Controlled studies after 1994

Since the authors’ 1994 review was published, only six new controlled trials have been published in peer-reviewed journals ( Tables 1–4 ) . Unfortunately, the results of these studies have not quelled the controversy surrounding the varicocele and the benefit of its repair. In 1998, Nieschlag and colleagues published an update of their previously reported controlled study (1995) in which they randomly assigned 120 infertile couples to receive surgical ligation, radiologic embolization, or no treatment. All male patients had a varicocele and abnormal semen parameters. The mean age of the men was 32 years, and the mean age of the female partners was 31 years. Patients who had other known causes of infertility (eg, cryptorchidism, infections, anovulation, tubal blockage) or any chronic comorbidities were excluded. Ninety-five patients completed the study. Pregnancy rate was the key outcome assessed; semen parameters and hormone concentrations were secondary variables. After 1 year there was no significant difference in pregnancy rates (25.2% in the treatment group versus 27.1% in the counseling group). Nevertheless, sperm concentration did increase significantly in the treated patients, whereas no significant changes in semen parameters occurred in the nontreatment group. Both study arms, however, included a large percentage of patients who had grade I varicoceles (48% in the treatment group and 57% in the control group). This fact is potentially significant, because several studies have noted that larger varicocele size may be associated with greater degrees of improvement after varicocele treatment ; thus the larger numbers of patients who had grade I varicoceles may have blunted any effect of treatment in this study. Nieschlag and colleagues subsequently published an update that included data on 30 more couples who completed the study with the same follow-up and outcomes. The additional participants did not affect the outcomes reported in the initial study. The population of this study was quite small, losing most patients (78) after randomization with a 38% drop-out rate. Nevertheless, this was a well-designed, carefully described, single-center study demonstrating a beneficial effect of varicocelectomy on semen parameters without an effect on pregnancy rates.

The study published by Madgar and colleagues is part of a multicenter trial (84,902) by the World Health Organization (WHO) that has yet to be published in its entirety. This prospective, randomized, controlled trial is well designed but differs significantly from the Neischlag study in that only men who had grade II or III varicoceles and sperm counts between 5 × 10 ⁶ and 20 × 10 ⁶ /mL were included in this study. Although 210 couples were evaluated within the study period, only 45 couples were assigned randomly to treatment with surgical ligation or to no treatment. The patients were followed for 12 months with pregnancy as the primary outcome variable. Patients in the observation group who did not achieve a pregnancy after 12 months underwent varicocelectomy. All patients were followed for 36 months after surgery. Semen parameters did not change in the nontreatment group during the observation period, but these patients experienced a significant improvement in sperm count, motility, and morphology after varicocelectomy. Only two pregnancies (10%) occurred in the nontreatment group during the observation year, but eight conceptions (44.4%) occurred in this group during the first year after surgery. The group of patients who underwent immediate surgery achieved 15 pregnancies (60%) in the first year. In the immediate-surgery group the sperm count, motility, and morphology improved significantly. At 36 months’ follow-up, the overall pregnancy rate in the group receiving immediate surgical treatment was 76%, whereas the delayed-treatment group achieved a pregnancy rate of 66.7%. This study is perhaps one of the best published regarding varicocele repair: the criteria for inclusion were strict regarding infertility and the presence of varicocele, and the authors simply compared surgical treatment and observation. As discussed later, few other studies have met these standards. Although this study did not use a true cross-over study design, the significant improvements in semen parameters and pregnancy outcomes after surgery in the observation group are remarkable. Unfortunately, the study population was small and had a high drop-out rate, a common problem in this young, busy, and mobile population .

