]
|
Figure 13. A, Fast Blue-labeled and CGRP-immunoreactive
neurons in GFN spinal nuclei of flutamide-treated neonatal rats (n
= 16 each sex) vs. control rats (n = 18 each sex).
Bars, mean ± SD Males, filled
columns; females: open columns; comparisons
between bars (#; +; * all significant (P < 0.001)
[Reproduced with permission from D. W. Goh et al.:
J Pediatr Surg 29:836–838, 1994 (97).] B, Fast
blue-labeled and CGRP-immunoreactive neurons in GFN spinal nucleus of
TS rats (n = 14 each sex) vs. control rats (n
= 10 each sex) [Reproduced with permission from D. W. Goh et
al.: J Reprod Fertil 102:195–199, 1994
(127).]
|
|
The effect of CGRP on the rodent gubernaculum has been investigated
extensively in vitro and in vivo. In male
neonatal rats under anesthetic, the gubernaculum, which has not yet
reached the scrotum, shows spontaneous rhythmic contractility, which is
enhanced by increased intraabdominal pressure and direct application of
human CGRP (134). In organ culture, neonatal rat gubernacula contract
rhythmically in a dose-responsive manner on exposure to CGRP (Fig. 14A), but not to other neuropeptides such as vasoactive
intestinal peptide, serotonin, somatostatin 8, met-enkephalin, TRH, or
neuropeptide Y (134). Control tissues such as female rat gubernacula,
skeletal muscle from the abdominal wall, and umbilical cord showed no
rhythmic contractions. Rhythmic contractility also was pronounced in
the neonatal mouse gubernaculum, with some endogenous contractions and
dose response to CGRP (135). By contrast, exposure to the inhibitory
analog of CGRP, CGRP(8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37), caused suppression of contractions (Fig. 14B). In vivo CGRP(8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37) also caused inhibition of
gubernacular migration, with weekly injections of 25 µl
10-4 molar CGRP(8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37) causing delayed descent. By 2 weeks
of age 43% of the saline-treated controls already had descended testes
whereas all experimental animals had UDT. At 3 weeks, 17% of testes in
experimental animals were still undescended (136).
View larger version (15K):
[in this window]
[in a new window]
|
Figure 14. A, Number of contractile gubernacula from normal
neonatal mice (n = 50 at each dose level) vs. CGRP
concentration. Rhythmic contractions were recorded by video camera
connected to a dissecting microscope. A positive correlation between
CGRP dose and contractility was present (r = 0.95;
P < 0.05). B, Number of contractile gubernacula
from normal neonatal mice (n = 50 at each dose) vs.
CGRP (*-37) concentration. A negative correlation was present
(r = 0.98; P 0.05) [Derived from
Ref. 135.]
|
|
The contractility of the mouse gubernaculum observed in
vitro is maximal in the first postnatal week, which coincides with
gubernacular migration from the inguinal canal to the scrotum (137).
CGRP causes an increase in both frequency and amplitude of
contractions, with most gubernacula also responding with increased
isotonic tension. In response to acetylcholine, the gubernacula showed
a single twitch contraction, but no rhythmic contractions (138). The
contractility of the gubernaculum requires influx of calcium ions via
dihydropyridine receptors in a manner similar to that seen in immature
cardiac or smooth muscle; acetylcholine receptors are not involved, as
nifedipine was unable to block contractions (139).
The gubernacular contractility in vitro can be altered by a
change in the status of the GFN before removal from the animal. Prior
transection of the GFN at birth sensitizes the normal rat gubernaculum
to exogenous CGRP (138).
The TFM gubernaculum has no endogenous contractility in
vitro but is hypersensitive to exogenous CGRP, consistent with
diminished release of CGRP from the GFN (135) (Fig. 15A). In addition, exogenous CGRP can induce elongation
of the gubernaculum in intact TFM mice postnatally (Fig. 15B) (140).
