Gender, age, disease severity, body mass index and diabetes may not affect response to subcutaneous tanezumab in patients with osteoarthritis after 16 weeks of treatment. A subgroup analysis of placebo-controlled trials.

AIM: To assess the impact of pre-specified patient characteristics on efficacy and safety of subcutaneous tanezumab in patients with osteoarthritis (OA).
METHODS: Data were pooled from two (efficacy; N = 1545) or three (safety; N = 1754) phase 3 placebo-controlled trials. Change from baseline to week 16 in Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) Pain, WOMAC Physical Function and patient global assessment of OA (PGA-OA) scores was compared between tanezumab (2.5 and 5 mg) and placebo groups via analysis of covariance. Treatment-emergent adverse events (TEAEs) were summarised descriptively. Analyses were done in patient subgroups (men or women; age <65, ≥65, or ≥75 years; body mass index [BMI] <25, 25 to <30, 30 to <35 or ≥35 kg/m2 ; diabetes or no diabetes; baseline WOMAC Pain score <7 or ≥7; and Kellgren-Lawrence [KL] grades 2, 3 or 4 in the index joint) and the overall population.
RESULTS: In all subgroups, improvements in WOMAC Pain were numerically greater and often statistically significant (P < .05) for both tanezumab groups compared with placebo. Results were similar for WOMAC Physical Function and PGA-OA. TEAE profiles were generally consistent across subgroups and similar to the overall population (ie slightly higher rates of TEAEs, serious TEAEs and severe TEAEs with tanezumab relative to placebo) with a few exceptions. Exceptions included women reporting slightly more TEAEs with tanezumab than men, and patients with diabetes reporting slightly more severe TEAEs with tanezumab than patients without diabetes. Additionally, TEAEs were more frequent with tanezumab than placebo in the age ≥65 and ≥75 years, but not the age <65 years, subgroups.
CONCLUSIONS: Efficacy and safety/tolerability of tanezumab may not be meaningfully impacted by gender, age, BMI, diabetes status, baseline pain severity or KL grade in the index joint. Conclusions are limited by low patient number in some subgroups. Clinicaltrials.gov: NCT02697773, NCT02709486, NCT01089725.

What's known

Response to pharmacological treatment, in terms of both efficacy and safety, can vary across patients with osteoarthritis.

Variability may be caused by differences in patient demographics, disease characteristics and comorbid conditions.

Based on large‐scale randomised, placebo‐controlled clinical trials, treatment with subcutaneous tanezumab has been shown to improve pain and function in patients with moderate‐to‐severe osteoarthritis and a history of inadequate response to standard analgesics for osteoarthritis.

Tanezumab, like other nerve growth factor antibodies, is associated with adverse events related to abnormal peripheral sensation (eg, paresthesia and hypoesthesia) and joint safety events, predominantly rapidly progressive osteoarthritis, in some patients.

What's new

We show that the efficacy and safety/tolerability of subcutaneous tanezumab after 16 weeks of treatment may not be meaningfully impacted by gender, age, disease severity, body mass index or diabetes in patients with moderate‐to‐severe osteoarthritis and a history of inadequate response to standard analgesics for osteoarthritis.

INTRODUCTION

Osteoarthritis (OA) represents a substantial global burden and is often associated with significant levels of pain and impairment of physical function. Current OA management concentrates on mitigating these symptoms through a combination of non‐pharmacologic (eg, weight loss, exercise, education, cognitive behavioural therapy) and pharmacologic (eg, acetaminophen, nonsteroidal anti‐inflammatory drugs [NSAIDs], opioids, duloxetine, intra‐articular corticosteroids or hyaluronic acid) approaches. , , , Managing OA‐related pain, however, is difficult and many patients express dissatisfaction with treatment because of inadequate efficacy or tolerability issues, highlighting an unmet need for new safe and effective therapies. , ,

Treatment response (efficacy and safety/tolerability) often varies between patients, even in highly controlled clinical trial settings, and may be attributed to differences in disease severity; patient characteristics such as age, weight and gender; and the presence of certain comorbidities. , , Therefore, examining the efficacy and safety of a particular therapy in subgroups of patients, based on baseline demographic or clinical characteristics, could help identify subgroups of patients who may or may not respond favourably to treatment.

Tanezumab is a monoclonal antibody against nerve growth factor that is in clinical development for the treatment of the signs and symptoms of moderate‐to‐severe OA in patients with inadequate treatment response or intolerability to standard OA analgesics (eg, acetaminophen, NSAIDs, opioids). Three randomised, placebo‐controlled, phase 3 studies have assessed the efficacy and safety of subcutaneous (SC) tanezumab in such patients. , , The current analysis pooled data from these trials to determine whether response (efficacy and safety/tolerability) to SC tanezumab is affected by patient characteristics including gender, age, body mass index (BMI), diabetes status, baseline pain severity and Kellgren‐Lawrence (KL) grade in the index joint.

