Outcomes of elderly patients with primary CNS lymphoma in different socioeconomic settings; a 12-year population-based analysis of combined modality treatment

Primary central nervous system lymphoma (PCNSL) is an aggressive lymphoma subtype characterized by brain, spinal cord, meninges, or eye involvement without systemic features. Approximately 50% of all patients with PCNSL are aged >60 years, yet the optimal treatment for elderly patients with PCNSL has not yet been established, translating to median survival rates of around 6–7 months. Herein lies an unmet need to evaluate the long-term survival outcomes of elderly patients diagnosed with PCNSL. Thus, Samhouri, et al., published an analysis in Anticancer Research evaluating patient outcomes to different modalities for the treatment of elderly patients with PCNSL, and aimed to characterize predictors affecting treatment decisions.¹ The study compared outcomes to chemotherapy, radiation, and a combination of both (combined modality treatment [CMT]) as first-line therapy, while examining socioeconomic disparities in lymphoma care.

Study design

• A comprehensive retrospective cohort analysis using data from the National Cancer Database of 6,537 patients diagnosed with PCNSL between 2004 and 2015 across the United States.
• Eligible patients were >65 years of age and human immunodeficiency virus negative.
• Survival outcomes were compared using a propensity score model, which considered age, sex, race, insurance status, median income, education, treatment facility type, type of area, comorbidity score, and distance from the treating facility.
• Primary outcome: overall survival (OS) in patients who received chemotherapy/radiation alone, CMT, or no treatment.

Results

• The majority of patients received chemotherapy alone as first-line treatment (Figure 1).
• Table 1 shows the baseline characteristics of patients analyzed.

CMT, combined modality treatment; PCNSL, peripheral central nervous system lymphoma.
*Adapted from Samhouri, et al.¹

Table 1. Adjusted baseline characteristics of patients selected for analysis*

*Data from Samhouri, et al.1
Characteristic, % (unless stated otherwise)	Chemotherapy alone	Radiation alone	CMT	No treatment	p value
Mean age, years	73.9	74.5	73.8	74.3	0.08
Female	52	54	52	53	0.93
Race					0.89
Non-Hispanic               White	88	88	89	89
Black	3	3	2	2
Hispanic	4	5	3	4
Other	5	5	5	5
Facility type					0.88
Community               cancer program	3	3	3	3
Comprehensive               community	35	37	35	35
Median income					1.00
Less than               $38,000	15	14	14	15
$38,000–               $47,999	22	22	23	22
$48,000–               $62,999	28	28	28	28
More than               $63,000	35	36	35	35
Percentage of at least high school education					1.00
≥29%	14	14	13	14
20%–28.9%	24	24	25	25
14%–19.9%	35	35	36	35
<14%	27	26	27	27
Insurance					0.95
Not insured	1	1	1	1
Private               insurance	12	12	13	12
Medicaid	1	1	1	1
Medicare	85	86	86	85
Others	1	1	0	1
Type of area					0.37
Metropolitan	83	85	84	83
Urban	15	15	14	15
Rural	2	1	1	2
Comorbidity Score					0.92
0	60	60	62	59
1	25	24	25	25
2	10	11	10	11
≥3	5	5	4	5
Mean distance, miles	39.8	80.0	30.88	35.07	<0.01

Treatment selection and socioeconomic factors

• Of the entire patient cohort, 51.5% and 13.1% received chemotherapy alone and radiation alone, respectively, whereas 12.8% and 22.2% received CMT and no treatment, respectively.
• Multivariable logistic regression analysis highlighted a number of factors associated with differential treatment:
  • The probability of receiving any treatment reduced with advancing age and higher comorbidity score.
  • Admission at an academic or research institution and having insurance were linked with more treatment administration.
  • Patients who had private insurance and Medicare underwent more treatment than those with no insurance.
• Of the treated patients, CMT use was significantly less in elderly patients, those undergoing treatment at an academic/research center, with higher income, and living in rural areas.

Survival

• Median follow-up: 7.1 months (IQR, 2.2–29.3).
• Patients who received CMT over single-modality treatment, in both the entire population and an age-adjusted analysis, had significantly longer OS (Table 2; p < 0.001).

Table 2. Survival rates by treatment category*

*Data from Samhouri, et al.1 †patients aged 75 years or older.
	All patients	Chemotherapy alone	CMT	Radiation alone	No treatment
Median OS, months	7.0	13.37	19.5	5.0	2.0
Age-adjusted median OS†, months	5.2	7.0	14.0	4.1	1.7
Predicted 12-month OS, %		40.6	53.2	25.2	14.3
Predicted 24-month OS, %		31.3	38.5	14.3	10.1

Conclusion

This retrospective study indicates that, although combining radiation with chemotherapy has fallen out of favor for elderly patients with PCNSL, there is an OS benefit of CMT over single modality treatment which was greatest in the initial year of treatment. Notably, the survival benefit then began to steadily decline to the point of overlap with other groups, indicating late toxicity and increased probability of death in patients receiving CMT. Further work should focus on identifying factors which contribute to increased toxicity and establishing an optimal radiation dosage and technique to minimize this. Furthermore, confirmatory studies that explore the survival advantage of CMT over single modality treatment should be conducted in randomized trials.

Disparities were identified among patients with PCNSL; treatment at an academic/research center, increased age, higher income, and living in rural areas were all associated with decreased CMT administration. Treatment in academic/research programs increased the likelihood of receiving any treatment and chemotherapy, both of which can drastically improve patient outcomes.

The inferior prognosis among patients receiving no treatment (median OS, 2 months) emphasizes the need to identify contributing factors to treatment availability. This study highlights the influence of social and economic factors such as age, gender, race, insurance status, and residence on treatment decisions, and thus patient outcome. Having insurance, especially Medicare or private insurance, was predictive of receiving therapy, while the probability of patients receiving any treatment at all reduced with advancing age and comorbidity score.