Quality-adjusted survival analysis of malignant glioma patients

Quality-adjusted life years in glioma patients: a systematic review on currently available data and the lack of evidence-based utilities

BACKGROUND

Cost-effectiveness studies gain importance in the context of rising health care expenses and treatment options. Especially in the neuro-oncological context, surgical therapy may increase overall survival, but restrain the patient by postoperative disability. Quality-adjusted life years, express treatment effects and are based on health utilities. In our study, we analyze the current evidence on health economic evaluations in glioma patients.

MATERIAL AND METHODS

We performed a systematic database search including Medline and Cochrane Library. Studies were critically appraised for statistical analyzes including glioma patients, health economic modeling and detailed health outcome. Study evidence was classified according to levels of evidence for therapeutic studies from the Centre for Evidence-Based Medicine (Oxford).

RESULTS

37 studies (1995-2018) were identified, 29 matched our inclusion criteria. Studies addressed surgical cost-efficiency and/or the standard treatment, postoperative chemotherapy (n = 6) and 5-ALA (n = 3). Only 16 studies used QALY as the outcome measure, most used overall survival or life years gained (LYG). Utilities were either based on one single study (Garside et al. in Health Technol Assess 11:iii-iv, ix-221) or derived from visual analogue scale (VAS). None assessed quality of life values for specific health statuses or utilities. Incremental cost-effectiveness ratios varied from 8325€ per QALY (5-ALA) to 518,342€ per LYG (tumor treating fields).

CONCLUSIONS

Only one study generated utility values to conduct cost-effectiveness analysis (CEA); most studies used indirect outcomes such as LYG or based their model on previously published data. Health economic evaluations lack specific utilities, further investigations are necessary to conduct reliable CEA in the neurosurgical context.

Cost-effectiveness analysis of a randomized study comparing radiosurgery with radiosurgery and whole brain radiation therapy in patients with 1 to 3 brain metastases

BACKGROUND

In this study, we compare 2 treatment options and determine cost-effectiveness and cost-utility.

METHODS

We carried out a decision analysis populated with data from patients with brain metastasis in a concurrent trial randomized to either stereotactic radiosurgery (SRS) and observation or SRS and whole brain radiation therapy. Outcomes included actual life years saved (LYS), quality-adjusted life years (QALYs), and incremental cost-effectiveness ratio (ICER). Costs used were from the healthcare perspective and utilities were captured through a time-trade-off method, using 10-year, 5-year, and 1-year time horizons. One-way sensitivity analyses were carried out to determine robustness of the decision analysis model.

RESULTS

Compared with SRS and whole brain radiation therapy, SRS and observation not only had a higher average cost ($74,000 vs $119,000, respectively) but also a higher average effectiveness (0.60 LYS vs 1.64 LYS, respectively) with an ICER of $44,231/LYS or $41,783/QALY (with utilities captured using a 10-year horizon). Slightly higher ICER estimates were achieved with utilities captured using the other time horizons ($43,280/QALY and $44,064/QALY, respectively). Sensitivity analysis showed that the following variables had the highest impact on the ICER: probability of no recurrence in recursive-partitioning analysis class 2 after SRS and observation; probability of being alive after SRS and observation in recursive-partitioning analysis class 2 and being treated for recurrence.

CONCLUSIONS

Compared with other interventions in the $50,000 to $100,000/QALY cost-effectiveness range, the application of SRS and observation, with subsequent neurosurgical management of recurrences, is shown to be a reasonable treatment modality for brain metastases.

Radiotherapy for malignant gliomas in the elderly

While fractionated external beam radiotherapy remains the principal treatment in patients with histologically verified malignant glioma, its use in older patients and in patients with adverse prognostic features has not been subject to randomized studies. Hypofractionated partial brain irradiation offers a short, well-tolerated treatment with a palliative benefit for patients with a predicted median survival of less than 6 months. To assess its true efficacy in terms of survival and quality-of-life gain requires a formal randomized comparison with supportive care either alone or in combination with palliative chemotherapy and with radical radiotherapy. On present evidence, palliative radiotherapy remains the appropriate treatment for older patients and those with adverse prognostic features, but the final choice should be based on the patient's wishes moderated by the clinician's perception of the gain of treatment.