In 2002, Krause and colleagues , in another prospective, randomized, controlled multicenter study, compared the efficacy of sclerotherapy versus no treatment of varicoceles. All subjects had abnormal semen parameters and had been infertile for at least 1 year. Varicoceles were diagnosed by palpation and Doppler sonography. Patients who had subclinical or symptomatic varicoceles were excluded, as were men who had comorbidities or sperm density below 2 × 10 ⁶ mL ⁻¹ . Patients were assigned randomly to treatment or observation groups. The authors provided an intention-to-treat analysis. Unfortunately, this study suffered from poor recruitment and difficult-to-understand shifts between the groups. An intention-to-treat analysis demonstrated that 300 patients were required for the study to be powered sufficiently, but only 67 patients were randomized. More than half the patients were lost to follow-up at 6 months. The authors acknowledge this study flaw and present an “as-treated analysis.” This analysis demonstrated that testicular volume increased significantly in treated groups, but there were no significant alterations in sperm parameters. The conception rate was 30% within 12 months after intervention in the treated patients compared with 16.2% in the untreated patients ( P = .189). Interestingly, the authors reported a cure rate (absence of reflux) of only approximately 50% in their patients. “Two high ligations of the varicocele were performed” with no explanation as to why or how these particular patients were analyzed in the study. Clearly, this study suffers from many methodologic flaws that prevent one from drawing any valid conclusions.

Several other controlled trials have been published since 1994. Unfortunately, these studies included patients who had subclinical varicoceles or normal semen analyses. Although the current guidelines do not recommend the treatment of varicoceles in these patients , these studies certainly are worthy of analysis in any critical analysis of varicocele outcomes. Yamamato and colleagues studied 85 patients diagnosed by thermography (infrared imaging) with thermal asymmetry (> 0.3°C) of the testes. Patients were diagnosed as having subclinical varicoceles (no varicocele was found on examination). Male patients had a mean age of 32 years and had been infertile for at least 1 year. Patients were assigned randomly to treatment with high ligation of the internal spermatic vein or to no treatment. There were no restrictions on semen parameters. Sperm density was the only variable that showed a significant increase ( P < .006) in the treatment group compared with controls. The authors conclude that subclinical varicocele repair is not warranted because it does not improve postoperative pregnancy rate. The authors, however, did not restrict the study participants to men who have abnormal semen parameters; indeed, the mean preoperative morphology was normal based on WHO standards (> 30%) . In addition, because scrotal thermography lacks specificity as a diagnostic tool, it has not been embraced as a diagnostic modality.

Grasso and colleagues also performed a randomized, controlled study comparing surgical treatment versus observation in patients who had grade I varicoceles diagnosed by bidirectional Doppler velocimetry. Sixty-eight patients 30 to 38 years old who had been infertile for more than a year were assigned randomly to spermatic vein ligation or to a 12-month observation period. Although patients included in this study did have abnormal semen parameters, the exclusion criteria were not described. Additionally, the age and fertility status of the female partners were not detailed. The authors discovered a significant decrease in sperm concentration in the first 6 months, but the count normalized to baseline at 12 months. There was no demonstrable trend in semen parameters in the group that underwent observation. The authors state that “paternity was verified in one patient in group 1 (surgery) and two patients in group 2 (observation).” It was not clear whether these were pregnancies or live births. The authors conclude that left spermatic vein ligation “has no influence on fertility in patients over the age of 30 years with grade I left varicocele.” This study has several difficulties. Again, the study population included only patients who had grade I varicoceles, and the sample size is quite small. In addition, because no information was provided regarding evaluation of the partners, one cannot be sure the persistent infertility was solely the result of male infertility, without other confounding factors.

Another trial by Unal and colleagues also studied subclinical varicoceles. This study randomly assigned 42 men who had subclinical varicoceles to high ligation of the spermatic vein or to 6 months of clomiphene citrate treatment. These men had a mean age of 32.7 ± 6.1 years, had been infertile for more than 1 year, and had a normal hormonal profile and normal testicular size. The fertility evaluation of the patients or partners was not described. Exclusion criteria also were not described. With the exception of motility, the baseline semen parameters for both groups were normal, based on WHO criteria . In the surgery group sperm density and motility percentage were improved significantly at 6 months compared with baseline. The clomiphene citrate group had no improvement in any parameter. There also was no statistical difference in semen parameters between the groups. Only two pregnancies were achieved in the surgery group, versus one in the clomiphene group, a difference that was not statistically significant. As in the report by Grasso , the fertility assessment of the study participants was not detailed. Additionally, there was no control group in this study. Participants were assigned randomly to one of two treatment arms. Although the evaluation of alternatives to invasive treatment methods of varicoceles is important, it is difficult to determine the value of an intervention or dismiss a treatment as ineffective without comparing it with a true control.

