Frequency and patterns of early recanalization after vasectomy

The methods of the FI and cautery studies have been previously described[8, 9] and are summarized in Additional file 1. Briefly, the FI trial[8] involved eight sites in seven countries. It was a randomized clinical trial comparing two occlusion techniques: ligation and excision with versus without FI. All surgeons used the no-scalpel vasectomy (NSV) approach to the vas and a standardized occlusion technique. The study was halted following a planned interim analysis that demonstrated a clear benefit from the use of FI[10]. Of the 841 men who were randomized in the FI study, 826 were included in the analysis reported here; 410 had FI and 416 did not. Fifteen men were excluded because they did not return for any semen analyses after vasectomy.

The cautery study[9] involved four sites in four countries. It was a prospective observational study designed to estimate the effectiveness of cautery as currently performed at each site and to describe trends in sperm counts after vas occlusion by cautery. Each surgeon used his or her customary cautery occlusion technique, which differed among the sites: two sites performed electrocautery alone and two sites used thermal cautery combined with FI. A small vas segment was excised in one site using electrocautery and in one site using thermal cautery. The other two sites did not remove any vas tissue. Three sites used the NSV approach to the vas. Of 400 men enrolled, 389 were included in the analysis reported here. Eleven men were excluded because they did not provide any semen samples after vasectomy.

Both studies conducted frequent semen analyses, beginning at two weeks after vasectomy. The FI trial conducted subsequent semen analyses every four weeks until a man had provided two consecutive azoospermic specimens, was declared a vasectomy failure, or reached the end of study follow-up at 34 weeks. After the first sample at two weeks, the cautery study conducted subsequent semen analyses at weeks 5, 8, 12, 16, 20 and 24 regardless of semen analysis findings.

Semen analysis methods for both studies were based on World Health Organization recommendations, but differed somewhat between the two studies. Freshly collected semen was examined in the FI trial and data were obtained on sperm concentration, motility, and viability. For the cautery study, two of the four sites did not routinely collect fresh specimens, so semen analysis data from those two sites were limited to sperm concentrations. In addition, specimens showing azoospermia or very low sperm concentrations were centrifuged in the FI trial but not in the cautery study. During both studies, the laboratories conducted periodic quality control tests.

Definitions of success and failure were different in each study. In the FI trial, success was defined as two consecutive azoospermic specimens at least two weeks apart. Failure was defined as 5 million or more motile sperm/mL at 14+ weeks or 100,000 sperm or more/mL with any motility at 26+ weeks. Men who did not meet the criteria for success or failure were classified as indeterminate. In the cautery study, success was defined as less than 100,000 sperm/mL in two consecutive specimens taken at least two weeks apart, and failure as not meeting the definition of success by 24 weeks or having more than 10 million sperm/mL at 12+ weeks. Men who had less then 12 weeks of follow-up without meeting the criteria for success or having been declared a failure by a study site clinician were classified as indeterminate.

The data collection forms, study monitoring, and laboratory quality control procedures were similar for both studies. One research site was common to both studies. Both studies were organized and managed by researchers and staff at FHI and EngenderHealth and received approval from FHI's Institutional Review Board (IRB) and by local IRB, when present.

Semen analysis results over time were used as a surrogate marker to determine early recanalization. Charts presenting the log of the sperm concentration at each follow-up time point were created for each participant. Participant identifiers, study sites, and vasectomy technique assignments were not included and charts were presented in random order. The percentage of motile sperm was included in the charts for the FI trial, but not for the cautery study. Data on cumulative number of ejaculations at each time point were included in the cautery study charts as was the participant's age, because age seemed to be associated with longer times to azoospermia[8]. The FI trial charts indicated participants' final vasectomy outcome (success, failure, indeterminate) because participants were discontinued from the FI trial when failure or success was determined. However, the cautery study charts did not include final vasectomy outcome as sperm concentration data were collected throughout the entire follow-up period. All charts from both studies are available in Additional files 2 (FI trial) and 3 (cautery study).

Since there are no established criteria for determining early recanalization based on semen analysis data, we used an adjudication process. Three experts in vasectomy research (ML, DS, and MB) independently reviewed the charts to determine if early recanalization had taken place according to their own criteria, which were set prior to reviewing the FI trial charts. These criteria (Table 1) were collected after review of the FI trial charts and were not discussed until the adjudication process for the cautery study was completed.

Classification of participants was done first for the FI trial and three months later for the cautery study as part of the procedures of each study. In the FI trial, reviewers were asked to provide their assessment on whether early recanalization had occurred by selecting either yes or no for each chart. In the cautery study an indeterminate category was included as well for those cases where the reviewers felt they could not classify a participant due to insufficient information because of missing data or loss to follow-up. After the adjudication process of the two studies was completed, discrepancies were discussed among the three reviewers and resolved by consensus. In the final consensus assessment, the indeterminate classification was not allowed, and such cases were reclassified as not indicative of recanalization by default. Consensus definitions agreed upon at the time of the final adjudication are presented in Table 1.

Charts best illustrating patterns of sperm clearance after vasectomy (success or failure, with and without early recanalization) were selected by consensus of the reviewers and are presented in Figures 1 and 2.

