Introduction

Testicular germ cell tumour (TGCT) is the most common malignant neoplasia in reproductive age [1]. The worldwide incidence of TGCTs shows significant variation between different geographical regions and ethnic groups. In Europe it ranges from 1.6 to 9.9 cases per 100,000 individuals [1]. These tumours are derived from germ cell neoplasia in situ (GCNIS) cells, which are arrested and transformed fetal gonocytes that failed to mature to spermatogonia [2]. Data in the literature indicate that both environmental and genetic factors acting on the primordial germ cells/gonocytes are implicated in the etiopathogenesis of this tumour [2,3,4,5].

For instance, familiarity and ethnic differences in TGCT incidence indicate a strong genetic component. Studies in families demonstrated that the familial risk of TGCT is 8-fold to 10-fold elevated for brothers of affected men and 4-fold to 6-fold elevated for fathers/sons of cases, compared with the general male population [6, 7]. In addition, it is well known that Asian and African populations have a low risk of TGCTs and men with African or Asian descent maintains their low risk even living in a geographic area of high risk [4, 8]. Based on population and genomic data Litchfield et al. [9] estimated the heritability of TGCT to be in a consistent range of 37–49%.

Besides genetic factors, a number of clinical predisposing factors are known. Among them, the presence of the components of the “testicular dysgenesis syndrome” [10], such as cryptorchidism, testicular hypotrophy/infertility, hypospadias and microlithiasis, confer higher risk for TGCT [11, 12].

Given the association between infertility and TGCT, genetic anomalies related to infertility has been studied as potential predisposing factors to TGCT.

The human Y chromosome harbors genes essential for testis development and function, such as SRY, which is indispensable for testis determination, and the genes residing in the azoospermia factor regions (AZFa, AZFb, and AZFc), which are involved in spermatogenesis [13]. AZF microdeletions are a well-established genetic cause of severe spermatogenic failure (5–10% of azoospermic and 2–5% of severe oligozoospermic men) [14]. Among the partial AZFc deletions, the clinically most relevant one is called gr/gr, discovered by Repping et al. [15]. This deletion removes half of the gene content and represents a significant risk factor for spermatogenic impairment. However, the entity of the risk for a gr/gr deletion carrier to be affected by oligozoospermia varies between different ethnicities and shows the highest values in the Mediterranean area: 5.8 (CI 2.0–16.9, p = 0.001) and 2.9 (1.0–8.0) in the Italian and Spanish ethnicity, respectively [13].

Three studies focused on the role of complete AZF deletions in TGCT, but a relationship with this disease could not be proven [16,17,18]. No association was found between increased risk of TGCT and Y haplogroups neither [19,20,21]. Concerning the partial AZFc deletions, the role of gr/gr deletions in TGCTs is still debated. In fact, only one out of three published studies observed an increased risk. Indeed, a large multicenter study by Nathanson and collaborators [22] show a 2-fold and 3-fold increased risk for sporadic and familiar TGCT in subjects carrying the gr/gr deletion, respectively. The two subsequent studies based on relatively small cohorts from England and Italy did not confirm the relationship between gr/gr deletions and TGCT [21, 23]. Worth noting that only one of the three studies considered sperm parameters in TGCT patients [21]. Due to the lack of information on the semen phenotype in cancer patients carrying the gr/gr deletion, it remains to be established whether this genetic defect truly represents a common genetic link between TGCT and impaired sperm production.

Only a single manuscript addressed the role of other partial AZFc deletions and in this study no correlation was found with TGCT [23].

We aimed at providing a comprehensive characterization of partial AZFc rearrangements in the largest European case–control study population available to date. In addition, we have addressed an important clinical question concerning the role of gr/gr deletion in TGCT in relationship with the semen phenotype.