Two additional randomized trials have been presented in abstract form but have yet to be published. A 2003 abstract by Dohle and colleagues randomly assigned 72 couples who had more than 1 year of infertility to treatment (“varicocele treatment”) or to watchful waiting. Men who had ultrasound-confirmed varicoceles and abnormal semen parameters were included. Men who had azoospermia were excluded. Female partners had normal gynecologic evaluations and were younger than 36 years old. After 1 year of follow-up there was an improvement in sperm density and motility. After 1 year the rate of spontaneous pregnancies was 36% in the treatment group compared with 9% in the control group. These results certainly are promising; one hopes that, when the results are published in a peer-reviewed journal, the study design and statistical analysis withstand critical review.

The WHO task force on the management and prevention of infertility began a prospective clinical trial to compare varicocele ligation and observation in 1984. The study included 248 couples in 12 countries. Men included in the study had a palpable varicocele and abnormal semen measurements; female partners were demonstrated to ovulate and to have proven fallopian tube patency. The study has not yet been published, but the results have been presented at several meetings. When the study was submitted for publication, it was determined to be too flawed to proceed with publication in a peer-reviewed journal . In this study the pregnancy rate was 34.8% in the intervention group compared with 16.7% in the observation group ( P < .003). There also were significant improvements in sperm concentration in the immediate-treatment group compared with the observed group, and this difference persisted for the entire first year of the study. The results of the study showed that surgical varicocele repair was 2.5 times more effective than delayed treatment. The study suffered from extensive loss to follow-up and bias, however. Not all patients were enrolled formally in the study before treatment. There also were procedural mistakes in the randomization of patients. The major flaw of this multi-institutional study was the varied interpretation of the study protocol by the different centers across the world .

Meta-analyses and statistical reviews

After reviewing the available randomized clinical trials in the literature, it is clear that difficulties with methodology and reporting persist. Only one of the published studies presents positive results. Is there, however, sufficient evidence to conclude that treatment of varicocele is not warranted? Evers and Collins arrived at this conclusion after their own review of the varicocele literature. They conducted an extensive search for controlled trials in the literature, including proceedings of annual meetings and hand searches through andrology journals. After reviewing eight studies (they included two studies previously described in the 1994 review by Schlesinger et al ), they concluded the Peto odds ratio was 1.10 (95% confidence interval [CI], 73–1.68) favoring treatment over observation. When only the three trials that included patients who had clinical varicoceles and abnormal semen analyses were analyzed, the Peto odds ratio was 1.75 (95% CI, 0.60–4.25) in favor of treatment. Although this analysis showed a trend in favor of treatment, it was not statistically significant ( P = .06). The authors state that their review “fails to offer evidence that treatment of a varicocele … improves the couple’s spontaneous pregnancy chances.”

A meta-analysis combines the results of several studies that address a set of related research hypotheses. A well-conducted analysis can provide a more precise estimate of a treatment effect. On the other hand, poorly conducted meta-analyses may be biased because of the inclusion of inadequate studies . Evers and Collins themselves state that the studies included for analysis were not of “high quality,” were significantly heterogeneous, and included studies with poor methodology. Ficarra and colleagues conducted their own review of the three randomized, controlled trials that included patients who had clinically diagnosed varicoceles and subnormal semen parameters. They acknowledged the high rate of drop-out and patients lost to follow-up and therefore searched the studies to perform an “as-treated” analysis. They demonstrated a pregnancy rate of 36.4% in the treatment group versus 20% in the control group ( P = .009). Although these numbers are favorable and contradict the conclusions of the Cochrane review , they again are drawn from data and patients from clinically and statistically heterogeneous studies.

