Evaluation of the efficacy and tolerance of pamidronate in complex regional pain syndrom type 1

Abstract

This study aimed to determine whether the delay between symptom onset and treatment initiation, the dose of pamidronate, and bone mineral density (BMD) influence the response to pamidronate treatment in complex regional pain syndrome type 1 (CRPS 1). A retrospective observational study included patients treated with pamidronate between 2013 and 2023. Treatment response was assessed based on symptom regression according to the Budapest criteria at one (M1) and four months (M4) post-treatment. Multivariate logistic regression identified factors associated with response. Among the 255 patients included, 14.5% responded at M1 and 67% at M4. Multivariate analysis showed that post-traumatic (OR 2.75, 95% CI [1.36–5.7], P = 0.0053) or idiopathic etiology (OR 5.47, 95% CI [1.86–18.92], P = 0.0037) compared to post-surgical etiology, and the presence of initial edema (OR 2.39, 95% CI [1.26–4.62], P = 0.0082), were associated with a better response at M4. BMD, treatment delay, and pamidronate dosage were not significantly associated with treatment response. These findings suggest that initial edema is a predictive factor for response to pamidronate in CRPS 1, with syndrome etiology also influencing outcomes. Increasing pamidronate dosage or infusion frequency does not seem to improve therapeutic efficacy.

Introduction

Complex Regional Pain Syndrome Type 1 (CRPS 1) is characterized by the onset of disproportionate localized pain occurring with or without a triggering event. This pain is accompanied by multiple symptoms, including swelling, excessive sweating, vasomotor disturbances, and sensory abnormalities¹. Bisphosphonates have emerged as a promising treatment option for CRPS-1, primarily due to their antiresorptive and anti-inflammatory properties that target bone metabolism abnormalities implicated in CRPS pathophysiology. Several randomized controlled trials (RCTs) and meta-analyses have reported that bisphosphonates, including pamidronate, neridronate, and alendronate, significantly reduce pain and improve functional outcomes in patients with CRPS-1 compared to placebo or standard care^2–9. Pamidronate, administered intravenously, has shown efficacy in reducing acute and chronic pain symptoms and improving mobility, with a relatively favorable safety profile. Neridronate is notably approved for CRPS-1 treatment in Italy since 2015, following robust clinical trials demonstrating its benefits¹⁰. The timing of bisphosphonate administration appears critical, with earlier treatment potentially correlating with improved outcomes, although this remains to be definitively established due to the limited size of available studies. Current evidence is limited by small sample sizes, heterogeneity in dosing regimens and treatment durations, and a lack of consensus on optimal protocols^10–15. Furthermore, the precise mechanisms underlying bisphosphonates’ analgesic effects in CRPS remain partially understood, involving modulation of osteoclastic activity, neurogenic inflammation, and central sensitization pathways¹. Moreover, while several studies have emphasized the central role of bone remodeling and osteoclastic activity in the pathophysiology of CRPS-1 and the mechanism of action of bisphosphonates, they do not explicitly evaluate bone mineral density as a predictive factor for treatment response^4,13. However, given that bisphosphonates primarily target bone resorption, it is plausible to hypothesize that variations in bone density may influence therapeutic efficacy. This study aims to determine whether the delay between symptom onset and treatment, the cumulative dose of pamidronate, and the patient’s bone mineral density predict the clinical response to intravenous pamidronate in patients with CRPS 1. We hypothesize that an earlier treatment, a higher cumulative dose, and lower bone mineral density would be associated with therapeutic response.

Methods

Study design

This retrospective, observational, real-practice study included all adult patients (≥ 18 years) who received at least one infusion of pamidronate for the treatment of CRPS 1 at the Rheumatology Department of CHU Amiens-Picardie between January 1, 2013, and December 31, 2023. Data were collected from patient records using the DxCare software of CHU Amiens-Picardie. This retrospective study was reviewed and approved by the institutional ethics committee of Amiens University Hospital (Comité de Protection des Personnes Nord‑Ouest II, Amiens, France). The committee confirmed that, under French law (Loi Jardé) and CNIL reference methodology MR-004 (registration number PI2024_843_0089), written informed consent was not required. All patients received an information letter and had the opportunity to opt out. The need to obtain informed consent was waived by the Comité de Protection des Personnes Nord‑Ouest II, Amiens, France. All methods were performed in accordance with the relevant guidelines.