METHODS

Data sources

Patient‐level data were pooled from two (efficacy: NCT02697773 and NCT02709486) or three (safety: NCT02697773, NCT02709486 and NCT01089725) randomised, placebo‐controlled, phase 3 trials of SC tanezumab (Table 1). , , Study NCT01089725 was terminated early because of a class‐wide partial clinical hold on anti‐NGF therapies and 90.5% of treated patients received only 8 weeks of treatment (ie, one dose of study medication at baseline). Thus, this study was excluded from the current efficacy analyses (which were based on week 16 data). However, a majority (70.7%) of treated patients remained in study NCT01089725 for >16 weeks for safety evaluation. Therefore, study NCT01089725 is included in the current safety analyses, which were based on the full study (treatment + follow‐up) periods and encompass safety data from all phase 3, placebo‐controlled OA studies of SC tanezumab conducted to date. In study NCT02697773, 87.5% and 79.9% of patients completed the treatment and full study (treatment + follow‐up) periods respectively. In study NCT02709486, 88.3% and 82.0% of patients completed the treatment and full study periods, respectively. All studies were conducted in compliance with the Declaration of Helsinki and the International Conference on Harmonisation Good Clinical Practice guidelines. Study protocols were approved by an institutional review board at each site, and all patients provided written informed consent.

TABLE 1

Summary of studies included in the analysis

Study	Key inclusion criteria	Duration	Treatment arms (n patients)	Rescue medication	NSAID use
NCT02697773	Age ≥18 y OA of the knee or hip a WOMAC Pain ≥5 WOMAC Physical Function ≥5 PGA‐OA of fair, poor or very poor History of inadequate pain relief with acetaminophen, and a history of inadequate pain relief with, or contraindication/intolerability to, oral NSAIDs, and either tramadol or an opioid (or unwilling to take opioids)	16‐wk treatment 24‐wk safety follow‐up	Placebo (232) SC tanezumab 2.5 mg (231) SC tanezumab 2.5/5 mg b (233)	Acetaminophen/paracetamol at doses ≤3000 mg/d for up to 3 d/wk during the treatment period	Limited use of NSAIDs was permitted on an occasional basis for self‐limiting conditions unrelated to OA. Aggregate use of NSAIDs during each 8‐wk SC dosing interval was not to exceed 10 d and total NSAID use was not to exceed 30 d between baseline visit and 16 wk after last SC dose
NCT02709486	Age ≥18 y OA of the knee or hip a WOMAC Pain ≥5 WOMAC Physical Function ≥5 PGA‐OA of fair, poor or very poor History of inadequate pain relief with acetaminophen, and a history of inadequate pain relief with, or contraindication/intolerability to, oral NSAIDs, and either tramadol or an opioid (or unwilling to take opioids)	24‐wk treatment 24‐wk safety follow‐up	Placebo (282) SC tanezumab 2.5 mg (283) SC tanezumab 5 mg (284)	Acetaminophen/paracetamol at doses ≤4000 mg/d for up to 5 d/wk during the treatment period and then daily during safety follow‐up	Limited use of NSAIDs was permitted on an occasional basis for self‐limiting conditions unrelated to OA. Aggregate use of NSAIDs during each 8‐week SC dosing interval was not to exceed 10 d and total NSAID use was not to exceed 40 d between baseline visit and 16 wk after last SC dose
NCT01089725	Age ≥18 y OA of the knee a WOMAC Pain ≥5 WOMAC Physical Function ≥4 PGA‐OA of fair, poor or very poor Unwilling or unable to take non‐opiate pain medication, or have a history of inadequate pain relief with non‐opiate pain medications, or be candidates for invasive interventions	16‐wk treatment c 8‐wk safety follow‐up	Placebo (72) SC tanezumab 2.5 mg (74) SC tanezumab 5 mg (63) SC tanezumab 10 mg (86) d IV tanezumab 10 mg (84) d	Acetaminophen/paracetamol at doses ≤3000 mg/d for up to 3 d/wk during the treatment period and then daily during safety follow‐up	Limited use of NSAIDs was permitted on an occasional basis for self‐limiting conditions unrelated to OA

Abbreviations: IV, intravenous; NSAID, nonsteroidal anti‐inflammatory drug; OA, osteoarthritis; PGA‐OA, patient global assessment of osteoarthritis; SC, subcutaneous; WOMAC, Western Ontario and McMaster Universities Osteoarthritis Index.

Diagnosed per American College of Rheumatology criteria with radiographic confirmation (Kellgren‐Lawrence grade ≥2).