Marmar and colleagues sought to improve the prior published meta-analysis by excluding studies that included patients who had subclinical varicoceles. They also attempted to decrease heterogeneity by including only patients treated with surgical varicocelectomy. They excluded any patients who had undergone assisted reproductive techniques (ART). They included only patients who had at least one abnormal semen parameter. Of 101 articles retrieved from a search that contained pregnancy data, only five studies were sufficiently free of bias to pass the authors’ rigorous review. The reviewers were blinded during evaluation: the methods, results, tables, and figures were separated from qualitative or quantitative reports of the results. Furthermore, each study was evaluated for four different categories of bias. These studies included randomized and observational studies. The odds of spontaneous pregnancy after varicocelectomy compared with no treatment were 2.87 (95% CI, 1.33–6.20). Tests for heterogeneity were not significant. The number needed to treat to achieve spontaneous pregnancy was 5.7. The statistical methodology in this study is more meticulous and thorough than in the other meta-analyses described previously. The authors excluded studies with large numbers of individual dropouts after randomization. The “scoring system” used to detect bias in the studies included in the meta-analysis was not clearly delineated, however. Each of the four categories of bias—selection, confounding, information, and “other”—was assigned a threshold score, and articles with scores below this threshold were excluded. This threshold score, however, was different for each category and was ill defined. Based on their meta-analysis techniques, the authors concluded that varicocelectomy improved spontaneous conception rates.

Meta-analyses and statistical reviews

After reviewing the available randomized clinical trials in the literature, it is clear that difficulties with methodology and reporting persist. Only one of the published studies presents positive results. Is there, however, sufficient evidence to conclude that treatment of varicocele is not warranted? Evers and Collins arrived at this conclusion after their own review of the varicocele literature. They conducted an extensive search for controlled trials in the literature, including proceedings of annual meetings and hand searches through andrology journals. After reviewing eight studies (they included two studies previously described in the 1994 review by Schlesinger et al ), they concluded the Peto odds ratio was 1.10 (95% confidence interval [CI], 73–1.68) favoring treatment over observation. When only the three trials that included patients who had clinical varicoceles and abnormal semen analyses were analyzed, the Peto odds ratio was 1.75 (95% CI, 0.60–4.25) in favor of treatment. Although this analysis showed a trend in favor of treatment, it was not statistically significant ( P = .06). The authors state that their review “fails to offer evidence that treatment of a varicocele … improves the couple’s spontaneous pregnancy chances.”

A meta-analysis combines the results of several studies that address a set of related research hypotheses. A well-conducted analysis can provide a more precise estimate of a treatment effect. On the other hand, poorly conducted meta-analyses may be biased because of the inclusion of inadequate studies . Evers and Collins themselves state that the studies included for analysis were not of “high quality,” were significantly heterogeneous, and included studies with poor methodology. Ficarra and colleagues conducted their own review of the three randomized, controlled trials that included patients who had clinically diagnosed varicoceles and subnormal semen parameters. They acknowledged the high rate of drop-out and patients lost to follow-up and therefore searched the studies to perform an “as-treated” analysis. They demonstrated a pregnancy rate of 36.4% in the treatment group versus 20% in the control group ( P = .009). Although these numbers are favorable and contradict the conclusions of the Cochrane review , they again are drawn from data and patients from clinically and statistically heterogeneous studies.

