|Home | About | Journals | Submit | Contact Us | Français|
Testis cancer is an increasing problem, especially in northern European male populations. However, survival has improved dramatically over one generation. Environmental factors may have a role in the aetiology with high oestrogen concentrations implicated. Testis cancer is subdivided between seminoma and non‐seminoma. At presentation, a testicular lump is the most common finding and radical inguinal orchidectomy is recommended for most. Further multidisciplinary management is determined by histological subtype and stage and involves chemotherapy, radiotherapy and surgery, with many patients only undergoing surveillance. There is increasing emphasis on reducing toxicity of treatments in long term survivors. Treatment refractory testis cancer remains a significant challenge.
Testis cancer is the most common solid malignancy of young men, but only accounts for about 1% of all cancers in men.1 The incidence has been increasing worldwide. Germ cell tumours (GCT) represent over 95% of these cancers and for the purpose of this review; references to testis cancer will imply GCTs.
Advances in the last 40 years in treatments of this group of patients have led to dramatic improvements in disease‐free survival. Better treatments are, however, still needed for high risk patients. There are many long term survivors and late toxicities from treatment are more prominent than in other solid malignancies. It is important to balance the risk of long term toxicities with effective treatment without reducing efficacy.
The incidence of testis cancer has been steadily increasing over the last 40 years.2 It appears to be most common in northern European populations with age standardised incidence rates between 4 and 10 per 100000, whereas in Asian, African and African American men, incidence rates are much lower, ranging between 0.2 to 1 per 100000.3 The peak incidence is between the ages of 15 to 35 years. Five year survival rates have increased significantly over the last 30 years from about 63% to more than 90%.3
The aetiology of testis cancer is not clearly understood. Rising incidence suggests a role for environmental factors, which is supported by an increasing frequency of other testicular problems such as declining sperm counts and increasing incidence of testicular maldescent. High oestrogen concentrations in utero have been suggested as a potential common factor. Sons of women who received the synthetic oestrogen, diethylstilboestrol, had an increased incidence of testicular abnormalities. Certain environmental oestrogens have been implicated in feminisation of male animals. To date there has been no direct link between oestrogen exposure and the risk of testis cancer. There are, however, several known risk factors (box 1).3
Approximately 2% of individuals with testis cancer report an affected first degree relative, and concordant twin studies have also demonstrated a higher incidence of testis cancer in monozygotic compared to dizygotic twins.4 Analyses of affected families have so far failed to identify a specific gene although linkage to the Xq27 region has been reported.5 Duplication or amplification of the short arm of chromosome 12 (12p) is seen in almost all cases of testis cancer, implying that a key gene is present in this area.6
GCTs can be subdivided into three main groups: infantile/prepubertal, adolescent/young adult, and spermatocytic seminoma. They originate from germ cells at different stages of development. By far the most common is the adolescent/young adult type and this is the type discussed in this review.
Intratubular germ cell neoplasia, unclassified type (ITGCN) is considered to be the precursor to invasive GCTs. It is also often referred to as carcinoma in situ. It is thought that primordial germ cells undergo abnormal cell division in response to environmental factors in utero giving rise to ITGCN.7 This is followed by the duplication of 12p, as well as several other chromosomal abnormalities, rendering these cells susceptible to stimulation by gonadotrophins and subsequent development of invasive tumours. Activation to pluripotency of neoplastic germ cells of these tumours gives rise to NSGCTs (non‐seminomatous germ cell tumours—that is, yolk sac tumour, embryonal carcinoma and choriocarcinoma) in a way similar to the reprogramming of a primordial germ cell to an embryonic germ cell. Factors causing reprogramming are unknown.
There are two main subtypes of GCTs arising in young men:
There are two main classifications used in the histopathological examination of testicular tumours, the British Testicular Tumour Panel classification and the World Health Organization system (table 11).). The former is widely used in the UK and Australia, and the latter is commonly used in North America and Europe.
It is common practice for pathologists to describe all cell types within the specimen as well as commenting on the presence or absence of lymphovascular invasion, involvement of tunica albuginea, tunica vaginalis, rete testis and spermatic cord. However, in practical terms, further clinical management depends on whether the tumour is seminoma or NSGCT and the presence or absence of lymphovascular invasion.