Patients received 2.5 mg at baseline and 5 mg at week 8. These patients were included in the 5 mg group in the current analyses.

Due to early termination, 90.5% of patients received only 8 wk of treatment (ie, one dose of study medication at baseline). As a result, study NCT01089725 was included in the current safety analysis but excluded from the 16‐wk efficacy analysis.

10 mg SC and IV arms were not included in the current analyses since the 10 mg dose was discontinued prior to studies NCT02697773 and NCT02709486.

OA of the knee or hip

SC tanezumab 2.5/5 mg (233)

OA of the knee or hip

OA of the knee

SC tanezumab 10 mg (86)

IV tanezumab 10 mg (84)

All studies enrolled patients aged ≥18 years with moderate‐to‐severe OA of the knee or hip and a history of inadequate response to other OA analgesics. Tanezumab was administered every 8 weeks for 16‐24 weeks at doses ranging from 2.5 to 10 mg. Patients in the tanezumab 2.5/5 mg group of study NCT02697773 were included in the tanezumab 5 mg treatment arm in the current efficacy analyses.* The tanezumab 10 mg treatment arms from study NCT01089725 were not included in the current safety analysis since the 10 mg was not assessed in post‐2015 trials NCT02697773 and NCT02709486. Co‐primary endpoints in each trial were change in Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC ) Pain, WOMAC Physical Function, and patient global assessment of OA (PGA‐OA) scores from baseline to end of the treatment period.

Subgroups of interest

In this study, efficacy and safety analyses were done in the overall pooled patient populations and in pre‐specified subgroups of interest that included men or women; age <65, ≥65, or ≥75 years; BMI of <25, 25 to <30, 30 to <35, or ≥35 kg/m2; diabetes or no diabetes; baseline WOMAC Pain score <7 or ≥7 (scores range from 0 to 10 with ≥7 representing severe pain); and KL grades 2, 3 or 4 in the index joint (KL grades of 0, 1, 2, 3 and 4 represent no, doubtful, minimal, moderate and severe OA, respectively). Index joint was defined as the most painful joint at screening with a qualifying WOMAC Pain score and radiographic KL grade as confirmed by a central reader. Patients were included in the diabetes subgroup if they had a medical history of type 1 or type 2 diabetes mellitus, hyperglycaemia or had a baseline haemoglobin A1c ≥6.5%.

Efficacy analysis

Efficacy was based on change, from baseline to week 16, in WOMAC Pain, WOMAC Physical Function and PGA‐OA scores using patient‐level data derived from studies NCT02697773 and NCT02709486. Week 16 was chosen since it was the longest treatment duration common to both studies. WOMAC Pain and Physical Function scores range from 0 to 10, with higher score indicating greater pain severity or function impairment, respectively. PGA‐OA scores range from 1 to 5, with higher scores indicating worse disease status.

Least squares (LS) mean changes from baseline to week 16 in these measures were assessed for the placebo, tanezumab 2.5 mg and tanezumab 5 mg treatment arms using an analysis of covariance (ANCOVA) model including terms for baseline score of the corresponding endpoint, baseline daily average pain score, index joint (hip or knee), treatment and study. A multiple imputation approach was used for missing data, dependent on the reason for missing data. For patients with missing data because of discontinuation prior to week 16 for lack of efficacy, for an adverse event, or death, imputation was based on sampling from a normal distribution using a mean value equal to the patient's baseline efficacy value and the SD (over all treatment groups) of the observed efficacy data at week 16. For patients with missing data for any other reason, imputation was based on sampling from a normal distribution using a mean value of the patient's last observed efficacy value and SD (over all treatment groups) of the observed efficacy data at week 16. The proportion of patients achieving ≥50% (substantial) and ≥30% (moderate) improvement from baseline to week 16 in WOMAC Pain was assessed using a logistic regression method with terms for baseline WOMAC Pain subscale score and baseline daily average pain score, and classification variables of index joint, treatment and study. A mixed last‐observation carried forward/baseline‐observation carried forward approach was used for missing data.

Treatment comparisons were based on LS mean differences from placebo (or odds ratios for 50% and 30% responder data), associated 95% confidence intervals (CIs) and P‐values. For all comparisons, nominal significance was declared if the two‐tailed test for the difference between treatment groups was significant at the 0.05 level.

Safety analysis

Safety was based on the occurrence of treatment‐emergent adverse events (TEAEs) over the full study (treatment + safety follow‐up combined) periods in studies NCT02697773 (16‐week treatment + 24‐week follow‐up), NCT02709486 (24‐week treatment + 24‐week follow‐up) and NCT01089725 (16‐week treatment + 8‐week follow‐up). TEAEs were coded using Medical Dictionary for Regulatory Activities v22.0, with severity assessed by site investigators, and were summarised descriptively in the overall patient population and in the subgroups of interest for the placebo, tanezumab 2.5 mg and tanezumab 5 mg treatment arms.

