Doi:10.1016/j.spinee.2007.10.02

Evidence-informed management of chronic low Scott Haldeman, DC, MD, PhD,Joanne Borg-Stein, , aDivision of Orthopaedic Surgery and Department of Epidemiology and Community Medicine, Faculty of Medicine, University of Ottawa, Ottawa, Canada cUS Spine & Sport Foundation, San Diego, CA, USA dSchool of Physical Therapy and Rehabilitation Sciences, College of Medicine, University of South Florida, Tampa, FL, USA eDepartment of Neurology, University of California, Irvine, CA, USA fDepartment of Epidemiology, University of California, LA, USA gResearch Division, Southern California University of Health Sciences, CA, USA hSpaulding Rehabilitation Hospital, Boston, MA, USA iNewton-Wellesley Hospital Spine Center, Newton, MA, USA jDepartment of Physical Medicine and Rehabilitation, Harvard Medical School, Boston, MA, USA Received 3 October 2007; accepted 13 October 2007 EDITORS’ PREFACE: The management of chronic low back pain (CLBP) has proven to be verychallenging in North America, as evidenced by its mounting socioeconomic burden. Choosing amongavailable nonsurgical therapies can be overwhelming for many stakeholders, including patients, healthproviders, policy makers, and third-party payers. Although all parties share a common goal and wish touse limited health-care resources to support interventions most likely to result in clinically meaningfulimprovements, there is often uncertainty about the most appropriate intervention for a particular pa-tient. To help understand and evaluate the various commonly used nonsurgical approaches to CLBP,the North American Spine Society has sponsored this special focus issue of The Spine Journal, titledEvidence-Informed Management of Chronic Low Back Pain Without Surgery. Articles in this specialfocus issue were contributed by leading spine practitioners and researchers, who were invited to sum-marize the best available evidence for a particular intervention and encouraged to make this informa-tion accessible to nonexperts. Each of the articles contains five sections (description, theory, evidenceof efficacy, harms, and summary) with common subheadings to facilitate comparison across the 24different interventions profiled in this special focus issue, blending narrative and systematic reviewmethodology as deemed appropriate by the authors. It is hoped that articles in this special focus issuewill be informative and aid in decision making for the many stakeholders evaluating nonsurgicalinterventions for CLBP. Ó 2008 Elsevier Inc. All rights reserved.
Chronic low back pain; Prolotherapy; Intraligamentous injection FDA device/drug status: investigational or not approved for this indica- Intraligamentous injection of solutions aimed at promot- ing connective tissue repair is commonly known as prolother- This study was funded by the National Institutes of Health (grant no.
1RY3AR 046342-01) which is a federal/state agency, and CAM Research apy, which has been defined as ‘‘the rehabilitation of an Institute, which is a nonprofit organization.
incompetent structure (such as a ligament or tendon) by the The authors SD and JM are consultants of CAM Research Institute; induced proliferation of new cells’’ Common synonyms JBS is a Consultant of Inflexxion Corporation.
for this therapy include regenerative injection therapy, * Corresponding author. 25 Corona Ave., Rockcliffe, ON, K1M 1K8, growth factor stimulation injection, nonsurgical tendon, lig- Canada. Tel.: (949) 466-8132; fax: (949) 266-8951.
ament and joint reconstruction, proliferant injection, prolo, 1529-9430/08/$ – see front matter Ó 2008 Elsevier Inc. All rights reserved.
doi:10.1016/j.spinee.2007.10.021 S. Dagenais et al. / The Spine Journal 8 (2008) 203–212 and joint sclerotherapy . Proponents of this intervention rationale provided by proponents of this latter method is generally dislike the term sclerotherapydwhich is associ- that it minimizes the inflammatory reaction and is therefore ated with the formation of unorganized scar tissuedrather than the more beneficial, organized connective tissue that It should be noted that these methods of prolotherapy were developed clinically based largely on anecdotal evi-dence and practitioner preference, many of whom eventu- ally modify these methods further and tailor treatments toindividual patient needs.
