Baseline clinical characteristics
Charts of 120 patients were identified (CHOP, n=108; EPOCH, n=12). The commonest histological diagnosis in the EPOCH group was DLBCL, 7(58%) while for the CHOP group was Diffuse Large Cell lymphoma (DLCL), 65(60%) (Table 1). Diagnosis of DLCL was obtained from the H&E stain with no additional IHC. All patients in the EPOCH group and a majority of patients in the CHOP group (n=105; 97.2%) were already receiving ART prior to initiation of chemotherapy. Only 3 patients (2.8%) in the CHOP group initiated ART after completion of their chemotherapy cycles. Ten patients (83%) in the EPOCH group were receiving tenofovir/lamivudine/efavirenz, one was receiving zidovudine/lamivudine/efavirenz, and the other was receiving tenofovir/lamivudine/nevirapine. Patients in the CHOP group received a variety of ART combination with 42(39%) patients receiving either tenofovir or zidovudine in combination with lamivudine and efavirenz. The rests of the patients received other combination of first line ART. No patient had ART interrupted while receiving chemotherapy. All patients in the EPOCH group and a majority of patients in the CHOP group (n=99, 92%) were receiving cotrimoxazole for PCP prophylaxis. The rest were either on dapsone, or none. Markers for HIV control were not documented as these were collected at a separate ART treatment centre, independent of the UCI.
TABLE 1: DEMOGRAPHIC FACTORS AND BASELINE CHARACTERISTICS
Variable
|
All patients n=120
|
CHOP,
n=108
|
EPOCH,
n=12
|
Sex, n(%)
|
|
|
|
Male
|
64 (53)
|
53 (49)
|
11(92)
|
Female
|
56 (47)
|
55 (51)
|
1 (8)
|
Mean age, years (SD)
|
40 (10)
|
40(10.2)
|
42(8.1)
|
Mean BMI, kg/m2 (SD)
|
22 (4.7)
|
21 (4.6)
|
24(4.4)
|
NHL type, n(%)
|
|
|
|
DLBCL
|
34 (28)
|
27 (25)
|
7(58)
|
DLCL (IHC not done)
|
69 (58)
|
65 (60)
|
4(33)
|
PBL
|
5 (4)
|
4 (4)
|
1(8)
|
NHL Other
|
9 (8)
|
9 (8)
|
0
|
Burkitt’s
|
3 (3)
|
3 (3)
|
0
|
ECOG score, n(%)
|
|
|
|
0
|
9 (8)
|
8 (7)
|
1(8)
|
1
|
29 (24)
|
22 (20)
|
7(58)
|
2
|
12 (10)
|
12 (11)
|
0
|
3
|
7 (6)
|
6 (6)
|
1(8)
|
Not assessed
|
63 (53)
|
60 (56)
|
3(25)
|
B-Symptoms
|
|
|
|
Yes
|
65 (54)
|
58 (54)
|
7(58)
|
No
|
55 (46)
|
50 (46)
|
5(42)
|
Comorbidity
|
|
|
|
Yes
|
17 (14)
|
14 (13)
|
3(25)
|
No
|
103(86)
|
94 (87)
|
9(75)
|
Stage
|
|
|
|
I
|
4 (3)
|
4 (4)
|
0
|
II
|
21(18)
|
17(16)
|
4(33)
|
III
|
54 (45)
|
48(44)
|
6(50)
|
IV
|
21 (18)
|
19(18)
|
2(17)
|
Not assessed
|
20 (17)
|
20(19)
|
0
|
NB: DLBCL - Diffuse large B-cell lymphoma; DLCL - Diffuse large cell lymphoma; ECOG - Easter Cooperative Oncology Group; IHC – Immunohistochemistry; PBL - Plasmablastic lymphoma; Comorbidity referred to the presence of any other diagnosis besides NHL.
Treatment completion
Only 51(47%) patients in the CHOP group and 8(67%) patients in the EPOCH group completed 6 or more cycles of chemotherapy. Those who completed 3-5 cycles were 31(29%) in the CHOP group and 1(8.3%) in the EPOCH group. Three patients in each group received less than 3 cycles of chemotherapy. Reasons for non-completion of chemotherapy cycles were serious adverse events (n=12, 10%), other reasons (n=4, 3%), and were not described in 104(87%) patients. Nineteen patients (18%) in the CHOP group and 3(25%) in the EPOCH group had serious adverse events detected, p=0.53. Most were laboratory adverse events like neutropenia (CHOP=13, 12%; EPOCH=2, 17%); p=0.65), anaemia (CHOP=12, 12%; EPOCH=1, 8%; P=0.71), and thrombocytopenia (CHOP=7, 6%; EPOCH=0; p=0.36). Others were sepsis (CHOP=1), treatment related death (EPOCH=1) and hepatic encephalopathy (CHOP=1), (Table 2). The lowest neutrophil count recorded was 0.12 x10^3/uL after the first cycle of chemotherapy in a patient treated with EPOCH.