Testis cancers commonly present as a unilateral lump or painless swelling noticed incidentally.8 Pain is less common, with a third of patients presenting with a dull ache, and acute pain is uncommon occurring in 10% of patients at presentation. Testis cancers uncommonly present with symptoms attributable to metastatic disease. These are summarised in box 2.
In any man with a solid firm mass within the testis, the diagnosis is testis cancer until proven otherwise. Prompt diagnosis and treatment offers the patient the best chance of a cure. Occasionally, metastatic disease can be rapidly progressive, warranting emergency treatment. An example of this is demonstrated by the chest radiograph appearances taken 4 days apart of a 32‐year‐old man with poor prognosis advanced germ cell cancer (fig 11).). The differential diagnosis of testis cancer is summarised in box 3.
Initial diagnostic evaluation of men with suspected testis cancer is with scrotal ultrasound. This is effective in distinguishing intrinsic from extrinsic testicular lesions.9 A plain film is carried out before surgery to exclude overt metastatic disease. The only other radiological investigation that is routinely performed is high resolution computed tomography (CT) scanning of the chest, abdomen and pelvis.
Three serum tumour markers must be measured in any man with suspected testis cancer: α‐fetoprotein (AFP), the β unit of human chorionic gonadotrophin (β‐HCG) and lactate dehydrogenase (LDH). In NSGCT the serum concentrations of AFP and β‐HCG are elevated in over 80% of men. In pure seminoma, AFP is not elevated and fewer than 20% of men have elevated β‐HCG. LDH is less specific but has independent prognostic value in patients with advanced GCTs. It is increased in 60% of NSGCTs and 80% of seminomas. Serum tumour markers alone are not diagnostic of testis cancer but very high values in men rarely occur outside of testis cancer.
Semen cryopreservation should be offered to all men diagnosed with testis cancer before starting treatment if they wish to preserve fertility. Ideally, a baseline sperm count and sperm banking should be performed before radiological diagnostic evaluation to avoid radiation exposure of sperm, but this is not always feasible. All cases of testis cancer should be discussed by a multidisciplinary team comprising specialist surgeons, oncologists, histopathologists, radiologists and specialist nurses.
Radical inguinal orchidectomy is recommended for all patients with suspected testis cancer to allow accurate histological evaluation as well as local tumour control. The procedure involves isolating and clamping the spermatic cord at the external inguinal ring, exteriorising the testis with its tunics, opening the tunica vaginalis, and inspecting and palpating the testis carefully.10 If the diagnosis is unclear, a biopsy is taken and examined under frozen section. Once the diagnosis is established, the inguinal canal is opened, the spermatic cord is divided at the level of the internal inguinal ring, and the testis is removed. The inguinal approach is preferable to the scrotal approach as there is a theoretical risk of lymphatic spread of testicular cancer cells to the scrotal skin and its lymphatic drainage. Complications include retroperitoneal haemorrhage, wound infection, seroma formation, local hypoaesthesia and persistent inguinal and scrotal neuralgia.
Partial orchidectomy may have a role in selected patients in whom the likelihood of a testicular neoplasm is low, based on ultrasonographic findings, age, physical examination and tumour markers.11 In men with bilateral tumours, consideration should be given to this approach, especially if maintaining fertility is important. However, in the majority of men undergoing surgery the radical approach is currently advisable, although interest is increasing in testis preservation techniques through clinical trials.
Patients should be offered the option of a testicular prosthesis and consideration should be given to contralateral testicular biopsy. About 5% of patients with testis cancer have ITGCN in their contralateral testis and in the majority of cases this will proceed to an invasive GCT.12 Biopsy is usually considered in high risk patients as defined by a small testis volume (<12 ml), history of cryptorchidism, and young age (<30 years).13 If ITGCN is found then the options available are low dose radiotherapy to prevent tumour progression, or surveillance and surgery once the need arises. Radiotherapy is not recommended for men who wish to preserve their fertility.
A clinical staging system that is widely used is the Royal Marsden Hospital classification (table 22).14 After surgery it is essential to monitor tumour markers and assess their rate of decline. If markers do not decline as predicted or start to rise postoperatively the contralateral testis should be examined for a metachronous primary. If this is not found, the patient should be treated as having metastatic disease, even if imaging studies do not corroborate this.