Diagnosis of CRPS 1 and pamidronate administration

The diagnosis of CRPS 1 was confirmed according to the Budapest criteria^16,17, which were assessed during a consultation in the Rheumatology Department that included a detailed clinical examination. No imaging was used to support the diagnosis. Local oedema was assessed clinically. Pamidronate infusion protocols varied among practitioners and included: 1–3 infusions of 60 or 90 mg over three days, a single infusion of 15 mg, 60 mg or 90 mg per month over one to three months, or, in rare cases, up to seven infusions over seven months depending on symptom persistence. All patients underwent a biological phosphocalcic assessment prior to pamidronate infusion, which included serum calcium, phosphate, vitamin D levels, and renal function evaluation. Parathyroid hormone (PTH) levels were not routinely measured. Vitamin D supplementation was prescribed concomitantly for all patients, and calcium intake was assessed and supplemented if found insufficient.

Patients

Inclusion criteria encompassed all patients ≥ 18 years who received at least one infusion of pamidronate for CRPS 1. The only exclusion criterion was any treatment in the year preceding the study that could affect bone metabolism, e.g. osteoporosis treatment. Patients were categorized into two groups based on their response to treatment: the “responder” group included patients whose Budapest criteria regressed after the first IV infusion of pamidronate at M1 and M4. The “non-responder” group included patients with persistent Budapest criteria. The reevaluation of the Budapest criteria to classify patients as responders or non-responders was performed during follow-up consultations in the Rheumatology Department, using clinical examinations consistent with those conducted at diagnosis.

Data collection

The following variables were collected: age at first pamidronate infusion; gender; etiology of CRPS 1: post-surgical, post-traumatic (fractures or injuries), post-stroke, idiopathic; location of CRPS 1; symptoms at each pamidronate infusion: pain, stiffness, edema, warmth, sweating; BMD status : osteoporosis or osteopenia, defined respectively by a T-score (Z-score for individuals < 50 years) between − 1 and − 2.5 (osteopenia) or below − 2.5 (osteoporosis); time from symptom onset to first pamidronate infusion: number of pamidronate infusions, dose in mg. None of the patients had chronic renal failure (CKD) or transplantation, conditions which contraindicate the use of bisphosphonates. Furthermore, none of the patients were receiving hormone therapy for breast or prostate cancer.

Statistical analysis

Comparisons between groups were conducted using the Student’s t-test for continuous variables, Fisher’s exact test for categorical variables, and the Mann-Whitney U test for non-normally distributed variables. Statistical significance was set at p < 0.05. Odds ratios (OR) with 95% confidence intervals (CI) were calculated. Multivariate analysis was performed using a logistic regression model, adjusting for variables significant in univariate analysis, as well as age and gender. All analyses were conducted using RStudio software.

Results

Study population

Patient characteristics at inclusion are presented in Table 1. A total of 255 patients were included in the study. The mean age was 53.7 ± 12.6 years, with women comprising the majority (76.8%). The predominant etiology of CRPS 1 was post-surgical, observed in 43.9% of patients. CRPS 1 mainly affected the lower limbs (73.33%), with preferential localization in the foot in 34.9% of cases. Multiple-site involvement was observed in 37 patients, 14 of whom had simultaneous foot and ankle involvement. No patient included had a history of CRPS 1 prior to 2013. Among the patients, 25.7% had undergone bone densitometry prior to their first pamidronate infusion. The mean time from symptom onset to the first pamidronate infusion was 23.5 months. Patients received a mean of 2.5 ± 1.4 infusions of pamidronate, with a mean cumulative dose of 215.8 ± 104.8 mg. A total of 37 patients, or 14.5%, were classified as responders at M1, while 171 patients, or 67.1%, were responders at M4.

