Assessment of the histological damage to the testis after torsion: to puncture or not to puncture?



We read with great interest the paper by Moritoki et al. [1]. We fully support the idea that measurable testicular features, such as intratesticular pressure (ITP), could prove valuable in the assessment of histological damage to the testis after torsion. We wish, however, to bring to attention the following methodological concerns.

Measurement of ITP using the Stryker pressure monitor requires insertion of a needle into the testicular parenchyma. The authors did not find histological lesions after testicular puncture in controls, and, therefore, they considered the effect of needle trauma to be negligible. By contrast, previous studies have shown that puncture of human testes, even with fine needles, may result in intratesticular bleeding, devascularization, haematomas and consequent inflammation, fibrosis or calcifications [2-4].

Moreover, it is understood (not clarified in the paper) that the testes of the experimental group were punctured three times to measure ITP (pre-torsion, pre-detorsion, post-detorsion). The difference of ITP before and after detorsion correlated with histological findings, i.e. two punctures at least were needed to produce a comparison. Multiple testicular punctures, however, increase the risk of complications [4]. This effect cannot be revealed by controls punctured only once. If, on the other hand, the needle was inserted once and kept in place throughout the measurements process, the trauma caused by the needle during detorsion of testis would be added to the initial trauma from the needle insertion. Again, this possible effect cannot be excluded by controls that underwent no twisting manipulations while punctured.

To puncture the testis for ITP measurement in order to predict subsequent spermatogenesis has a risk similar to that of a needle biopsy. It is evident to us that it is better not to puncture a torted testis that already bears a >25% probability of atrophy. A non-traumatic method for predicting the histological status of the testis would be preferable. To this end, we have previously developed a mechanical device that provides atraumatic measurements of testis rigidity (hardness) [5], a variable affected by changes in intratesticular fluid pressure and/or elasticity of the non-liquid testicular components [6]. In rat testes after torsion, where, as Moritoki et al. showed, ITP is elevated [1], we found that rigidity measures were dramatically increased within 1 h [5] (we are currently analysing the correlation with histology).

In addition, we have the following objections about the statistical analysis. First, the comparisons between the four groups of ITP values (controls, pre-torsion, pre-detorsion, post-detorsion) were made using multiple (six) pairwise comparisons without adjustment of the significance level for the number of tests; the adjusted P value here should be 0.05/6 = 0.0083 [7]. Lack of adjustment increased the probability of false rejection of each of the six null hypotheses being tested from 0.05 to 1 − (1 − 0.05)6 = 0.27 (type I errors) and, accordingly, the chance of declaring a false statistical significance in each of these comparisons (false-positive results) [7].

Second, ITP values derived from the same testes before torsion, before and after detorsion are obviously not independent of each other and should be subject to paired data set analysis. The Friedman test, the non-parametric test for more than two groups of paired data, would be suitable for comparisons between these three groups [7]. The unpaired analysis performed here may have caused some real differences between groups to be missed (type II errors, false-negative results) [7], and/or may have caused some groups to be false different (type I errors) [8].

Finally, in the scatter plots of Fig. 4, there is an evidently outlying observation (markedly distant point [>25 cm H2O] from the rest of data [<15 cm H2O]). This outlier could have influenced correlation coefficients [7]. Presentation of correlation results including and excluding this outlier would be a more proper approach to ensure validity of results [7].

In conclusion, we think that puncture of the testis for application of the Stryker pressure monitor may lead to biased histology results, attributable to the puncture per se; in addition, results here seem to suffer some statistical pitfalls. Consequently, results do not support the authors' suggestion that ITP can be an index to determine orchiectomy, as already discussed in the editorial comment by Dr Bayne [9].