Histology-specific nomogram for primary retroperitoneal soft tissue sarcoma

Original Article

Histology-Specific Nomogram for Primary Retroperitoneal Soft Tissue Sarcoma Ilaria Ardoino, PhD1; Rosalba Miceli, PhD1; Mattia Berselli, MD2; Luigi Mariani, MD, PhD1; Elia Biganzoli, PhD1,3; Marco Fiore, MD2; Paola Collini, MD4; Silvia Stacchiotti, MD5; Paolo Giovanni Casali, MD5; and Alessandro Gronchi, MD2

BACKGROUND: This study was conducted to develop a histology-specific nomogram to predict postoperative overall survival (OS) at 5 and 10 years in primary retroperitoneal soft tissue sarcoma (STS). METHODS: Data registered at a single institution (National Cancer Institute, Milan, Italy) prospective sarcoma database were used. In the present analysis, patients with primary localized retroperitoneal STS resected with curative intent between 1985 and 2007 were included. A parametric piecewise exponential survival multivariate model was used for nomogram development, and internal validation was performed with standard methodologies. Known prognostic variables, such as age, tumor burden, histologic variant (as reviewed by a sarcoma pathologist), grade, and surgical margins were considered as putative predictors. RESULTS: Among the 192 patients analyzed, within 10 years from surgery, 114 patients were alive, with a median follow-up time of 55 months (interquartile range, 25-104 months). Among the investigated factors, only histologic subtype did not reach significance at the 10% level. The relative hazard increased while increasing tumor size up to about 25 cm, and decreased thereafter. CONCLUSIONS: A histology-specific nomogram for retroperitoneal STS is now available. It can be used for better assessing the risk of the single patient and then making individualized decisions within the specific subset of retroperitoneal sarcomas. Cross-check external validation should C 2010 American Cancer Society. be performed. Cancer 2010;116:2429–36. V KEYWORDS: sarcoma, retroperitoneal sarcoma, nomogram, prognosis, survival.

Retroperitoneal soft tissue sarcoma (STS) is an uncommon disease. Only 10% of patients with STS have a tumor arising from the retroperitoneum.1 In view of the finding that these sarcomas present with peculiar characteristics with respect to other STSs,2,3 their prognosis may not be well predicted by using the same criteria as for STS in general.4,5 In most cases, a staging system or prognostic indices may not be fully satisfactory as prediction tools, as these do not allow quantification of individual risk. Furthermore, the staging system specifically designed for STS is based on prognostic factors relevant for extremity and superficial trunk tumors, and is largely inadequate for retroperitoneal STS when considering a cutoff at 5 cm for tumor size, depth, and presence/absence of lymph node metastases.4,5 Several reports have identified histologic grade and completeness of macroscopical resection as the major prognosticators for survival.2,3,5-13 The roles of histologic subtype and size of the tumor and age of the patient have been more controversial and less properly explored. Other recent treatment-related factors, such as the extension of surgical resection and the use of radiation therapy, seem promising as far as local control is concerned, but without a clear-cut benefit regarding survival detected so far.14,15 Nomograms represent an outcome prediction tool that may usefully extend and accomplish standard staging systems on an individualized basis. They should be mainly used to support the decision-making process at the patient’s bedside.16 As an additional advantage, a nomogram allows evaluation of the relative contribution of each prognostic factor in predicting outcomes. A nomogram designed for sarcomas in general was developed some years ago,17 but the majority of study patients had extremity or trunk STS, with retroperitoneal STS comprising 10 and 20 cm; and >20 cm). The first 2 factorial axes explain 68% of the amount of total variability. The plane of the first 2 factorial axes is depicted in Figure 5. The first factorial axis appears to distinguish mainly between 2 groups of patients. One group is characterized by very large and low-grade sarcomas, mainly liposarcomas, and to a lesser extent, older age. The second group is very heterogeneous; it is associated with smaller and highly aggressive (grade 3) sarcomas, with different histologic subtypes. The second factorial axis mainly distinguishes between patients with different histologic features: leiomyosarcomas versus solitary fibrous sarcomas and malignant peripheral nerve sheath tumors.

Piecewise Regression Model In the piecewise exponential regression model, the loghazard function was modeled as an additive function of the baseline log-hazard, and the covariates effect in the same fashion as a Cox regression model.

