Cost and survival analysis of metastatic cerebral tumors treated by resection and radiation

Drivers of hospitalization cost after craniotomy for tumor resection: creation and validation of a predictive model

Background

The economic sustainability of all areas of medicine is under scrutiny. Limited data exist on the drivers of cost after a craniotomy for tumor resection (CTR). The objective of the present study was to develop and validate a predictive model of hospitalization cost after CTR.

Methods

We performed a retrospective study involving CTR patients who were registered in the Nationwide Inpatient Sample (NIS) database from 2005–2010. This cohort underwent 1:1 randomization to create derivation and validation subsamples. Regression techniques were used for the creation of a parsimonious predictive model.

Results

Of the 36,433 patients undergoing CTR, 14638 (40.2%) underwent craniotomies for primary malignant, 9574 (26.3%) for metastatic, and 11414 (31.3%) for benign tumors. The median hospitalization cost was $24,504 (Interquartile Range (IQR), $4,265-$44,743). Common drivers of cost identified in the multivariate analyses included: length of stay, number of procedures, hospital size and region, and patient income. The models were validated in independent cohorts and demonstrated final R² very similar to the initial models. The predicted and observed values in the validation cohort demonstrated good correlation.

Conclusions

This national study identified significant drivers of hospitalization cost after CTR. The presented model can be utilized as an adjunct in the cost containment debate and the creation of data-driven policies.

Electronic supplementary material

The online version of this article (doi:10.1186/s12913-015-0742-2) contains supplementary material, which is available to authorized users.

The usefulness of stereotactic radiosurgery for radioresistant brain metastases

Objective

We investigated the effectiveness of stereotactic gamma knife Radiosurgery (GKR) for radioresistant brain metastases with the impact upon histology.

Methods

Between April 2004 and May 2011, a total of 23 patients underwent GKR for 67 metastatic brain tumors from 12 renal cell cancers, 5 sarcomas and 6 melanomas. The mean age was 56 years (range, 18 to 79 years). Most of the patients were classified as the Radiation Therapy Oncology Group recursive partitioning analysis class II (91.3%). The synchronous metastasis was found in 6 patients (26.1%) and metachronous metastasis in 17 patients (73.9%). We analyzed the local control rate, intracranial progression-free survival (PFS) and overall survival (OS).

Results

The mean tumor volume for GKR was 2.24 cc and the mean prescription dose was 19.4 Gy (range, 10 to 24) to the tumor margin. Out of metachronous metastases, the median duration to intracranial metastasis was 3.3 years in renal cell cancer (RCC), 2.4 years in melanoma and 1.1 years in sarcoma (p=0.012). The total local control rate was 89.6% during the mean 12.4 months follow-up. The six-month and one-year local control rate was 90.2% and 83% respectively. Depending on the pathology, the control rate of RCC was 95.7%, sarcoma 91.3% and melanoma 80.5% during the follow-up. The common cause of local failure was the tumor bleeding in melanoma. The median PFS and OS were 5.2 and 8.4 months in RCC patients, 6.5 and 9.8 months in sarcoma, and 3.8 and 5.1 months in melanoma.

Conclusion

The GKR can be one of the effective management options for the intracranial metastatic tumors from the radioresistant tumors. The melanoma showed a poor local control rate compared to other pathologies because of the hemorrhage.

Comparing the cost-effectiveness of two brain metastasis treatment modalities from a payer's perspective: stereotactic radiosurgery versus surgical resection

OBJECTIVES

This study aims to identify the cost-effectiveness of two brain metastatic treatment modalities, stereotactic radiosurgery (SRS) versus surgical resection (SR), from the perspective of Germany's Statutory Health Insurance (SHI) System.

METHODS

Retrospectively reviewing 373 patients with brain metastases (BMs) who underwent SR (n=113) and SRS (n=260). Propensity score matching was used to adjust for selection bias (n=98 each); means of survival time and survival curves were defined by the Kaplan-Meier estimator; and medical costs of follow-up treatment were calculated by the Direct (Lin) method. The bootstrap resampling technique was used to assess the impact of uncertainty.