Materials

Subjects

We prospectively evaluated 497 patients with TGCT and 2030 control subjects without TGCT who were referred to the University Hospital of Careggi (Italy), Fundació Puigvert (Spain) and University of Rome “La Sapienza” (Italy). The control group included 1068 normozoospermic and 962 non-normozoospermic men with no history of TGCT, no family history of TGCT, no history of cryptorchidism or other abnormal andrological findings and no genetic abnormalities (karyotype alterations and complete AZF deletions). The subjects belonging to these two control groups have been studied previously in relationship with male infertility (for review see Krausz and Casamonti [13]); for this study 24 novel normozoospermic and 18 non-normozoospermic controls were added. TGCT patients came to visit for semen cryopreservation before chemo/radiotherapy. Semen samples were collected by masturbation after 3–5 days of sexual abstinence and semen parameters were investigated according to the WHO guidelines (World Health Organization) [24]. A complete clinical history and andrological exams (including physical examination, scrotal ultrasound and hormonal analysis) were performed. Among TGCT patients, 353 (71%) were affected by seminoma, while 144 (29%) were affected by non-seminoma. Fifty-seven TGCT patients had history of cryptorchidism, 3 had hypospadias and 4 had family history of TGCT. Overall, 268 (54%) TGCT subjects were normozoospermic (total sperm count ≥ 39 × 106) and 229 (46%) were non-normozoospermic (total sperm count < 39 × 106). Among the latter, 35 were affected by azoospermia. This study was approved by local ethics committees and all participants provided written informed consent.

Methods

Molecular analysis

We used standard salting-out method to extract genomic DNA from peripheral blood samples [25]. The analysis for Y chromosome AZF microdeletions was performed by multiplex PCR (using markers from AZFa, AZFb, and AZFc regions according to the EAA/EMQN [14]). The screening for partial AZFc rearrangements was performed in Florence and Barcelona centers by using a multistep method described by Giachini et al. [26]. The first step consisted in multiplex PCR plus/minus based on the amplification of Specific sequence-tagged sites (STSs): sY1291 and sY1191. We identified the gr/gr deletion by the absence of sY1291. The b2/b3 deletion is characterized by the absence of sY1191 and the b1/b3 deletion by the absence of sY1291 and of sY1191. To confirm the multiplex PCR results, we repeated the analysis in all samples with suspected deletion by using a simplex PCR for sY1291 and sY1191. In addition, two subsequent confirmatory steps have been performed in each deletion carrier (as described in Giachini et al. [26].): (i) gene dosage analysis; (ii) RFLP analysis. All the partial AZFc rearrangements identified in the present study were submitted in the LOVD v.3.0 database, https://databases.lovd.nl/shared/variants (variant IDs 461227–461232).

AZFc gene dosage

CDY1 and DAZ gene dosage was performed in all TGCT patients and in 1424 control subjects in order to confirm gene-copy loss in deletion carriers and to define gene duplications in the AZFc region. In brief, we simultaneously amplified the AZFc locus to be quantified (CDY1 and DAZ) and a homologous locus outside the AZFc interval (CDY2 and DAZL, respectively) as an internal standard with a known number of copies, using a single primer pair in a Quantitative Fluorescence-PCR (QF-PCR) reaction with the maximum of 24 cycles (end point of exponential phase). The fluorescence PCR products were separated on automatic sequencer (ABI PRISM ® 3730 XL DNA Sequencer, Applied Biosystem, Milan, Italy). For copy number quantification we compared the peak area corresponding to the CDY1 and DAZ loci to their homologs CDY2 and DAZL, respectively.

RFLP analysis

For DAZ qualitative analysis, we chose the sequence family variant (SFV) at STS sY587 in intron 10, which discriminates DAZ1/2 from DAZ3/4. For CDY1, we used a SFV situated 7750 bp upstream of the CDY1 translation start codon (CDY7750), which distinguishes CDY1A from CDY1B. SFVs were scored by PCR followed by enzyme digestion using the following enzymes: DraI for DAZ sY587 and PvuII for CDY1–7750.

Y chromosome haplogroups definition

Subjects with partial AZFc deletions were genotyped for six binary markers (M168, M145, M9, M45, M96, LLy22g). In brief, multiplex PCR of the regions spanning the six binary makers was followed by enzymatic clean-up of PCR products and multiplex single-base primer extension (SNaPshot Multiplex System, Applied Biosystems) [27]. DNA purification and analysis of primer extension products by ABI PRISM ® 3730 XL DNA Sequencer. The combination of genotypes allowed us defining specific clades or haplogroup clusters: (i) AB; (ii) DE*DE; (iii) CFGHI or J; (iv) E; (v) KLMNO*OST; (vi) N1; (vii) P*QR (see Karafet et al. [28]). A single PCR of the 12f2 marker was performed to differentiate the J haplogroup from the CFGHI cluster.