Prolotherapy has been used to treat chronic low back pain (CLBP) for over 60 years and was originally adap-ted from sclerotherapy, which involves injection of irritant solutions to induce acute inflammation, stimulate connec- The treatment procedure for prolotherapy varies a great tive tissue growth, and promote formation of collagen tissue deal among practitioners. The treatment begins with the pa- Sclerotherapy was commonly used to close the lumen tient gowned and lying prone on a treatment table. Through in varicose veins, and in nonsurgical abdominal hernia re- manual palpation, the physician first identifiesdand possi- pair . Based on the hypotheses that joint hypermobility bly marks with a pendlandmarks in the lumbosacral spine could be attributed to incomplete connective tissue repair area such as the iliac crest, sacroiliac joints, and interverte- after an injury, this treatment approach was later applied bral spaces to use as anatomical reference points for the to chronic musculoskeletal conditions suspected of being injections. The skin is then cleaned with alcohol and beta- related to ligament or connective tissue laxity. The use of dyne to minimize the risk of infection. Some practitioners prolotherapy for CLBP was promoted in the 1950s by a gen- begin the treatment by injecting subcutaneous wheals of eral surgeon named George S. Hackett, who published large local anesthetic into the areas to be injected with prolother- case series claiming very high rates of success for a condi- apy solution to minimize patient discomfort .
tion that had few valuable surgical options at that time.
Prolotherapy injections are often administered with 2.5 in., 20-gauge spinal needles to deliver a small bolus of solution into the following areas: posterior sacroiliac lig- A survey of 908 primary care patients receiving opioids aments, iliolumbar ligaments, interspinous ligaments, supraspinous ligaments, and posterior intervertebral facet reported that 8.3% had used prolotherapy in their lifetime, capsules. Some practitioners target ligaments that are ten- and 5.9% had used it in the past year; those who reported der to manual palpation or otherwise suspected of causing using prolotherapy did do an average of 3.5 times in the pain , whereas others prefer to inject all of the larger posterior ligaments that are accessible and possibly associ-ated with CLBP To access these ligaments, the needleis typically inserted in the midline directly above an inter- vertebral space and oriented laterally to avoid accidentally Although there are no formal subtypes of prolotherapy, injecting into the spinal canal. The needle is then inserted there is substantial heterogeneity in the treatment protocols until the tip contacts bone and the plunger is partially with- used by different practitioners. The approaches to prolo- drawn to confirm the absence of blood in the aspirate that therapy generally differ according to the type of solution would indicate possible blood vessel puncture. The desired injected, the volume and frequency of injections, and use bolus of solutiondtypically 0.5 to 2.0 mldis then injected at each site Practitioners often target several ligaments The method used in studies by Ongley et al. typi- from a single needle insertion point by reorienting the cally involves six weekly injections of 20 to 30 ml of a solu- needle in situ. The total amount of solution injected during tion containing dextrose 12.5%, glycerin 12.5%, phenol 1%, one session of prolotherapy depends on the number of and lidocaine 0.25% into multiple preselected lumbosacral structures that are targeted and the bolus delivered to each ligaments. Injections are usually accompanied by spinal ma- structure; 10 to 30 ml is common. Radiological guidance nipulation and instructions for the patient to perform repeated (eg, fluoroscopy) is seldom used in prolotherapy.
standing lumbar flexion-extension exercises for several After the injection procedure, the patient is briefly ob- weeks. This approach may also use corticosteroid injections served and then sent home with instructions to perform re- into tender lumbosacral areas before the first session of petitive spinal range of motion (ROM) exercises such as standing lumbar flexion and extension to maintain their Prolotherapy methods used in other studies involved flexibility as the acute inflammation produced by the injec- a greater number of injections at longer intervals tions results in lumbosacral stiffness and soreness over the (biweekly-monthly) with a smaller gauge needle, a lower next few days. Patients are also advised to use over-the- injected volume (10–20 ml), and a solution containing only counter (eg, acetaminophen) or prescription (eg, hydroco- dextrose 10% to 20% and lidocaine 0.2% to 0.5% The done) analgesics as needed for any increased pain in the S. Dagenais et al. / The Spine Journal 8 (2008) 203–212 few days after the injections. Instructions are also given to current procedural terminology manual. Although it has avoid anti-inflammatory medication, because the acute in- not been determined whether it is acceptable to bill under flammatory reaction provoked by this treatment is consid- these codes, practitioners occasionally use related injection ered beneficial . This injection procedure is typically codes to describe various aspects of this procedure, such as: repeated weekly, biweekly, or monthly until six to eight 20550 (injectiondsingle tendon sheath, or ligament), treatments have been administered. Treatment intervals 20551 (injectiondsingle tendon origin/insertion), 20552 are dictated by patient response to treatment and the recov- (injectiondsingle or multiple trigger point), 27096 (injec- ery time required for any resulting acute inflammation to tiondprocedure for sacroiliac joint), 64776 (injectiond diminish before the next injection. Patients may also re- anesthetic agent and/or steroid, paravertebral facet joint or quire periodic injections if pain relapses or reinjury occurs.