Table 1.

Characteristics of patients with CRPS-1 treated with pamidronate at inclusion.

Patient population (n = 255)	n (%)
Mean age +/- SD	53.71 +/- 12.66
Female	196 (76.86%)
CRPS 1
Etiology
Post-surgical	122 (43.92%)
Post-traumatic	106 (41.57%)
Idiopathic	35 (13.73%)
Post-stroke	2 (0.78%)
Localization
Upper/lower limb	67 (26.27%)/187 (73.33%)
Foot	89 (34.9%)
Ankle	49 (19.22%)
Multiple	37 (16.97%)
Shoulder	26 (10.02%)
Knee	26 (10.02%)
Wrist	11 (4.31%)
Hand	8 (3.14%)
Elbow	5 (1.96%)
Symptoms at baseline
Pain	253 (99.22%)
Stiffness	172 (67.45%)
Edema	139 (54.51%)
Warmth	118 (46.46%)
Sweating	32 (12.60%)
Bone densitometry at baseline	58 (25.75%)
Time onset of symptoms – perfusion in months +/- SD	23.52 +/- 122.17
Pamidronate
Mean number of perfusions +/- SD	2.51 (+/- 1.47)
Mean dose of pamidroante in mg +/- SD	215.82 +/- 104.8
Mean duration of protocol +/- SD	2.58 +/- 1.44
Mean cumulative dose per month +/- SD Responders at M1 Responders at M4	84.53 +/- 12.17 37 (14.5%) 171 (67.1%)

Pamidronate infusion protocols

Patients included in the study received different pamidronate infusion protocols. The most used protocol was 3 infusions of 90 mg spaced 1 month apart, given to 127 patients (49.08%). This was followed by a single 90 mg infusion, applied to 46 patients (18.03%), and two 90 mg infusions spaced 1 month apart, used for 28 patients (10.98%). Additionally, 28 patients (10.98%) received 3 infusions of 60 mg spaced 1 month apart, while 10 patients (3.92%) were treated with a single 60 mg infusion. Less frequently used protocols included: 4 infusions of 90 mg spaced 1 month apart for 6 patients (2.35%), 2 infusions of 60 mg spaced 1 month apart for 4 patients (1.57%), and 15 infusions of 60 mg for 2 patients (0.78%). Other rare protocols included: 2 infusions of 60 mg followed by 2 infusions of 90 mg spaced 1 month apart for 1 patient (0.39%), 3 infusions of 15 mg spaced 1 month apart for 1 patient (0.39%), 5 infusions of 90 mg spaced 1 month apart for 1 patient (0.39%), and 7 infusions of 90 mg spaced 1 month apart for 1 patient (0.39%).

Comparison of patient characteristic responders at M1 or M4

Comparisons of patient characteristics between responders and non-responders are available in Table 2. CRPS 1 etiologies differed significantly between responder and non-responder groups at M1 (p = 0.0049) and M4 (p = 0.0054). Non-responders mainly had a postsurgical etiology, while responders were more affected by post-traumatic causes. At M4, the localization of CRPS 1 also differed significantly between the groups (p = 0.0375), with responders showing a higher association with ankle and wrist involvement, while non-responders had a greater prevalence of knee and hand involvement. Furthermore, the presence of edema at baseline was significantly more common among responders at M4 (p = 0.0043).

Table 2.

Comparisons of patient characteristics between responders and non-responders to pamidronate.