RESULTS

Survival time means of SR and SRS were 13.0, 18.4 months, respectively (P=0.000). Medians of free brain tumor time were 10.4 months for SR and 13.8 months for SRS (P=0.003). Number of repeated SRS treatments significantly influenced the survival time of SRS (R(2)=0.249; P=0.006). SRS had a lower average cost per patient (€9964 - SD: 1047; Skewness: 7273) than SR (€11647 - SD: 1594; Skewness: 0.465), leading to an incremental cost effectiveness ratio of €-3740 per life year saved (LYS), meaning that using SRS costs €1683 less than SR per targeted patient, but increases LYS by 0.45 years.

CONCLUSION

SRS is more cost-effective than SR in the treatment of brain metastasis (BM) from the SHI perspective. When the clinical conditions allow it, early intervention with SRS in new BM cases and frequent SRS repetition in new BM recurrent cases should be advised.

Treatment costs for glioblastoma multiforme in Nova Scotia

BACKGROUND

Glioblastoma Multiforme (GBM) is the most common and malignant brain tumor in adults. The median survival in patients harboring this neoplasm is 12 months irrespective of any form of therapy. Health care costs of illnesses with high mortality rates, such as GBM, are of particular interest in times of constrained health care resources. No information regarding costs for the treatment of patients with GBM is available in Canada. The aim of this study was to conduct an analysis of the costs of treatment of GBM in Nova Scotia.

METHODS

Patients with histologically proven GBM during a three year period (1996-1998) in Nova Scotia were included in the study. Analysis was based on hospital costs supplemented by data on additional medical services received following discharge for the primary intervention.

RESULTS

The mean cost of medical care per patient from the time of diagnosis to death was $17,149. The highest costs were related to hospitalization with ward costs alone accounting for 48% of all costs. Radiotherapy costs were 25%, surgery costs were 16% and chemotherapy costs were 7% of all costs. Costs for diagnostic procedures were 6% of the total costs.

CONCLUSION

Our data reflect the costs and practice pattern in the treatment of GBM in Nova Scotia and may be of value as an initial attempt to analyze costs of treatment of GBM in Canada.

Brain tumours in Sweden 1996: care and costs

OBJECTIVES

Brain tumours cause considerable concern due to a high mortality and there are increasing efforts to provide adequate care, sometimes outside hospitals. Health care utilisation, direct costs of care, and the indirect social cost of morbidity and early mortality caused by brain tumours in Sweden in the year 1996 was analysed.

METHODS

Quantification of ambulatory care, care in hospital, long term and palliative/terminal care, drug consumption, temporary as well as long term morbidity, and mortality from comprehensive national data sources. Direct costs were calculated using 1996 charges. Indirect costs were calculated by sex and age specific salaries. A sensitivity analysis considered the impact of alternative estimates of each item.

RESULTS

Indirect costs were 75% of the total and were caused mainly by early mortality. Direct costs were predominantly for care in hospital, long term care, and home health care. Among direct costs, astrocytomas III-IV and meningiomas accounted for 42% and 30% respectively.

CONCLUSIONS

The cost of illness from brain tumours reflects the characteristics of these malignancies. Despite their low incidence rate, the economic impact caused by high mortality among young persons is a predominant trait. Costs of acute hospital care and also long term care and home care are considerable.

Dose optimization and indication of Linac radiosurgery for brain metastases

PURPOSE

The authors have examined treatment effects of linear accelerator based radiosurgery for brain metastases. Optimal doses and indications were determined in an attempt to improve the quality of life for terminal cancer patients.