RESULTS

Efficacy

Patient demographics

The overall efficacy population included 1545 patients (placebo = 514, tanezumab 2.5 mg = 514, tanezumab 5 mg = 517). The population was predominantly women (67.3%), white (80.5%) and had an approximate mean age of 63 years (Table 2). OA disease duration ranged from 7.9 to 8.7 years across groups, with a majority (84.1%) of patients having knee as the index joint. Mean baseline WOMAC Pain, WOMAC Physical Function and PGA‐OA scores were 6.9, 7.0 and 3.5, respectively, in all treatment groups.

TABLE 2

Demographics of the overall efficacy population (NCT02697773 and NCT02709486)

Characteristic	Placebo (n = 514)	Tanezumab 2.5 mg (n = 514)	Tanezumab 5 mg (n = 517)
Gender, n (%)
Male	161 (31.3)	171 (33.3)	173 (33.5)
Female	353 (68.7)	343 (66.7)	344 (66.5)
Mean (SD) age, years	62.5 (9.8)	63.2 (9.4)	63.4 (9.9)
Race, n (%)
White	403 (78.4)	423 (82.3)	418 (80.9)
Black or African American	60 (11.7)	43 (8.4)	50 (9.7)
Asian	47 (9.1)	43 (8.4)	42 (8.1)
Other	4 (0.8)	5 (1.0)	7 (1.4)
Mean (SD) disease duration, years	8.7 (8.1)	7.9 (7.8)	8.3 (7.2)
Mean (SD) baseline WOMAC Pain a	6.9 (1.1)	6.9 (1.1)	6.9 (1.1)
Mean (SD) baseline WOMAC Physical Function b	7.0 (1.1)	7.0 (1.0)	7.0 (1.1)
Mean (SD) baseline PGA‐OA c	3.5 (0.6)	3.5 (0.6)	3.5 (0.6)
Index joint, n (%) d
Hip	80 (15.6)	83 (16.1)	83 (16.1)
Knee	434 (84.4)	431 (83.9)	434 (83.9)

Abbreviations: PGA‐OA, patient global assessment of osteoarthritis; SD, standard deviation; WOMAC, Western Ontario and McMaster Universities Osteoarthritis Index.

Scores range from 0 to 10, with higher scores indicating greater pain severity.

Scores range from 0 to 10, with higher scores indicating greater functional impairment.

Scores range from 1 = very good to 5 = very poor.

Index joint was defined as the most painful joint at baseline with a qualifying WOMAC Pain score and Kellgren‐Lawrence grade as confirmed by a central reader.

Change in WOMAC Pain, WOMAC Physical Function and PGA‐OA

Improvements in WOMAC Pain from baseline to week 16 were numerically greater and often statistically significant (P < .05) in both tanezumab groups compared with placebo, irrespective of gender, age, BMI, diabetes status, baseline WOMAC Pain score or KL grade in the index joint (Figure 1; the number of patients in each subgroup can be seen within the figure). The only improvements that did not reach the level of significance for tanezumab vs placebo were the 2.5 mg dose in the male subgroup, the 2.5 mg dose in the age ≥75 years subgroup, the 5 mg dose in the <25 kg/m2 BMI subgroup, both doses in the ≥35 kg/m2 BMI subgroup, the 2.5 mg dose in the patients with diabetes subgroup and both doses in the KL grade 2 in the index joint subgroup. In the overall patient population (all subgroups combined), improvements in WOMAC Pain from baseline to week 16 were significantly greater in both tanezumab groups compared with the placebo group. LS mean (standard error) change from baseline was −3.1 (0.12) for tanezumab 2.5 mg and −3.2 (0.12) for tanezumab 5 mg compared with −2.5 (0.12) for placebo (both tanezumab groups P < .0001 vs placebo). Improvements in WOMAC Physical Function (Figure S1) and PGA‐OA (Figure S2) also favoured tanezumab over placebo in all the subgroups of interest and often reached the level of significance.

