Practice patterns and outcomes in older patients with DLBCL in England: A real-world data analysis

The urgent cancer referral pathway established by the National Health Service (NHS) in 2000 enables specialist referral of patients from general practice (GP) within 2 weeks, providing earlier diagnosis and improving outcomes for patients. However, limited population-level data exist on referrals, treatment patterns and effects on outcomes by the route of diagnosis in patients with diffuse large B-cell lymphoma (DLBCL), a non-Hodgkin lymphoma (NHL) that commonly presents in patients over 65 years of age. The standard treatment of DLBCL is rituximab, cyclophosphamide, doxorubicin, vincristine, and prednisolone (R-CHOP), though alternative R-CHOP-like regimens have been used in unfit patients, including those with comorbidities and the very elderly (≥80 years).

Based on the results from a phase II Lymphoma Study Association (LYSA) trial (NCT01195714),¹ R-miniCHOP, an attenuated R-CHOP-based regimen, has become a standard approach in the treatment of elderly patients with DLBCL. While R-miniCHOP is widely accepted, however, the ideal dose intensity that balances disease control with acceptable toxicity in elderly patients remains unclear. Here we summarize the key findings by Hounsome et al.² published in the British Journal of Cancer, examining the factors that influence treatment patterns and outcomes in elderly patients with DLBCL using population-level data.

Study design

This was a retrospective study using population-level data from the National Cancer Registration and Analysis Service (NCRAS), part of Public Health England (PHE). The dataset included patients in England diagnosed between 2013 and 2015 with DLBCL and aged ≥ 65 years. The route to diagnosis was a derived measure of the way in which a patient arrived at their cancer diagnosis, based on several datasets.

The outcomes of interested included:

• Overall survival (OS), defined as time from DLBCL diagnosis until death from any cause.
• Lymphoma-specific survival (LSS), defined as time from DLBCL diagnosis until death from DLBCL.
• Intended dose intensity (IDI) across the whole population, with specific focus on very elderly patients (≥ 80 years).
  • IDI was determined by strict first cycle total dosage analysis of individual drug components of R-miniCHOP (vincristine 1 mg, doxorubicin ≤55 mg, and cyclophosphamide ≤880 mg).

Results

Baseline characteristics

A total of 9,186 patients were included; 68% of these patients were younger (65–79 years) and gender distribution was balanced (Table 1).

• Men were more likely to be younger than elderly (55% vs 49%; p < 0.001) and had lower levels of deprivation (56% in quintile 1 vs 51% in quintile 5; p = 0.008).
• The percentage of patients diagnosed with advanced stage disease decreased with increasing age (63% in 65‒79 years vs 56% in ≥80 years; p < 0.001).
• Emergency presentations were common in elderly patients compared with younger patients (37% vs 33% respectively; p < 0.001) and in advanced stage disease.
• The most common route to diagnosis for early-stage disease was via the NHS urgent cancer referral pathway.
• Of patients for whom the treatment regimen was known, 48% received R-CHOP; other regimens included:
  • Rituximab, cyclophosphamide, vincristine, prednisolone (R-CVP) (6%)
  • Rituximab, cyclophosphamide, etoposide, vincristine, prednisolone (R-CEOP) (2%)
  • Rituximab, prednisolone, mitoxantrone, etoposide, bleomycin, vincristine (R-PMitCEBO) (1%)
  • Rituximab, etoposide, prednisolone, vincristine, cyclophosphamide, doxorubicin (R-EPOCH) (0.5%)
  • Rituximab, gemcitabine, cyclophosphamide, vincristine, prednisolone (R-GCVP) (0.5%)