2433

Original Article Table 3. Ten-Year and 12-Year Survival Probabilities for a Patient Aged 60 Years With a 20-cm Retroperitoneal Soft Tissue Sarcoma According to Most Common Histologies and Grade

Histology/ Grade

Current Study

Nomogram for STS in General17

65% 15% 0

63% 27% 27%

65% 35% 0

52% 16% 16%

65% 20% 0

28% 3% 3%

Liposarcoma Grade 1 Grade 2 Grade 3

Leiomyosarcoma Grade 1 Grade 2 Grade 3

MPNST Grade 1 Grade 2 Grade 3

STS indicates soft tissue sarcoma; MPNST, malignant peripheral nerve sheath tumor.

The equation of the model is: logðlij Þ ¼ b0 þ b1 t1 þb2 t02 þ b2 t03 þ b4 Agei þb5 Sizei þ b6 Size0i þ b7 GradeIIi þb7 GradeIIIi þ b8 HistoIIi þ b9 HistoIIIi þb10 HistoIVi þ b11 HistoIXi þ b13 SurgMari þlogðtij Þ

where lij is the expected number of deaths for subject i in the interval j and (t1, t0 2, t0 3) is the restricted cubic spline basis for the time variable, allowing for a direct estimate of the baseline hazard. The variables Age, Size, Size0 , GradeII, GradeIII, HistoII, HistoIII, HistoIV, HistoIX, and SurgMar represent individuals’ covariate values, properly coded as in the design matrix; in particular, Size0 is the nonlinear component of the 3-knotrestricted cubic spline, and tij is the length of time subject i spent in interval j. Coefficients were then estimated by maximizing the equivalent Poisson likelihood function.

DISCUSSION We have developed a nomogram to predict OS of patients with primary retroperitoneal STS who have undergone a curative resection. Nomograms have been developed for several diseases, including STS. Kattan et al were the first to build a

2434

nomogram for STS at any site.17 This model was externally validated by using a series of patients from the University of California, Los Angeles18 and, for patients affected by extremity STS, with a series from our institution.19 Other attempts to improve prognostication by using nomograms in STS have been made, based on the histological subtype.29,30 The advantage of nomograms is to give the clinician a practical tool to take into consideration several variables together at the patient’s bedside. Recently, a site-specific nomogram for retroperitoneal sarcoma has been published.31 However, it currently provides 5-year OS, which may be a limitation for retroperitoneal sarcomas. Of course, there may be an added value for a site-specific nomogram. This is suggested by Table 3, where the 10- to 12-year OS is calculated for different histologies and malignancy grades in a typical retroperitoneal sarcoma patient (age, 60 years; size, 20 cm). Compared with the non–site-specific general nomogram for STS, prognosis assessed with our site-specific tool would be comparatively better for G1-2 leiomyosarcomas and malignant peripheral nerve sheath tumors (MPNSTs), and worse for G2-3 liposarcomas. This makes sense clinically, and is consistent with the results reported in the different retrospective series.3-15 Actually, some histologies are virtually lacking in the retroperitoneum (eg, myxoid/round cell or pleomorphic liposarcoma), whereas others are more often found there (eg, solitary fibrous tumor). Then, the meaning of tumor site is different for limb sarcomas as compared with retroperitoneal ones, inasmuch as all retroperitoneal sarcomas are of course deep-seated. Focusing on retroperitoneal STS only, we were also able to encompass other variables, such as grading expressed in a 3-tier as opposed to a 2-tier system. In addition, specific histologies that are typical of retroperitoneal sarcomas have been included. Finally, size was included as a continuous variable, encompassing the finding that retroperitoneal sarcomas having a large size are more likely to be low-grade liposarcomas. The tumor size score axis on the nomogram, with the scale that wraps around, reproduces the same nonmonotonic shape as the relative hazard curve (Fig. 3), increasing up to 20 to 25 cm and decreasing thereafter. It reflects how often giant tumors have a more indolent behavior and therefore a lower risk of death, at least up to 10 years. As a matter of fact, mainly low-grade liposarcomas can reach sizes >30 cm, because their very slow and indolent growth remains largely asymptomatic until they become clinically detectable. This is also the main reason that—at variance with STS at any other site—several reports, including 2 from

