Predictive value of normal sperm morphology: a structured literature review

Abstract

The aim of the study was to conduct a structured review of the literature published on the use of normal sperm morphology, as an indicator of male fertility potential in the in-vitro fertilization (IVF) situation, and to establish the universal predictive value of this semen parameter. Published literature in which normal sperm morphology was used to predict fertilization and pregnancy, during the period 1978-1996, was reviewed. A total of 216 articles were identified by the sourcing methodology, but only 49 provided data that could be tabulated and analysed. Of these, only18 provided sufficient data for statistical analysis. Fifteen studies used the strict criteria to evaluate sperm morphology, two used World Health Organization (WHO) guidelines and one used both the strict criteria and the WHO guidelines. All the studies (n = 10) using the 5 and 14% normal sperm morphology thresholds (strict criteria) produced positive predictive values for IVF success. In the prediction of pregnancy, 82% (9/11) and 75% (6/8) of the studies produced positive predictive values when using the 5% and 14% thresholds respectively. Aggregating the data produced around the 5% normal sperm morphology threshold (strict criteria), the overall fertilization rates were 59.3% (1979/3337; per oocyte) for the < or = 4% group and 77.6% (10345/13327; per oocyte) for the >4% group, and the overall pregnancy rates were 15.2% (60/395; per cycle) and 26.0% (355/1368; per cycle) respectively. The no-transfer rates across the 5% threshold were 24.0% (86/359; per cycle) in the < or = 4% group compared to 7.4% (80/1088; per cycle) in the >4% group. The inclusion of an accurately evaluated normal sperm morphology count as an integral part of the standard semen analysis makes this analysis still the most cost-effective means of evaluating the male factor.

Dimensions of human ejaculated spermatozoa in Papanicolaou-stained seminal and swim-up smears obtained from the Integrated Semen Analysis System (ISAS®)

“…When used both with standardized methods and with variables for analysis [21,23,24], these machines provide high repeatability and precision compared with subjective npg morphological evaluation [25,26]. However, problems of cell recognition, image digitization [2], and the effects of staining procedures [11] cannot be avoided and the process is time-consuming.…”

Semen analysis with regard to sperm number, sperm morphology and functional aspects

“…The article by Morgentaler et al [16] was quoted by Mortimer and Menkveld [17], who classified it as one of the 'occasional publications with negative reports' and by Eliasson [2]. In the 'structured reviews' by Coetzee et al [18] and Van Waart et al [19], and in the fifth edition of the WHO manual [6], one cannot find any reference to this publication.…”

“…These facts ought to be viewed in the light of the claims that the 'Tygerberg strict criteria' are supported by a 'myriad of publications' (e.g., [20]). Section 2.13.2 of the new WHO manual [6] states, 'By the strict application of certain criteria of sperm morphology, relationships between the percentage of normal forms and various fertility endpoints (TTP, pregnancy rates in vivo and in vitro) have been established (Eggert-Kruse et al [21]; Jouannet et al [22]; Coetzee et al [18]; Toner et al [23]; Menkveld et al [24]; Van Waart et al [19]; Garrett et al [25]; Liu et al [26]) that may be useful for the prognosis of fertility'. Hopefully these authors used 'strict application of (their own) criteria', but that is not to be confused with the application of 'strict criteria'.…”

“…Section 2.13 of the manual says: 'Using these guidelines ('strict criteria'-my comment), the range of percentage normal values during analysis of fertile and infertile men is likely to be in the range of 0%-30%, with few samples exceeding 25% normal spermatozoa [6]. This low value will inevitably produce low thresholds discriminating between fertile and infertile populations; indeed with this method reference limits and thresholds of 3%-5% normal forms have been found in studies of in vitro fertilization (Coetzee et al [18]), intra-uterine insemination (Van Wart et al [19]) and in vivo fertility (Van der Merwe et al [27])'. When these three supporting references are 'decoded for identity', one finds that all involved are well known for their work in favour of the 'Tygerberg strict criteria', for their engagement in assisted reproduction technology (ART) and for their many joint publications.…”

“…In addition, the methods should maximize the proportion of candidates for the ART treatments. Coetzee et al [18] wrote, 'The inclusion of an accurately evaluated normal sperm morphology count as an integral part of the standard semen analysis makes this analysis still the most costeffective means of evaluating the male factor'. Staining the spermatozoa with the 'Quick-diff' is much faster and cheaper than with the Papanicolaou method.…”