FIGURE 1

Change in WOMAC Pain† from baseline to week 16 (NCT02697773 and NCT02709486). Symbols: ○ = tanezumab 2.5 mg; ● = tanezumab 5 mg. †Scores range from 0 to 10, with higher scores indicating greater pain severity. ‡Patients were included in the diabetes group if they had a medical history of hyperglycaemia, type 1 or type 2 diabetes mellitus, or a baseline haemoglobin A1c ≥6.5%. §In the index joint; defined as the most painful joint at screening with a qualifying WOMAC Pain score and KL grade as confirmed by a central reader. N represents the number of patients, with available baseline data, included in the analysis. *P < .05 vs placebo. BMI, body mass index; CI, confidence interval; KL, Kellgren‐Lawrence; LS, least squares; WOMAC, Western Ontario and McMaster Universities Osteoarthritis Index

Proportion of patients with ≥50% or ≥30% improvement in WOMAC Pain

The proportion of patients achieving ≥50% improvement in WOMAC Pain from baseline to week 16 was numerically greater and often statistically significant (P < .05) in both tanezumab groups compared with placebo, irrespective of gender, age, BMI, diabetes status, baseline WOMAC Pain score or KL grade in the index joint (Figure 2). The only proportion of 50% responders that did not reach the level of significance for tanezumab vs placebo were the 2.5 mg dose in the age ≥75 years subgroup, the 5 mg dose in the <25 kg/m2 BMI subgroup, both doses in the ≥35 kg/m2 BMI subgroup, both doses in the patients with diabetes subgroup and both doses in the KL grade 2 in the index joint subgroup. In the overall patient population (all subgroups combined), the proportion of patients achieving ≥50% improvement in WOMAC Pain was significantly greater in both tanezumab (2.5 mg = 51.9%, 5 mg = 51.8%; both P < .0001) groups compared with the placebo (36.8%) group in the overall population. Similar to the 50% responder threshold, the proportion of patients achieving ≥30% improvement in WOMAC Pain in the tanezumab (2.5 mg = 68.0%, 5 mg = 69.4%; both P < .0001) groups was significantly greater than the placebo (55.6%) group in the overall population and in many subgroups of interest (Figure S3).

FIGURE 2

Proportion of patients achieving ≥50% improvement in WOMAC Pain† from baseline to week 16 (NCT02697773 and NCT02709486). Symbols: ○ = tanezumab 2.5 mg; ● = tanezumab 5 mg. †Scores range from 0 to 10, with higher scores indicating greater pain severity. ‡Patients were included in the diabetes group if they had a medical history of hyperglycaemia, type 1 or type 2 diabetes mellitus, or a baseline haemoglobin A1c ≥6.5%. §In the index joint; defined as the most painful joint at screening with a qualifying WOMAC Pain score and KL grade as confirmed by a central reader. N represents the number of patients, with available baseline data, included in the analysis. *P < .05 vs placebo. BMI, body mass index; KL, Kellgren‐Lawrence; WOMAC, Western Ontario and McMaster Universities Osteoarthritis Index

Safety

The overall safety population included 1754 patients (placebo = 586, tanezumab 2.5 mg = 602, and tanezumab 5 mg = 566). Demographics of the overall safety population (Table 3) were similar to those of the overall efficacy population.

TABLE 3

Demographics of the overall safety population (NCT02697773, NCT02709486 and NCT01089725)

Characteristic	Placebo (n = 586)	Tanezumab 2.5 mg (n = 602)	Tanezumab 5 mg (n = 566)
Gender, n (%)
Male	186 (31.7)	199 (33.1)	198 (35.0)
Female	400 (68.3)	403 (66.9)	368 (65.0)
Mean (SD) age, years	62.3 (10.2)	62.9 (9.5)	63.2 (10.1)
Race, n (%)
White	463 (79.0)	494 (82.1)	458 (80.9)
Black or African American	70 (11.9)	54 (9.0)	56 (9.9)
Asian	49 (8.4)	47 (7.8)	44 (7.8)
Other	4 (0.7)	7 (1.2)	8 (1.4)
Mean (SD) disease duration, years	8.9 (8.4)	8.0 (7.9)	8.4 (7.5)
Mean (SD) baseline WOMAC Pain a	7.0 (1.1)	7.0 (1.2)	7.0 (1.2)
Mean (SD) baseline WOMAC Physical Function b	7.0 (1.1)	7.0 (1.1)	7.0 (1.1)
Mean (SD) baseline PGA‐OA c	3.5 (0.6)	3.5 (0.6)	3.5 (0.6)
Index joint, n (%) d
Hip	80 (13.7)	88 (14.6)	78 (13.8)
Knee	506 (86.3)	514 (85.4)	488 (86.2)

Abbreviations: PGA‐OA, patient's global assessment of osteoarthritis; SD, standard deviation; WOMAC, Western Ontario and McMaster Universities Osteoarthritis Index.

Scores range from 0 to 10, with higher scores indicating greater pain severity.

Scores range from 0 to 10, with higher scores indicating greater functional impairment.

Scores range from 1 = very good to 5 = very poor.

Index joint was defined as the most painful joint at baseline with a qualifying WOMAC Pain score and Kellgren‐Lawrence grade as confirmed by a central reader.