The International Germ Cell Cancer Collaborative Group (IGCCG) prognostic classification is the generally accepted prognostic tool used for treatment decisions and eligibility for clinical trials in metastatic disease (table 33).15 It relies on the extent of disease and tumour markers together with the primary site. This was in part developed because of the variations in classifications and staging systems used throughout the world, making it difficult to compare trial data. Supplementary prognostic information is gained by observing the rate of decline of tumour markers after starting chemotherapy. Further management of GCTs depends on their clinical stage and the pathological classification of the tumour.
About 85% of men present with stage I disease and a further 10% present with stage II disease. The options post‐orchidectomy involve active surveillance, radiotherapy or single agent chemotherapy.
About 15% of men with stage I disease will relapse within 4 years.16 Salvage rates are high so active surveillance has the advantage of avoiding unnecessary treatment and adverse effects. However, men undergoing surveillance may need multi‐agent chemotherapy on relapse. Regular and reliable attendance is also necessary. In one pooled analysis of over 600 men undergoing surveillance for stage I seminoma for an average of 7 years, the only significant risk factors for risk of relapse were tumour size 4 cm and rete testis invasion.17
Traditionally radiotherapy was administered to the ipsilateral renal hilum and pelvic lymph nodes and the bilateral para‐aortic nodes as well as the regional lymph nodes of the involved testis. This so called “dog‐leg” field yielded excellent results with 5 year relapse‐free survival rates in excess of 94%.18 However, significant gastrointestinal toxicity and increased risk of second malignancies have led to strategies to reduce the radiation field without compromising on efficacy.
More recently, the radiation field has been limited to the para‐aortic nodes (“PA strip”). The evidence for this was provided by a number of trials including a prospective trial by the Medical Research Council (MRC) Testicular Tumour Working Group.19 Three year survival was similar as were the total number of relapses, with a slightly higher pelvic relapse rate in the smaller field group. There was a significant reduction in short term morbidity in the smaller field group.
A recent MRC trial comparing two different doses of irradiation (30 Gy in 15 fractions compared to 20 Gy in 10 fractions) in stage I seminoma demonstrated no difference in disease‐free survival with a median follow up of 5 years.20 Quality of life data indicated better tolerance for lower dose radiotherapy suggesting that the dose of radiation can be safely reduced.
Single agent carboplatin has recently become established as an alternative to adjuvant radiotherapy in stage I seminoma. A pooled analysis of phase II trials using two cycles of adjuvant carboplatin demonstrated a relapse rate of 2.9%.21 A joint MRC and European Organisation for Research and Treatment of Cancer (EORTC) trial recently reported results of a randomised phase III trial comparing adjuvant radiotherapy to a single course of carboplatin dosed at an area under the concentration × time curve (AUC) of 7.22 Almost 1500 men with stage I seminoma were randomised and with a median follow up of 4 years there was no significant difference in relapse‐free survival. There was a trend of relapse in the para‐aortic nodes in the carboplatin arm. Longer term follow up is needed to determine whether carboplatin can safely replace radiotherapy, but many centres are now offering it as an alternative. Whether patients should be given one or two cycles remains controversial as most phase II trials have shown lower relapse rates when two cycles are given.
All options should be considered for patients with stage I disease, although less radiotherapy is being given and increasing numbers of patients are undergoing surveillance or chemotherapy.
Stage II disease is defined as metastatic disease that is confined to the infradiaphragmatic lymphatics. Treatment usually consists of radiotherapy or platinum based chemotherapy post‐orchidectomy. The optimal treatment depends on the volume of nodal disease, with chemotherapy being offered to patients with bulkier disease due to the increased risk of renal damage and out of field recurrences with radiotherapy. The definition of bulky disease varies, but in general this is nodal disease >5 cm in greatest dimension.
Non‐bulky disease is usually treated with radiotherapy alone to the para‐aortic and high ipsilateral iliac lymph nodes.23 Five year survival rates average about 90% and, with the benefit of effective salvage treatments, overall cure rate is also about 90%.
There is some evidence supporting the use of chemotherapy in this setting. A recent Spanish study demonstrated a progression‐free survival rate of 91% and overall survival rate of 98%, with no late toxicities using bleomycin, etoposide and cisplatin (BEP) or etoposide and cisplatin (EP).24 Randomised trials are needed comparing this approach to radiotherapy before any definitive conclusions can be made.