Statistical analysis

SPSS software (version 25.0 Chicago, IL, USA) was used. We tested the significance of the observed differences in the incidence of the AZFc rearrangements between the TGCT patients and the control group using Fisher´s exact test: our null hypothesis was that the incidence is the same in the TGCT patients and in the control group. For comparisons of medians between groups of different genotypes, the non-parametric Mann–Whitney U-test for independent samples was applied. After Bonferroni correction considering two independent variables, (i.e., deletions and duplications), a p-value ≤ 0.025 was considered statistically significant for each test.

Results

Analysis of complete AZF deletions and partial AZFc rearrangements in TGCT patients

In this study we aimed at: (i) defining the frequency of complete AZF deletions, partial AZFc deletions, partial AZFc duplications and DAZ dosage alteration; (ii) providing a detailed molecular and phenotypical characterization of the rearrangements carriers.

Complete AZF deletions

None were found in the entire cohort of TGCT patients (n = 497).

Partial AZFc deletions

14/497 patients (2.8%) carried the gr/gr deletion (chrY:g.(24876071_25505070)_(25505734_25316178)del) (hg19) (LOVD3 DB-ID:chrY_000067, ID Variant:0000461229), whereas b2/b3 (chrY: g.(24524070_24872541)_(24876071_25505070)del) (hg19) (LOVD3 DB-ID: chrY_000068, ID Variant: 0000461230), b1/b3 (chrY: g.(23978074_24354700)_(25316178_25505734)del (hg19) (LOVD3 DB-ID: chrY_000135, ID Variant: 0000461228) and atypical (chrY: g.(25505596_26191376)_(26194116_26979967) del) (hg19) (LOVD3 DB-ID: chrY_000134, ID Variant: 0000461227) deletions were found in only 5 subjects (1.0%) with a frequency of 0.6%, 0.2%, and 0.2%, respectively (Table 1).

Table 1 Frequency of AZFc rearrangements and DAZ dosage alterations in TGCT patients (reported in subgroups based on histotypes and total sperm count) and in controls (reported in subgroups based on total sperm count)
Full size table

1) gr/gr deletion: this deletion type was more frequent in normozoospermic compared to non-normozoospermic TGCT patients (3.4% vs. 2.2%), and in seminoma vs. non-seminoma patients (3.1% vs. 2.1%) without reaching statistical significance (p = 0.305 and p = 0.385, respectively). The highest frequency was observed in normozoospermic seminoma patients (4.1%) (Table 1).

The histology and semen phenotypes together with the gr/gr deletion subtypes for the 14 gr/gr deletion carriers are reported in Table 2. As for histology, 78.6% were seminomas, whereas only 21.4% were non-seminomas. Only 2 patients (14.3%) were azoospermic, 3 (21.4%) were oligozoospermic and the large majority were normozoospermic (64.3%). The combined gene dosage and RFLP analysis allowed a more detailed molecular characterization of the deletion subtypes based on the number and type of CDY1 and DAZ copies. All patients, except one, showed a reduced CDY1 and DAZ gene copy number in respect to the reference sequence, i.e., 1 copy of CDY1 and 2 copies of DAZ. The increased CDY1 and DAZ copy number in one patient was due to deletion followed by multiple duplications. In fact, RFLP analysis confirmed the presence of 3 CDY1A and 6 DAZ3/4 copies. This subject was affected by non-seminoma and the semen phenotype was severe oligozoospermia.

Table 2 Phenotype/genotype description of the 14 TGCT patients with gr/gr deletion and of the 5 TGCT patients with other partial AZFc deletions
Full size table

Based on the type of the deleted CDY1 and DAZ copies, four different subtypes were identified: CDY1A+DAZ3/4 (7/14, 50%), CDY1A+DAZ1/2 (4/14, 28.6%), CDY1B+DAZ1/2 (2/14, 14.3%), and CDY1B+DAZ3/4 (1/14, 7.1%). Summing frequencies up, CDY1A was missing in 11 (78.6%), while CDY1B was missing only in 3 (21.4%) patients. Concerning DAZ copy number, DAZ3/4 was deleted in 8 (57.1%) and DAZ1/2 in 6 (42.9%) subjects. All oligo/azoospermic patients affected by seminoma were lacking CDY1A whereas those with normozoospermia had CDY1A copy deletion in 66.7%. Altogether, no relationship was found between the type of deleted gene copy and the cancer histology or the semen parameters.

