Original article
Decision analysis defining optimal management of clinical stage 1 high-risk nonseminomatous germ cell testicular cancer with lymphovascular invasion

https://doi.org/10.1016/j.urolonc.2018.03.021Get rights and content

Highlights

  • •

    Treatment values of BEP ×1, RPLND, and surveillance were calculated in a decision analysis model.

  • •

    Values were based on patient preference for long-term morbidity and survival.

  • •

    BEP ×1 or RPLND outperformed surveillance in a patient with high-risk CS1 NSGCT.

  • •

    Surveillance would only be equivalent to RPNLD or BEP ×1 if postorchiectomy cure rate exceeds 82%.

Abstract

Background

Risk of recurrent disease for men with clinical stage 1 high-risk nonseminomatous germ cell testicular cancer (CS1 NSGCT) with lymphovascular invasion (LVI) after orchiectomy is 50% and current treatment options (surveillance [S], retroperitoneal lymph node dissection [RPLND], or 1 cycle of BEP [BEP ×1]) are associated with a 99% disease specific survival, therefore practice patterns vary. We performed a decision analysis using updated data of long-term complications for men with CS1 NSGCT with LVI to quantify and assess relative treatment values.

Methods

Decision analysis included previously defined utilities (via standard gamble) for posttreatment states of living from 0 (death from disease) to 1 (alive in perfect health) and updated morbidity probabilities. We quantified the values of S, RPLND, and BEP ×1 via the rollback method. Sensitivity analyses including a range of orchiectomy cure rates and utility values were performed.

Results

Estimated probabilities favoring treatment with RPLND (0.97) or BEP ×1 (0.97) were equivalent and superior to surveillance (0.88). Sensitivity analysis of orchiectomy cure rates (50%–100%) failed to find a cure rate that favored S over BEP ×1 or RPLND. Varying utility values for cure after S from 0.92 (previously defined utility) to 1 (perfect health), failed to find a viable utility state favoring S over BEP ×1 or RPLND. An orchiectomy cure rate of ≥82% would be required for S to equal treatment of either type.

Conclusions

We demonstrate that for surveillance to be superior to treatment with BEP ×1 or RPLND, the orchiectomy cure rate must be at least 82%, which is not expected in a patient population with high-risk CS1 NSGCT.

Introduction

Primary treatment of clinical stage I nonseminomatous germ cell tumor (CS1 NSGCT) is guided by expert opinion and varies by institution [1], [2], [3], [4], [5]. Acceptable treatment options include surveillance (S), retroperitoneal lymph node dissection (RPLND), or 1 cycle of chemotherapy with bleomycin, etoposide and cisplatin (BEP ×1) [6]. Risk-adapted approach further classifies CS1 NSGCT into high risk if there is evidence of lymphovascular invasion (LVI) or the embryonal carcinoma component is >50%. However, although risk of recurrence under surveillance for high-risk CS1 NSGCT increases from 30% to 50%, disease specific survival after adjuvant or salvage treatment reaches 99% and therefore treatment selection remains variable [7], [8], [9].

Current controversy is whether to treat or observe high-risk CS1 NSGCT. Previous study by Nguyen et al. [10] investigated the merit of a decision analysis to mathematically distinguish the value of each treatment for CS1 NSGCT taking into account oncological outcomes, treatment-related toxicities and patient preferences. They found that treatment is preferred over surveillance only when the risk of relapse after orchiectomy is 46%–54%, that is, for high-risk CS1 NSGCT. We performed a decision analysis in men with high-risk CS1 NSGCT as defined by the presence of LVI while using updated morbidity data including long-term complications to compare treatment values.

Section snippets

Decision tree modeling

Decision tree model was used to calculate and compare the values of each treatment (S, RPLND, or BEP ×1) in CS1 NSGCT with LVI. We incorporated probabilities of survival, recurrence, and treatment-related outcomes obtained from updated literature, including recently published data regarding recurrence risks and long-term treatment-related morbidity [11], [12], [13], [14], [20]. Model’s purpose was to investigate long-term outcomes, therefore morbidities that occurred during treatment period and

Results

Expected initial treatment values for RPLND and BEP ×1 in relationship to long-term treatment morbidities and patient preference were greater (RPLND = 0.97; BEP ×1 = 0.97) than for surveillance (0.88). Treatment either with RPLND or BEP ×1 outperformed surveillance after orchiectomy regardless of the calculation method of patient preference utilities, Standard Gamble or Rating Scale (Table 2).

In sensitivity analyses, varying orchiectomy cure rate from 50% to 100%, a threshold value between

Discussion

We performed a decision analysis of estimated outcome probabilities using updated long-term morbidity rates from historical series specific to the patient population with high-risk CS1 NSGCT and previously accepted patient preference utilities. From this analysis, we found that treatment with either BEP ×1 or RPLND mathematically outperforms surveillance following orchiectomy when integrating all of possible chance occurrences and survival related morbidities. In other words, the estimated

Conclusion

In this analysis, we evaluated relative value of approaches to treatment for CS1 NSGCT and found that treatment with RPLND or BEP ×1 is superior mathematically to surveillance. These findings should be considered in concert with patient preferences for specific side effects during treatment discussions for this patient population with a high risk of recurrence postorchiectomy.

Acknowledgments

David F. Penson, MD, MPH: validation and review of the article.

References (31)

  • P. Maroto et al.

    Multicentre risk-adapted management for stage I non-seminomatous germ cell tumours

    Ann Oncol

    (2005)
  • R.J. Motzer et al.

    Testicular cancer, Version 2.2015

    J Natl Compr Canc Netw

    (2015)
  • L. Wood et al.

    Canadian consensus guidelines for the management of testicular germ cell cancer

    Can Urol Assoc J

    (2010)
  • J. Oldenburg et al.

    Testicular seminoma and non-seminoma: ESMO Clinical Practice Guidelines for diagnosis, treatment and follow-up

    Ann Oncol

    (2013)
  • G. Daugaard et al.

    Surveillance for stage I nonseminoma testicular cancer: outcomes and long-term follow-up in a population-based cohort

    J Clin Oncol

    (2014)

    • Cited by (6)

    View full text
    © 2018 Elsevier Inc. All rights reserved.