Acute Toxicity Evaluation of Proliferol: A Dose-Escalating,Placebo-Controlled Study in Swine

Abstract

Prolotherapy is one of the many treatments available for chronic musculoskeletal disorders. A commonly used drug contains dextrose 12.5%, glycerin 12.5%, phenol 1.0%, and lidocaine hydrochloride 0.25% in aqueous solution (recently termed Proliferol). For chronic low back pain, this is injected into lumbosacral ligaments to stimulate connective tissue repair. Despite generally positive clinical results, the toxicity of this drug is not well characterized and was assessed in 48 (24 male, 24 female) Yucatan miniature swine randomly assigned to low (1×), medium (5×), or high (10×) dose or saline placebo. Outcomes included clinical observations, clinical chemistry, hematology, coagulation, urinalysis, toxicokinetics, and full gross and microscopic histopathology after 24 hours or 14 days. Findings attributable to Proliferol after 24 hours included dose-response elevations in alanine aminotransferase, aspartate aminotransferase, lactate dehydrogenase, and creatine kinase, which returned to normal after 14 days. There were no remarkable findings in hematology, coagulation, or urinalysis. Urine concentrations of lidocaine and phenol both peaked after 8 hours. Histopathology findings after 24 hours included hemorrhage, inflammation, necrosis, and vascular changes in the ligaments and adjacent soft tissues at the sites of injection. After 14 days, there was evidence of repair under way, with fibrosis and skeletal muscle regeneration at the injection sites.

Prolotherapy is a treatment used to treat a variety of musculoskeletal disorders. ¹ This intervention requires multiple injections of various drugs administered at bone-ligament or bone-tendon junctions. The drugs used in prolotherapy vary, but many share the same proposed mechanisms of action. ² Commonly used ingredients such as hypertonic (eg, 10%–25%) dextrose and glycerin, as well as low concentrations of phenol (eg, 0.5%–1.5%), are thought to provoke a localized acute inflammatory response at the injection site. ³ This response is purported to initiate an innate healing response, resulting in an increased production of fibroblasts, leading to collagen production, repairing injured ligaments or tendons, and improving symptoms from related musculoskeletal disorders. ³ One of the most common indications for prolotherapy is chronic low back pain (CLBP) due to suspected ligament or connective tissue injury. ¹

For CLBP, a frequently used drug in prolotherapy consists of dextrose 12.5%, glycerin 12.5%, phenol 1.0%, and lidocaine hydrochloride 0.25%, recently termed Proliferol. ⁴ A local anesthetic is added to minimize the discomfort associated with these injections and to provide temporary relief of symptoms. Similar drugs have been purchased from compound pharmacies or prepared by physicians and used clinically in the United States since the 1950s. Several observational studies have reported generally positive results when using such drug solutions for CLBP, but evidence from randomized controlled trials (RCTs) is conflicting due to clinical heterogeneity and methodological weakness among the study designs. ^5,6

Prior to conducting clinical trials in an attempt to clarify the uncertainty surrounding this intervention for CLBP, it was necessary to provide evidence related to the safety of this drug to regulatory agencies in the United States. Despite prolonged clinical use in humans without any apparent drug-related serious adverse events (SAEs) or drug-related toxicity, it was determined that the acute local and systemic toxicity of this drug should be formally assessed in 2 animal species according to related International Conference on Harmonisation (ICH) guidelines. ² Furthermore, it was suggested that such preclinical studies should attempt to mimic the current clinical use of this drug in humans. The primary purpose of this study was therefore to assess the acute local and systemic toxicity of Proliferol in swine following injection into the lumbar spine and sacroiliac joints. The secondary purpose was to assess the reversibility of any observed changes following a 14-day recovery period.

Materials and Methods

Study Design

A dose-escalating, placebo-controlled acute local and systemic toxicity study was conducted at Charles River Laboratories, Preclinical Services Arkansas (Redfield, Arkansas), which is registered with the US Food and Drug Administration (FDA). The study was approved by the facility’s Institutional Animal Care and Use Committee (IACUC).