	M1			M4
Characteristics	Responders (n = 37)	Non responders (n = 218)	p	Responders (n = 171)	Non responders (n = 84)	p
Age	50.97 (+/−10.08)	54.04 (+/−12.99)	0.1077	54.2 (12.34)	52.09 (13.14)	0.2285
Female	28 (75.68)	165 (76.74)	1.000	132 (77.19)	60 (75)	0.8243
CRPS 1
Etiology			0.0049			0.0054
Post-surgery	7 (18.92)	105 (48.84)		63 (36.84)	48 (60)
Post-traumatic	21 (56.76)	82 (38.14)		79 (46.2)	24 (30)
Idiopathic	8 (21.62)	27 (12.56)		28 (16.37)	7 (8.75)
Post-stroke	1 (2.7)	1 (0.47)		1 (0.58)	1 (1.25)
Localization			0.2272			0.0375
Upper/lower limb	10 (27.03)/27 (72.97)	55 (25.58)/159 (73.95)	0.9040	42 (24.56)/128 (74.85)	23 (28.75)/57 (71.25)	0.6272
Foot	12 (32.43)	77 (35.81)		58 (33.92)	31 (38.75)
Ankle	8 (21.62)	40 (18.6)		39 (22.81)	9 (11.25)
Multiple	3 (8.11)	37 (17.21)		25 (14.62)	14 (17.5)
Shoulder	4 (10.81)	21 (9.77)		17 (9.94)	8 (10)
Knee	4 (10.81)	22 (10.23)		15 (8.77)	11 (13.75)
Wrist	2 (5.41)	8 (3.72)		10 (5.85)	0 (0)
Hand	1 (2.7)	7 (3.26)		3 (1.75)	5 (6.25)
Elbow	3 (8.11)	2 (0.93)		4 (2.34)	1 (1.25)
Symptoms at baseline
Pain	36 (97.3)	214 (99.53)	0.68	169 (98.83)	80 (100)	0.8341
Stiffness	14 (66.67)	158 (83.6)	0.11	112 (81.75)	60 (83.33)	0.8503
Edema	15 (68.18)	124 (65.61)	1.00	103 (72.54)	35 (51.47)	0.0043
Warmth	12 (57.14)	65 (36.72)	0.11	79 (61.24)	41 (60.29)	1.00
Sweating	5 (25)	27 (17.88)	0.64	23 (20.72)	9 (15.25)	0.51
Bone densitometry			0.6796			0.0851
Normal	2 (50)	23 (41.82)		22 (47.83)	3 (23.08)
Osteopenia	2 (50)	23 (41.82)		16 (34.78)	9 (69.23)
Osteoporosis	0 (0)	9 (16.36)		8 (17.39)	1 (7.69)
Time onset of symptoms – perfusion			0.42			0.27
In months	15.54 (22.3)	25.2 (132.7)	0.3238	12.17 (15.47)	48.83 (215.69)	0.1329
< 6 months	15 (40.54)	69 (32.09)		60 (35.09)	23 (28.75)
6 months − 1 year	10 (27.03)	81 (37.67)		64 (37.43)	27 (33.75)
> 1 year	12 (32.43)	65 (30.23)		47 (27.49)	30 (37.5)
Pamidronate
Mean number of perfusions at M1/M3	1.11 (0.46)	2.78 (1.44)	< 0.001	2.32 (1.03)	2.99 (2.06)	0.0069
Mean dose in mg at M1/M3	107.03 (49.77)	234.63 (100.95)	< 0.001	245.81 (125.99)	202.11 (91.55)	0.00534

The results in bold are statistically significant (p <  0.05).

Regarding treatment protocols, responders received a significantly lower number of infusions compared to non-responders, both at M1 (p < 0.001) and M4 (p = 0.0069). Additionally, responders were administered lower doses of pamidronate at both M1 (p < 0.001) and M4 (p = 0.00534).

Factors influencing treatment response at M1 and M4

The results of the odds ratio estimation in the univariate analysis are presented in Table 3. The outcomes from the multivariate logistic regression for the variables of interest are shown in Table 4. The multivariate analysis included the following variables: age, sex, number of infusions, etiology, and the presence of edema.

Table 3.

Estimation of odds ratios of characteristics of patients with CRPS I treated with pamidronate in univariate analysis.

