Introduction

This thesis aimed at contributing to the pathophysiology and treatment of varicoceles. The research project encompasses five prospective studies and one retrospective study. The specific objectives of these clinical studies were to assess: (i) the safety, efficacy and tolerance of two types of glue and a non-gluing liquid embolic agent, in the endovascular treatment of varicoceles, (ii) to assess radiation burden for patient during varicocele embolization, (iii) to develop a standardized method to read the phlebographic anatomy of varicoceles and to test a different ontogenesis of varicoceles in adolescents and adults, and (iv) lastly to develop a standardized method to measure the intravascular pressure in the internal spermatic vein and to determine whether infertility in varicoceles can be explained by the principle of elevated hydrostatic pressure.

Liquid embolics

Sclerosing agents (sodium tetradecyl sulfate 3% (STS), sodium morrhuate, dextrose and aethoxysklerol (1–2 or 3%)) are the most commonly used liquid agents in Europe (not authorized in the USA for spermatic veins) [21]. During Valsalva maneuver, the sclerosing agent is injected as a pure liquid or mixed with air as foam at the level of the inguinal ring. Manual compression of the external inguinal ring is mandatory to prevent the agent penetrating into the PP, hence avoiding inflammation and thrombosis.

Tissue-adhesives (Cyanoacrylates) were first introduced by Kunnen, et al. in 1980 [22]. Glue is a less frequently used liquid agent which requires a coaxially catheter system for application (Figure 1) [10, 18, 23]. After verifying flow control with contrast agent injections during different table inclinations, the glue is sequentially injected through the microcatheter and pushed into the ISV as well as into the collaterals. As soon as the glue comes in contact with blood the polymerization process starts, and a permanent occlusion of the vessel occurs. The first glue used was a mixture of contrast agent and isobutyl-2-cyanoacrylate, IBCA (Bucrylate, Ethicon). Later on, IBCA was replaced by NBCA (n-Butyl-2-Cyanoacrylate or Embucrilate; Histoacryl transparent, Braun, Tuttlingen, Germany) because of possible carcinogenicity [24]. Today NBCA (Histoacryl) or NBCA-MS (Glubran2, General Enterprise Marketing, Viareggio, Lucca, Italy) is used with a mixture of Lipiodol Ultrafluid (Guerbert, France) which is necessary for the fluoroscopic visualization of the glue. Both, a higher percentage of Lipiodol and the prior injection of glucose 10% slow down the polymerization rate.

Figure 1 

ISV embolization with glue. Right: Selective venography in erect position with the diagnostic catheter (arrow) in the outflow of the internal left spermatic vein. Internal spermatic vein insufficiency (small arrows) is proved with visualization of the pampiniform plexus (arrowhead). Right middle: Microcatheter (arrow) venography performed in horizontal supine position, revealed a small paraspermatic collateral (small arrows) originating from the lateral duplication of the inguinal internal spermatic vein. This anatomy forced us to reposition the microcatheter in the lateral bifurcation of the internal spermatic vein to be sure that the small collateral is also occluded during controlled injection of glue. Glue embolization will be started in the lateral branch at the level of the coxofemoral joint. During withdrawing of the microcatheter, the glue is pushed into the lateral branch, the small collateral and finally into the medial branch and the main internal spermatic vein then up to the level of the crista iliaca. Left middle: Embolization cast in the left spermatic vein with the glue located between the coxo-femoral joint and the crista iliaca including the small paraspermatic branch. Left: Control venography with the diagnostic catheter in the internal spermatic vein (arrow) demonstrates contrast agent up to the glue cast (small arrows) and the absence of contrast to the pampiniform plexus (arrowhead).

Onyx® (ev3, Irvine, California, USA) is a biocompatible ethylene-vinyl alcohol copolymer (EVOH) dissolved in dimethyl sulfoxide (DMSO) that in contact with blood solidifies as a kind of plastic. Onyx proved to be superior to tissue-adhesives for the occlusion and penetration of side-branches in cerebral embolization [25, 26]. Onyx embolization of varicoceles is feasible but still under investigation in our institution. Technical problems, combined with significant patient discomfort and a high radiation dose preclude, at the moment, its clinical use.