Test Animals and Study Groups

Test animals were 4- to 7-month-old Yucatan minipigs (Sinclair Research Center, Auxvasse, Missouri), identified with ear tags, examined, quarantined, and reexamined by the study veterinarian prior to day 1. This test animal was chosen as a counterpart to the rat used in a previous acute toxicity study to comply with ICH guidelines requiring evaluation in 1 rodent and 1 nonrodent species. Rats were deemed too small to receive spinal intraligamentous injections and were assessed using spinal intramuscular injections. ⁷ To mimic the method of administration used with similar drug solutions in humans, it was necessary to select a nonrodent species with a large enough spine to allow injection into specific ligaments and connective tissues in the lumbosacral region. The Yucatan minipig was chosen based on a successful pilot study in which intraligamentous injections were successfully administered by a physician familiar with this treatment in humans. ⁴ A total of 48 animals (24 male, 24 female) were used in the study based on industry conventions for the type of study, size of animal, and discussions with the FDA. Animals were randomly assigned to study groups based on body weight using a computer-generated sequence to ensure similar (ie, ±10%) mean body weights among groups.

Animal Care and Handling

Study animals were housed individually in chain-link runs with vinyl-coated elevated grating, fed Purina Certified Pig Diet (St Louis, Missouri) twice daily (200 g for animals <25 kg, 350 g for animals >25 kg), and provided with filtered tap water ad libitum. Environmental controls were set at 16°C to 27°C (61°F to 81°F), with a relative humidity of 30% to 70%, 10 air changes per hour, and a 12-hour light/dark cycle.

Drug Preparation and Testing

The drug used in this study was Proliferol, which contains dextrose 12.5%, glycerin 12.5%, phenol 1.0%, and lidocaine hydrochloride 0.25% in sterile aqueous solution. A single batch of this drug was prepared according to current Good Manufacturing Practices (cGMP) by a registered manufacturer (Bell-More Labs, Hampstead, Maryland). Batch release testing was performed by a registered analytical laboratory (ACTA Laboratories, Foothill Ranch, California) and included drug ingredient assays, bacterial endotoxin, sterility, particulate matter, and degradants (ie, 5-hydoxymethylfufural and 2,6-dimethylalanine); methods were adapted from the United States Pharmacopeia (USP). ⁸

Study Groups

Clinical studies using drugs similar to Proliferol for CLBP have reported positive results using doses of 20 mL, which was selected as the low-dose reference point for this study. ¹ Doses for study animals were calculated based on a drug solution volume to body weight basis (ie, mL/kg). Based on a typical human body weight of 80 kg, the low dose was therefore 0.25 mL/kg, the medium dose was set at 5× the low dose (1.25 mL/kg, equivalent to 100 mL in humans), and the high dose was set at 10× the low dose (2.5 mL/kg, equivalent to 200 mL in humans). The control group received the same volume of physiologic saline (0.9% NaCl) as the high-dose group (2.5 mL/kg) in an attempt to distinguish injection effects (eg, needlestick trauma) from drug effects (eg, drug-related toxicity). The 4 study groups are outlined in Table 1.

Drug Administration

Animals were fasted overnight. The lumbosacral region was clipped and disinfected with alcohol and betadyne. Surgical anesthesia was achieved with intramuscular Atropine, Telazol, and Acepromazine, followed by isoflurane gas. Injections were given by a medical physician experienced with this intervention in humans. The targeted injection sites included bilateral lumbar interspinous ligaments (ie, from the inferior surface of the spinous process above to the superior surface of the spinous process below), supraspinous ligaments (ie, along the posterior surface of several adjacent spinous processes), and the iliolumbar ligaments (ie, from the transverse processes of the lower lumbar vertebrae to the superior surface of the posterior iliac crest), as well as the posterior intra-articular facet joint capsules, bilateral sacroiliac joints and ligaments, and the sacrotuberous ligaments.

Necropsy

Half of the animals in each study group (3 males, 3 females) were sacrificed after 24 hours, and half were sacrificed following a 14-day recovery period. These time points were selected based on a pilot study and discussions with the FDA. ⁴ Animals were euthanized via Telazol followed by sodium pentobarbital overdose and exsanguination.

