Introduction The incidence of testicular cancer is
The incidence of testicular cancer is increasing, with age-standardized rates of 5.6 cases/100,000 men diagnosed in the United States and 7.0 cases/100,000 men diagnosed in the United Kingdom annually [1,2]. Approximately sixty percent are seminomas (peak incidence 35 years), and the rest predominantly non-seminomatous germ-cell tumours (NSGCT, peak incidence 25 years). Most patients present with tumour located within the testis (stage 1 disease), and long-term survival is high (US five-year survival >95%, UK ten-year survival >98%) [3,4]. Patients presenting with recurrent or metastatic disease at diagnosis can be defined by prognostic category according to the site of the primary tumour and metastases, and the tumour marker level. It remains a highly curable cancer: good and poor prognostic disease have five-year survival of 92% (NSGCTs) / 86% (seminoma) and 48% respectively . The high rate of relatively young cancer survivors necessitates a review of the most efficient means of surveillance for recurrent disease.
During active surveillance following curative treatment for stage I disease, 10–20% of seminomas and 15–50% of NSGCT recur, most within two years . No consensus exists on the optimal follow-up schedule to detect recurrence, but European guidelines recommend intensive multimodal follow-up including physical examination, tumour markers, chest radiograph and either Abdominal-Pelvic Computed Tomography (CT) or Magnetic Resonance Imaging (MRI) scans . The biomarkers human choriogonadotropin (HCG), α-fetoprotein (AFP) and lactate dehydrogenase (LDH) are measured at a frequency determined by tumour stage and treatment received: approximately 3–4 times/year during the first three years and annually thereafter.
Previous reviews have discussed the diagnostic performance of these biomarkers for detecting recurrence of testicular germ cell tumours. These include a 2007 review focused on germ cell tumour surveillance outcomes including survival and quality of life  and more recent reviews that 2-Guanidinoethylmercaptosuccinic Acid looked at biochemical markers for the diagnosis of germ cell tumours more generally (primary as well as recurrent disease) [6,9,10]. A striking feature of these reviews is that, although they may represent contemporary clinical opinion, only the first of these is systematic in nature, and none presents complete diagnostic accuracy data on the performance of biomarkers used to detect testicular cancer recurrence. Therefore, although biomarkers are recommended in international guidelines , the evidence base underpinning their use in detecting recurrence is unclear. A systematic diagnostic test accuracy review , rather than a potentially selective narrative overview, is lacking.
This diagnostic test accuracy review aims to clarify this issue by quantifying the accuracy of blood AFP, HCG and LDH as triage tests for further investigation for testicular cancer recurrence in adults. If sufficiently accurate, these biomarkers (alone, or in combination) could provide a cost-effective means of reducing unnecessary follow-up investigations and hospital episodes, and the radiation burden [12,13], particularly as the value of clinical examination and chest radiographs are questioned as tools to detect recurrence . We also aimed to summarise the available evidence regarding cut-offs used to define a positive biomarker test for the purpose of diagnosis of testicular cancer recurrence.
Material and methods The review protocol was registered in advance on the PROSPERO website, registration number CRD42017074683 . Elevated biomarkers at the time of initial diagnosis of the primary tumour are of prognostic value and form part of the International Germ Cell Consensus Classification staging criteria . Biomarker levels are expected to normalise following orchiectomy and persistently elevated levels suggest there may be residual disease. For clarity, we have separated the prognostic role of biomarkers prior to treatment from their application in the diagnosis of recurrence. In this review we included only patients treated with curative intent and regarded biomarker elevation during the surveillance period as a potential indicator of disease recurrence.