	M1			M4
	OR	95% CI	p	OR	95% CI	p
Age	0.98	0.95–1.01	0.1731	1.01	0.99–1.03	0.2170
Gender	1.06	0.44–2.32	0.8873	0.89	0.48–1.67	0.7027
Etiology (vs. post-surgical)
Post-traumatic	3.86	1.63–10.18	0.0035	2.51	1.4–4.57	0.0023
Idiopathic	4.44	1.48–13.74	0.0078	3.03	1.28–8.08	0.0163
Post-stroke	15.03	0.55-407.48	0.0649	0.76	0.03–19.69	0.8489
Symptoms at baseline
Pain	0.17	0.01–4.31	0.211	0
Stiffness	0.39	0.15–1.11	0.063	0.9	0.41–1.88	0.7760
Edema	1.13	0.45–3.06	0.810	2.48	1.36–4.57	0.0030
Warmth	0.44	0.17–1.08	0.075	1.04	0.57–1.9	0.8970
Sweating	1.54	0.47–4.35	0.446	1.45	0.64–3.53	0.3871
Bone densitometry (vs. normal)
Osteopenia	1	0.11–8.94	1.0000	0.24	0.05–0.96	0.0566
Osteoporosis	0			1.09	0.12.23.81	0.9434
Time onset of symptoms – perfusion (vs. < 6 months)
6 months–1 year	0.57	0.23–1.34	0.198	0.9	0.47–1.75	0.7755
> 1 year	0.85	0.36–1.95	0.700	0.6	0.31–1.16	0.1323
Pamidronate
Mean number of perfusions at M1/M3	0.07	0.02–0.15	6.5e-09	0.65	0.47–0.84	0.0043
Mean dose in mg at M1/M3	0.98	0.97–0.98	2.0e-10	1.0	0.99-1	0.0054

The results in bold are statistically significant (p <  0.05).

Table 4.

Logistic regression of patient characteristics with CRPS-I treated with pamidronate.

	M1			M4
	OR	95% CI	p	OR	95% CI	p
Age	0.97	0.92–1.02	0.211	1.04	1.01–1.06	0.0734
Gender	1.03	0.25–3.86	0.965	0.97	0.46–2.08	0.92853
Etiology (vs. post-surgical)
Post-traumatic	1.28	0.34–4.95	0.709	2.75	1.36–5.7	0.00530
Idiopathic	0.62	0.09–3.71	0.607	5.47	1.86–18.92	0.00367
Symptoms : edema	0.63	0.18–2.18	0.455	2.39	1.26–4.62	0.00820
Pamidronate : Mean number of perfusions at M1/M3	0.09	0.03–0.21	< 0.001	0.69	0.49–0.89	0.01158

The results in bold are statistically significant (p <  0.05).

The analysis of etiologies, compared to a post-surgical etiology, shows significant variations between M1 and M4. At M1, in univariate analysis, post-traumatic etiology was associated with an OR of 3.86 (95% CI [1.63–10.18], P = 0.0035), and the absence of etiology found (idiopathic CRPS1) with an OR of 4.44 (95% CI [1.48–13.74], P = 0.0078), indicating a statistically significant association with an increased probability of being a responder. However, in multivariate analysis, after adjustment for confounding variables, these associations were no longer significant: post-traumatic etiology had an OR of 1.28 (95% CI [0.34–4.95], P = 0.709), and the absence of etiology an OR of 0.62 (95% CI [0.09–3.71], P = 0.607). In contrast, at M4, post-traumatic etiology and absence of etiology show a significant association in univariate and multivariate analysis. Post-traumatic etiology is associated with an increased probability of being a responder, with an OR of 2.51 (95% CI [1.4–4.57], P = 0.0023) in univariate and an OR of 2.75 (95% CI [1.36–5.7], P = 0.0053) in multivariate. Similarly, absence of etiology is associated with an increased probability of being a responder with an OR of 3.03 (95% CI [1.28–8.08], P = 0.0163) in univariate and an OR of 5.47 (95% CI [1.86–18.92], P = 0.00367) in multivariate. These results indicate that, compared with a post-surgical etiology, post-traumatic and idiopathic etiologies increase the probability of being a responder to treatment at M4 by 2.75 and 5.47 times, respectively.