Similarly, in the flutamide-treated rat the neonatal gubernaculum has
reduced endogenous contractility but is hypersensitive to exogenous
CGRP (141) (Fig. 16A). By contrast, the neonatal TS rat
gubernaculum remains inert in organ culture, even on exposure to
exogenous CGRP (141) (Fig. 16B). In vivo the TS gubernaculum
fails to respond to exogenous CGRP (W. H. Park and J. M. Hutson,
unpublished). The inert TS gubernaculum has a decreased number of CGRP
receptors, but these can be restored to normal number and the
gubernaculum can be restored to normal contractility by prior
transection of the GFN (142). This is consistent with the hypothesis
that cryptorchidism in this mutant may be secondary to excess CGRP
release from the nerve that disrupts inguinoscrotal migration by
down-regulation of CGRP receptors.
View larger version (21K):
[in this window]
[in a new window]
|
Figure 15. A, Contractile reponse to CGRP on neonatal
gubernacula from androgen-resistant (TFM) mice (n = 10 per dose)
[Derived from Ref. 141.] B, Length of the procesus vaginalis in
15-day-old normal male, female, and androgen-resistant male (TFM) mice
(n = 10) after injections of CGRP into the groin postnatally
(hatched bar) or control (black bar)
[Derived from Ref. 140.]
|
|
View larger version (33K):
[in this window]
[in a new window]
|
Figure 16. A, Neonatal rat gubernacular contractile response
to CGRP 715 nmol/liter (hatched bars) vs.
no added CGRP (black bars) in normal control rats and
rats exposed to flutamide in utero. Flutamide
induced a significant increase in contractility to CGRP compared
with untreated rats (P < 0.05) [Derived from Ref.
141.] B, Neonatal rat gubernacular contractile reponse to CGRP 715
nmol/liter (hatched bars) vs. no added
CGRP (black bars) in sham-operated and GFN-transected
normal and TS rats. Contractile response with CGRP is enhanced in
control rats after denervation (+, P < O.01).
Sham-operated TS rats have suppressed response to CGRP (*,
P < 0.01), but this recovers significantly after
denervation (#, P < 0.01). [Derived from Ref.
142.]
|
|
Specific binding sites for CGRP have been found on the developing
cremaster muscle fibers within the gubernaculum (143). Sections of
gubernacula from neonatal male rats were incubated with
125I-labeled CGRP (human) and various cold neuropeptides,
with quantification of the autoradiographs by computer densitometry.
The binding was concentrated over the cremaster muscle whereas the
mesenchymal component was not labeled. A single class of high-affinity
receptors was found, which was maximal during the first week after
birth, coincident with gubernacular migration (130). CGRP and CGRP
[8–37] showed high-affinity specific binding and calcitonin and CGRP
[28–37] bound with low affinity (Fig. 17A).
Serotonin, substance P, vasoactive intestinal peptide, and somatostatin
showed no specific binding. When binding was compared between normal
rats and the flutamide-treated and TS rats, the flutamide-treated rat
had higher CGRP binding in the gubernaculum, while the TS rat had lower
binding (Fig. 17B). Prior transection of the GFN increased the binding
capacity of the gubernaculum (130). Furthermore, GFN transection in the
TS rat was able to restore CGRP binding, which initially was
significantly lower than in normal rats, to the normal levels (142).
View larger version (16K):
[in this window]
[in a new window]
|
Figure 17. A, Specificity of binding displacement of
125I-labeled human CGRP (hCGRP) to neonatal rat
gubernaculum by unlabeled hCGRP (open circles) rat CGRP
([8–37]) (filled triangles), rat CGRP [28–37]
(filled squares), salmon calcitonin (filled
circles), vasoactive intestinal peptide (open
squares), somatostatin (open triangles),
substance P (+), serotonin (x). Binding (B) is expressed relative to
binding in the absence of unlabeled hCGRP (Bo). [Reproduced with
permission from J. Yamanaka et al.:
Endocrinology 132:1–5, 1993 (130). © The Endocrine
Society] B, Total 125I-labeled hCGRP (50 pmol/liter)
binding with various concentrations of unlabeled hCGRP.