Results

There are currently two commercially available tissue-adhesives that can be used for the embolization of varicoceles. The two adhesives differ only minimally in chemical formula, but polymerize in a different way. The more recently developed Glubran2, available since 2001, would have the same occlusion features as the older glue, hystoacryl transparent. The manufacture of Glubran2 obtained a CE label for intravascular use and alleged that Glubran2 has a higher stability and a lower toxicity.

We tested both glues (Histoacryl transparent, Glubran2) for differences in efficacy, safety and tolerance in a double-blind, prospective, randomized study. We found that both glues can be handled in the same way and that the embolic result is similar.

Technical success was 100%, this means that in all patients we could effectively block the reflux in the insufficient ISV. The procedure was very safe with no complications caused by the glue.

Concerning patients comfort we were the first to delineate a late inflammatory reaction to the cyanoacrylates. During the week after embolization, 59% of patients reported some discomfort, which was in 35% at least a bearable pain. There was no significant difference in discomfort between the two glues in the post-embolization week. Moreover, there was some mild acute inflammatory pain reaction during the embolization in 20% of the patients. There was no difference between both glue groups. Given these findings, patients should be informed that they might experience discomfort or mild pain during and after embolization of their varicocele. As this inflammatory reaction is self-limiting and mostly mild, we would not advise to prescribe NSAIDs prophylactically.

Because there is no difference between Histoacryl transparent and Glubran2, both glues can be used effectively and safely. However, it is medicolegally sensible to choose Glubran2 because it has a CE approval for endovascular use.

Another liquid embolic agent, which is non-adhesive, is ethylene-vinyl alcohol also called “Onyx”. Onyx is a plastic dissolved in dimethyl sulfoxide. After diffusion of the solvent, Onyx solidifies to a rigid cast, which can be pushed and extended in different directions. For these reasons, Onyx embolization has become the primary intravascular therapy for cerebral arterio-venous malformations (c-AVM) and dural fistulas (d-AVF). Because of the anatomical similarities between AVM and complex insufficient ISV, we thought that Onyx could be a potential embolic agent in complex varicoceles. Moreover, we know from extirpated AVM species that Onyx does not induce a post-embolization inflammatory reaction.

In a pilot study we tested Onyx as an embolic agent in ISV embolization. We found that Onyx was efficient and indeed better tolerated in the post-embolization period. However, there was an acute pain reaction during the injection of Onyx in nine of ten patients, prompting to stop the injection temporarily in six patients. The technique to inject Onyx as we do with glue, allowing only little reflux, demonstrated to be suboptimal and ensued a higher radiation dose. Further research should concentrate on adaptation of the embolization technique as used for c-AVM and on reducing the acute pain reaction by injection of intravenous anesthetics.

The first basic principle of radiation protection is justification. This means that the clinical benefit of the treatment has to outweigh the possible radiation-induced risks. The treatment of varicoceles to prevent infertility or to improve fertility, is very controversial. Other studies and our study, show that the gonadal dose in varicocele embolizations is definitely far below the lower threshold corresponding to the deterministic effect of temporary sterility (150 mSv).

Based on the in-vivo measured testes doses, we could estimate the risks of hereditary effects as being very low (mean value of 3.5 10^–6). With respect to the risk of cancer mortality a mean value of 0.06% was calculated. The latter value is much lower than other typical interventional X-ray procedures. According to the NCRP publication 168, the obtained risks in our study population are requiring a minor to moderate benefit to justify the varicocele embolization procedure. The low dose and risk values obtained are probably linked to a set of specific radiation reducing measures. The latter optimization includes the use of pulsed fluoroscopy, the application of gonadal shielding and the careful collimation of the radiation beam.

Apart from the patient, the interventional radiologists are also at risk, especially those, which perform a high amount of procedures. Optimizing patient radiation exposure, will also reduce the patients scatter dose and hence the staff dose.