The analysis of the Y chromosome haplogroups in the deletion carriers did not reveal a specific haplogroup associated with gr/gr deletion. The most frequent haplogroup in our cohort was P*QR (36%). No differences were observed in haplogroup distribution in relationship with the semen phenotype (Table 2).

2) b2/b3, b1/b3 and atypical deletions: similar to gr/gr, these deletions were slightly more frequent in normozoospermic compared to non-normozoospermic TGCT patients (1.1% vs 0.9%), and in seminoma (1.1%) than in non-seminoma patients (0.7%). The highest frequencies were observed in normozoospermic non-seminoma patients (1.4%) and non-normozoospermic seminoma (1.3%) (Table 1).

The molecular characterization and the associated semen and histology phenotypes of the 5 patients carrying the b2/b3, b1/b3 and atypical deletions are reported in Table 2. All patients with b2/b3 deletion had seminoma. Two patients (66.7%) were non-normozoospermic. Two patients showed reduced CDY1 and DAZ gene copy number (1 CDY1 copy and 2 DAZ copies), the other patient presented 2 CDY1 and 4 DAZ copies as a consequence of partial deletion followed by duplication, i.e., RFLP analysis revealed the deletion of the CDY1A and DAZ3/4 copies. The three b2/b3 deletion carriers showed a heterogeneous situation concerning the type of deleted CDY1 copy: CDY1A was deleted in 2 patients and CDY1B in 1 patient. Interestingly enough, all patients presented a deletion of DAZ3/4 copy. The b1/b3 deletion carrier had seminoma with normozoospermia and presented 2 CDY1 copies and 4 DAZ copies (deletion of DAZ1/2 followed by duplication). The patient with atypical deletion had non-seminoma and was normozoospermic and he showed CDY1B copy deletion without DAZ copy deletion, i.e., presenting DAZ1/2 and DAZ3/4 copies.

Concerning the Y haplogroup analysis, it revealed that 3 out of 5 patients belong to the CFGHI cluster (60%) whereas the remaining two presented haplogroup E and J, respectively (Table 2).

Partial AZFc duplications

(Chry:g.(24876071_25505070)_(25505734_ 25316178)dup) (hg19) (LOVD3 DB-ID: chrY_000086, ID Variant: 0000461232); (chrY:g.(24524070_24872541)_(25316578_28457316)dup) (hg19) (LOVD3 DB-ID: chrY_000074, ID Variant: 0000461231): these rearrangements are defined when both DAZ and CDY1 present increased copy number and it was observed in 22 TGCT patients (4.6%). We identified patients with 3 or 4 copies of CDY1 and 6 or 8 copies of DAZ. Partial AZFc duplications were observed more frequently in non-normozoospermic compared to normozoospermic TGCT patients (5.9% vs 3.5%) and in seminoma (4.7%) vs. non-seminoma patients (4.3%), without reaching statistical significance (p = 0.159 and p = 0.523, respectively). The highest frequency was observed in non-normozoospermic seminoma patients (5.9%) (Table 1).

DAZ dosage alteration

We defined it as any rearrangement (partial AZFc deletion or duplications or isolated DAZ duplication) causing an alteration in DAZ copy number from the canonical 4 copies (greater or less). In the 497 patients with TGCT, 51 (10.3%) presented alterations in DAZ dosage. DAZ dosage alteration was similar in normozoospermic and non-normozoospermic TGCT patients (10.4% vs. 10.0%) and in seminoma (10.8%) vs. non-seminoma patients (9.0%). The highest frequency was observed in normozoospermic seminoma patients (11.7%) (Table 1).