(squares, flutamide rat; circles, normal
rat; triangles, trans-scrotal (TS) rat. Each
point represents 10–12 sections. [Reproduced with
permission from M. Terada et al. J Urol 152:759–762,
1994. © Williams and Wilkins.]
|
|
A number of issues remain unresolved when considering a possible
role for CGRP. For instance, CGRP is described as a neuromuscular
transmitter, which would not be expected to be important in migration
of the human gubernaculum, where there is very little cremasteric
muscle (37). The data from the above mentioned studies, however, show
that CGRP acts independently of acetylcholine receptors. In addition,
CGRP has so many effects in different systems, including chemotaxis of
cells (144, 145, 146), that alternative mechanisms other than contractility
may be applicable in the human. In some sexually dimorphic nuclei, CGRP
release is increased by androgen blockade (122), but the cryptorchid
rodent models show a clear association between androgen resistance or
blockade and increased gubernacular sensitivity to CGRP, which is
consistent with decreased release from the GFN. It remains unknown why
flutamide-treated rats have about 50% of testes descended, despite
almost complete blockade of prostatic development (96).
Houle and Gagne (147) attempted recently to induce premature
testicular descent in postnatal mice by exogenous CGRP. The
neurotransmitter inhibited descent in this model, which was similar to
our own findings (148). Normal production of CGRP by the GFN, as well
as failure to replicate regular release of the neuropeptide by the GFN,
may account for failure to induce descent prematurely. When a flutamide
rat was used, the direction of gubernacular migration could be diverted
by ectopic CGRP administration (148).
|
II. Cryptorchidism
|
|---|
A. Etiology
Cryptorchidism will be caused by any anomaly that disrupts normal
testicular descent (149). The normal mechanism of descent is such a
complex interaction of hormonal and mechanical or anatomical factors
that it is not surprising that the cause of UDT is multifactorial
(Table 3). Although the common causes of UDT are not
known, the transabdominal phase is rarely affected, as in most
instances testes have descended to or beyond the inguinal canal (150).
Intraabdominal testes comprise a small group of about 5–10% of cases
perhaps because the transabdominal phase merely requires holding the
testis near the groin during growth of the embryo. By contrast, the
active migration of the gubernaculum during the inguinoscrotal phase is
much more prone to mishap. The UDT is commonly located near the neck of
the scrotum or just outside the external inguinal ring; the latter
position, when just lateral to the pubic tubercle, is described as the
superficial inguinal pouch. Here the testis, within its tunica
vaginalis, is constrained anteriorly by the superficial abdominal wall
fascia (Scarpa) and posteriorly by the external oblique muscle.
The commonest cause of UDT is believed to be a defect in prenatal
androgen secretion secondary to either deficient pituitary gonadotropin
stimulation or low production of gonadotropin by the placenta (37, 76, 91). Work from our own laboratory has suggested that androgen
deficiency may be manifested by abnormal physiology of the GFN (49),
which may disrupt migration of the gubernaculum. Androgen deficiency
during the second and third trimester would necessarily be mild or
transient, as no genital anomaly, save inadequate growth of the
epididymis, is present at birth (151, 152, 153, 154). Furthermore, endocrine
assessment at birth in babies with UDT has not revealed a consistent
anomaly (64, 155, 156). Only later in infancy is a deficiency of
androgen production (2–4 months) or MIS production (4–12 months)
noted (11, 12, 63, 157). Whether postnatal endocrinopathy in UDT is
evidence of a primary defect in the testis or is secondary to
temperature-dependent degeneration is controversial, although many
authors believe that it is secondary (147).
Recognizable endocrine disorders, such as defects in testosterone
synthesis or receptor function, do cause UDT but are rare (17, 85).
Defects in MIS production or the MIS receptor also cause UDT, in the
context of the PMDS (71, 75, 86, 88). As discussed previously,
controversy centers around whether in PMDS the Müllerian ducts
and/or broad ligaments mechanically prevent descent (69) or whether UDT
is caused by failure of gubernacular development (88). The
hypermobility of the gonads and genital tract in PMDS is evidence
against a simple mechanical block.
Ectopic testes lying beyond the normal line of descent are rare and may
occur secondary to particular anatomical defects. Rodent studies show
that transection of the gubernaculum can lead to accidental descent
down the contralateral canal (transverse ectopia) (40, 42). In the
perineal testis the migration process itself has occurred to a normal
extent but is misdirected (Fig. 18); we have proposed
that this may be caused by the GFN being in the wrong site, thereby
controlling migration to a different place (49).