In everyday practice, varicoceles are treated because of symptomatology (pain and congestion of the scrotum) and the assumed association with infertility. In adolescents, varicoceles are mostly asymptomatic. Treatment of adolescent’s varicocele is advised to anticipate testicular hypotrophy and future fertility. To justify this prophylactic treatment, it is generally assumed that varicocele in children are an early stage of the varicocele in adults. Although the adult varicocele is usually symptomatic, treatment is requested in first place because of sperm dysfunction.

At a certain moment in my practice, I had the impression that the embolization of a varicocele in a child seemed a technically less challenging procedure than in an adult. I thought that the easier procedure was due to the absence of competent valves and a straighter anatomy in adolescents. These observations could point at a different phlebographic anatomy between adolescents and adults, and indirectly disclose a different pathophysiology of varicoceles.

To investigate this hypothesis, we set up a retrospective study to compare phlebographic radio- anatomical landmarks between adolescents and adults. For this purpose, we used the selective phlebographies that were performed prior to the embolization of the varicocele (Table 2a) (Figure 2a).

After analysis and comparison of about 500 phlebograms, we demonstrated that our suspicion was not unfounded. Adults had significantly more often a competent outflow valve than adolescents. Moreover, adult varicocele had a more complex anatomy consisting of a significantly higher number of collaterals and of a complex outflow in the renal vein. The phlebographic complexity of adult varicocele could be the remnant of an insufficient embryonic venous drainage during the ontogenesis of the left ISV.

Left-sided varicocele in many adolescents seems to be more a congenital abnormality rather than a progressive chronic disease: we found a higher rate of insufficient outflow valves, congenital renospermatic bypasses and the compressive nutcracker phenomenon.

On the other hand, we have arguments that right-sided varicocele might be a progressive disease. The incidence of an insufficient ISV increase with age (2.4 times more often in adults (33%) than in adolescents (14.7%). The diameter of the ISV also increases significantly with age: a fact that could point at a degenerative process with reflux and dilatation (Table 2b) (Figure 2b).

Many theories have been developed about how varicoceles can cause infertility. The causal association is most probably multifactorial, combining of lifestyle factors, genetic factors and the effect of reflux of blood into the PP.

The reflux of blood into the ISV is a direct result of the insufficiency of valves, a mechanism we can observe during the phlebography of varicocele patients. By reflux of warm blood, temperature in the testicle rises. Furthermore, in a varicocele patient standing upright, the hydrostatic pressure at the PP and the testicle increases according the height of the fluid column in the ISV. Venous drainage at the PP will then be impaired, and the cooling of the testicle less efficient. This hydrostatic counterpressure might interfere with the arterial and nutritional supply of the testicle and decrease spermatogenesis and testicle growth.

In an editorial of Andrologia, Gat titled that, “Erect posture of humans leads to infertilit.” [46]. In two other articles, he described that the increased pressure at the PP is only caused by the height of the blood column in the insufficient ISV. He calculated this hydrostatic pressure by the expected distance between PP and renal vein according the formula P = h × g × ρ, resulting in a value of 31.5 mm Hg at the left and a value of 27 mm Hg at the right side. This hydrostatic pressure is, according to Gat, high enough to cause testicular insufficiency and infertility in varicoceles. This is the theoretical point of view.

Pressure measurements in the PP or inguinal ISV of varicocele patients were previously performed by direct measurements through a needle or catheter in the PP [47, 48, 49]. The pressures that were measured in these studies were inconsistent and conflicting, in our opinion because of methodological errors. We aimed to measure pressures in the caudal ISV directly and to find out whether infertility can be explained by increased pressure.