Patients with bulky stage II seminoma either at presentation or at relapse tend not to respond to radiotherapy as well, with 5 year disease‐free survival rates of about 65%; with the benefit of salvage treatment, 5 year overall survival is about 77%.25 Chemotherapy has been investigated as first line treatment following orchidectomy. The optimal chemotherapy regimen has not been defined but four courses of EP or three courses of BEP are usually recommended.
Management of post‐treatment residual masses tends to be determined by their size. If <3 cm then surveillance alone is adequate; if 3 cm then management is not as simple, with some recommending surgical resection.26 There is little evidence supporting adjuvant treatment after combination chemotherapy for advanced seminomas and some of these masses show fibrosis alone.27 Positron emission tomography (PET) scanning is often utilised to determine whether there is any viable tumour in such cases. In the multicentre SEMPET study 51 patients who had CT confirmed residual masses post‐chemotherapy for bulky seminoma underwent PET scanning.28 All 19 cases with residual lesions >3 cm and 35 of 37 cases with lesions <3 cm were correctly predicted by PET scanning. Specificity and sensitivity for FDG (fluoro‐deoxy‐D‐glucose) PET were 100% and 80% compared with 74% and 70% for CT scans, respectively.
The options for men with stage I NSGCT in the UK consist of surveillance with treatment given at relapse or chemotherapy. In the USA and parts of Europe, primary retroperitoneal lymph node dissection (RPLND) is also considered. Surveillance tends to be reserved for highly motivated men with low risk disease. Patients at high risk of relapse can be identified by certain histological factors that are summarised in box 4.29 In clinical practice, the presence of vascular invasion alone is widely accepted as the parameter on which to base the decision to administer adjuvant chemotherapy. In patients with none of these risk factors, relapse rates are between 10–15%, while if several risk factors are present then risk of relapse approaches 50%.30
The main advantage of surveillance is the avoidance of unnecessary treatment for the 70% of patients with stage I disease who do not relapse after orchidectomy. However, there are disadvantages to this approach: intensive follow up requiring additional radiation exposure and uncertainty for the patient, and an additional chemotherapy cycle if the patients do relapse compared to two cycles in the adjuvant setting.
Adjuvant chemotherapy consists of two cycles of BEP with an etoposide dose of 360 mg/m2. Recurrence rates are reduced from 50% to <5%. The Anglian Germ Cell Cancer Group published a retrospective review of 382 patients with stage I NSGCT treated between 1978 and 2000; all men before 1986 were followed by surveillance, and from 1986 onwards those with >30% risk of relapse were offered adjuvant chemotherapy.31 Relapse occurred in 30% on surveillance and mortality rate was 2.6%. Only 4% of men who received adjuvant treatment relapsed and 1.4% died from progressive disease.
This approach has the advantages of reducing the anxiety of recurrence and intensive surveillance for the patient as well as reducing the risk of recurrence at all sites. Short term and long term toxicities are of concern and are addressed later in this review.
In the USA, RPLND is the preferred treatment option in this patient group. This approach is considered to be safe especially with nerve sparing surgery, and additional staging information is obtained as current radiological techniques are inadequate in evaluation of retroperitoneal nodal disease in almost a third of patients.32 Over‐treatment with chemotherapy is also avoided. There is a risk of retrograde ejaculation and hence infertility, but this is <5% with nerve sparing techniques in specialist centres.33 Criticisms of the surgical approach include the increased morbidity of surgery and the fact that 10% of patients relapse outside of the retroperitoneum.10 After RPLND the options available are either surveillance with chemotherapy at relapse or adjuvant chemotherapy.
Five year survival rates for GCTs now exceed 95% and even patients with disseminated disease are very curable—a remarkable statistic for a solid malignancy. Figure 22 demonstrates complete response in a patient with an advanced, widespread metastatic GCT. The main reason for this improvement has been treatment advances since the early 1970s. The pivotal study, published in 1981, combined cisplatin with bleomycin and vinblastine resulting in cure rates exceeding 50%.34 Vinblastine was replaced with etoposide, improving efficacy and limiting toxicities, and the three‐drug combination of BEP has become the mainstay of treatment for men with advanced seminoma and NSGCTs.35 Cure rates are around 80% and no regimen to date has been shown to be superior; the scheduling and number of cycles varies according to prognosis. Wherever possible, men should be referred to centres with expertise in management of GCTs.