Adverse events (AEs)

TEAEs in the overall safety population are summarised in Table 4. TEAE rates were largely similar in the tanezumab groups (2.5 mg = 62.8%, 5 mg = 62.0%) relative to the placebo group (60.9%). Rates of serious (placebo = 3.6%, tanezumab 2.5 mg = 5.3%, tanezumab 5 mg = 5.5%) and severe (placebo = 3.9%, tanezumab 2.5 mg = 4.7%, tanezumab 5 mg = 6.4%) TEAEs were slightly higher with tanezumab relative to placebo. Rates of treatment discontinuations caused by TEAEs, however, were lower in the tanezumab groups (2.5 mg = 1.3%, 5 mg = 0.9%) than the placebo group (2.0%). Among common TEAEs (TEAEs occurring in ≥3% of patients in any treatment group), rates of peripheral oedema (placebo = 0.3%, tanezumab 2.5 mg = 1.2%, tanezumab 5 mg = 3.2%) and paresthesia (placebo = 1.2%, tanezumab 2.5 mg = 2.5%, tanezumab 5 mg = 3.0%) were at least twice as high in both tanezumab groups than in the placebo group.

TABLE 4

Summary of TEAEs the overall safety population over the full study (treatment + follow‐up) period (NCT02697773, NCT02709486 and NCT01089725)

Patients, n (%)	Placebo (n = 586)	Tanezumab 2.5 mg (n = 602)	Tanezumab 5 mg (n = 566)
With any TEAE	357 (60.9)	378 (62.8)	351 (62.0)
With any serious TEAE	21 (3.6)	32 (5.3)	31 (5.5)
With any severe TEAE	23 (3.9)	28 (4.7)	36 (6.4)
Discontinued treatment due to TEAE	12 (2.0)	8 (1.3)	5 (0.9)
Discontinued study due to TEAE	5 (0.9)	10 (1.7)	2 (0.4)
Common TEAEs b
Arthralgia	95 (16.2)	91 (15.1)	83 (14.7)
Nasopharyngitis	49 (8.4)	61 (10.1)	47 (8.3)
Back pain	32 (5.5)	42 (7.0)	34 (6.0)
Headache	33 (5.6)	34 (5.6)	24 (4.2)
Osteoarthritis	19 (3.2)	22 (3.7)	24 (4.2)
Pain in extremity	16 (2.7)	26 (4.3)	21 (3.7)
Musculoskeletal pain	23 (3.9)	31 (5.1)	20 (3.5)
Fall	21 (3.6)	35 (5.8)	19 (3.4)
Upper respiratory tract infection	13 (2.2)	18 (3.0)	19 (3.4)
Joint swelling	13 (2.2)	17 (2.8)	18 (3.2)
Peripheral oedema	2 (0.3)	7 (1.2)	18 (3.2)
Paraesthesia	7 (1.2)	15 (2.5)	17 (3.0)

Abbreviation: TEAE, treatment‐emergent adverse event.

This is a total of 40, 48 and 24 wk for NCT02697773, NCT02709486 and NCT01089725, respectively.

Reported in ≥3% of patients in any treatment group. Bolding indicated the event was reported at a higher frequency in both tanezumab groups relative to the placebo group.

A summary of AEs in the subgroups of interest can be found in Tables [Link], [Link], [Link], [Link], [Link], [Link]. In general, the profile of TEAEs (proportion of overall TEAEs, serious TEAEs, severe TEAEs, discontinuations caused by TEAEs and common TEAEs) in most subgroups of interest was broadly similar to the profile in the overall patient population, and only a few differences were noted. For example, women reported numerically more TEAEs with tanezumab (2.5 mg = 64.8%, 5 mg = 66.0%) than men (2.5 mg = 58.8%, 5 mg = 54.5%). In addition, overall TEAEs were more frequent in both tanezumab groups than the placebo group in the age ≥65 and age ≥75 years, but not the age <65 years, subgroups. Finally, patients with diabetes reported more severe TEAEs with tanezumab (placebo = 2.2%, tanezumab 2.5 mg = 7.8%, tanezumab 5 mg = 12.0%) than patients without diabetes (placebo = 4.3%, tanezumab 2.5 mg = 4.0%, tanezumab 5 mg = 5.2%), though overall rates of discontinuations caused by TEAEs were similar (1.1%‐2.2%).

DISCUSSION

In this analysis of over 1500 patients with moderate‐to‐severe OA of the knee and hip and a history of inadequate response to other standard OA analgesics, improvements in pain, function and PGA‐OA favoured SC tanezumab (2.5 and 5 mg) over placebo irrespective of gender, age, BMI, diabetes status, baseline pain severity or KL grade in the index joint. The TEAE profile of tanezumab was mostly consistent among the subgroups and broadly similar to the profile in the overall patient population.