View larger version (104K):
[in this window]
[in a new window]
|
Figure 18. Perineal ectopic testis, caused by misdirected
rather than deficient gubernacular migration.[Reproduced with
permission from S. W. Beasley, J. M. Hutson, and N. A. Myers:
Pediatric Diagnosis: An Illustrated Guide to Disorders of
Surgical Significance. Chapman and Hall Publishers, New York.]
|
|
UDT is common in inherited syndromes with multiple anomalies.
Microcephaly is a common feature, which is consistent with hypothalamic
or pituitary dysfunction (158). Peripheral mechanical anomalies are
also common and may lead to external compression of the inguinoscrotal
region (159, 160, 161). Specific urological disorders, such as prune belly
syndrome and posterior urethral valves, have a high frequency of UDT
(162, 163, 164, 165, 166, 167). The possible mechanisms proposed include massive but
transient prenatal urinary tract obstruction or a mesodermal defect.
Abdominal wall defects predispose to UDT (15% in gastroschisis; >30%
in exomphalos) (168). Three proposed mechanisms include lower abdominal
pressure (38), traumatic disruption of the gubernaculum, or a defect in
the hypothalamus (158). Neural tube defects have a very high frequency
of UDT as discussed previously (113, 169).
B. Frequency
UDT occurs in up to 4–5% of males at birth (170), but less than
half of these still have an abnormality by 6–12 months. In the last
30–40 yr the frequency of UDT at 1 yr of age appears to have increased
in the United Kingdom from 0.96% to 1.58% (170, 171, 172, 173). Elsewhere in
the world an increasing frequency has been noted also, but there are
few reliable studies with sufficient population number and standardized
methods of assessment (174). The very high frequency of UDT reported
for premature infants is related to birth before completion of
inguinoscrotal migration (175, 176). Longitudinal study of the Oxford
region cohort demonstrated that postnatal descent is common, but only
during the first 12 weeks after term (171, 172).
Orchidopexy rates, when used as a measure of the prevalence of UDT,
suggest that cryptorchidism is present in up to 5% of males by the
time they reach puberty (171, 172). The apparent discrepancy between
the number of infants with UDT by 1 yr of age and the higher number
undergoing surgery subsequently has provoked some observers to blame
unnecessary surgical intervention (171, 172, 177). High rates of
operative treatment, however, are documented in many countries, even
those with public health systems and no obvious financial gain to the
surgeon (171, 172). Recent studies suggest that surgery is common for
"retractile" or "ascending" testes, which may in part account
for an increased incidence.
C. Are some UDT acquired?
The cremaster muscle controls testicular temperature and protects
the gonad from trauma (178). Low temperature stimulates the cremasteric
reflex, in which the testis is pulled up out of the scrotum into the
inguinal region, thereby insulating the testis by the subcutaneous fat.
The GFN mediates the reflex by temperature receptors in its cutaneous
branch, which supplies the inner surface of the thigh. At birth the
cremasteric reflex is weak, while later in childhood it is quite
pronounced until 10–11 yr of age when it begins to wane. The high
frequency of a prominent reflex in 3- to 10-yr-old boys is associated
with low androgen levels during this time, prompting Farrington (179)
to suggest that the reflex is modified by androgens.
The clinical dilemma with retractile testes is determining whether they
are normal or abnormal. There is little consensus (180), although some
appear to become more abnormal in position with increasing age (181).
One explanation proposed recently is that at least some "retractile
testes" are testes with acquired maldescent, secondary to failure of
normal elongation of the spermatic cord (182).
The "ascending testis" is defined as a testis that resides in the
scrotum in early infancy but is too high later in childhood (183).
Delayed descent in the first 12 weeks after birth, with subsequent
ascent out of the scrotum, is a common feature (171, 172). Ascending
testes have been described by many authors (181, 183, 184, 185), but whether
or not they are the same as "retractile" testes is not known.