Performing pressure measurements through a microcatheter in the ISV was not an easy task. Initial results were not at consistent. We tested and calibrated our method ex-vivo and standardized the measurement method. We compared pressures obtained via a microcatheter (external pressure transducer) with those registered with two different commercial pressure wires. In-vitro (in an experimental tube) and in-vivo (endovascular in the ISV) tests demonstrated that the direct measurements (pressure wires) could be reproduced by the microcatheter measurements (external pressure transducer) (Figure 3). Measurements were extrapolated to a 90° erect position (Figure 4)

For the first time, an in-vivo experiment done in 42 patients showed that the absolute mean pressure in the inguinal left ISV is 33 mm Hg. According to the hypothesis of Gat, this pressure would only exist of a hydrostatic pressure component. However, the height of the blood column that produces this hydrostatic pressure would reach far above the renal vein, even higher than the right atrium. We disclosed that the pressure in the renal vein and at the outflow of the ISV did not change with patient’s position. It was constant at about 12 mm Hg, corresponding to the systemic venous pressure. The absolute measured pressure is therefore composed of a hydrostatic (21 mm Hg), and a systemic pressure (12 mm Hg) component. This hydrostatic pressure was compatible with the indirect calculation from the measured height of the ISV (distance between inguinal ISV and renal vein).

As the normal capillary pressure of the testicle can be estimated at less than 20 mm Hg, then the hydrostatic pressure in upright standing position alone appears to be high enough to act as a counterpressure. This counterpressure could impair arterial supply and spermatogenesis in varicoceles.

Only standardized pressure measurements in fertile patients without varicoceles would be able to answer whether erect posture in varicocele patients leads to infertility.

Conclusion and perspectives

The current commercially available tissue-adhesives, Histoacryl transparent and Glubran2, are equally efficient, safe and tolerable for transcatheter embolization of varicoceles. Glue can cause a mild, self-limiting pain in 1/3 of the patients in the week after embolization, independently of what product was used. Because Glubran2 has a CE label for intravascular use, it is medico-legal justified to use this glue. Worldwide, glue is not a popular embolic agent for the treatment of varicoceles, probably because it requires more skills or preconception persists about the potential complications. Controlled randomized trials, reporting the success and complications rate, as well as the pregnancy rate could lead to an FDA approval for intravascular use and contribute to the attractiveness of glue.

Onyx has the potential of an efficient and safe agent for the embolization of varicoceles. Unfortunately, the acute pain reaction as well as the radiation exposure during embolization have to find a solution before Onyx can be recommended. Finally, we have to explore in which anatomical situation Onyx can constitute an advantage over the other embolics.

Radiation exposure is safe during embolizations of varicoceles with highly viscous liquid products. We report a gonadal dose that is far below the lower threshold corresponding to deterministic effects. In future, by the use of the flat panel detectors, the radiation exposure may be reduced even further.

In left-sided varicocele children’s spermatic vein showed more often valve insufficiency, renospermatic bypasses and nutcracker syndrome. In adults, reflux more often is caused by collateral connections. Left varicocele seems rather a congenital abnormality than an evolutionary disease as commonly believed. At the right varicocele, we have shown that the diameter of the ISV increases in adults and that a bilateral varicocele occurs 2.4 times more. As far as the right varicocele is concerned, we do have arguments for an evolutionary chronic disease. In a subsequent study, we could further investigate and discover links which might predict whether certain children may profit from a prophylactic embolization.

This thesis demonstrated in-vivo that the mean absolute pressure in the inguinal ISV in the upright position is significantly high enough to build a counter pressure to the capillary pressure, and hence impair the spermatogenesis. In addition, we have invalidated Gat’s hypothesis and proven that not only the hydrostatic pressure is responsible for this elevated pressure, but the sum of the systemic and the hydrostatic pressure. We could investigate whether clinical presentation, thermography, CDUS, phlebographic anatomy or age are associated with different pressure levels. Here again, the detection of certain sub-groups could change the therapy policy.

However, we still ignore whether the pressures we obtained, are really pathological since we cannot compare with a healthy control population. We could measure the pressure in the ISV of a control group of fertile men without varicocele. It will not be an easy task to convince the ethics committee. Moreover, it will be a technical challenge to catheterize healthy men retrograde through several competent valves down to the inguinal ISV. Only then we could answer the question “does erect posture in varicoceles lead to infertility?”