facet joint nerve), 99070 (supplies and materialsdeg, listdrugs, trays, supplies, or materials). The health care proce-dural coding system specifically for prolotherapy is M0076, which is not recognized for Medicare purposes. The cost of Prolotherapy is mainly administered by medical or oste- the procedure varies considerably by practitioner. A single opathic physicians, though in rare instances a physician’s treatment typically costs $250 to $500 excluding any re- assistant, nurse practitioner, or naturopath can administer quired diagnostic testing, imaging, rehabilitative equipment, injections where state licensure permits deep injections or treatment facility fees. A series of six treatments is there- Practitioners who perform this treatment are expected fore approximately $1,500 to $3,000. In the United States, to have advanced knowledge of spinal anatomy, including this treatment is not directly covered by major medical in- the attachment points for all connective tissue structures surance companies and is not generally covered by workers’ that may be injected, and the locations of any surrounding compensation insurance. In some cases, it may be covered blood vessels or nerves that may be inadvertently injected.
by automobile insurance medical payment riders.
Extensive experience with spinal injections is usually A survey of patients with chronic pain (including CLBP) recommended before learning prolotherapy and many who had used prolotherapy reported that insurance coverage clinicians who perform these treatments have specialized paid part of the cost in 88% of cases . Patients had paid training in anesthesiology, pain management, or physical some cost out of pocket in 19% of cases; mean out-of-pocket medicine and rehabilitation Additional postgraduate costs in the past year for prolotherapy were $365 .
training specifically in prolotherapy is typically offeredthrough continuing medical education courses sponsored by related organizations and medical schools, includingthe University of Wisconsin. Organizations involved in pro- Although individual ingredients such as dextrose and li- lotherapy training include the American Association of docaine are approved for injection by the FDA, they are not Orthopaedic Medicine (AAOM) , American College approved for this indication. Drug solutions injected during prolotherapy are typically prepared by compound pharma- (ACOSPM) Hackett Hemwall Foundation , and cies or individual practitioners, and thus not subject to reg- American Institute of Prolotherapy .
In general, this treatment is performed in private clinics. It may occasionally be performed in a medical imaging facilityif radiological guidance is required. A minority of practi- tioners prefer to administer oral or intravenous (IV) sedation before treatment to calm the patients and facilitate an other-wise uncomfortable procedure; this requires an ambulatory As with many other treatments available for CLBP, the surgical center. This procedure is rarely performed in hospi- mechanism of action for prolotherapy is not well under- tals. Prophylactic oral analgesics may also be administered stood. This treatment was derived from sclerotherapy for before the treatment instead of, or in addition to, sedation.
varicose veins, where injected irritants provoke a localized Based on membership in related associations, it is estimated acute inflammatory reaction, connective tissue prolifera- that there are approximately 500 to 1,000 practitioners offer- tion, and lumen closure. In prolotherapy, four types of so- ing this treatment in private practices in various cities lutions have been identified according to their suspected throughout the United States. Groups such as the AAOM mechanism of action: 1) osmotic (eg, hypertonic dextrose); and ACOSPM maintain on-line membership directories with 2) irritant/hapten (eg, phenol); 3) particulate (eg, pumice contact information of practitioners who offer this treatment.
flour); and 4) chemotactic (eg, sodium morrhuate) Os-motics are thought to dehydrate cells, leading to cell lysisand release of cellular fragments, which attracts granulo- cytes and macrophages; dextrose may also cause direct gly- There is no current procedural terminology code specif- cosylation of cellular proteins Irritants possess ically for this procedure listed in the latest edition of the phenolic hydroxyl group that are believed to alkylate S. Dagenais et al. / The Spine Journal 8 (2008) 203–212 surface proteins, which renders them antigenic or damages periods). The identity of anatomic structures responsible for them directly, attracting granulocytes and macrophages nociception may be confirmed by observing temporary pain . Particulates are believed to attract macrophages, lead- relief after local anesthetic injections, although the pre- ing to phagocytosis . Chemotactics are structurally re- sumption that these findings can be used to identify struc- lated to inflammatory mediators such as prostaglandins, tures to be injected with prolotherapy has yet to be leukotrienes, and thromboxanes, and are believed to validated. As is often the case with other interventions, the indication for prolotherapy may be CLBP that has failed Other mechanisms of action for prolotherapy that do not to respond to other, more conservative treatments.