Despite early orchiopexy in infancy, large numbers of older boys are
still presenting with cryptorchidism at ages well beyond that
recommended for correction of congenital UDT (174). Persisting patency
of the processus vaginalis, which prevents normal elongation of the
spermatic cord, was proposed by Atwell (184) as the cause. Pathological
spasticity of the cremaster muscle occurs in children with cerebral
palsy and spastic diplegia, leading to a high incidence of secondary
maldescent (186).
To explain why UDT occurs in older boys as well as at birth,
gonadal position is proposed to depend not only on normal prenatal
gubernacular migration, but also postnatal elongation of the spermatic
cord (187). Acquired, as compared with congenital, UDT could then be
caused by failure of cord elongation (Fig. 19) because
the distance from the inguinal canal to the scrotum increases with age
(178). Clarnette and Hutson (187) have suggested that failure of
complete disappearance of the processus vaginalis may be a common
cause of acquired UDT; this would link inguinal hernia (widely patent
processus vaginalis), hydrocele (narrow processus vaginalis), and
ascending/retractile testes (partially obliterated lumen but
persistence of processus vaginalis) on a common spectrum.
View larger version (21K):
[in this window]
[in a new window]
|
Figure 19. Failure of the spermatic cord to elongate in
proportion to body growth may be a cause of "ascending" or
"retractile" testes. The testis that is fully descended in infancy
(A) assumes a relatively higher position later in childhood (B).
[Reproduced with permission from J. M. Hutson and S. W. Beasley:
Descent of the Testis, 1992 (149)].
|
|
D. Risks of infertility/malignancy
Secondary degeneration of the cryptorchid testes is presumed to be
related to a higher temperature (35–37°C) compared with the normal
location in the scrotum (33°C) (4, 188, 189). Steroidogenesis becomes
deranged within a few months of birth (155, 190), although it is
uncertain, at least in humans, whether this is a primary or secondary
defect. It is also unknown whether it is temporary or persistent.
Animal studies, however, show clearly that UDT causes secondary
degeneration particularly of germ cells (191, 192, 193). Serum levels of MIS
are normal in neonates with UDT but are lower than normal at 4–12
months of age (11, 63, 64, 157) (Fig. 20). Decreased
secretion of androgens and/or MIS are proposed as the cause for
postnatal failure of normal germ cell maturation.
View larger version (50K):
[in this window]
[in a new window]
|
Figure 20. A, Mean (±SE) levels of serum MIS at
different ages [Reproduced with permission from M. L. Baker et
al.: J Clin Endocrinol Metab 70:11–15, 1990 (63). © The
Endocrine Society.] B, Levels of serum MIS (± SD) in
cryptorchid boys (filled bars) vs.
paired, age-matched controls (open bars) [Reproduced
with permission from J. Yamanaka et al.: J Pediatr Surg
26:621–623, 1991.]
|
|
Recent studies show that germ cell development becomes abnormal
in early infancy (194, 195, 196) in biopsies of the UDT, confirming
previously held suspicions that the gonocytes fail to transform
to type A spermatogonia. Deficiency of spermatogenesis was
considered congenital at one time (178), but in recent years evidence
has accumulated to support a secondary degeneration of germ cells
(197, 198, 199). Huff et al. (194, 195, 196) have shown that the
maturation of gonocytes to type A spermatogonia, which is the first
step in postnatal spermatogenic development, is deficient in
cryptorchid infants. Where the gonocytes fail to mature into
spermatogonia, they persist for a while in the testis and then
degenerate, leading to deficiency in total germ cell numbers (200).
Although controversial, persisting gonocytes could be the source of
carcinoma in situ (CIS) cells later in adolescence (201).
The higher risk of infertility in men with a past history of
cryptorchidism is believed to be secondary to temperature-induced
degeneration (4). Rats rendered cryptorchid surgically at birth have
reduced fertility (202, 203), at least as measured by sperm analysis.
Assessment by paternity rates, however, fails to reflect the same
degree of anomaly. The cause of the UDT, whether surgical or hormonally
induced by postnatal estrogen treatment, did not alter the prognosis
for fertility (204). Testes that are intracanalicular or
intraabdominal are associated with a poorer prognosis for fertility
than inguinal testes (205). Significantly, retractile testes are
reported to lead to reduced sperm counts in adulthood (206, 207),
although this is not found in paternity studies (208, 209).