The variations in response according to the locations were assessed separately in univariate analysis at M1 and M4. None of the locations analyzed demonstrated a significant association as a factor influencing the response to pamidronate infusions.

The presence of initial edema is significantly associated with an increased probability of being a responder to treatment at M4, although this association is not observed at M1. In univariate analysis at M1, the odds ratio (OR) is 1.13 (95% CI [0.45–3.06], P = 0.810), indicating no significant association. In contrast, at M4, edema is associated with an OR of 2.48 (95% CI [1.36–4.57], P = 0.0030), reflecting a significant increase in the probability of being a responder. In multivariate analysis, the association remained significant at M4, with an OR of 2.39 (95% CI [1.26–4.62], P = 0.0082), while at M1, the OR was 0.63 (95% CI [0.18–2.18], P = 0.455), confirming the absence of a significant effect. These results suggest that patients with initial edema were 2.39 times more likely to be responders to treatment at M4 compared with those without edema.

Analysis of the association between bone mineral density (BMD) and response to anti-osteoporosis treatment did not reveal a significant association. In univariate analysis at M1, no association was observed, with an OR of 1.00 for osteopenia (95% CI [0.11–8.94], P = 1.000) and an OR of 0 for osteoporosis (result not calculable). At M4, osteopenia has an OR of 0.24 (95% CI [0.05–0.96], P = 0.0566), indicating a marginally non-significant association. Similarly, for osteoporosis, the OR was 1.09 (95% CI [0.12–23.81], P = 0.9434), with no significant association with response to treatment.

Regarding the influence of the delay between the onset of symptoms and the first pamidronate infusion, the univariate analysis of the data in months (continuous variable) did not find a significant association (OR 0.99, 95% CI [0.98–0.1], P = 0.33). Similarly, the comparison of a delay of more or less than 3 months did not find a significant difference in the response to pamidronate (OR 1.31, 95% CI [0.43–4.85], P = 0.66). The same was true for delays of 6 months or one year, with an odds ratio of 1.48 (95% CI [0.84–2.66], P = 0.18) and 1.58 (95% CI [0.9–2.77], P = 0.11), respectively.

Significance analysis of the number of infusions showed an inversely proportional association between the mean number of infusions and the probability of treatment response, both in univariate and multivariate analysis, for M1 and M4. In univariate analysis, at M1, the odds ratio (OR) was 0.07 (95% CI [0.02–0.15], P = 6.5 × 10⁻⁹), indicating a significant decrease in the probability of being a responder with an increasing number of infusions. At M4, the OR was 0.65 (95% CI [0.47–0.84], P = 0.0043), showing a similar but less marked trend. In multivariate analysis, the results confirmed this association. At M1, the OR is 0.09 (95% CI [0.03–0.21], P < 0.001), while at M4 it is 0.69 (95% CI [0.49–0.89], P = 0.0116). Thus, a higher mean number of infusions is significantly associated with a reduced probability of treatment response during both evaluation periods.

The analysis of the association between mean pamidronate dose and treatment response was performed in univariate for M1 and M4. At M1, mean dose was associated with an odds ratio (OR) of 0.98 (95% CI [0.97–0.98], P = 2.0 × 10⁻¹⁰), indicating a statistically significant association but without notable clinical relevance. At M4, the OR was 1.0 (95% CI [0.99–1.00], P = 0.0054), showing no significant or clinically relevant association. Due to the lack of clinically significant effect observed in the univariate analysis, the variable “dose” was not included in the multivariate analysis.

Tolerance of pamidronate

No benign or serious adverse effects were observed during the study, either clinically or biologically, in particular no hypocalcemia or worsening of renal function, confirming the good tolerance of the treatment evaluated.