involve localized acute inflammation have also been pro- This treatment is considered to be contraindicated in pa- posed. For instance, macrophages that respond to particulates tients with non-musculoskeletal pain (eg, referred visceral are believed to secrete polypeptide growth factor, leading to pain), metastatic cancer, systemic inflammation, spinal an- fibroplasia Other possibilities include neurolysis of no- atomical defects that preclude deep injections (eg, spina bi- ciceptive fibers (denervation) because of the presence of phe- fida), morbid obesity, inability to perform posttreatment nol, lysis of connective tissue adhesions because of the ROM exercises, bleeding disorders, low pain threshold, volume of solution injected, and neovasculogenesis or neo- chemical dependency, or whole body pain In addition, neurogenesis during the inflammatory cascade .
there is some indication from preclinical studies that very A number of studies have been conducted in animals to high doses of a prolotherapy solution containing dextrose elucidate the mechanism of action for solutions used in pro- 12.5%, glycerin 12.5%, phenol 1.0%, and lidocaine 0.25% lotherapy, including rat, guinea pig, and rabbit models .
may produce a temporary increase in hepatic enzymes such In general, these studies involved ligament or tendon injec- as alanine transaminase and aspartate aminotransferase tions and histological or biomechanical examination of Although these findings are preliminary, it may be prudent to connective tissue. For example, rabbit medial collateral lig- not administer high doses of these solutions to patients with ament mass, thickness, and weight-to-length ratio increased significantly 7 weeks after 5% sodium morrhuate injections The ideal patient for this treatment is generally consid- . However, there were no differences in traumatized rat ereddpurely on the basis of clinical opiniondan otherwise Achilles tendon tensile strength 7 weeks after injection of healthy adult aged 30 to 50 years with mechanical CLBP, various solutions used in prolotherapy (eg, dextrose, sodium no serious comorbidities, no psychopathology, signs and morrhuate) Other studies reported changes consistent symptoms of lumbosacral ligament or tendon involvement, with local acute inflammation . In three patients with confirmation of pain structures by local anesthetic, and CLBP of suspected ligamentous origin, biopsies of posterior a positive but temporary response to manual therapy.
sacroiliac ligaments were taken 3 months after 6 weekly in-jections with dextrose 12.5%, glycerin 12.5%, phenol1.25%, lidocaine 0.25%, spinal manipulation, and repeated trunk flexion exercises . Electron microscopy revealed a significant increase in cellularity and active fibroblasts,with a 60% increase in average fiber diameter .
A computerized search of Medline, Embase, and CI- NAHL was performed by combining the Cochrane BackReview group strategy for identifying controlled trials related to low back pain (LBP) with intervention specific Other than a thorough history and physical examination index terms and free text: or Glucose/ad, tu, th [Administra- to rule out the possibility of serious pathology related to tion & Dosage, Therapeutic Use, Therapy], or Glycerol/ad, CLBP, no advanced diagnostic testing is required before tu, th [Administration & Dosage, Therapeutic Use, Ther- prolotherapy. Plain film radiography is often used to evalu- apy], or Phenol/ad, tu [Administration & Dosage, Thera- ate spinal anatomy and pathology before injections.
peutic Use], or Lidocaine/ad, tu, th [Administration &Dosage, Therapeutic Use, Therapy], or Sclerosing Solu-tions/ad, tu, th [Administration & Dosage, Therapeutic Use, Therapy], or prolotherapy/ti, ab. Only articles pub- This treatment is generally used for nonspecific mechan- lished in English from 1997 to 2007 were considered. Ref- ical CLBP resulting from ligament or tendon injury from erences of pertinent articles were also used to identify trauma, repetitive sprain injury, or collagen deficiency studies. Search results were combined and screened for el- . It is challenging to identify this subgroup of CLBP pa- igibility by two independent reviewers (JM, JBS); conflicts tients, as there are no direct, noninvasive methods of assess- were resolved by a third reviewer (SD). Studies were eligi- ing lumbosacral ligament health. Diagnosis is therefore ble if they were clinical practice guidelines, systematic re- made on a clinical basis according to pain referral patterns, views, or randomized controlled trials (RCTs) pertaining to superficial ligament palpation, joint palpation, or history prolotherapy for CLBP (longer than 3 months) and reported (eg, pain aggravated by maintaining a position for extended clinically relevant outcomes such as pain or function.