Marmar and colleagues sought to improve the prior published meta-analysis by excluding studies that included patients who had subclinical varicoceles. They also attempted to decrease heterogeneity by including only patients treated with surgical varicocelectomy. They excluded any patients who had undergone assisted reproductive techniques (ART). They included only patients who had at least one abnormal semen parameter. Of 101 articles retrieved from a search that contained pregnancy data, only five studies were sufficiently free of bias to pass the authors’ rigorous review. The reviewers were blinded during evaluation: the methods, results, tables, and figures were separated from qualitative or quantitative reports of the results. Furthermore, each study was evaluated for four different categories of bias. These studies included randomized and observational studies. The odds of spontaneous pregnancy after varicocelectomy compared with no treatment were 2.87 (95% CI, 1.33–6.20). Tests for heterogeneity were not significant. The number needed to treat to achieve spontaneous pregnancy was 5.7. The statistical methodology in this study is more meticulous and thorough than in the other meta-analyses described previously. The authors excluded studies with large numbers of individual dropouts after randomization. The “scoring system” used to detect bias in the studies included in the meta-analysis was not clearly delineated, however. Each of the four categories of bias—selection, confounding, information, and “other”—was assigned a threshold score, and articles with scores below this threshold were excluded. This threshold score, however, was different for each category and was ill defined. Based on their meta-analysis techniques, the authors concluded that varicocelectomy improved spontaneous conception rates.

Varicocelectomy: “stacking therapies”

The Cochrane review included only studies that reported pregnancy rates as an outcome measure. The pregnancy rate is, of course, the most important variable when considering an intervention to improve fertility, but pregnancy rates are affected by a multitude of other factors that often are not well studied or characterized by the published studies. Additionally, today ART is becoming more common and more accessible to infertile couples. As described previously, in their review of the literature the present authors often found that individual semen parameters often improved, but pregnancy rates did not. Indeed, semen analysis has a limited ability to distinguish fertile from infertile males, and pregnancies credited to increases in semen parameters may be unrelated to these increases . Similarly, the lack of pregnancy does not in itself indicate that varicocelectomy did not have a beneficial effect. Improvements in individual semen parameters may not be insignificant. Galarneau and Nagler have coined the term “progressive stacking of therapy” to capture the full beneficial effect of varicocelectomy. The concept of “stacking of therapy” states that although semen parameters may not improve enough after varicocelectomy to result in spontaneous pregnancy, they may be sufficient for ARTin patients who before surgery did not have adequate semen parameters to access these treatments. One study showed that in 31% of patients semen parameters improved enough for the couples to shift from being candidates for in vitro fertilization candidates to being eligible for intrauterine insemination or achieving spontaneous pregnancy . The overall pregnancy rate achieved after varicocele treatment was 36.6%. The treatment of male infertility with varicocelectomy followed by intrauterine insemination has been shown to be more cost effective than in vitro fertilization and intracytoplasmic sperm injection .

Effect of varicocelectomy on individual semen parameters

Most of the studies demonstrating improved semen quality and pregnancy rates are uncontrolled . Although these studies are not the most powerful tools with which to evaluate the efficacy of varicocele treatment, they do contain valuable information. Noncontrolled studies may be closer to real clinical practice than randomized, controlled trials in which more patients may refuse randomization . These studies generally have larger series of patients with longer follow-up. The information provided by these studies is summarized and discussed in the following sections.

Effect of varicocelectomy on sperm density

In addition to the randomized, controlled trials described previously in this article (see Tables 1–4 ), 19 studies between 1997 and 2007 included an assessment of sperm density ( Table 5 ) . The data comprised 2988 patients. The types of varicocele repair differed across the studies. Infertility was the indication for most repairs, but one study included only patients who had persistent or recurrent varicocele. Indications for varicocelectomy in other studies included pain and testicular atrophy . Three of the randomized, controlled trials demonstrated an improvement in sperm density after treatment of varicocele. The majority of the uncontrolled studies demonstrated an improvement in sperm density after treatment. Only two studies demonstrated no improvement . Two studies had mixed results . The controlled study by Onazawa and colleagues compared 31 patients who underwent repair versus 18 who underwent conservative treatment. Although there was a control arm, this was not a randomized study because patients were allowed to choose their method of treatment. Therefore this controlled, nonrandomized study is included in this section for analysis. The patients undergoing varicocele repair had a significant improvement in sperm density in this study. The trial by Zini and colleagues revealed an improvement in sperm density in 37 men who underwent repair, but this result did not reach statistical significance.