Discussion

Our observational, retrospective and longitudinal study focused on patients with type I CRPS treated with pamidronate between 2013 and 2023. The patient population in our study was consistent with those described in the literature^10,18.

In our study, no significant association was found between the time from symptom onset and clinical response to pamidronate. The mean delay between symptom onset and the first pamidronate infusion was 23.5 months. This interval can be explained by several factors, including delays in referral to specialized care, the time required to plan intravenous treatment, and the frequent use of first-line non-pharmacological approaches such as rehabilitation and physiotherapy, as bisphosphonates are not formally approved for CRPS-1 in France and considered in second-line therapy. The literature data on the association between symptom onset and response to pamidronate are variable. A network meta-analysis published in 2014 in Pain Medicine reported increased efficacy of pamidronate when the symptom delay was less than 12 months, whereas calcitonin appeared to be more effective when delay exceeded 12 months³. Furthermore, a retrospective study by Varenna et al. involving 194 patients identified the duration of symptoms as a predictive factor of response to bisphosphonates (clodronate, pamidronate, neridronate), with the response rate progressively decreasing and reaching zero beyond 12 to 24 months after the onset of symptoms¹⁹. Additionally, a discontinued clinical trial involving 57 patients concluded that neridronate was ineffective for symptom durations of up to 2 years²⁰. Conversely, some studies have shown efficacy with longer symptom durations, despite the absence of a control group. For example, the study by Cortet et al. (1997²¹), reported efficacy of pamidronate in patients with a mean duration of symptoms of 15 months. Similarly, Maillefert et al. (1995²²), observed a favorable response in patients with a minimum of 6 months of symptoms. Finally, Kubalek et al. (2001²³), described efficacy with a mean duration of symptoms of 42 months. The underlying pathophysiological hypothesis is that treatment efficacy would be optimal when scintigraphy shows fixation at an early stage, which could indicate a higher local concentration of the bisphosphonate²⁴. In our study, due to its retrospective nature, bone scintigraphy was not systematically performed, as the diagnosis of CRPS-1 is based on the Budapest criteria, which are exclusively clinical. As a result, imaging data were not consistently available for analysis. Moreover, an early stage of the disease would be favorable, because neuropeptides and neurogenic inflammation, targets of bisphosphonates, only influence CRPS at the beginning of the onset of symptoms^25,26. These data highlight a heterogeneity of results, probably linked to the characteristics of the populations studied and the methodologies used, thus highlighting the need for additional analyses to better understand the impact of the delay of symptoms on the response to treatment.

In our study, no significant association was observed between the mean dose of pamidronate administered and the response to treatment. However, the literature reports an efficacy of pamidronate at various dosages: 60 mg/day for 3 days in the study of Kubalek et al.²³, 1 mg/kg/day for 1 to 3 days in the study by Cortet et al.²¹, 30 mg/day for 3 days in the study by Maillefert et al.²², cumulative dose of 180 mg IV in the study by Young et al.¹² and a single infusion of 60 mg in the study by Robinson et al.¹¹. In our retrospective real-life cohort, the dosage and number of pamidronate infusions varied among patients, as treatment decisions were left to the physician’s discretion based on clinical judgment and usual practice in the absence of specific guidelines. To our knowledge, this study is the only one to include a repeated infusion at one month in cases of persistent pain and/or functional impairment, reflecting real-life management strategies. To assess whether this variability may have influenced outcomes, we specifically analyzed the potential impact of cumulative dose and number of infusions on treatment response. Our data showed that neither cumulative dose nor additional infusions beyond the initial administration were associated with improved clinical outcomes. Interestingly, a greater number of infusions was associated with poorer response, probably reflecting more severe or refractory cases rather than a causal relationship. These findings suggest that a single dose of 60 to 90 mg pamidronate may be sufficient, and that increasing the dose or repeating infusions does not appear to enhance efficacy.