Men with testicular cancer are separated into prognostic groups derived from the IGCCC as previously discussed. The distribution of patients in each category is summarised in table 44.
Long term outcome of chemotherapy in this risk group is excellent with long term relapse‐free survival in excess of 90%. Modifications to the use of four cycles of BEP have been made by decreasing the number of cycles from four to three. In 2001 the EORTC published data confirming that for good prognostic disease three cycles of 3 day BEP (500 mg/m2) was sufficient with a progression‐free survival at 2 years of 90.4%.36 An earlier US study also demonstrated similar results.37
Studies have also been performed omitting bleomycin to avoid the serious pulmonary toxicity associated with this drug. A US study compared three cycles of BEP with three cycles of EP in good prognosis patients with metastatic disease,38 and a European study compared four cycles of EP with four of BEP.39 The consensus based on these trials is that the omission of bleomycin leads to an inferior outcome. A more recent trial by the same European group demonstrated equivalence of three versus four cycles of BEP and of 5 days versus 3 days per cycle in good prognosis germ cell cancer.40 The dose of etoposide should be 500 mg/m2 per cycle. In general, most cancer centres recommend three cycles of 3 day BEP in this patient group, with four cycles of EP an alternative for patients with compromised lung function.
For this group of patients four cycles of 5 day BEP is the standard. Five year survival for intermediate and poor risk patients is 80% and 48%, respectively.15 There is significant interest in improving the outcome using different regimens of chemotherapy, but as this group of patients comprises a minority of all patients with metastatic disease, generating adequate randomised data can be challenging. The MRC are currently investigating the CBOP/BEP (carboplatin, bleomycin, vincristine and cisplatin followed by BEP) regimen. A phase II study demonstrated an 87.6% 5 year survival which has led to the phase III study comparing with BEP currently recruiting.41
Men in these risk groups often have residual tumours seen on imaging studies at the completion of chemotherapy. These patients with residual disease, particularly in the retroperitoneum but also mediastinum and neck, should be considered for further surgery to render them disease‐free. Figure 33 demonstrates an example of multi‐modality treatment of a patient who underwent initial chemotherapy followed by retroperitoneal surgery. The resected specimen revealed mainly necrotic tissue, but some cells stained positive for AFP so he underwent retroperitoneal radiotherapy. This case is an excellent example of the multidisciplinary management of testis cancer.
Furthermore, pathological review of the resected specimen confirming viable tumour would suggest a significant risk of relapse.42 These patients can be offered further chemotherapy or radiotherapy; however, there is some debate over how beneficial this approach actually is.
Following first line chemotherapy, up to 50% of men with intermediate or poor risk germ cell cancer will require salvage treatment for relapsed disease. Men with stage I NSGCT or seminoma who relapse while undergoing surveillance can usually be salvaged with standard dose cisplatin based chemotherapy or RPLND followed by chemotherapy, and are largely cured by their subsequent treatment.
Relapsed germ cell cancer is still a chemosensitive disease and potentially curable in approximately 30% of cases. Several different relapse regimens have been investigated using cytotoxics that have demonstrated activity in the relapsed setting. The optimum salvage regimen still needs to be defined, but most patients retain platinum sensitivity at relapse. Vinblastine, etoposide and cisplatin (VIP) used as a salvage regimen has a complete response rate of 50% and long term survival of 30%.43 More recently paclitaxel has been added to other active drugs, notably ifosfamide and cisplatin, with a 19–77% complete response rate44,45 and an 85% 2 year survival rate.45 Suggested prognostic factors at relapse are inadequate response to initial treatment, progression‐free interval <2 years, and non‐testicular primary.
High dose chemotherapy with autologous stem cell rescue has been investigated as a second or third line treatment. Several phase II studies or retrospective analyses have demonstrated efficacy with acceptable toxicity. Longstanding complete remissions have been reported in 15–25% of patients. A retrospective series from Einhorn's group reported a 57% disease‐free rate with a median follow up of 39 months.46
However, a recent study suggested no difference in outcome between conventional standard treatment (cisplatin, ifosfamide and etoposide or vinblastine) and high dose treatment (carboplatin, etoposide and cyclophospamide).47 Similar complete and partial response rates were seen in each arm and there were no significant differences seen in overall survival.