Table 1. Baseline characteristics*

ECOG PS, Eastern Cooperative Oncology Group performance status; GP, general practice; NHS, National Health Service. *Adapted from Hounsome et al.2
Baseline characteristics, %	Age 65–79 (n = 6,203)	Age ≥80 (n = 2,983)	All ages (n = 9,186)
Sex
Male	55	49	53
Female	45	51	47
Disease stage
Early	37	44	39
Advanced	63	56	61
ECOG PS
0	42	29	39
1	35	38	35
2	23	34	26
Number of comorbidities
0	72	66	70
1	13	15	13
2	9	9	9
≥3	6	10	7
Deprivation quintile
1 – Least deprived	23	24	23
2	24	24	24
3	21	21	21
4	19	19	19
5 – Most deprived	13	12	13
Route to diagnosis
Emergency	34	38	35
GP referral	26	23	25
Inpatient elective	1	1	1
Other outpatient	10	8	9
NHS urgent cancer referral pathway	29	30	29

Survival outcomes

Three-year OS was:

• Higher in younger patients (65–79 years) than in those aged ≥80 years (57% vs 32% respectively; p < 0.001). OS for all patients was 49%.
• 64%, 58%, 50%, 39%, and 23% in age groups 65–69, 70–74, 75–79, 80–84, and ≥85 years, respectively (p < 0.001 for trend).
• 65%, 44%, and 26% in patients receiving the listed treatment regimen, other regimens, and when no treatment was recorded, respectively (p < 0.001).
• 74%, 68%, 68%, and 53% for the NHS urgent cancer referral pathway, GP referrals, other outpatients, and emergency presentation, respectively.
• Superior for NHS urgent cancer referral pathway (65–79 years, 73%; ≥80 years, 44%) and inferior by the emergency route (65–79 years, 39%; ≥80 years, 15%).

Patients who had early or advanced stage DLBCL and presented as an emergency had inferior OS compared with any other subgroup.

Univariate (UVA) and multivariate (MVA) analysis

The UVA and MVA analyses for the 4,392 patients treated with R-CHOP are presented in Table 2.

• Patients presenting through the NHS urgent cancer referral pathway had a 22% lower risk of death on UVA compared to those presenting through routine GP referral (p < 0.01). This significant effect was also retained in MVA (hazard ratio [HR] 0.81, p < 0.01).
• Advanced stage disease, male gender, 5-year age increment, deprivation quintile and comorbidity increment were independently associated with inferior OS.
• The comorbidity index became insignificant as a predictor of LSS (HR 0.97; p = 0.60) in an adjusted regression analysis.
  • Patients referred via the NHS urgent cancer pathway showed improved LSS (HR 0.71; p = 0.02) while those with an emergency presentation had significantly worse LSS in MVA (HR 1.95; p < 0.01).

Table 2. Univariate and multivariate analyses for OS for all R-CHOP-treated patients*

GP, general practice; HR, hazard ratio; NHS, national health service; Ref, reference HR. *Adapted from Hounsome et al.2 †Statistically significant p values are in bold.
Variable	Patients (n)	Univariate		Multivariate
		HR	p value†	HR	p value†
Route to diagnosis
GP referral	1110	Ref	—	Ref	—
Emergency presentation	1160	1.68	<0.01	1.63	<0.01
Inpatient elective	51	0.78	0.36	0.83	0.50
Other outpatient	374	0.97	0.81	0.96	0.70
NHS urgent cancer referral pathway	1635	0.78	<0.01	0.81	<0.01
Unknown	61	0.45	0.01	0.52	0.04
Age (5-year increment)	1265	1.05	<0.01	1.05	<0.01
Sex, Female	2081	0.83	<0.01	0.83	<0.01
Deprivation quintile (increment of quintile)	1014	1.04	0.05	1.03	0.09
Stage
Early	1497	Ref	—	Ref	—
Advanced	2342	1.73	<0.01	1.73	<0.01
Unknown	553	1.66	<0.01	1.61	<0.01
Number of comorbidities (increment of index)	3252	1.18	<0.01	1.15	<0.01

Outcomes based on treatment regimen

R-CHOP

Three-year OS was:

• 68% in patients receiving either full or attenuated R-CHOP at dosing greater than R-miniCHOP (n = 4,392) and did not differ across the years analyzed (p = 0.82).
• Superior in younger patients (65–79 years) compared to patients aged ≥80 years (70% vs 54%, respectively; p < 0.001).
• Highest in female patients aged 65–79 and lowest in men aged ≥80 years (73% vs. 51%; p < 0.001).
• Inferior in patients with advanced stage disease compared to early-stage disease (62% vs 77%, respectively (p < 0.001).
• Significantly inferior in patients with emergency presentation while patients referred via the NHS cancer referral pathway showed the highest OS rate (54% vs 75%; p < 0.001).
• Statistically inferior in more deprived groups (p = 0.01).
• A diagnosis to treatment interval (DTI) of 0‒29 days was significantly associated with shorter OS compared to DTI of 30+ days (p < 0.001).