To our knowledge, no previous studies have explored the relationship between bone mineral density (BMD) and response to pamidronate treatment. This hypothesis is based on the mechanism of action of pamidronate, which acts by inhibiting osteoclast activity and bone remodeling. The pain observed in complex regional pain syndrome type 1 (CRPS 1) may be related to osteoclast hyperactivity, which is inhibited by bisphosphonates, suggesting a potential link with BMD^4,13,24. However, our analysis did not reveal a significant association between BMD and response to pamidronate treatment.

Other criteria associated with treatment response were post-traumatic or idiopathic (vs. post-surgical) etiology and the presence of initial clinical edema. Regarding etiologies associated with treatment response, this is consistent with the data available in the literature: according to Adami et al., predictive factors for treatment response included the absence of a found etiology, and Varenna et al. identified fractures rather than surgical causes as predictive factors for response^19,27. Bisphosphonates appear to be more effective in treating CRPS-1 of post-traumatic origin compared to post-surgical cases, likely due to the prominent role of local bone remodeling and osteoclastic activity in post-traumatic CRPS, which are the primary targets of bisphosphonate therapy, whereas recovery after surgery is often more challenging due to additional nerve injury and scar tissue factors^1,19. Regarding edema, to our knowledge, no other study has investigated edema as an independent factor in response to treatment. However, many studies have demonstrated a significant reduction in edema after treatment with pamidronate or other bisphosphonates, suggesting a beneficial effect of these treatments on this clinical aspect^10,12,18,27. Beyond clinical edema, it could be interesting in future studies to investigate the potential influence of bone marrow edema in MRI on the efficacy of bisphosphonates in CRPS-1, as bone has a central role in initiating and sustaining the early phases of the disease, and bisphosphonates are known to directly inhibit mononuclear cells within the bone marrow^25,28.

Regarding CRPS-1 localization, in our study, it primarily affected the lower limbs, with a notable predominance in the foot. This distribution contrasts with literature data, where CRPS is most reported in the upper limbs, especially the wrist and hand, for approximately 60% of cases^29,30. This difference can be explained by the characteristics of our study population, which was largely drawn from an orthopedic surgery department, where lower limb injuries and surgical procedures are more prevalent. Furthermore, 10% of patients in our cohort had CRPS-1 at the knee, localization that remains debated. The prevalence of CRPS-1 following total knee arthroplasty has previously been estimated at up to 21% using the Orlando criteria for CRPS-1 diagnosis, which are less specific than the currently accepted Budapest criteria. When using the Budapest criteria, the prevalence of CRPS at the knee appears to be below 1%, or even negligible according to some authors. This reflects the diagnostic complexity in the context of persistent neuropathic pain or vasomotor disturbances frequently encountered postoperatively at the knee, which may lead to potential overdiagnosis of CRPS^31,32.

This study is notable for its relatively large sample size (n = 255), as well as its real-practice approach, including infusion protocols and variable doses of pamidronate. To our knowledge, this is the first study to explore the influence of bone mineral density (BMD) on the response to pamidronate treatment, as well as to analyze the impact of the number of infusions and doses administered on therapeutic efficacy. However, a major limitation lies in its retrospective nature, which may introduce bias in the interpretation of the results.

In conclusion, the response to pamidronate treatment in patients with complex regional pain syndrome (CRPS) appears to be modulated by the etiology of the syndrome and the presence of initial edema. In contrast, increasing the number of infusions or the dose administered does not appear to improve therapeutic efficacy. Furthermore, no significant association was identified between bone mineral density (BMD), time to treatment initiation and response to pamidronate.

Author contributions

M.D. wrote the main manuscript text. All authors reviewed the manuscript.

Data availability

The data supporting the findings of this study are available from the corresponding author upon reasonable request.

Declarations

Conflict of interest

The authors declare no conflicts of interest. No specific funding was received from any bodies in the public, commercial or not-for-profit sectors to carry out the work described in this article.