METHODS AND MATERIALS

Ninety-two patients with 162 lesions were treated with Linac radiosurgery for brain metastases between April 1993 and September 1998. To determine prognostic factors, risk factors for recurrence, and appearance of new lesions, univariate and multivariate analyses were performed. To compare the local control between the high-dose (minimum dose > or =25 Gy: prescribed to the 50-80% isodose line) and low-dose (minimum dose <25 Gy) irradiated groups, matched-pairs analysis was performed.

RESULTS

Median survival time was 11 months. In univariate analysis, extracranial tumor activity (p<0.001) and Karnofsky Performance Status (KPS) (p = 0.036) were two significant predictors of survival. In multivariate analysis, the status of an extracranial tumor was the single significant predictor of survival (p = 0.005). Minimum dose was the single most significant predictor of local recurrence in univariate, multivariate, and matched-pairs analyses (p<0.05). As to the appearance of new lesions, activity of extracranial tumors was a significant predictor (p<0.05). Side effects due to radiosurgery were experienced in 4 of 92 cases (4.3%).

CONCLUSIONS

We concluded that brain metastases patients should be irradiated with > or =25 Gy, when extracranial lesions are stable and longer survival is expected. Combined surgery and conventional radiation may have little advantage over radiosurgery alone when metastatic brain tumors are small and extracranial tumors are well controlled. When extracranial tumors are progressive, the rate of appearance of new lesions in other nonirradiated locations becomes higher. In such cases, careful follow-up is required and a combination with whole brain irradiation should be considered.

[Could the evaluation of the cost of complications be a worthwhile means to improve radiotherapy?]

At the present time, the current improvement of technical and dosimetric aspects of radiation oncology has to be evaluated in terms of potential benefit for the patient and the society. For this last point of view, specially designed economic analyses must be performed in order to justify the number of resources involved by these technical improvements. If the question is how the current technical procedures could reduce the risk of undesirable side-effects, the response cannot be immediately drawn from the literature. This paper emphasizes the possibility to evaluate the role of side-effects as endpoints of economic analyses when using special models in medical decision making such as Markov's. Only few oncologic situations are reliable to properly analyze the relationship between sophisticated radiation techniques and the incidence of post-radiation complications. These situations should be selected when prospective economic analyses are planned in the field of radiation therapy.

Patterns of relapse and late toxicity after resection and whole-brain radiotherapy for solitary brain metastases

BACKGROUND

This retrospective analysis was performed in order to evaluate the pattern of relapse and the risk of late toxicity for solitary brain metastases treated with surgery and whole-brain radiotherapy and to correlate the results with those from radiosurgical trials.

PATIENTS AND METHODS

From a total of 66 patients, 52 received 10 x 3 Gy and 10 were treated with 20 x 2 Gy whole-brain radiotherapy after resection of their brain metastases.

RESULTS

The actuarial probability of relapse was 27% and 55% after 1 and 2 year(s), respectively. The local relapse rate (at the original site of resected brain metastases) was rather high for melanoma, non-breast adenocarcinoma, and squamous-cell carcinoma. No local relapse occurred in breast cancer and small-cell carcinoma. Failure elsewhere in the brain seemed to be influenced by extracranial disease activity. Size of brain metastases and total dose showed no correlation with relapse rate. Occurrence of brain relapse was not associated with a reduced survival time, because 10/15 patients who developed a relapse received salvage therapy. Of the patients, 11 had symptoms of late radiation toxicity (the actuarial probability was 42% after 2 years).

CONCLUSIONS

Most results of surgical and radiosurgical studies are comparable to ours. Several randomized trials investigate surgical resection versus radiosurgery, as well as the effects of additional whole-brain radiotherapy in order to define the treatment of choice. Some data support the adjuvant application of 10 x 3 Gy over 2 weeks as a reasonable compromise when local control, toxicity, and treatment time have to be considered.

