Follow-up of non-palpable testicular incidentalomas under 1 cm: does growth rate differentiate malignant and non-malignant lesions?

In this retrospective study, a consecutive series of 130 patients is investigated in whom small, non-palpable testicular tumours with negative tumour markers were initially followed up, with interval surgery undertaken only for growing lesions (n = 28), or patient/urologist preference (n = 11). Our data confirm that a conservative approach is effective in reducing unnecessary operations; surgery was avoided in 91/130 (70%) lesions.

For the 11 patients who were operated even though the lesion was stable during follow-up, all had benign tumours or non-neoplastic lesions. Conversely, 18/28 (64%) growing lesions were malignant, and evaluation of growth indicators helped identification of malignant tumours. The best cut-off values to differentiate between malignant and non-malignant histology were 3.47 × 10⁻³% volume/day, ≤ 179 days, > 10 × 10⁻³mm/day, and > 5 × 10⁻³mm/day for SGR, DT, ∆D_max, ∆D_av, respectively. An increase of the maximum diameter of the lesion of about one mm in three months and about two mm in six months, respectively, suggests malignancy and could lead to orchiectomy without further delay.

There is increasing evidence that orchidectomy constitutes overtreatment in patients with small incidentalomas [10], responsible for adverse fertility, hormonal, and mental health outcomes [11, 12]. Several investigations show that the majority of small testicular incidentalomas are benign or non-neoplastic lesions. Bieniek et al found six malignant lesions in 120 sub-centimetric testicular incidentalomas [3]. All malignant lesions were > 5 mm on initial imaging. Scandura et al investigated a series of 81 testicular incidentalomas < 1 cm and found that two-thirds of them, and all those < 5 mm were benign [10], with the conclusion that patients who undergo an orchidectomy for lesions < 5 mm are “Victims of Modern Imaging Technology”. In a recent systematic review, virtually all lesions < 3 mm and 86.6% of lesions < 5 mm were non-malignant [1]. Despite this, there remains a possibility that very small incidentalomas can represent small cancers. This is demonstrated in the current study, where the smallest malignant tumour identified had a volume of 176 mm³ and a maximum diameter of 4 mm.

Urologists are increasingly aware that a reasoned approach is necessary to reduce overtreatment while maintaining oncological safety. In these patients, radical orchidectomy should be avoided wherever possible, and in some cases, surgery may be avoided altogether. US surveillance is appropriate for the majority of testicular incidentalomas, where tumour markers are negative, and there are no clearly suspicious US characteristics [10].

Management strategies to reduce unnecessary surgery for small testicular incidentalomas include both improvement of lesion characterisation with multiparametric US [13‐19] and magnetic resonance imaging (MRI) [20] and management based on the growth characteristics of the lesion at US. The Scrotal and Penile Imaging Working Group of the ESUR suggests active surveillance in these lesions, with US examinations every three months in the first year and then annually [9]. The World Federation of Ultrasound in Medicine and Biology recommends surgery for lesions that grow or display increasing vascularity on interval imaging, while stable lesions can be managed conservatively [17].

According to the ESUR recommendations, patients with small testicular incidentalomas should be monitored every three months for 12 months and then annually. How long lesions should be followed remains unclear. To the best of our knowledge, there is no generally accepted recommendation [17].

In principle, slow-growing, or stable seminomas cannot be excluded [21]. We therefore prefer to recommend follow-up of stable lesions indefinitely until a larger body of evidence is available to provide more robust guidance on the total duration of follow-up that is necessary.

In our series, 11 non-growing lesions were eventually operated due to patient and/or referring urologist preference. 10/11 of them were non-neoplastic lesions, one was a benign tumour. Similarly, in the series of Toren et al six non-growing lesions were operated; four were non-neoplastic lesions, 2 were benign tumours [4]. These data show that even though non-growing small incidentalomas are virtually all benign leaving a small testicular lesion in place presents a cause of anxiety for both urologists and patients and will result in overtreatment if surgery is performed. It is therefore desirable to provide reassurance and education to both urologists and patients about the safety of surveillance.

Although, in principle, malignancy cannot be excluded in non-operated incidentalomas this is unlikely.

Our data suggest that malignant small, non-palpable incidentalomas are mostly seminomas (17/18 in our series, 94%). All were identified and removed within 18 months after discovery, and all were stage IA tumours.

In clinical stage I seminoma tumour size has a bearing on the risk of tumour recurrence [22]. Therefore, it could be argued that a delay in the operation could be detrimental for the patient. For these small tumours, the increased risk of tumour recurrence is minimal. According to Chung et al, an increase in size from 1 cm to 2 cm raises the recurrence risk from 9% to 11% [23]. Therefore, the follow-up period should not adversely affect the patient’s prognosis.

