A randomized comparative study of methylcobalamin, methylcobalamin plus pregabalin and methylcobalamin plus duloxetine in patients of painful diabetic neuropathy.

CONTEXT: Diabetic neuropathy affects 10.5%-32.2% of diabetic population posing clinical burden onto society. AIMS: We aimed to study the efficacy, safety, and tolerability of methylcobalamin, methylcobalamin plus pregabalin, and methylcobalamin plus duloxetine in patients of painful diabetic neuropathy. SETTINGS AND
DESIGN: It is a prospective, randomized, open-label, interventional, and parallel-group study done in patients of painful diabetic neuropathy.
MATERIALS AND METHODS: A total of 100 patients were recruited and randomized to three study groups A, B, and C on methylcobalamin, methylcobalamin and pregabalin, and methylcobalamin and duloxetine, respectively. Patients were assessed at day 0 and 4, 8, and 12 weeks. The tuning fork test, monofilament test, Thermal Sensitivity testing, and Visual Analog Scale (VAS) were used to analyze vibration, pressure, thermal sensitivity, and pain. STATISTICAL ANALYSIS USED: The results are expressed as mean ± standard deviation. Appropriate statistical methods were used to calculate P value (<0.05 - significant).
RESULTS: The increase in number of patients with vibration perception is 11.6%, 37.9%, and 41.4%; pressure sensation is 7.6%, 37.9%, and 37.9%; and thermal sensitivity is 15.4%, 31.1%, and 37.9% in Groups A, B, and C, respectively. The decrease in VAS scores is 0.58 ± 0.14, 3.82 ± 0.05, and 4.17 ± 0.48 in Groups A, B, and C correspondingly. The adverse effects reported in Groups A, B, and C are 0%, 6.9%, and 10.3%, respectively.
CONCLUSIONS: Group C is more efficacious when compared to Groups A and B while Group B is safer.

Introduction

Diabetes mellitus is a metabolic disorder resulting in increased glucose levels due to decreased insulin secretion or action or both. Among type 1 and type 2, type 2 diabetes mellitus is more common. As a consequence of chronic hyperglycemia, the autonomic and somatic nerves get damage and produce diabetic neuropathy as its long-term complication which affects 10.5%–32.2% of Indian diabetic population.[1] As per the International Association of Pain, “Neuropathic pain is the pain due to anomaly of peripheral somatosensory system of diabetic patient” affecting nearly one-fourth of diabetic population.[2] International guidelines recommend the Food and Drug Administration-approved drugs, duloxetine and pregabalin, to deal with the excruciating pain.[3] Role of methylcobalamin is also suggested in pain inhibition. However, the lacunae in the knowledge still exist as to which treatment is best. In my research, I have tried to cover these gaps by comparing methylcobalamin, methylcobalamin plus pregabalin, and methylcobalamin plus duloxetine in patients of painful diabetic neuropathy.

Materials and Methods

This was a prospective, interventional, randomized, open-label, and parallel-group study conducted in the department of pharmacology in collaboration with department of medicine in a tertiary care hospital after obtaining the approval of the institutional ethics committee (ethical committee approval letter no: 018/TH/D-26/2018 Batch). All the patients who reported to OPD and IPD of department of medicine with more than one year history of type 2 diabetes were recruited for the study. They were accessed for painful diabetic neuropathy. Recruitment was done for one and half year between January 1, 2019, to July 31, 2020. Patients with decreased or absent vibration and touch and temperature sense and who scored above 4 on the Visual Analog Scale (VAS) assessment were enrolled. The patients with type 1 diabetes, altered sensorium or with any other cause including drug induced neuropathy were excluded from the study

Informed consent in vernacular language was taken. Patients were randomized using computer-generated sequence into three study groups A, B, and C with Group A receiving methylcobalamin (1500 mcg/day) alone, Group B receiving pregabalin (150 mg/day) and methylcobalamin, and Group C receiving duloxetine (60 mg/day) and methylcobalamin.

Baseline investigations were performed which included hemoglobin level, total leukocyte count, fasting blood sugar, postprandial blood sugar, glycated hemoglobin (HbA1c), blood urea, serum creatinine, serum bilirubin, serum glutamic-oxaloacetic transaminase, serum glutamic-pyruvic transaminase, total cholesterol levels, serum low-density lipoprotein levels, serum triglyceride levels, and serum high-density lipoprotein levels. Baseline Vitamin B12levels were not performed as all the three groups were supposed to receive methylcobalamin in equal doses. Special investigations were performed at day 0 and then at weeks 4, 8, and 12 which included 128 Hz tuning fork test for vibration sensation,[4] Semmes-Weinstein monofilament test for touch sensation,[5] thermo tube test for temperature sensation, and 11-point VAS used for pain assessment.[6]

The results were analyzed and mean ± standard deviation was calculated. Statistical tests such as ANOVA, Fisher's exact t-test, and two-tailed t-test were used and P value was calculated. P <0.05 was taken as statistically significant.