Table 1Systematic reviews of prolotherapy for CLBP (Wilkinson,unpublished)21 observational studies Ongley, 1987; Klein, 1993; Dechow, Mathews, 1987; Ongley,1987; Klein, 1993; Mathews, 1987; Ongley, 1987; Klein, 1993; Observational studies reported subjective Protocols very differentand results cannot be Observational studies provided few details, had Possible dose-response relationship because poor methodology, and were poorly reported two studies with negative results administered only 3Â10 ml compared with three studies RCTs had different protocols and could not be Two studies with cointerventions had positive Two RCTs had positive results with 6Â20–30 Three studies with prolotherapy alone had Three RCTs had negative results with dextrose because three injections seemedinferior to six Temporary (24–96 h) increase in pain and Other AEs were leg pain, nausea, diarrhea, Dextrose alone likely not beneficial for LBP No evidence of efficacy for prolotherapy alone Evidence of partial prolonged pain relief for Appears safe with newer dextrose/glycerin/ Safety comparable with other spinal injections SCI5science citation index; CBRG5Cochrane back review group; SMT5spinal manipulative therapy; CLBP5chronic low back pain; RCT5randomized controlled trial; LBP5low back pain; S. Dagenais et al. / The Spine Journal 8 (2008) 203–212 was a crossover and it was not possible to attribute reported We identified five reviews related to prolotherapy for results to prolotherapy The details from these five RCTs are given in and briefly discussed below.
it failed to report its methodology and there was no indica- In an RCT by Mathews et al. 22 individuals with tion that a systematic search approach had been followed CLBP were randomly assigned to receive prolotherapy The four included systematic reviews are summarized (n516) or control (n56) injections. Prolotherapy injections consisted of a solution containing dextrose 10%, glycerin A review was published in The Spine Journal by our 10%, phenol 1%, and procaine 0.3%. There were three ses- group in 2005 and identified the same five RCTs that were sions injecting 10 ml every 2 weeks into lumbosacral liga- uncovered in this search . Two of those RCTs had positive ments. Control injections consisted of a solution containing results and noted improvements in pain or disability with six 0.05% procaine, also with three sessions injecting 10 ml ev- weekly sessions injecting 20 to 30 ml of solutions containing ery 2 weeks into tender spots. Outcomes were assessed at 2 weeks and 1, 3, 6, and 12 months and included a six-point therapy (SMT), exercise, and other cointerventions .
numerical rating scale (1–4 designated ‘‘not recovered’’ Three RCTs had negative results and reported improvements and 5–6 designated ‘‘recovered’’), pain intensity (visual an- but no differences between the prolotherapy and control: one alog scale [VAS]), and medication use. At 3 months, 10 out study used only dextrose/lidocaine one used only of 16 patients in the experimental group (63%) were recov- three sessions of dextrose/glycerin/phenol/lidocaine, and ered, compared with 2 out of 6 (33%) in the control group.
one study with a small sample size used dextrose/glycerin/ There were no significant differences between the groups at phenol/procaine without SMT. The review concluded there was no evidence to support prolotherapy with dextrose In an RCT by Ongley et al. 82 individuals with CLBP alone. Furthermore, a dose-response relationship was postu- were randomly assigned to receive prolotherapy (n540) or lated for solutions containing dextrose/glycerin/phenol/ control (n542) injections, along with various cointerven- lidocaine, in which the lower doses (eg, 3Â10 ml) appeared tions. Prolotherapy injections for the experimental group to be less effective than the larger doses (eg, 6Â20 ml).
consisted of a solution containing dextrose 12.5%, glycerin Rabago et al. conducted a review on prolotherapy 12.5%, phenol 1.25%, lidocaine 0.25%. There were six ses- for musculoskeletal pain, including CLBP and identified sions injecting 20 ml every week into lumbosacral ligaments four of the five RCTs mentioned in the previous review.