• DTI was shorter in patients with emergency presentation (0–14 days and 15–29 days, 35% each; p < 0.001).
• DTI was statistically longer for NHS urgent cancer referral pathway compared to GP referrals (0–14 days, 10% vs 14%; 15–29 days, 33% vs 29%; p < 0.001).

R-miniCHOP

Of the patients treated with R-CHOP regimens, 313 received R-miniCHOP, and patients aged 65‒79 years who received R-miniCHOP had a higher comorbidity burden compared to those who were treated with standard R-CHOP (comorbidity index ≥1 in 35% vs 25%).

Three-year OS was:

• 57% in patients receiving R-miniCHOP.
• 59% vs 54% in patients aged 65–79 and ≥80 years, respectively, and 2-year OS was 60% in patients aged ≥80 years.
• Inferior in male patients aged ≥80 years and in those presenting through the emergency route.

R-CHOP vs R-miniCHOP in patients aged ≥80 years

The baseline characteristics in both groups were well balanced, except that more patients ≥85 years received R-miniCHOP (Table 3).

• OS curves completely overlapped, with a 54% 3-year OS for both groups.
• Advanced stage disease (HR 1.31; p=0.01), emergency presentation (HR 1.82; p < 0.01) and age (HR 1.06 per age bracket; p < 0.01) were adverse prognostic features, and female sex (HR 0.82; p =0.04) was a positive prognostic factor.
• Treatment with R-CHOP or R-miniCHOP did not influence OS (R-CHOP reference, HR 0.95; p = 0.68).

Table 3. Baseline characteristics according to dose intensity of R-CHOP*

R-CHOP, rituximab, cyclophosphamide, doxorubicin, vincristine, and prednisolone; R-miniCHOP, doxorubicin 25 mg/m2, cyclophosphamide 400 mg/m2 and vincristine 1 mg capped-dose. *Adapted from Hounsome et al.2
Baseline characteristics	R-CHOP		R-miniCHOP
	Number of patients	%	Number of patients	%
Age
80-84 years	564	76	106	67
≥85 years	182	24	52	33
Sex
Male	392	53	73	46
Female	354	47	85	54
Deprivation quintile
1 – Least deprived	185	25	41	26
2	183	25	36	23
3	152	21	30	19
4	127	17	22	14
5 – Most deprived	83	11	20	13
Stage at diagnosis
Early	311	42	68	43
Advanced	335	45	69	44
Comorbidity index
0	583	72	110	70
1	89	12	20	13
2	71	10	17	11
≥3	48	6	11	7

Other regimens (non-anthracycline-based)

Three-year OS was:

• 43% for R-CVP and 54% for R-CEOP.
• Higher in patients aged 65–79 years compared to those aged ≥80 years (54% vs 49%; p = 0.42).
• 64%, 55%, and 39% for GP referral, NHS urgent cancer referral, and emergency presentation, respectively.

Conclusion

This large, retrospective, real-world data study demonstrated that patients referred via the NHS urgent cancer referral pathway achieved superior survival outcomes and elderly patients aged ≥ 80 years had significantly inferior outcomes, especially male patients. In addition, patients presenting as an emergency at diagnosis had inferior OS when compared with standard GP referrals. Comparable OS was demonstrated for patients aged ≥80 years treated with R-CHOP and R-miniCHOP. The study was limited by incomplete data, influence of unmeasured confounding factors, lack of specific comorbidity classification, and under reporting of DLBCL as a specific cause of death. Further efforts to improve the availability of primary care data, research investigating the gender discrepancy and the use of novel-chemotherapeutic targeted agents in elderly patients with DLBCL are warranted.