Cancer

May 15, 2010

Nomogram for Retroperitoneal STS/Ardoino et al

our institution, have failed to find meaningful cutoffs for size in retrospective series.2-15 As far as the completeness of surgical resection is concerned, our nomogram reflects that obviously the inability to grossly remove the tumor has a strong prognostic impact, because surgery is the primary treatment of this disease and cannot be substituted for by any other treatment. If the tumor cannot be completely removed for whatever reason, the patient has a very high risk of dying from it.2,3,5-12 In our model, margins are split between R0 þ R1, taken together, and R2. In fact, the difference between R0 and R1 is of questionable value in retroperitoneal sarcomas. Our nomogram does not distinguish between multivisceral resections and more conventional surgery, in the lack of definitive evidence that survival is affected thereby. As far as the histological subtype is concerned, we found that liposarcoma, although a more indolent tumor than others, had a very strong negative effect, second only to that of the ‘‘Other’’ category (mainly constituted by, eg, undifferentiated pleomorphic sarcoma, synovial sarcoma, and angiosarcoma; ie, tumors that are very aggressive). This is in accordance with previous reports from our institution,3,14,32 which showed how difficult it is to eradicate a retroperitoneal liposarcoma. Patients affected by leiomyosarcoma, although having a higher risk of metastasis, tend to have a higher chance of cure, being much less likely to recur locally. That is, those who do not develop distant metastases do not die of the disease, whereas the patients affected by liposarcoma, although less likely to develop distant disease, often die from locoregional recurrences in the long run. This finding may change in the future, when long-term follow-up of more aggressively treated patients will be available. We may speculate that patients affected by liposarcoma—if treated by aggressive surgery (ie, resection of the tumor en bloc with surrounding uninvolved viscera)—may have a higher chance of cure, if the more aggressive approach proves to be effective even in the long run.32 Conversely, solitary fibrous tumor, a diagnosis that has become more frequent, had the best outcome, although the hazard ratio CIs (see Table 2) are wide because of the small numbers. In this regard, a major limitation of nomograms is the subgroups with a low proportion of patients. As far as retroperitoneal STS is concerned, low-grade leiomyosarcoma virtually does not exist. Similarly, low-grade MPNST, or ‘‘Other’’ sarcomas, are very rare in the retroperitoneum. Finally, high-grade

Cancer

May 15, 2010

solitary fibrous tumors are rare. Therefore, the prediction for these tumors should be made with care, given the low precision achievable with small numbers. Nevertheless, these rare combinations of histological subtype and malignancy grade account for roughly 5% of the overall retroperitoneal STS presentations. We can then rely on the model having good accuracy in at least 95% of cases, although leaving a degree of uncertainty in the remaining presentations. The use of a flexible regression model, such as the piecewise, allows development a prognostic tool for individual outcome prediction, directly incorporating the baseline hazard function. Our nomogram needs to be externally validated, although internal validation was conducted. Cross-check comparisons with the recently proposed site-specific nomogram could also be planned. In conclusion, a histology-specific nomogram for retroperitoneal STS is now available. It can be used for better assessing the risk of the single patient and then making individualized decisions within the specific subset of retroperitoneal sarcomas. It could also be considered as a stratification tool when undertaking clinical trials. External validation studies are warranted to confirm the robustness of the model.

CONFLICT OF INTEREST DISCLOSURES The authors made no disclosures.

REFERENCES 1. Brennan M, Singer S, Maki RG, O’Sullivan B. Soft tissue sarcoma. In: DeVita VT, Lawrence TS, Rosenberg SA, eds. Cancer. Principles and Practice of Oncology. Philadelphia, PA: Lippincott Williams & Wilkins; 2008:1741-1794. 2. Lewis JJ, Leung D, Woodruff JM, Brennan MF. Retroperitoneal soft-tissue sarcoma: analysis of 500 patients treated and followed at a single institution. Ann Surg. 1998;228:355-365. 3. Gronchi A, Casali PG, Fiore M, et al. Retroperitoneal soft tissue sarcomas patterns of recurrence in 167 patients treated at a single institution. Cancer. 2004;100:2448-2455. 4. Lahat G, Tuvin D, Wei C, et al. New perspectives for staging and prognosis in soft tissue sarcoma. Ann Surg Oncol. 2008;15:2739-2748. 5. Nathan H, Raut CP, Thornton K, et al. Predictors of survival after resection of retroperitoneal sarcoma: a population-based analysis and critical appraisal of the AJCC Staging System. Ann Surg. 2009;250:970-976. 6. Stoeckle E, Coindre JM, Bonvalot S, et al. Prognostic factors in retroperitoneal sarcoma: a multivariate analysis of a series of 165 patients of the French Cancer Center Federation Sarcoma Group. Cancer. 2001;92:359-368. 7. Catton CN, O’Sullivan B, Kotwall C, Cummings B, Hao Y, Fornasier V. Outcome and prognosis in retroperitoneal soft tissue sarcoma. Int J Radiat Oncol Biol Phys. 1994;29: 1005-1010.