with IV sedation. Cointerventions included 50 mg triamcin- Two of those studies were generally positive, and two were olone injection into the gluteus medius and SMT for the ex- generally negative. Authors of that review noted that the perimental group, lidocaine 0.5% injection and sham SMT RCTs had poor patient selection and operator differences for the control group, and standing lumbar flexion-extension stretching exercises for both groups. At 6 months, there were Two reviews on prolotherapy injections for CLBP have statistically significant differences (p!.05) for all outcomes been published by the Cochrane Collaboration, one by Yell- in favor of the experimental group. Additionally, 35 out of 40 and et al. in 2004, and an update by Dagenais et al. (88%) patients in the experimental group had greater than in 2007. Study eligibility for both reviews included RCTs 50% improvement in disability score at 6 months, compared and quasi-RCTs on prolotherapy for CLBP, and both reviews with 16 out of 41 (39%) patients in the control group assessed study quality using Cochrane Back Review group (p!.05). There were 15 out of 40 (38%) patients in the exper- criteria. The efficacy results of prolotherapy for CLBP re- imental group versus 4 out of 41 (10%) patients in the control ported in the updated review were as follows: prolother- group with a disability score of 0 at 6 months (p!.05). Radi- apy protocols vary a great deal and results cannot be ating leg pain present at baseline was resolved in 10 out of 12 combined; there is a possible dose-response relationship (83%) patients in the experimental group versus 2 out of 12 with prolotherapy, because negative results were noted in (17%) patients in the control group (p!.05) at 6 months.
two RCTs with lower doses of the administered drug (eg, Given the various cointerventions, however, it is difficult to 3Â10 ml) compared with three studies with higher doses partition the positive outcomes of the experimental group (eg, 6Â20–30 ml); two RCTs with prolotherapy adminis- tered with cointerventions had positive results, whereas three In an RCT by Klein et al. , 79 individuals with RCTs with prolotherapy administered alone had negative re- CLBP were randomly assigned to receive prolotherapy sults. Thus, the authors concluded that there is no evidence of (n539) or control (n540) injections. Prolotherapy injec- efficacy for prolotherapy alone, whereas there is evidence of tions for the experimental group consisted of a solution partial prolonged pain relief for prolotherapy combined with containing dextrose 12.5%, glycerin 12.5%, phenol 1.2%, exercise, SMT, and other interventions.
lidocaine 0.5%. There were six sessions injecting 30 ml ev-ery week into lumbosacral ligaments, facets, and SI joints, along with IV sedation and lidocaine wheals. Control injec- We identified six RCTs related to prolotherapy for tions consisted of a solution containing saline 0.45% and li- docaine 0.25%, along with IV sedation. Outcomes were Table 2Randomized controlled trials of prolotherapy for CLBP Prior spinal surgeryRadiculopathyUnresolved litigation/work comp Experimental: 30 out of 39Control: 21 out of 40 Control: 3.87No SS changes in VAS disability, menstrual irregularities, likelyfrom triamcinolone assessment of toxicology (CBC,sed rate, urinalysis, chem panel,thyroid function) patients (8 Experimental, 5Control) with irritable foci whoreceived triamcinolone reportedSS changes favoring Experimentalgroup for all outcomes S. Dagenais et al. / The Spine Journal 8 (2008) 203–212 assessed at 6 months included the Roland Morris Disability and/or stiffness at the injection site, which is consistent Questionnaire, pain intensity (VAS), pain grid, lumbar with drug’s purported mechanism of action (ie, acute in- ROM, lumbar isometric strength, and the proportion of sub- flammation) In a recent survey of prolotherapy practi- jects with greater than 50% improvement in disability and tioners (n5171) who were members of the AAOM or pain intensity. Both groups had significant improvement ACOSPM, side effects with the highest median estimated (p!.05) in pain intensity, pain grid scores, and disability at prevalence were pain (70%), stiffness (25%), and bruising 6 months. When defining success with treatment as a 50% (5%) following prolotherapy for spinal pain. Other side ef- improvement in pain intensity or disability, 30 out of 39 fects reported in the literature were increased transient leg (77%) patients in the experimental group versus 21 out of pain, headache, nausea, diarrhea, minor allergic reactions, 40 (53%) in the control group were successful (p!.05).