The subgroups in these analyses were selected to determine the consistency of treatment response across a number of demographic and clinical variables. Evidence suggests that men and women experience pain, and respond to some treatments, differently. , , Though the underlying mechanisms are unknown, this may be because of the inherent differences in drug metabolism, levels of drug receptors, psychological factors, the endocrine system or processing pathways in men and women. , , Advanced age is also known to affect response to small molecule analgesics because of physiological changes associated with aging, including increased drug absorption caused by slowing of the gastrointestinal tract, decreased drug metabolism caused by decreased hepatic function and reduced renal excretion. , In addition to age, body weight and BMI can also affect the pharmacokinetics of drug therapies by affecting absorption, distribution and clearance. Likewise, diabetes can potentially affect treatment response via changes in drug pharmacokinetics caused by altered drug absorption (via changes in blood flow in subcutaneous adipose tissue or muscle tissue) or excretion (via decreased renal function). The presence of diabetic peripheral neuropathy, which alters nociceptive signalling, could also affect analgesic responses in patients with diabetes. Finally, disease severity may also affect response to treatment. There is some evidence in knee OA, for example that radiologic severity of disease (assessed by KL grading) may be a possible predictor of efficacy response to intra‐articular corticosteroid injection. , It should be noted, however, that OA severity as assessed by radiographic KL grading does not necessarily correlate with OA‐related pain severity. , Thus, we assessed both objective (KL grade) and subjective (baseline pain scores in the index joint) measures of OA severity.

Improvements in WOMAC Pain, WOMAC Function and PGA‐OA favoured tanezumab over placebo in all subgroups and many, but not all, comparisons reached the level of statistical significance. Because of the low number of patients in many of the subgroups, however, it is not surprising that not all comparisons vs placebo were significant. Low patient numbers were particularly evident in the age ≥75 years (56‐64), patients with diabetes (86‐92) and <25 kg/m2 BMI (64‐72) subgroups. LS mean differences vs placebo in these subgroups with low patient numbers were similar to (and often greater than) the observed LS mean differences vs placebo observed in subgroups with greater patient numbers, but there was greater variability (ie, 95% CIs) in the low patient number subgroups that contributed to the lack of statistical significance. In addition, a higher placebo response in the KL grade 2 subgroup (LS mean change from baseline in WOMAC Pain = −2.9) relative to the KL grade 3 subgroup (LS mean change from baseline in WOMAC Pain = −2.6) and, particularly, the KL grade 4 subgroup (LS mean change from baseline in WOMAC Pain = −1.8) may have contributed to the lack of significant effect for both doses of tanezumab in the KL grade 2 subgroup. This observation also suggests that there may be an inverse correlation between the magnitude of the placebo effect and the degree of structural OA severity. The LS mean change for the tanezumab groups was more similar across all KL grades, ranging from −2.9 to −3.4.

Overall, conclusions on the clinical significance of treatment effect in the various subgroups are difficult to make because of the limited number of patients in some subgroups and the fact that subgroups were not directly compared with each other. It is notable, however, that statistically greater proportions of tanezumab‐treated patients achieved ≥30% and ≥50% improvement in WOMAC Pain compared with placebo‐treated patients in most subgroups. These 30% and 50% thresholds represent moderate and substantial, respectively, improvements in pain for patients with chronic pain conditions and suggest that the benefits observed in this analysis were clinically meaningful in many of the subgroups.

Though the TEAE profile of tanezumab among the subgroups of interest was broadly consistent with the overall population, there were a few instances where the profile appeared somewhat different across subgroups. Woman reported more TEAEs with tanezumab than men. It is possible that this may represent a modest, but real, difference between the genders as patient numbers were high in each subgroup and previous studies suggest that women experience (or report) adverse drug reactions at a higher frequency than men across all drug classes. TEAE rates were higher among tanezumab‐treated patients than placebo‐treated patients in older (aged ≥65 and ≥75 years) patients, but not in younger (aged <65 years) patients. This could represent a difference in tanezumab's safety profile among older patients since the risk of AEs to drug treatments in general, as well as the risk for drug‐drug interactions, is increased with advanced age. A more likely reason for this discrepancy, however, may be because of the higher TEAE rates among placebo‐treated patients in the age <65 years subgroup (65.9%) relative to the age ≥65 years (50.7%) and age ≥75 years (54.7%) subgroups. Further, TEAE rates among tanezumab‐treated patients were more comparable across the three age subgroups (58.6%‐65.7%) and similar to the TEAE rate among all tanezumab‐treated patients in the overall safety population (62.0%‐62.8%).