Results and Observations

A total of 100 patients involving 32, 34, and 34 patients in Groups A, B, and C, respectively, were enrolled (due to COVID-19, most of the OPDs remain closed accounting for less sample size). Only 26, 29, and 29 patients in Groups A, B, and C correspondingly completed the study forming a total of 84 patients (16 patients dropped, out of which 4 patients had no relief to symptoms, 5 had adverse effects, and 7 lost to follow-up due to pandemic situation and some other reasons). The mean age of patients is 57.8 years, with 38% of males and 62% of females. Sixty-two percent of the cases belong to urban background while 38% reside in a rural area. The groups are comparable with respect to age, sex, region, and baseline characters [mean values are given in Table 1] which include fasting blood glucose levels, postprandial blood glucose levels, HbA1c, lipid profile, liver function tests, and renal function tests.

Table 1

Baseline parameters in study groups

Parameter	Mean±SD			P
	Group A	Group B	Group C
FBG (mg/dl)	193.1±71.52	192.87±86.6	198.36±69.15	0.954
PPBG (mg/dl)	244.1±82.93	262.84±95.76	259.51±91.51	0.721
HbA1c	7.56±1.66	8.85±2.6	7.93±1.82	0.171
Blood urea (mg/dl)	16.48±4.17	16.02±6.05	15.23±4.12	0.635
Serum creatinine	0.82±0.21	0.84±0.23	0.86±0.17	0.712
Serum bilirubin	0.72±0.21	0.72±0.18	0.81±0.20	0.837
SGOT (mg/dl)	30.91±7.93	29.82±7.21	29.77±9.47	0.735
SGPT (mg/dl)	35.15±13.9	34.51±9.44	33.75±10.2	0.947
TC (mg/dl)	194.80±39.1	201.6±34.4	198.95±42.8	0.808
LDL-C (mg/dl)	115.18±36.2	118.78±37.1	115.52±37.2	0.762
TGs (mg/dl)	171.32±52.97	173.9±56.96	191.32±62.09	0.344
HDL-C (mg/dl)	49.8±9.7	47.86±9.9	46.62±7.3	0.632

All values are expressed in mean±SD. SD=Standard deviation, FBG=Fasting blood glucose, PPBG=Postprandial blood glucose, HbA1c=Glycosylated hemoglobin, SGOT=Serum glutamic-oxaloacetic transaminase, SGPT=Serum glutamic-pyruvic transaminase, TC=Total cholesterol, LDL-C=Low-density lipoprotein, TGs=Triglycerides, HDL-C=High-density lipoprotein

Tuning fork test

Among the three study groups, a significant increase in number of patients with vibration perception was observed in Group C (41.4%) followed by Group B (37.9%), with no significant improvement in Group A (11.6%) by the end of week 12, as illustrated in Table 2. When the difference between groups is observed at 4 weeks, the variation of Group A from that of Groups B and C is 9.8% each, so both Groups B and C show equal response with no variation, but when compared at the end of week 8, the change between Groups A and C is greater (26.7%) than Groups A and B (16.4%) with a difference of 10.3% between Groups B and C. Thus, Group C shows earlier and a better response. By the end of week 12, Groups A and C show maximum variation (29.8%, P = 0.016) followed by Groups A and B (26.3%, P = 0.032) with a minimal variation between Groups B and C (3.5%, P = 0.787). Thus, Group C appears to be most efficacious for the recovery of vibration sense by the end of the study duration. Group B can serve as an alternative, but Group A has no significant response.