In an RCT by Dechow et al. , 74 individuals with Adverse events (AEs) related to prolotherapy for CLBP CLBP who were referred by a general practitioner were include mainly needlestick injuries similar to those reported randomly assigned to receive prolotherapy (n536) or with other common spinal injection procedures There control (n538) injections. Prolotherapy injections for the are previous reports of severe headache indicative of lumbar experimental group consisted of a solution containing dex- puncture, leg pain with neurological features, disturbed trose 12.5%, glycerin 12.5%, phenol 1.2%, and lidocaine sleep because of psychological trauma from injections, and 0.5%. There were three sessions injecting 10 ml every week severe cough. No fatalities related to this treatment have into L4–L5 ligaments with single needle insertion point, been reported in the literature for a period of almost 50 years along with IV sedation. Control injections consisted of a in which this treatment has been offered Rare AEs in- solution containing saline 0.45% and lidocaine 0.5%.
clude pneumothorax, disc injury, meningitis, hemorrhage, Outcomes were assessed at 1, 3, and 6 months and included and nerve damage. In a recent survey of practitioners the McGill pain questionnaire, ROM (modified Schober (n5171) who had each provided a median of 2,000 prolo- test), modified somatic perception questionnaire, modified therapy treatments for spinal pain, 470 AEs were reported Zung questionnaire, Oswestry Disability Index, pain grid, . Of these 470 AEs, 70 were considered severe: 65 re- and pain intensity (VAS). At each follow-up time point, quired hospitalization and 5 resulted in permanent injury.
there were no significant differences between the treatment The vast majority (80%) of AEs were related to needle in- and control groups for any of the outcomes.
juries rather than drug toxicity. They included spinal head- An RCT by Yelland et al. was designed to concur- ache (n5164), pneumothorax (n5123), nerve damage rently assess the efficacy of prolotherapy and exercise. In (n554), hemorrhage (n527), spinal cord insult (n57), and this study, 110 individuals with CLBP were randomly as- signed to one of four groups: prolotherapy injections with Practitioners often report that the predictors of negative exercise (n528) or without exercise (n526) or control in- outcomes with prolotherapy are similar to other treatments jection with exercise (n527) or without exercise (n529).
for CLBP and include tobacco use, obesity, inability to per- Prolotherapy injections consisted of a solution containing form posttreatment ROM exercises, serious comorbidities, dextrose 20% and lidocaine 0.2%. There were six sessions psychopathology, and an incorrect diagnosis.
injecting a maximum of 30 ml every 2 weeks into lumbosa-cral ligaments. Control injections consisted of a solution containing saline 0.9%. Exercise consisted of standing lum-bar flexion/extension stretches 4Â/day, whereas the non- Prolotherapy is one of a number of treatments recom- exercise groups continued normal activity. Outcomes mended for the treatment of CLBP. It has a prolonged were assessed at 2.5, 4, 6, 12, and 24 months. At each fol- history of use, a reasonable but not proven theoretical basis, low-up time point, there were no significant differences in a low complication rate, and conflicting evidence of effi- any of the outcomes among the groups. In their analyses, cacy. A possible dose-response effect or the combination the authors disregarded assignment to exercise or normal with other interventions such as SMT may explain the con- activity and compared subjects who received prolotherapy flicting results of RCTs. Two of the RCTs in which prolo- with those who received saline injections.
therapy was administered using six weekly injections of 20to 30 ml dextrose/glycerin/phenol/lidocaine with SMT and exercise had positive results, suggesting this particular in- Preclinical studies are currently under way by our group tervention protocol is worth considering for patients with to support a planned Phase 1 clinical trial of a drug solution CLBP who are refractory to other approaches. At this time commonly used in prolotherapy for CLBP.
there is no evidence of efficacy for prolotherapy injectionsalone without cointerventions.
There is sufficient interest and utilization of this proce- dure to warrant further investigation. Future studies are The most common side effect related to prolotherapy is needed to support or refute the positive results obtained a temporary (12–96 hours postinjection) increase in pain in some of the prior RCTs while addressing some of the S. Dagenais et al. / The Spine Journal 8 (2008) 203–212 methodological weaknesses by minimizing differences be- [14] American College of Osteopathic Sclerotherapeutic Pain Manage- tween the intervention and control groups. Other studies ment. What is sclerotherapy? ACOSPM, 2007. Available at: Last accessed November 20, 2007.
are also needed to establish the safety of common prolo- [15] The Hackett Hemwall Foundation. Welcome to the Hackett Hemwall therapy solutions, and determine the optimal dose and num- [16] The American Institute of Prolotherapy. The American Institute of prolotherapy. The American Institute of Prolotherapy, 2007. Avail-able at: Last accessed November [17] Banks AR. A rationale for prolotherapy. J Orthop Med 1991;13:54–9.
[1] Gove P. Webster’s Third New International Dictionary, Unabridged.