Table 2: TREATMENT OUTCOMES
Variable
|
CHOP
n=108
|
EPOCH
n=12
|
RESPONSE TO TREATMENT, n(%)
|
|
|
Complete response (CR)
|
29 (27)
|
5 (42)
|
Partial response (PR)
|
14 (13)
|
2 (17)
|
Progressive disease (PD)
|
15 (14)
|
1 (8)
|
No response assessment
|
50 (46)
|
4 (33)
|
ADVERSE EVENTS
|
|
|
Neutropenia, n(%)
|
13 (12)
|
2 (17)
|
|
Grade
|
<2
|
|
0
|
0
|
|
|
3
|
|
6 (6)
|
2 (17)
|
|
|
4
|
|
7 (6)
|
0 (0)
|
Anaemia, n(%)
|
|
|
13 (12)
|
1 (8)
|
|
Grade
|
<2
|
|
4 (4)
|
1 (8)
|
|
|
3
|
|
9 (8)
|
0 (0)
|
Thrombocytopenia, n(%)
|
7 (6)
|
0 (0)
|
|
Grade
|
<2
|
|
6 (3)
|
0 (0)
|
|
|
3
|
|
1 (1)
|
0 (0)
|
Other Adverse Events, n(%)
|
4 (3)
|
0 (0)
|
Sepsis
|
1 (1)
|
0 (0)
|
Death
|
0 (0)
|
1 (8)
|
Hepatic Encephalopathy
|
1 (1)
|
0 (0)
|
Note: Nadir levels of neutrophil, haemoglobin and platelet counts were recorded after each chemotherapy cycles; adverse events were classified using the NCI CTCAE v5.0
Treatment response and survival
Overall treatment response rate was 40% in the CHOP group and 59% in the EPOCH group. Complete response (CR) was achieved in 29(27%) patients in the CHOP group and 5(42%) patients in the EPOCH group. Partial response was observed in 16(13%) patients in the CHOP group and 2(17%) patients in the EPOCH group. (Table 3).
The entire study population had a one year (12 months) overall survival (OS) rate of 56.7% (95% CI, 45.4–66.5), (Figure 1, Panel A). Patients treated with CHOP had a one year OS of 54.5% (42.8–64.8) and those treated with EPOCH of 80.2% (95% CI, 40.3–94.8), but the difference was not statistically significant (hazard ratio, 0.43 (95% CI, 0.10-1.78), (Figure 1, Panel). Subset analysis for patients with DLBCL showed a one year OS rate of 56.1% (95% CI, 33.0–74.0) in the CHOP group and 100% in the EPOCH group. Predictors of survival were analysed using patients’ age, sex, type of chemotherapy received, completion of 6 or more cycles of chemotherapy, type of lymphoma, stage of lymphoma, presence of B-symptoms and comorbidities. At univariable analysis, factors that were associated with favourable survival were ECOG performance score of 3-4, BMI 18.5-24.9 kg/m2 and completion of 6 or more cycles of chemotherapy. However, at multivariable analysis, only BMI 18.5-24.9 kg/m2 (normal BMI), (p=0.03) and completion of 6 or more cycles of chemotherapy, (p<0.001) were favourably associated with survival, Table 3.
Table 3: PREDICTORS OF SURVIVAL
Variable
|
Univariable
|
Multivariable
|
CHR (95%CI)
|
P-Value
|
AHR (95%CI)
|
P-Value
|
Age in years
|
1.0 (0.98 - 1.05)
|
0.32
|
1.0 (0.96 - 1.04)
|
0.95
|
Sex: Male
|
1.2 (0.64 - 2.24)
|
0.57
|
1.6 (0.76 - 3.32)
|
0.22
|
Study arm: EPOCH
|
0.4 (0.10 - 1.78)
|
0.24
|
0.6 (0.13 - 2.84)
|
0.52
|
ECOG
|
|
|
|
|
0 – 2
|
|
|
|
|
3 – 4
|
3.4 (1.10 - 8.82)
|
0.01
|
1.7 (0.55 - 5.26)
|
0.36
|
Not assessed
|
2.3 (1.01 - 5.05)
|
0.05
|
1.5 (0.60 - 3.80)
|
0.39
|
BMI
|
|
|
|
|
<18.5 Kg/m2
|
|
|
|
|
18.5 – 24.9 kg/m2
|
0.4 (0.20 - 0.83)
|
0.01
|
0.4 (0.18 - 0.89)
|
0.03
|
>25 kg/m2
|
0.5 (0.23 - 1.33)
|
0.19
|
0.6 (0.21 - 2.02)
|
0.45
|
Lymphoma stage
|
|
|
|
|
Early stage
|
|
|
|
|
Late stage
|
1.8 (0.76 - 4.04)
|
0.19
|
1.6 (0.58 - 4.60)
|
0.35
|
Not assessed
|
1.5 (0.49 - 4.32)
|
0.50
|
1 (0.28 - 3.94)
|
0.95
|
Type of lymphoma
|
|
|
|
|
Diffuse large B-cell lymphoma
|
|
|
|
|
Diffuse Large cell lymphoma
|
1.2 (0.58 - 2.63)
|
0.58
|
1.0 (0.39 - 2.43)
|
0.95
|
Plasmablastic lymphoma
|
2.1 (0.58 - 7.75)
|
0.25
|
1.4 (0.30 - 6.02)
|
0.69
|
NHL Other
|
2.5 (0.84 - 7.28)
|
0.10
|
1.7 (0.52 - 5.74)
|
0.37
|
Burkitt’s
|
1.0 (0.12 - 7.45)
|
0.96
|
0.3 (0.03 - 3.45)
|
0.33
|
Presence of B-symptoms
|
0.9 (0.47 - 1.63)
|
0.67
|
0.9 (0.44 - 1.96)
|
0.85
|
Presence of comorbidity
|
0.7 (0.24 - 1.91)
|
0.46
|
0.9 (0.28 - 2.69)
|
0.81
|
Chemotherapy cycles received
|
|
|
|
|
< 6
|
|
|
|
|
> 6
|
0.2 (0.11 - 0.43)
|
<0.001
|
0.2 (0.10 - 0.47)
|
<0.001
|
Note: AHR: Adjusted Hazard Ratio, BMI: Body Mass Index, CHR: Crude Hazard Ratio, ECOG: Eastern Cooperative Oncology Group