Table 2

Tuning fork test

Visit	Score	Group A (n=26), n (%)	Group B (n=29), n (%)	Group C (n=29), n (%)
1st visit (baseline)	0	26 (100)	29 (100)	29 (100)
	1	0	0	0
4th week (follow-up)	0	25 (96.1)	25 (86.3)	25 (86.3)
	1	1 (3.9)	4 (13.7)	4 (13.7)
8th week (follow-up)	0	24 (92.3)	22 (75.8)	25 (86.2)
	1	2 (7.7)	7 (24.1)	10 (34.4)
12th week (follow-up)	0	23 (88.4)	18 (62.1)	17 (58.6)
	1	3 (11.6)	11 (37.9)	12 (41.4)
Fisher’s exact test statistic (baseline to 12th-week follow-up), 95% CI		0.235 (NS), −3.3309-29.0507	p<0.05* (significant), 18.6910-55.9677	0.0001* (highly significant), 21.6854-59.2815

Score 0=Absent or reduced vibration sensation; Score 1=Vibration sensation present, CI=Confidence interval, NS=Not significant

The Semmes monofilament test

Further moving to the touch/light pressure sensation, Figure 1 reveals the enhancement in number of patients with respect to pressure sensation since day 1 to week 12 of the study. The change in number of patients over 4 weeks is 0% in Group A, 10.3% in Group B, and 6.8% in Group C which progressed to 3.8% in Group A, 24.1% in Group B, and 13.7% in Group C by the end of week 8. The final change over 12 weeks in number of patients over 3-month period is 7.6% (P = 0.755), 37.9% (P = 0.0025), and 37.9% (P = 0.007) in Groups A, B, and C, respectively. Thus, response is early in Group B but better with Group C. This is better educed with the help of intergroup variations where variation between Groups A and B is 5.9%, A and C is 2.4%, and B and C is 3.5% at the beginning of the study. Groups B and C showed a similar response of 4.4% with respect to Group A by week 4. The difference increased to 14.4% in Group B and 7.5% in Group C with respect to Group A, with a difference of 6.9% among themselves. On comparing Group A with Group C and B, the difference get increased to 27.9% (P = 0.0383) and 24.4% (P = 0.038) respectively. From this, it is inferenced that although Group B gives an early result, Group C is more efficient.

Figure 1

The Semmes monofilament test

The thermo tube test

This test elicited the response to the hot and cold sensation, as depicted in Table 3, where all the three groups increased the number of patients who responded to it. At the first follow-up, no patient in Group A recovered while 6.9% of the patients in Group B and 24.1% of the patients in Group C got better with the treatment. The next follow-up, i.e., at week 8, the number raised to 7.7% in Group A, 13.8% in Group B, and 31.0% in Group C with a final lift up to 15.4% in Group A, 31.1% in Group B, and 37.9% in Group C. Thus, Groups B and C both show a significant efficacy in recovering hot and cold sensation. On comparing Group A with Group B and C at the start of the study, the variation observed was 2.2% and 1.2%. The change became 9.1% between Groups A and B and 22.9% between Groups A and C with 13.8% change between Groups B and C. Thus, Group C showed a sudden improvement, but by the end of week 8, a few more patients in Group A responded, so the difference with Groups B (8.3%) and C (22.1%) shortens. However, the diversity of Groups B and C (13.8%) maintained itself. However, by the end of week 12, difference in improvement amongst Groups B and C was just 3.4% (P = 0.792). Although Groups B and C differed from Group A by 17.9% (P = 0.168) and 21.3% (P = 0.103), respectively, at the end of week 12, this did not significantly contribute to show any superiority. Henceforth, no group is significantly better than others in improving the temperature sense.

Table 3

Thermo tube test

Visit	Score	Number of cases
		Group A (n=26), n (%)	Group B (n=29), n (%)	Group C (n=29), n (%)
1st visit (baseline)	0	23 (88.4)	25 (86.2)	26 (89.5)
	1	3 (11.5)	4 (13.7)	3 (10.3)
4th week (follow-up)	0	23 (88.4)	23 (79.3)	19 (65.5)
	1	3 (11.5)	6 (20.6)	10 (34.4)
8th week (follow-up)	0	21 (80.7)	21 (72.4)	17 (58.6)
	1	5 (19.2)	8 (27.5)	12 (41.3)
12th week (follow-up)	0	19 (73.1)	16 (55.1)	15 (51.7)
	1	7 (26.9)	13 (44.8)	14 (48.2)
Fisher’s exact test statistic (baseline to 12th-week follow-up), 95% CI		0.291 (NS), −6.1670-35.0052	0.009*, 6.8419-50.6776	p<0.0015*, 13.8333-56.5387

0=Absent sensation, 1=Present sensation, CI=Confidence interval, NS=Not significant