We computed crude agreement, kappa statistic, and its 95% confidence interval to study the level of agreement among the three reviewers based on their initial independent review. The frequency of early recanalization after reviewers' consensus was calculated. We produced a cross-sectional tabulation to examine the relationship between early recanalization, vasectomy occlusion technique, and final vasectomy outcome. Association between the risks of early recanalization and study center and vasectomy related adverse events in the FI trial were evaluated with the Cochran-Mantel-Haenszel Chi-square test. The probabilities of semen analysis showing any motile sperm, when data available, or 1 million sperm/mL or more at each follow-up time point (weeks) were calculated by early recanalization status and vasectomy occlusion technique.

The major strength of our study is that it was based on results from prospective studies with standardized and excellent follow-up, including early and frequent post-vasectomy semen analyses. However, the primary purpose of this secondary analysis was not to compare the studies and techniques but to generate hypotheses based on observation. Thus, a number of limitations must be emphasized.

As study design of the two studies was different, comparisons between the four occlusion techniques are based on non randomized groups, apart from the two groups in the FI trial. There were also slight methodological differences between the two studies, such as sample size, timing of semen analyses, and definitions of success and failure. In addition, although all participating sites in both studies followed standardized methods for performing semen analysis, semen samples showing azoospermia or very low sperm counts were centrifuged in the FI trial but not in the cautery study. Non-centrifuged semen samples categorized as azoospermic commonly have some sperm if examined after centrifugation, but the numbers would be too low to have interfered with chart adjudication relative to whether or not there was a recanalization[23]. In laboratories that do not centrifuge specimens, sperm concentrations below 100,000/mL are likely to be read as azoospermic (D Sokal, manuscript in preparation).

The level of expertise of participating surgeons in the two studies may explain differences between the results observed with the four occlusion techniques. Some of the surgeons involved in the FI trial were not well-experienced with the FI technique and different methods of FI were used in the cautery study. Despite some training provided before starting the study, technical errors may partly explain the high failure and recanalization risk encountered with this technique. In fact, FI may be difficult to adequately master[19]. In the cautery study, excellent results obtained with thermal cautery and FI combined may be due to the extensive expertise of the two participating surgeons with this technique. Differences in the level of expertise of participating surgeons would not however influence the findings and conclusions related to the failure/recanalization ratio within each individual technique.

Independent review of charts by three experienced vasectomy clinicians/researchers with consensus on discordant cases was used to estimate the risk of presumed early recanalization after vasectomy. Although this method minimizes the risk of bias, it does not exclude it. Recanalization cannot be proven on clinical grounds and no histology was done to confirm recanalization.

Since reviewers knew the outcome of the cases while reviewing the charts from the FI trial, the proportion of failures explained by early recanalization may have been overestimated. Furthermore, knowing the outcome may have led to underestimating the numbers of transient early recanalizations that are followed by subsequent azoospermia and vasectomy success. However, failure or success was included as a criterion to identify early recanalization by only one reviewer and not retained as a consensus criterion (Table 1).

Semen analysis data from two of the four cautery study sites were limited to sperm concentrations, so we could not consider sperm motility when adjudicating charts from the cautery study. The risk of early recanalization may have been underestimated without this information. Using the results from the two sites where motility was assessed and the consensus criteria that include motility (see Table 1), early recanalization was presumed in 1% (1/99) and 18% (18/97) of cases with cautery with and without FI, respectively, whereas according to the classification obtained without motility, these figures were 0% (0/99) and 9% (9/97). Some argue that motility found as early as three weeks after vasectomy is either due to a surgical error or to early recanalization[13, 24‐26].

This possible underestimation does not change our primary results showing that cautery combined with FI is associated with the lowest risk of early recanalization and simple ligation and excision with the highest. The risk associated with cautery alone and ligation and excision with FI is lying in between.

Other factors could also have led to a slight underestimation of early recanalization risk. In the few instances where sperm counts never fell to sub-fertile levels, it was assumed that failure was due to a surgical error. However, it is possible that early recanalization could have occurred within the first two weeks after the procedure, before any semen analysis was performed. In addition, in cases where lack of data meant a participant could not be classified definitively as a presumed early recanalization or not, he was considered as not having had a recanalization.

Adjudication of the charts from the cautery study was done about three months after adjudication of the FI trial charts. Although individual criteria were collected after adjudication of the FI trial charts and not discussed until after completing the adjudication of the cautery study charts, the criteria and interpretation of charts may have changed in time from "learning" about recanalization. This potential bias was minimized using multiple independent reviewers and consensus was needed in only a small and similar proportion of charts in both studies (13% and 14%).

We were unable to asses the risk of late recanalization in this study. Late recanalization is defined as the reappearance of motile sperm after the vasectomy was declared a success, often only discovered by an unexpected pregnancy. This situation suggests that recanalization can occur at any time after vasectomy and not only a few weeks after vasectomy as observed in our study. However, late recanalization is believed to be a very rare phenomenon occurring in only about one out of 2000 to 3000 men [26‐28].