Hypofractionated radiotherapy for poor prognosis malignant glioma: matched pair survival analysis with MRC controls

PURPOSE

To assess the survival benefit of palliative hypofractionated radiotherapy in patients with poor prognosis high grade glioma by a matched comparison to conventionally treated controls.

METHOD

Ninety-two elderly and/or disabled patients with high grade glioma with poor prognostic features received palliative partial brain radiotherapy to a dose of 30Gy in six fractions over 2 weeks. Patients were matched for WHO histological grade, performance status and age from a cohort of patients treated with conventionally fractionated radiotherapy to a dose of 60Gy in 30 fractions in an Medical Research Council (MRC) BR05 trial.

RESULTS

Patients treated with hypofractionated radiotherapy had a median survival of 5 months with a 1-year survival rate of 12% from diagnosis. The median survival of case-matched controls was estimated to be 2.5-4.5 months longer. Following hypofractionated radiotherapy, Barthel score was improved or remained stable in 68% of patients.

CONCLUSION

Hypofractionated partial brain radiotherapy is a well-tolerated regimen with palliative benefit. Comparison with matched controls suggests lesser survival benefit than would be obtained with radical radiotherapy. However, this is compensated by lower intensity and duration of irradiation induced side effects. It is postulated that there may not be a significant difference in good quality survival or 'quality adjusted survival' between the two regimens and this requires testing in prospective trials.

Quantitative measurement of quality outcome in malignant glioma patients using an independent living score (ILS). Assessment of a retrospective cohort

PURPOSE

Although a number of tools have been developed to measure 'quality of life' in patients with malignant glioma, there remains no completely satisfactory technique that incorporates a quality of life measure into survival analysis. We propose that a patient's ability to maintain independent activity offers a way to accomplish this goal.

PATIENTS AND METHODS

An independent living score (ILS) is generated by awarding points on a monthly basis based on Karnofsky score and weighing the score based on the particular month of the clinical course. The ILS has a large range for any given survival, and can discriminate important treatment effects to which standard survival analyses are completely insensitive. Using this score and several variations, we were able to retrospectively analyze a patient cohort to assess what correlated with ILS.

RESULTS

We found a strong correlation with survival of all the measures tested. Interestingly, we found that patients for whom a total resection was performed and those who were most intensively treated had significantly higher ILS values, suggesting that not only did more aggressive treatment improve survival but that it did not simply increase survival at the expense of the time a patient remained independent.

CONCLUSION

Since the general course for patients with malignant glioma is one of increasing disability and loss of independence, we feel that these measures can serve as a way to distinguish between those therapies that increase survival at the expense of quality of life versus those that do not. Consideration should be given to incorporating these measures into prospective trials.

Quality of life in patients with glioblastoma multiforme participating in a randomized study of brachytherapy as a boost treatment

OBJECT

Until recently the assessment of outcome in patients treated for glioma has emphasized length of survival with the evaluation of quality of life (QOL) limited to unidimensional, mostly physical, measures. The authors report the multidimensional assessment of QOL as part of a randomized clinical trial of brachytherapy as a boost in the initial treatment of patients with glioblastoma multiforme.

METHODS

A questionnaire previously developed by the senior authors and psychometrically validated was completed by patients on randomized entry into the study and at follow-up review every 3 months thereafter. The questionnaire was presented in a linear-analog self-assessment format. Karnofsky Performance Scale (KPS) scores were also recorded on each occasion. No differences were found between patients in either arm of the study (conventional radiation therapy consisting of 50 Gy in 25 fractions or conventional radiation plus a brachytherapy boost of a minimum peripheral tumor dose of 60 Gy) in KPS and QOL scores during the 1st year of follow-up review. However, there was a statistically significant deterioration in patients' overall KPS scores during the 1st year of follow up compared with baseline scores. Of QOL items evaluated, statistically significant deteriorations were found in self care, speech, and concentration, and on subscale analyses, cognitive functioning and physical experience (symptoms) deteriorated significantly during the 1st year of follow up, compared with baseline values. The correlation between QOL and KPS scores was low.

CONCLUSIONS

Future studies in patients harboring malignant gliomas must incorporate measures assessing QOL because traditional measures focusing on physical or neurological functioning give an incomplete assessment of the patient's experience.