Many patients with testis tumours are young and a proportion receive significant doses of chemotherapy and radiotherapy. As cure rates tend to be high, many are expected to achieve a normal life expectancy. Consequently, toxicities from treatments are increasingly important and have a bearing on optimising management. The most common long term toxicities are cardiovascular disease, second cancers, and reduction in fertility.
In a Royal Marsden cohort study of almost 1000 patients, there was a significantly increased risk of a cardiac event in all patients who had received treatment for testis cancer.48 The relative risk was 2.74 for radiotherapy and 2.59 for chemotherapy. In a retrospective analysis of 453 men with stage I or II seminoma treated with post‐orchidectomy radiotherapy, the standardised mortality ratio (SMR) for deaths from all causes was only significant beyond 15 years of follow up (SMR 1.89).49 For cardiac deaths the SMR was also significant beyond 15 years (1.95). The mechanism for cardiac mortality has been postulated to be related to irradiation either to the mediastinum or para‐aortic nodes.
The same study also reported a cancer SMR of 1.9 but the study was based on a cohort treated between 1951 and 1999, so whether the results can be translated to modern radiotherapy techniques has been questioned.48 The risk of second malignancy with chemotherapy is also a concern. In an international study, the relative risk of second malignancy at 10–15 years was 2.42.50 In the Royal Marsden study the relative risk of second malignancy was 1.6, which was not statistically significant.48
Approximately 50% of men with testis cancer have some degree of underlying impairment of spermatogenesis.51 It is not known why this occurs but it has been suggested that common aetiological factors are responsible for both low semen quality and testis cancer. All patients should be checked for hypogonadism (luteinising hormone and β‐HCG detected on assay) before initiation of treatment.
It is known that after chemotherapy, concentrations of follicle stimulating hormone and luteinising hormone rise and testosterone concentrations fall. It also commonly causes azoospermia, but most patients recover. In a recently published series from the Royal Marsden, 13% of patients had hypogonadism, and in patients who were normospermic before treatment, 80% had spermatogenesis 5 years after treatment.52 Another multicentre study reported a reduction in fertility rates of about 30% after cancer treatment, with radiation likely to have the biggest impact.53 Patients are usually advised not to attempt conception for 6 months after adjuvant treatment, but the evidence to support this is lacking. It is likely that irradiated sperm creates a damaged zygote that is miscarried before the woman detects pregnancy.
Testis cancer patients are more likely to be cured than those with most other solid cancers so there is a trend towards minimising exposure to toxic treatments. Recent trials using less radiotherapy and carboplatin in early stage seminoma have demonstrated this, as well as increasing use of surveillance strategies in all types of testis cancer. More accurate staging and assessment of response to treatments using PET scanning is evolving.
Perhaps the biggest challenge for oncologists treating testis cancer is the treatment of cisplatin refractory germ cell cancer. Recent trials have demonstrated promising results with combination chemotherapy, with response rates of over 30%.54,55,56 The best results have been seen with gemcitabine and oxaliplatin.54,55 Three drug combinations such as gemcitabine–oxaliplatin–paclitaxel or paclitaxel–gemcitabine–cisplatin are currently being developed in clinical trials, but significant toxicity in these heavily pre‐treated patients is a major limiting factor. It is hoped that successful development of new regimens in this patient group will allow new options for earlier stages of the disease. Patients should be strongly encouraged to participate in clinical trials. As our understanding of the molecular mechanisms of testis cancer development increases it is hoped that novel targeted treatments will be further explored.
We thank Dr Paul Rogers, consultant medical oncologist at the Royal Berkshire Cancer Centre, for providing the images for fig 1 and 2.
AFP - α‐fetoprotein, AUC, area under the concentration × time curve
β‐HCG - β‐human chorionic gonadotrophin
BEP - bleomycin, etoposide and cisplatin
CT - computed tomography
EORTC - European Organisation for Research and Treatment of Cancer
EP - etoposide and cisplatin
GCT - germ cell tumour
IGCCG - International Germ Cell Cancer Collaborative Group
ITGCN - intratubular germ cell neoplasia
LDH - lactate dehydrogenase
MRC - Medical Research Council
NSGCT - non‐seminomatous germ cell tumour
PET - positron emission tomography
RPLND - primary retroperitoneal lymph node dissection
SMR - standardised mortality ratio
VIP - vinblastine, etoposide and cisplatin
Competing interest statements: None declared