Comparison of the phenotypic characteristics between TGCT patients with and without gr/gr deletion, partial AZFc duplications, and DAZ dosage alteration

Genotype/phenotype correlation is reported in Table 3. For all AZFc rearrangements the most frequent histology was seminoma both in carriers and non-carriers ranging between 70 and 78%. The normozoospermic phenotype was more frequently encountered in gr/gr deletion carriers (64.3%) than in non-carriers (53.6%). On the contrary, partial AZFc duplication carriers were more frequently non-normozoospermic than non-carriers (59.1% vs. 45.8%, respectively). No statistical differences were found between carriers and non-carriers of the distinct rearrangements as far as testis volume, hormonal dosage values and routine sperm parameters are concerned (Table 3 and Supplementary Table 1). History of cryptorchidism was more frequent in TGCT patients with gr/gr deletion than in those without deletion (27.3% vs. 15%; p = 0.27). History of cryptorchidism was absent in partial AZFc duplications. No familial cases of TGCT and history of hypospadias were reported in carriers of the distinct rearrangements.

Table 3 Comparison of the phenotypic characteristics between TGCT patients with and without gr/gr deletion, partial AZFc duplications and DAZ dosage alteration. a) Histology and semen phenotype; b) testis volume, hormonal dosage and clinical risk factors. The frequency between subgroups did not result significantly different for any of the analyzed parameters
Full size table

Comparison of the frequency of AZFc rearrangements between TGCT patients and controls

Partial AZFc deletions

The frequency of all partial AZFc deletions carriers was higher in our TGCT cohort than in the control group (3.8% vs. 2.5%), without reaching statistical significance (Table 4). Partial AZFc deletions confer significant TGCT risk in the following comparisons performed against the normozoospermic controls: i) the entire TGCT group (OR = 3.5, 95% CI = 1.7–7.3, p = 0.001); ii) normozoospermic TGCT (OR = 4.1, 95% CI = 1.8–9.3, p = 0.001). In addition, partial AZFc deletions were observed more frequently in non-normozoospermic TGCT compared to normozoospermic controls (3.1% vs. 1.1%), although not reaching statistical significance (Table 4).

Table 4 Comparison of the frequency of gr/gr deletion, partial AZFc duplications and DAZ dosage alteration between TGCT patients and control subjects (total and subgroups based on total sperm count: normozoospermia vs. non-normozoospermia)
Full size table

1) gr/gr deletion: although the frequency of this deletion type was higher in TGCT patients than in the control group (2.8% vs. 2.0%), this difference was not statistically significant (Table 4). It is worth noticing that, in the control cohort, non-normozoospermic patients presented a significantly higher frequency (3.1%) compared to normozoospermic subjects (0.9%) (OR = 3.4; 95% CI = 1.7–7.0; p = 0.000). The comparison of the deletion frequency between the entire TGCT group (2.8%) and the normozoospermic control group (0.9%) shows a 3-fold increased risk for TGCT in the deletion carriers (OR = 3.1; 95% CI = 1.4–7.0; p = 0.006) (Table 4). An even higher risk was observed when normozoospermic TGCT patients (3.4%) were compared to normozoospermic controls (0.9%) (OR = 3.7; 95% CI = 1.5–9.1; p = 0.006) (Table 4).

2) b2/b3, b1/b3, and atypical deletions: the frequency of these deletions was higher in the TGCT group (1%) in respect to the controls (0.5%) but without reaching statistical significance, p = 0.191.

Partial AZFc duplications

These rearrangements were observed in 22 TGCT patients (4.6%), while they were detected in 51/1373 controls (3.7%). We did not found significant differences between the TGCT group and the control group, even when the comparison was done in subgroups based on seminal parameters (Table 4).

DAZ dosage alteration

The frequency was higher in TGCT patients than in the control group (10.3% vs. 7.8% p = 0.056). It is worth noticing that in the control cohort, non-normozoospermic patients presented a significantly higher frequency (9.9%) compared to normozoospermic subjects (6.0%) (OR = 1.7; 95% CI = 1.2–2.6; p = 0.004). The dosage alteration frequency between the entire TGCT group (10.3%) and the normozoospermic control group was statistically significantly higher (OR = 1.8; 95% CI = 1.2–2.7; p = 0.004) (Table 4). The comparison of normozoospermic controls vs. normozoospermic TGCT patients showed that DAZ dosage alteration confers a significant TGCT risk (OR = 1.8; 95% CI = 1.1–3.0; p = 0.013) (Table 4). Concerning the non-normozoospermic TGCT patients, a higher frequency of the DAZ dosage alteration in this subgroup of patient vs. normozoospermic control group was found, with a p-value close to significance (p = 0.027) (Table 4).