Radiosurgery for the treatment of brain metastases

Surgical resection and whole brain radiotherapy (WBRT) have been the mainstays of the treatment of cerebral metastases. This approach results in a median survival of about 10 months. Several recent publications and our own experience suggest that a similar median survival can be achieved with stereotactic radiosurgery using either the Leksell Gamma Knife or the linear accelerator radiosurgical techniques. In addition, radiosurgery can effectively treat metastatic tumors in surgically inaccessible sites, e.g., the brainstem. Radiosurgery can also effectively treat multiple intracranial metastases in widely separated areas of the brain. In fact, we have shown that patients with multiple metastases have similar lengths and qualities of survival as do patients with single metastases treated with stereotactic radiosurgery. The most important predictor of success in radiosurgical treatment of cerebral metastases is the neurological status of the patient, usually expressed as the Karnofsky Performance Status (KPS). The histological type of primary cancer is not an outcome predictor. Even so-called "radioresistant" tumors (e.g., melanoma, renal cell) respond favorable to radiosurgery. A great benefit of radiosurgery is the virtual lack of perioperative complications and the minimal interference with quality of life compared either to surgery or to fractionated whole brain radiotherapy. Long-term complications of radiosurgery are infrequent and primarily relate to failure of local tumor control (10%) and radiation-induced edema or necrosis. The later usually can be controlled with corticosteroids, but occasionally, craniotomy may be required to treat life-threatening mass effects. We believe that radiosurgery is the treatment of choice for most cerebral metastases. Only large lesions (> 3.5-4 cm diameter) and those which require immediate decompression to treat life-threatening mass effects require surgical treatment. Radiosurgery also may be used to treat residual disease after surgical resection. We have shown that WBRT does not increase the efficacy of radiosurgery in the treatment of cerebral metastases, and, therefore, we prefer to avoid both the short- and long-term morbidity of that treatment, if possible.

Economic analyses in health care: an introduction to the methodology with an emphasis on radiation therapy

Payers are increasingly interested in knowing whether they are receiving value for the dollars they spend on health care. Because economic analysis will be used as a means of evaluating radiation therapy, it is important that radiation oncologists understand the basic methodology employed in such analyses. This review article describes the four basic types of economic analyses: cost minimization, cost effectiveness, cost utility, and cost benefit. Specification of alternative therapies, choice of perspective of the analysis, measurements of costs and benefits, and the role of discounting and sensitivity analyses are discussed. Published economic analyses that pertain directly to treatment with radiation therapy are reviewed. Finally, we close with a brief discussion of the potential areas for future economic outcomes research in radiation oncology.

Cost-effectiveness of head CT in patients with lung cancer without clinical evidence of metastases

OBJECTIVE

To estimate the cost-effectiveness of CT for detecting brain lesions in patients with lung cancer without clinical evidence of metastases.

DESIGN

Decision analysis model comparing two different strategies for detecting brain metastases: brain CT routinely (CT-first) or brain CT only when patients develop neurologic signs and/or symptoms (CT-deferred).

PATIENTS

Hypothetical cohort of patients with lung cancer with an unremarkable screening clinical evaluation for metastases.

MEASUREMENTS

Net costs are calculated as the difference in costs between the two limbs of the decision tree. Net benefits are expressed as the difference in calculated years of life expectancy between the two strategies. Net costs are divided by net benefits, yielding the marginal cost per quality adjusted year of added life expectancy (C/QALY) for the CT-first strategy.

RESULTS

In the baseline analysis, the C/QALY for the CT-first strategy is about $70,000. Improving the clinical evaluation as a screen for detecting brain metastases markedly increases the C/QALY. Increasing the cost of brain CT magnifies this effect. More effective treatment for asymptomatic brain metastases and better accuracy of CT for identifying resectable and unresectable brain metastases lower C/QALY.

CONCLUSIONS

Although a threshold cost-effectiveness has not been defined for identifying "cost-effective" diagnostic procedures, the marginal C/QALY of the CT-first strategy is substantially higher than many accepted medical interventions. At current costs, the routine use of brain CT is not warranted in patients with lung cancer who have normal findings on a standardized clinical evaluation for metastases.