The benefit of treatment intensification is age and histology-dependent in patients with locally advanced non-small cell lung cancer (NSCLC): a quality-adjusted survival analysis of radiation therapy oncology group (RTOG) chemoradiation studies

PURPOSE

Currently, chemoradiation treatment strategies in locally advanced NSCLC are essentially the same irrespective of tumor histology or patient age. The purpose of this study is to analyze the impact of age, histology, Karnofsky performance status (KPS), and specific toxicities on the median survival time (MST) and quality-adjusted survival (QTime) for each treatment strategy.

METHODS AND MATERIALS

Nine hundred seventy-nine patients with Stage II-IIIB inoperable NSCLC were enrolled on 6 prospective Phase II and III studies from 1983 to 1995. Treatment regimens ranged from standard RT (SRT) to 60 Gy, hyperfractionated RT (HRT) to 69.6 Gy, induction chemotherapy (ICT) of cisplatin (CIS) and vinblastine (VBL) followed by SRT, ICT + concurrent CT (CCT) + SRT, and CCT + HRT; CCT consisted of etoposide or VBL + CIS. Toxicities assessed were skin, mucous membrane, lung, esophagus, neurologic, hematologic, and upper GI. QTime was calculated by weighting the time spent with a specific toxicity, as well as local or distant tumor progression. Each toxicity was weighted with increasing severity as the toxicity increased in grade.

RESULTS

As expected, patients with the worst KPS (50-70) had the lowest MST (7.8 months) and QTime (6.7 months). Patients <70 years had improved survival with more aggressive therapy (i.e., ICT + SRT or CCT + HRT), while patients > 70 years achieved the best QTime with standard RT (SRT) alone. In patients with squamous cell carcinoma, those treated with ICT + CCT + SRT had dramatically improved MST (25.7 months) and QTime (21.8 months) compared to the other treatment regimens (11.7-12.8 and 10.7-12 months, respectively). Patients with adenocarcinoma, however, generally manifested incrementally better MST and QTime as the therapies intensified. Within the concurrent chemoradiation arms, the upper GI and lung toxicities had the greatest impact on QTime.

CONCLUSION

This quality-adjusted survival analysis suggests that there is a critical relationship between the type of histology and its optimal treatment, age and the ability to tolerate intensive therapy, and the need to reduce lung and upper GI toxicities.

Does assessment of quality of life in comparative cancer trials make a difference? A discussion

This paper discusses two accompanying manuscripts addressing the question whether the assessment of quality of life in comparative cancer trials makes a difference. A number of thoughts and comments stimulated by the manuscripts are presented. It is concluded that, when comparing treatment regimens for a disease such as cancer, three measures can be considered in turn. First, the primary endpoint must remain: How long do patients survive? The logical next question is: How well do patients function? Then we have the third question: How well do patients feel? Certainly, these questions are interrelated, but our decisions will be the most rational (and we will provide the most useful information to patients) if we keep these distinctions in mind. Whatever the endpoint of interest, however, the appropriate way to compare different therapeutic interventions reliably is the randomized trial.

Assessing quality of life in comparative cancer clinical trials: does it make a difference?

Many randomized Phase III trials of cancer therapy involve quality of life assessments. This paper argues for graphical presentation of quality of life measures in order to visualize changes over time.

Do quality of life assessments make a difference in the evaluation of cancer treatments?

The question posed by this set of quality of life papers is whether or not quality of life assessments in cancer clinical trials help evaluate the effects of cancer treatment on patient functioning. In this discussion, missing data problems, particularly those commonly found in advanced stage disease trials, are highlighted. Researchers are encouraged to investigate the extent of bias associated with missing data and to select analysis approaches accordingly. In the worst case, it may not be possible to analyze data longitudinally; descriptive or graphical portrayals of the data may be more appropriate. The importance of instrument reliability (minimizing measurement error) is emphasized for clinical trials research, particularly with respect to enhancing a trial's ability to detect quality of life differences by treatment arm. One strategy for addressing missing data is evaluated with respect to its impact on the measurement properties of the quality of life questionnaire. Clinical trials groups have been successful in obtaining quality of life data in multi-site settings and patients, by and large, appreciate the effort to include a systematic and standardized report of the effects of treatment on their functioning.