Visual Analog Scale score

Pain assessment was done using VAS. The mean VAS scores at each visit are demonstrated in Figure 2. The mean VAS scores showed a reduction of 0.19 ± 0.05, 1.24 ± 0.05, and 1.51 ± 0.02 in Groups A, B, and C after 1 month treatment and the score further decreased to 0.39 ± 0.02, 2.14 ± 0.37, and 2.69 ± 0.06 respectively by the end of week 8. The final decrease by 12 week period is 0.58 ± 0.14 (P = 0.092), 3.82 ± 0.05 (P < 0.001), and 4.17 ± 0.48 (P < 0.0001) in Groups A, B, and C, respectively. Hence, it is proved that Group C produces a greater amount of pain reduction. This is further evidenced by comparing the three groups. Table 4 depicts the intergroup variation of mean scores and confirmed Group C to be the best treatment modality for reducing pain. The change in Group C observed was 3.61 ± 0.8 (P < 0.0001) which was the highest amongst the three groups. Both the groups are highly efficient in controlling pain, but Group C is better than Group B with a variation of 0.93 ± 0.72 (P = 0.029).

Figure 2

Visual Analog Scale score

Table 4

Intergroup difference of mean values of Visual Analog Scale score

Visits	A/B	A/C	B/C
At day 1	0.02±0.67	0.02±0.42	0.04±0.19
At 4 week	1.03±0.61	1.34±0.39	0.31±0.22
At 8 week	1.73±0.8	2.32±0.5	0.59±0.7
At 12 week	2.68±0.08	3.61±0.8	0.93±0.72

Adverse effects

None of the patients in Group A showed adverse effects while 6.9% of the patients in Group B and 10.3% of the patients in Group C observed adverse effects in the form of dizziness, headache, and constipation which is the major reason for the loss of follow-up. Group C is the best treatment option for painful diabetic neuropathy. Group B is another alternative and can be substituted in conditions where Group C cannot be given or safety is a major concern. Although no adverse effects are observed in Group A, this is not a good treatment option as no significant results are observed.

Discussion

Painful diabetic neuropathy is a major issue affecting the quality of life and thus day-to-day activities.[7] Therefore, the need of the hour is to have a treatment which can both provide pain relief as well as improve the signs and symptoms of neuropathy. Moreover, we use a variety of drugs clinically alone or in combination, but which among them is best is still unknown. Hence, the present study was done to fill these gaps in knowledge.

The study has shown female predominance which is consistent with the study of Daniel et al.[8] The females generally are emotionally weak with a low threshold for pain which can explain the female preponderance. The study was conducted on a sample size of 100 patients. Three groups were formed. Group A received methylcobalamin, Group B received methylcobalamin plus pregabalin, and Group C received methylcobalamin plus duloxetine in patients of painful diabetic neuropathy. All the three groups showed efficacy in improving vibration perception, touch sensation, temperature sensitivity, and pain. However, methylcobalamin had no significant effect in treating painful diabetic neuropathy which was similar to the findings of Jayabalan and Low.[9] Thus, combinations are better over methylcobalamin which show a similarity with my study.[10] Many studies have elaborated the role of pregabalin and methylcobalamin combination in the treatment of painful diabetic neuropathy which is in accordance with my study where the combination is found efficacious.[111213] Duloxetine, another treatment option, has been studied extensively in management of neuropathic pain, but its beneficial effect in combination with methylcobalamin is still a mystery.[14] The present study has enlightened combinations to be better treatment modalities than methylcobalamin alone. Moreover, the present study has proved a combination of duloxetine and methylcobalamin superior to pregabalin and methylcobalamin despite few adverse effects. Pregabalin and Duloxetine have been studied by many researchers for the treatment of neuropathic pain.[1516] However, no research work in my knowledge has critically assessed these combinations for the management of painful diabetic neuropathy.

Although the present study is novel in respect of the combinations compared, few limitations are there which need to be focused in future studies. The study was an open-label study, thereby accommodating the risk of bias. In addition, the study was conducted on a small sample size, single center, and for a short time period. Hence, long-term efficacy and safety cannot be studied. Further studies are required in this field so as to facilitate the treatment of neuropathic pain in majority of patients without producing much adverse effects.

Conclusions

Although only combination groups show significant results, combination of duloxetine and methylcobalamin proves to be better than combination of pregabalin and methylcobalamin in decreasing progression and pain in patients of painful diabetic neuropathy. With respect to the safety, all the three groups were comparably well tolerated. Methylcobalamin did not show any adverse effect. However, combination of pregabalin and methylcobalamin seems to be a good alternative to combination of duloxetine and methylcobalamin in patients not tolerating the latter. Although the study had few limitations with respect to sample size, open-label, and single centric study, further studies need to be done to evaluate various beneficial effects and safety issues of methylcobalamin, pregabalin, and duloxetine in varying combinations.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.