Discussion

Testicular germ cell tumors (TGCTs) derive from a pre-invasive precursor lesion, called germ cell neoplasia in situ (GCNIS) and show an incidence that peaks in young adulthood [1, 2]. Basic research studies and epidemiological trends support the hypothesis that TGCT is of fetal origin and it is due to failure of normal fetal programming of the differentiation of primordial germ cells through a gonocyte stage into spermatogonia [2]. The differentiation process of primordial germ cells/gonocytes is regulated through several germ cell-specific genes including Y chromosome-linked ones. The alteration of germ cell differentiation might set the basis both for the development of testicular tumour and for impaired spermatogenesis.

The AZF genes on the long arm of the Y chromosome are important regulators of spermatogenesis and genes belonging to the AZFc region are specifically expressed in male germ cells. The most frequent, known molecular genetic cause of oligo/azoospermia are complete AZF deletions [14] and their role in TGCT has been ruled out by previous studies [16,17,18, 21]. On the contrary, the role of gr/gr deletion (removing half of the AZFc gene content) is still debated due to contrasting results in the literature [21,22,23]. The largest multiethnic study by Nathanson et al. [22] tested for the gr/gr deletion in 1842 TGCT cases and 2599 unaffected men. TGCT cases were significantly more likely to carry the gr/gr deletion compared to unaffected males (OR = 2.1; 95% CI = 1.3–3.6; p = 0.005) and this association was stronger in men with seminoma than for those with non-seminoma. In a subsequent study [23], which evaluated 263 TGCT cases (167 with family history and 96 sporadic cases), the authors were not able to demonstrate a significant association neither for gr/gr deletion nor for other rare partial AZFc deletions. In these two studies the fertility status and/or sperm parameter of the patients were not available. The only study in which semen parameters were reported is based on a small Italian cohort (n = 118) in which no gr/gr deletion carriers were found among TGCT patients [21]. Hence, it remains to be established whether the association between gr/gr deletion and TGCT is restricted to a specific seminal subgroup of patients (subjects with reduced sperm count).

Our study has been designed to evaluate the role of partial AZFc deletions/duplications and specifically the DAZ gene dosage variation in TGCT in the largest European study population available to date.

Our first objective was to elucidate the role of gr/gr deletion as genetic risk factor for TGCT in two Mediterranean populations with full semen/andrological characterization. When dealing with Y chromosome-related risk factors, matching for ethnicity and geographic origin is fundamental in order to avoid population stratification bias [29]. Consequently, much care was taken for geographic and ethnic matching in our study. Semen parameters were known both for cases and controls, allowing us to evaluate the role of gr/gr deletion as TGCT genetic risk factor in combination with spermatogenic efficiency. The similar frequency of gr/gr deletion in the TGCT group (2.8%) and the entire control group (2.0%) is due to the relatively high frequency of gr/gr deletion in patients with idiopathic oligo/azoospermia (3.1%). In fact, in our previous studies we demonstrated that gr/gr deletion is significantly associated with oligozoospermia in the Spanish and Italian populations [13, 27, 30]. Surprisingly enough, our most important finding is related to the normozoospermic group: while only 0.9% of controls were carrier of gr/gr deletion, this percentage was higher in normozoospermic TGCT patients (3.4%) and even higher in normozoospermic seminoma patients (4.1%). gr/gr deletion in normozoospermic subjects confers almost 4-fold increased risk for developing testis tumor.