2435

Original Article 8. Ballo MT, Zagars GK, Pollock RE, et al. Retroperitoneal soft tissue sarcoma: an analysis of radiation and surgical treatment. Int J Radiat Oncol Biol Phys. 2007; 67:158-163. 9. Singer S, Corson JM, Demetri GD, Healey EA, Marcus K, Eberlein TJ. Prognostic factors predictive of survival for truncal and retroperitoneal soft-tissue sarcoma. Ann Surg. 1995;221:185-195. 10. Van Doorn RC, Gallee MP, Hart AA, et al. Resectable retroperitoneal soft tissue sarcomas. The effect of extent of resection and postoperative radiation therapy on local tumor control. Cancer. 1994;73:637-642. 11. Hassan I, Park SZ, Donohue JH, et al. Operative management of primary retroperitoneal sarcomas: a reappraisal of an institutional experience. Ann Surg. 2004;239:244250. 12. Van Dalen T, Plooij JM, van Coevorden F, et al. Longterm prognosis of primary retroperitoneal soft tissue sarcoma. Eur J Surg Oncol. 2007;33:234-238. 13. Anaya DA, Lahat G, Wang X, et al. Establishing prognosis in retroperitoneal sarcoma: a new histology-based paradigm. Ann Surg Oncol. 2009;16:667-675. 14. Gronchi A, Lo Vullo S, Fiore M, et al. Aggressive surgical policies in a retrospectively reviewed single-institution case series of retroperitoneal soft tissue sarcoma patients. J Clin Oncol. 2009;27:24-30. 15. Bonvalot S, Rivoire M, Castaing M, et al. Primary retroperitoneal sarcomas: a multivariate analysis of surgical factors associated with local control. J Clin Oncol. 2009;27:31-37. 16. Iasonos A, Schrag D, Raj GV, Panageas KS. How to build and interpret a nomogram for cancer prognosis. J Clin Oncol. 2008;26:1364-1370. 17. Kattan MW, Leung DH, Brennan MF. Postoperative nomogram for 12-year sarcoma-specific death. J Clin Oncol. 2002;20:791-796. 18. Eilber FC, Brennan MF, Eilber FR, Dry SM, Singer S, Kattan MW. Validation of the postoperative nomogram for 12year sarcoma-specific mortality. Cancer. 2005;101:22702275. 19. Mariani L, Miceli R, Kattan MW, et al. Validation and adaptation of a nomogram for predicting the survival of patients with extremity soft tissue sarcoma using a 3-grade system. Cancer. 2005;101:402-408.

2436

20. Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc. 1958;53:457-481. 21. Greenacre M. Correspondence analysis in medical research. Stat Methods Med Res. 1992;1:97-117. 22. Boracchi P, Biganzoli EM, Marubini E. Joint modelling of cause specific hazard functions with cubic splines: an application to a large case series of breast cancer patients. Comp Stat Data Anal. 2003;42:243-262. 23. Cox DR. Regression models and life tables. J R Stat Soc. 1972;34:187-220. 24. Trojani M, Contesso G, Coindre JM, et al. Soft tissue sarcomas of adults: study of pathological prognostic variables and definition of a histopathologic grading system. Int J Cancer. 1984;33:37-42. 25. Harrell FE Jr. Regression modelling strategies with applications to linear models, logistic regression, and survival analysis. New York, NY: Springer; 2001. 26. Harrell FE Jr, Lee KL, Mark DB. Multivariable prognostic models: issues in developing models, evaluating assumptions and adequacy, and measuring and reducing errors. Stat Med. 1996;15:361-387. 27. Antolini L, Boracchi P, Biganzoli EM. A time-dependent discrimination index for survival data. Stat Med. 2005;24: 3927-3944. 28. Coradini D, Daidone MG, Boracchi P, et al. Time-dependent relevance of steroid receptors in breast cancer. J Clin Oncol. 2000;18:2702-2709. 29. Dalal KM, Kattan MW, Antonescu CR, Brennan MF, Singer S. Subtype specific prognostic nomogram for patients with primary liposarcoma of the retroperitoneum, extremity, or trunk. Ann Surg. 2006;244:381-391. 30. Canter RJ, Qin LX, Maki RG, Brennan MF, Ladanyi M, Singer S. A synovial sarcoma-specific preoperative nomogram supports a survival benefit to ifosfamide-based chemotherapy and improves risk stratification for patients. Clin Cancer Res. 2008;14:8191-8197. 31. Anaya DA, Lahat G, Wang X, et al. Postoperative nomogram for survival of patients with retroperitoneal sarcoma treated with curative intent. Ann Oncol. 2010;21:397-402. 32. Mussi C, Collini P, Miceli R, et al. The prognostic impact of dedifferentiation in retroperitoneal liposarcoma: a series of surgically treated patients at a single institution. Cancer. 2008;113:1657-1665.

Cancer

May 15, 2010