[18] Linetsky FS, Miguel R, Torres F. Treatment of cervicothoracic pain Springfield, MA: Merriam-Webster, 2002.
and cervicogenic headaches with regenerative injection therapy. Curr [2] Dagenais S, Haldeman S, Wooley JR. Intraligamentous injection of sclerosing solutions (prolotherapy) for spinal pain: a critical review [19] Liu YK, Tipton CM, Matthes RD, Bedford TG, Maynard JA, of the literature. Spine J 2005;5:310–28.
Walmer HC. An in situ study of the influence of a sclerosing solution [3] Munavalli GS, Weiss RA. Complications of sclerotherapy. Semin in rabbit medial collateral ligaments and its junction strength.
Connect Tissue Res 1983;11:95–102.
[4] Kim SR, Stitik TP, Foye PM, Greenwald BD, Campagnolo DI. Crit- [20] Harrison ME. The biomechanical effects of prolotherapy on trauma- ical review of prolotherapy for osteoarthritis, low back pain, and tized Achilles tendons of male rats. Brigham Young University; 1995.
other musculoskeletal conditions: a physiatric perspective. Am J Phys [21] Klein RG, Dorman TA, Johnson CE. Proliferant injections for low back pain: histologic changes of injected ligaments and objective [5] Gedney EH. Hypermobile joint. Osteopath Prof 1937;4:30–1.
measurements of lumbar spinal mobility before and after treatment.
[6] Hackett GS. Prolotherapy in whiplash and low back pain. Postgrad J Neurol Orthop Med Surg 1989;10:141–4.
[22] Dagenais S, Ogunseitan O, Haldeman S, Wooley JR, Zaldivar F, [7] Tisi PV, Beverley C, Rees A. Injection sclerotherapy for varicose Kim RC. Acute toxicity pilot evaluation of proliferol in rats and veins. Cochrane Database Syst Rev 2006. CD001732.
swine. Int J Toxicol 2006;25:171–81.
[8] Fleming S, Rabago DP, Mundt MP, Fleming MF. CAM therapies [23] Rabago D, Best TM, Beamsley M, Patterson J. A systematic review among primary care patients using opioid therapy for chronic pain.
of prolotherapy for chronic musculoskeletal pain. Clin J Sport Med BMC Complement Altern Med 2007;7:15.
[9] Ongley MJ, Klein RG, Dorman TA, Eek BC, Hubert LJ. A new ap- [24] Dagenais S, Yelland MJ, Del MC, Schoene ML. Prolotherapy proach to the treatment of chronic low back pain. Lancet 1987;2: injections for chronic low-back pain. Cochrane Database Syst Rev [10] Klein RG, Eek BC, DeLong WB, Mooney V. A randomized double- [25] Mathews JA, Mills SB, Jenkins VM, et al. Back pain and sciatica: blind trial of dextrose-glycerin-phenol injections for chronic, low controlled trials of manipulation, traction, sclerosant and epidural back pain. J Spinal Disord 1993;6:23–33.
injections. Br J Rheumatol 1987;26:416–23.
[11] Yelland MJ, Glasziou PP, Bogduk N, Schluter PJ, McKernon M.
[26] Dechow E, Davies RK, Carr AJ, Thompson PW. A randomized, dou- Prolotherapy injections, saline injections, and exercises for chronic ble-blind, placebo-controlled trial of sclerosing injections in patients low-back pain: a randomized trial. Spine 2004;29:9–16.
with chronic low back pain. Rheumatology (Oxford) 1999;38: [12] Dagenais S, Ogunseitan O, Haldeman S, Wooley JR, Newcomb RL.
Side effects and adverse events related to intraligamentous injection [27] Wilkinson HA. Injection therapy for enthesopathies causing axial of sclerosing solutions (prolotherapy) for back and neck pain: spine pain and the ‘‘failed back syndrome’’: a single blinded, ran- a survey of practitioners. Arch Phys Med Rehabil 2006;87:909–13.
domized and cross-over study. Pain Physician 2005;8:167–73.
[13] AAOM. AAOM organizational statement. AAOM, 2007. Available [28] Yelland MJ, Mar C, Pirozzo S, Schoene ML, Vercoe P. Prolotherapy injections for chronic low-back pain. Cochrane Database Syst Rev

Source: http://anesthesiology.queensu.ca/assets/LBP%20Prolotherapy_Dagnais.pdf