The specific genes removed can vary between gr/gr deletions, and this has been suggested as the most direct modulating factor for the associated semen phenotype [13]. In this regard, mainly the DAZ and CDY1 copies, which lie within the deleted AZFc region, have been examined as possible predictors for pathogenicity. The loss of DAZ1/DAZ2 and CDY1A has been proposed as more deleterious than the removal of DAZ3/DAZ4 and CDY1B [26, 31, 32], however this was not fully proved by a large multicenter study [33]. Concerning DAZ gene copies, different members of the DAZ gene family have different number of RNA recognition motifs (RRMs) and DAZ repeats, which may confer different functional activity to the four DAZ copies [33]. Hence, we were interested in investigating on the deletion “haplotypes” (based on different combinations of DAZ and CDY1 copies) in TGCT patients with and without altered spermatogenesis and with different tumor histology. Moreover, we questioned whether the deletion haplotypes in TGCT patients were different from that of the controls especially concerning the deleted DAZ copies. 11 (78.6%) patients presented a CDY1A deletion, and data break-down showed that this type of deletion was the most frequent in all semen and histology phenotypes. A similarly high frequency of CDY1A deletion was found in the control group (77.8%). Concerning the subtype of DAZ gene copy deletion, there was no predominant copy loss in specific sperm count-based subgroups or histology types. The same finding was observed in the control group. This data indicates that the type of deleted CDY1 or DAZ copies are irrelevant to sperm output or histology type in TGCT patients carrying the gr/gr deletion. Most importantly, the deletion “haplotype” was not different between TGCT cases and tumour-free controls.

Beside gr/gr deletions, three other types of deletions were identified in our cohorts (b2/b3, b1/b3 and atypical deletions). The frequency of these deletions was extremely rare both in patients and controls (0.6% vs. 0.4%, 0.2% vs. 0.0%, 0.2% vs. 0.2%, respectively), hence their role, if any, in TGCT remains to be established.

Given that partial AZFc deletions perturb AZFc gene dosage, we were interested in analyzing the role of an increased gene dosage (>4 for DAZ and >2 for CDY1, respectively) due to partial duplications. Although case/control association studies failed to reach to a univocal conclusion in respect to the hypothetical “optimal” AZFc gene dosage needed for normal spermatogenesis [27, 30, 34, 35], it seems that a positive selection for an “optimal” gene dosage does exist (four DAZ and two CDY1 copies) during evolution [36]. Hence, our working hypothesis was a possible predisposing effect of AZFc dosage variation (not only deficit but also an excess) to TGCT development. Partial AZFc duplications frequency was 4.6% in TGCT patients while it was 3.7% in controls. Therefore, on the contrary of gr/gr deletion, we were unable to detect a significant effect of partial AZFc duplications on TGCT development.

We have separately evaluated the DAZ gene dosage variation (as a consequence of partial AZFc deletions, duplications or isolated DAZ duplication event). We found that patients with TGCT have a higher frequency of alterations of the DAZ dosage with respect to the entire control group (10.3% vs. 7.8%) (Table 4). The comparison of the entire TGCT group vs. normozoospermic controls showed a significant difference (10.3% vs. 6.0%, p = 0.004) indicating that DAZ dosage alteration in normozoospermic subjects confers almost 2-fold increased risk for developing TGCT (OR = 1.8, 95% CI = 1.2–2.7).

The biological link between AZFc gene dosage variation (especially its reduction due to deletions) and tumourigenesis remains elusive. One hypothesis could be that DAZ gene dosage may influence the differentiation of the gonocytes. In fact, it has been demonstrated that DAZ is present in both the nuclei and cytoplasm of fetal gonocytes and in spermatogonial nuclei [37]. Hence, based on its expression profile, DAZ gene dosage variation may play a role in TGCT etiology. Alternatively, we can also speculate that the “fragility” expressed as deletions/duplications of the Y chromosome is a marker for general “genomic instability” potentially leading to tumors, including TGCT.

In conclusion, our data provide evidence that the gr/gr deletion is a risk factor for TGCT also in the Italian and Spanish populations and that normozoospermic gr/gr deletion carriers have almost 4-fold increased risk to develop TGCT. The observed significant association between gr/gr deletion and normozoospermic TGCT, suggests that this deletion is an independent risk factor from impaired spermatogenesis. In line with this finding, a recent epidemiological study reported an increased risk for TGCT in first-degree relatives of subfertile normozoospermic men [38]. Diagnosing gr/gr deletion in a TGCT patient has a potential clinical relevance to his male relatives, especially to the brothers in whom preventive measures such as regular autopalpation or ultrasound scan of the testis can be proposed. In some countries, gr/gr deletion screening is already part of the routine genetic diagnosis of oligozoospermic men [14]. This implies that similarly to brothers of gr/gr deletion carriers, also infertile patients diagnosed with this type of deletion should receive careful long-term andrological follow-up.