UDT is associated with a risk of malignant tumor of the testis in
adulthood. At one time the risk was thought to be 35- to 50-fold
greater than normal (210), but more recent calculations of relative
frequency suggest the risk is closer to 5- to 10-fold (211, 212). The
relative frequency of a history of UDT in men with testicular tumors is
15-fold for unilateral UDT and 33-fold for bilateral UDT (213, 214, 215).
The increased risk of malignancy is considered to be caused by germ
cell degeneration and dysplasia within the UDT (196, 216, 217). An
alternative proposal, however, is that there is an intrinsic defect in
the testis (218), as CIS germ cells may be found in neonates with
dysgenetic testes and ambiguous genitalia. CIS germ cells
arebelieved to be premalignant (201), but their origin is
uncertain.
E. Role of hormone therapy
Medical treatment of UDT with various hormone preparations has
been attempted since the 1940s (219). The underlying basis for hormonal
therapy is the view that a deficiency of the
hypothalamic-pituitary-gonadal axis is the common cause of UDT (76).
Whether or not UDT is secondary to androgen deficiency, treatment of
cryptorchid boys with human CG (hCG) or LHRH has been of mixed success,
from 10%–20% (220, 221, 222, 223).
Despite overall poor results (147), hormone treatment is effective in
certain special groups. When the testis lies near the neck of the
scrotum or when the anomaly is bilateral, the results of treatments are
better. Older children and those with retractile testes also respond
better to hormone therapy (223). As retractile testes are known to
descend spontaneously at puberty (224, 225), it has been suggested that
their favorable response to hormone stimulation is secondary to
induction of precocious puberty (226).
The effectiveness of hormonal therapy is difficult to determine because
congenital UDT has not been separated from its apparently acquired
variants, ascending and retractile testes. Nearly all studies include
boys between 5 and 12 yr, in whom an acquired anomaly could be present,
while few studies examine the effectiveness of hormone treatment solely
in infants, where congenital UDT is likely. Certainly, infants with
unilateral UDT in the superficial inguinal pouch have the lowest
success rate, which is consistent with the widely held view that
congenital UDT is relatively resistant to hormone treatment. Although
hormone treatment is in common use in many European and American
centers, elsewhere it is uncommon. A further factor in its waning
popularity may be the need for multiple intramuscular injections of hCG
(227, 228). Nasal application of LHRH several times a day for 1 month
also has been tried (220, 223, 229, 230) but this method is not
available in some countries, including Australia.
Hormonal therapy may be advantageous in distinguishing retractable
testes from congenital UDT (223). In addition, hormone stimulation may
restore normal testicular function after surgery, as measured by
increased numbers of germ cells and Leydig cells (231).
F. Timing of surgery
The optimal time for surgical correction of UDT remains unknown,
although numerous studies suggest that early intervention is
beneficial. Early correction of UDT does prevent degeneration, as is
relatively easily shown in animal studies (204, 232, 233, 234, 235, 236). In humans,
degeneration of germ cells is first seen at 6–12 months (194, 195, 196).
Morphological changes are detected at the EM level at 1–2 yr (198),
light microscopic signs of degeneration appear by 3–4 yr (237), and
clinical atrophy of the testis is detected by 5–7 yr. Based on EM
evidence of degeneration (197), orchiopexy in humans has been
recommended at 1–2 yr of age. The recent recognition of germ cell
maldevelopment by 6 months of age, however, suggests that even earlier
surgery may be needed to prevent germ cell death or dysplasia: a number
of centers are now offering orchidopexy at 6–18 months of age.
|
Footnotes
|
|---|
Address reprint requests to: J. M. Hutson, M.D., General Surgery, Royal Children’s Hospital, Parkville Victoria 3052, Australia.
1 Supported by National Health and Medical Research Council of Australia
Grant 950610. 
|
References
|
|---|
-
Bedford JM 1978 Anatomical evidence for the
epididymis as the prime mover in the evolution of the scrotum. Am
J Anat 152:483–508[CrossRef][Medline]
-
Williams MPL, Hutson JM 1991 The phylogeny of
testicular descent. Pediatr Surg Int 6:162–166
-
Top
Abstract
I. Normal Development