Joint safety events, particularly rapidly progressive OA, are associated with nerve growth factor antibodies such as tanezumab in some patients. , , Though a detailed analysis of joint safety is beyond the scope of this paper, we note that OA was reported as an AE (representing new [non‐index joint] or worsening [index joint] cases of OA) more often among tanezumab‐treated patients than among placebo‐treated patients in the overall safety population and in many subgroups of interest.

Our findings regarding the use of SC tanezumab in patients with moderate‐to‐severe OA and a history of inadequate response to other analgesics agree with, and build upon, a previous analysis of intravenous (IV) tanezumab in patients with OA. Like the current SC analysis, the previous IV analysis demonstrated that tanezumab provided significant improvement of pain, function and global disease status in subpopulations of patients based on age, baseline pain severity, diabetes status and BMI without identifying new safety risks in those subpopulations. Our findings are also in broad agreement with a recent pharmacokinetic analysis of IV and SC data suggesting that tanezumab dosing does not need to be adjusted based on factors such as gender, age or BMI.[submitted manuscript ]

Limitations of this study include the low number of patients in many subgroups and the post‐hoc nature of our analyses. In addition, comparisons were with placebo, and different subgroups were not directly compared with each other (eg, men vs women). Likewise, the two doses of tanezumab were not directly compared with each other and conclusions on relative efficacy are limited. Efficacy was based on changes in pain after 16 weeks (two doses) of treatment, and findings should not be generalized to longer treatment durations. Finally, it should be noted that the overall efficacy and safety populations were largely women (approximately two‐thirds of all patients) and white (over 80% of patients). This may limit the ability to generalize our findings to other OA populations, particularly those of predominantly non‐white racial groups. However, strengths of the study include overall large patient population from which the subgroups were derived and similarity in trial design for the studies included in the analysis.

CONCLUSIONS

In summary, treatment with SC tanezumab (at doses of 2.5 or 5 mg) every 8 weeks provided improvements over placebo (often reaching the level of statistical significance) in pain, function and overall OA disease status in all patient subgroups, which were based on gender, age, BMI, diabetes status, baseline pain severity or KL grade in the index joint. The overall TEAE profile of tanezumab was mostly consistent among the subgroups, broadly similar to the profile in the overall patient population, and no new safety risks were identified in the subgroups. These findings suggest that tanezumab efficacy is maintained regardless of comorbidity, disease severity and selected clinical criteria in patients with moderate‐to‐severe OA and a history of inadequate response to standard OA analgesics. Conclusions, however, are limited by the small number of patients in some subgroups.

DISCLOSURE

TJS reports clinical research study support from Pfizer, Eli Lilly and Company, Regeneron, Galapagos, Taiwan Liposome Corporation, and Anika Therapeutics and has served as a consultant or on an advisory board for Pfizer, Eli Lilly and Company, Glaxo‐Smith Kline, AstraZeneca, Noven, Galapagos and Merck. FB has received grants through his institution from TRB Chemedica, MSD and Pfizer; has worked as a consultant to Novartis, MSD, Pfizer, Eli Lilly and Company, UCB, AbbVie, Roche, Servier, Sanofi‐Aventis, Flexion Therapeutics, Expanscience, GSK, Biogen, Nordic, Sandoz, Regeneron, Gilead, Bone Therapeutics, Regulaxis, Peptinov and 4P Pharma; has served as an instructor for Sandoz; and has served as a speaker for Novartis, MSD, Pfizer, Eli Lilly and Company, UCB, AbbVie, Roche, Servier, Sanofi‐Aventis, Flexion Therapeutics, Expanscience, GSK, Biogen, Nordic, Sandoz, Regeneron, Gilead and Sandoz. AJK owns stock in Pfizer, Amgen, Gilead and GSK; has served as a consultant to AbbVie, Janssen, Novartis, Pfizer, Regeneron, SUN Pharma Advanced Research, Boehringer Ingelheim, and Gilead; and has served as a speaker for Celgene, Horizon, Merck, Novartis, Pfizer, Regeneron, Flexion and AbbVie. LV and EJ are employees of, and own stock in, Eli Lilly and Company. RY, EW, LT and DS are employees of, and own stock/options in, Pfizer.

ROLE OF THE FUNDER

The study was sponsored by Pfizer Inc (manufacturer of tanezumab) and Eli Lilly and Company. Authors from Pfizer Inc and Eli Lilly and Company contributed to study design; data collection (Pfizer), management (Pfizer) and interpretation of data; and preparation, review and approval of the manuscript. All authors had full access to study data and final responsibility for submission.

Fig S1

Click here for additional data file.

Fig S2

Click here for additional data file.

Fig S3

Click here for additional data file.

Table S1

Click here for additional data file.

Table S2

Click here for additional data file.

Table S3

Click here for additional data file.

Table S4

Click here for additional data file.

Table S5

Click here for additional data file.

Table S6

Click here for additional data file.