In our series, only two patients had lesions > 2 cm in diameter when operated. They were a benign leydigoma, and a seminoma in a monorchid patient, who declined operation until the lesion reached a maximum diameter of 28 mm; this patient did not receive further treatment and is free from relapse after 10 years of surveillance.

We believe that based on these data and disease prevalence, a non-aggressive management of small testicular incidentalomas is justified, and should be considered safe when follow-up includes self-examination and tumour marker assessment, in order to reduce the risk of interval tumour growth [9].

Slow-growing seminomas can occasionally be encountered which cannot be distinguished from benign tumours based on growth characteristics [21]. In our series, a slow-growing seminoma was encountered (SGR 2 × 10⁻³%/day, DT 316 days). The lesion was stable during the first six months of follow-up and then began to increase in size. The patient initially resistant to surgery could have been eligible for testis-sparing surgery, but orchidectomy was performed.

In our series, SGR, ∆D_max, and ∆D_av, were significantly different for malignant and non-malignant lesions, while DT did not reach a statistical significance. The limitations of DT for quantifying tumour growth have been documented for pulmonary nodules [24].

The variability of DT is much larger than that for SGR, making the latter a more suitable measure of volume change. In particular, it has been shown that DT largely overestimates the difference in the growth rate of slowly growing tumours and underestimates the difference in the growth rate of rapidly growing tumours, while SGR uniformly indicates the difference between growth rates throughout all ranges [7, 25]. Consequently, the growth rate of testicular incidentalomas should be expressed by SGR, or percentage volume increase per unit time, when technically feasible. In our clinical practice, however, an overall increase in the greater diameter ≥ 1 mm during two consecutive follow-up investigations was considered an unequivocal growth. Serial evaluations are necessary due to the large intrinsic errors in the measurement of the diameters of sub-centimetric lesions.

For the same reason, SGR and DT are not routinely evaluated at the time of the examination, since errors in calculating the diameters are multiplied. This study has shown that linear parameters such as ∆D_max, and ∆D_av, which are immediately derived from the measurement of the diameters, are effective in providing a rough estimation of tumour growth, sensitive enough for prompt identification of fast-growing lesions. An increase in the maximum diameter of the lesion of about one mm in three months and about two mm in six months, in particular, suggests malignancy and could lead to orchiectomy without further delay. The use of high-end equipment and higher frequency probes improves spatial and contrast resolution. It could improve the accuracy of the measurements and make prompt identification of slow-growing lesions easier. Also, 3D imaging with automatic volume calculation could be an option to reduce variability.

This study has several limitations. The most important is the retrospective design and the relatively small number of operated patients. The timing of follow-up was not the same in all patients, as 29 patients had one or more missed examinations during follow-up. As is commonly accepted for small tumours, in this study exponential growth has been postulated for calculating SGR and DT [7], but this assumption requires confirmation with large prospective studies.

A second limitation regards the intrinsic inter- and intra-observer variability of measuring sub-centimetric nodules at US. Even though ultrasound technology continues to progress and the technical standards for measurements of distance have improved, a significant error cannot be excluded, when considering size differences of less than 1 mm. To reduce the impact of inaccurate measurement, one follow-up evaluation was not considered enough to affirm that the lesion was growing, but at least two consecutive measurements were deemed necessary, in which the greater diameter increased in both, and the overall increase was ≥ 1 mm. This means, for a nodule of 5 mm, an increase of the greatest diameter of 20%.

Another limitation of this study is that only one lesion was a non-seminomatous tumour, and only two non-neoplastic lesions were growing. This likely reflects the real prevalence of disease, suggesting that among small non-palpable incidentalomas with negative tumour markers that are growing, benign and malignant tumours prevail, malignancy is more frequent (64% in our series), and most malignant tumours are seminomas (94% in our series).

The very low incidence of non-seminomatous tumours in this series is difficult to explain since the overall prevalence of seminomas is approximately 55–60% [26]. However, about 85% of seminomas are diagnosed as clinical stage I disease as compared with about 60% among non-seminomas [27, 28]. This may be explained by a faster progression of non-seminomatous tumours [29]. Our findings may have a similar explanation. If non-seminomatous tumours grow faster, it may be less likely to identify them when they are non-palpable, < 1 cm lesions. Moreover, since about 58% of clinical stage I non-seminomatous tumours have positive tumour markers [27], some of these lesions could also have been excluded for this reason.

In conclusion, growing parameters allow an educated assessment of benign and malignant small testicular incidentalomas even though overlap exists. These results are promising, but the strength of the evidence suffers from the retrospective design of the study. Prospective multicentre follow-up studies using multiparametric US are necessary to substantiate our results. Further improvement in lesion characterisation is now offered by the increasing capabilities of multiparametric US. Changes of lesions characteristics in the different US modes during follow-up, considered together with growing parameters, are expected to further improve lesion characterisation, leading to an evidence-based paradigm shift in the management of small testicular incidentalomas.