Table 5

Uncontrolled varicocelectomy trials: sperm density

Study	Year	# of Patients	Comment	Intake sperm density (million/mL)	Postoperative sperm density (million/mL)	Statistical significance
Al-Kandari et al	2007	40	Ivanisevich	22.4 ± 4	40 ± 6	P < .01
		40	Laparascopic	21 ± 5	41 ± 6	P < .01
		40	Microsurgical subinguinal	20 ± 5	42 ± 7	P < .01
Hsieh et al	2006	254	Ivanisevich	24.2 ± 18	41 ± 28	P <.05
Marmar and Benoff	2005	60	Marmar	11.92 ± 8.80	19.93 ± 12.00	P < .01
Libman et al	2006	157	Bilateral varicocelectomy	20.7 ± 2.0	27.3 ± 2.5	P = .005
		212	Unilateral varicocelectomy	19.0 ± 1.7	24.8 ± 2.4	P = .002
Hussein et al	2006	104	Microsurgical subinguinal	13 ± 17.9	17.4 ± 13	P < .0001
Zucchi et al	2006	22	Ivanisevich	13.4 ± 4.4	14.0 + 4.5	P < .5
		21	Antegrade sclerotherapy
Orhan et al	2005	82	Microsurgical inguinal	30 ± 8.3	33 ± 8.9	P = .01
		65	Microsurgical subinguinal	29 ± 8.5	32 ± 10.0	P = .05
Zini et al	2005	37	Marmar	34.6 ± 6.0	38.4 ± 7.6	P = .54
Gat et al	2005	101	Embolization	0.22 ± 0.30	9.28 ± 1.2	P < .001
		32	Azoospermia	0	3.81 ± 1.69	P < .03
		31	Virtual azoospermia	0.054 ± 0.007	10.31 ± 1.87	P < .001
		38	Severe oligoazoospermia	0.54 ± 0.04	12.11 ± 1.85	P < .001
Watanabe et al	2005	50	High ligation	15.9 ± 19.8	28.8 ± 32.5	P < .01
		33	Laparoscopic	21.9 ± 22.2	52.3 ± 55.4	P < .01
		61	Marmar	23.5 ± 29.7	59.4 ± 50.2	P < .01
Grober et al	2004	54	Recurrent or persistent varicoceles/micro-surgical repair	15.8	26.0	P = .02
Polito et al	2004	426	Percutaneous scleroembolization	27.55 ± 4.41	34.08 ± 3.53	P < .01
Kibar et al	2002	90	Subinguinal	22.1 ± 4.2	38.3 ± 6.1	P = .00002
Onozawa et al	2002	18	Observed	36.6 ± 31.9	41.9 ± 45.7	P > .05
		26	Palomo	26.0 ± 28.8	34.6 ± 29.2	P = .09
Cayan et al	2000	232	High ligation	30.97 ± 2.46	34.57 ± 3.58	P < .001
		236	Microsurgical high ligation	29.7 ± 1.21	36.62 ± 1.58	P < .001
Jungwirth et al	2000	272	Subinguinal microsurgical	51.7 ± 4.2	64.3 ± 5.6	P < .0001 versus control
Scherr et al	1999	26	Unilateral microsurgical	98.5 ± 94.8	167.6 ± 200.3	P = .0519
		65	Bilateral microsurgical	69.6 ± 90.0	136.9 ± 157.2	P = .00003
Vazquez-Levin et al	1997	33	Microsurgical	35.2 ± 6.2	45.8 ± 8.0	P < .01
Seftel et al	1997	30	Subinguinal	18.9 ± 15.1	41.7 ± 33.1	P < .0001

Only gold members can continue reading. Log In or Register to continue

Share this:

• Click to share on Twitter (Opens in new window)
• Click to share on Facebook (Opens in new window)

Related posts:

Hormonal Evaluation of the Infertile Male: Has It Evolved? Assessing Sperm Function Ejaculatory Duct Obstruction Restructuring Reconstructive Techniques—Advances in Reconstructive Techniques Intrauterine Insemination and Male Subfertility Intracytoplasmic Sperm Injection (ICSI) – What are the Risks?

Stay updated, free articles. Join our Telegram channel

Join