Outcome Measures

Acute and local systemic toxicity were assessed with a variety of outcome measures, and with 3 exceptions, analysis of all specimens was performed at the testing facility. Blood samples for haptoglobin were analyzed at Antech Diagnostics (Morrisville, North Carolina), urine samples for toxicokinetics were analyzed at Pacific Toxicology Laboratories (Chatsworth, California), and bone marrow smears were analyzed at Can-Bio Pharma Consulting (Rockwood, Ontario). Reference ranges were provided by the testing laboratory and previously reported normative data. ^9,10

Clinical observations

Daily observations were made for mortality, moribundity, and body temperature; weekly observations were made for body weight. Study technicians were advised to note symptoms related to abnormal gait or overt pain behavior. Ophthalmic examinations, including macroscopic examination of the anterior portion and indirect ophthalmoscopy of the posterior segment, were conducted before drug administration and prior to necropsy.

Clinical chemistry

Blood samples were obtained via the anterior vena cava before drug administration and prior to necropsy. The following clinical chemistry parameters were evaluated: total protein, albumin, globulin, albumin/globulin ratio, glucose, cholesterol, triglycerides, total bilirubin, direct bilirubin, indirect bilirubin, urea nitrogen, creatinine, creatine kinase (CK), alanine aminotransferase (ALT), aspartate aminotransferase (AST), alkaline phosphatase, gamma-glutamyltransferase (GGT), lactate dehydrogenase (LDH), calcium (Ca), phosphorus (P), sodium (Na), potassium (K), and chloride (Cl).

Hematology and coagulation

The following hematology and coagulation parameters were evaluated: red blood cell (RBC) count, hemoglobin concentration, hematocrit, mean corpuscular volume (MCV), mean corpuscular hemoglobin (MCH), RBC morphology, platelet count, white blood cell (WBC) count, neutrophil count, lymphocyte count, monocyte count, eosinophil count, basophil count, activated partial thromboplastin time (aPTT), prothrombin time (PTT), and other cells (as appropriate).

Urinalysis

Samples for routine urinalysis were collected overnight before drug administration and prior to necropsy with a cage pan catch, withholding feed during the collection periods. Urine samples were evaluated for the following parameters: pH, protein, glucose, ketone, bilirubin, urobilinogen, leukocytes, specific gravity, color, clarity, occult blood, nitrite, volume, and microscopic examination of sediment.

Toxicokinetics

To provide an initial assessment of toxicokinetics, we also collected additional urine samples with a cage pan catch for a 2-hour period prior to each of the following time points after drug administration: 2 hours, 8 hours, and 24 hours. Those samples were analyzed for phenol with high-pressure liquid chromatography (HPLC) and lidocaine with gas chromatography flame ionization detection (GC-FID) concentration using methods validated by the testing facility.

Gross pathology and histopathology

Organ weights were recorded for the adrenal glands, brain, epididymis, heart, kidneys, liver, lungs, ovary, spleen, testes, thymus, thyroid gland, and uterus; paired organs were weighed together. The following tissues were inspected visually prior to being fixed: adrenal gland, administration site 1 (needle insertion point with surrounding skin, subcutis, and muscle at left sacroiliac joint), administration site 2 (needle insertion point with surrounding skin and subcutis at L5–L6), animal identification, aorta, bone (femur), bone (sternum), bone marrow (sternum), brain (brain stem, cerebellum, cerebrum), cervix, epididymis, esophagus, eye, gallbladder, gross lesions, heart, interspinous and supraspinous ligaments (L4–L5 and L5–L6), intestine (cecum, colon, rectum, duodenum, ileum, jejunum), kidney, lacrimal gland, liver, lung, lymph node (mandibular, mesenteric, bronchial), mammary gland (inguinal), nasal cavities (one middle section examined histopathologically), optic nerve, sciatic nerve, ovary, pancreas, pituitary gland, prostate gland, salivary gland (mandibular, unilateral), sacroiliac joint and immediate posterior (dorsal) joint ligament (left side), seminal vesicle, skeletal muscle, skin (inguinal), spinal cord (cervical, lumbar, thoracic), spleen, stomach (cardiac, fundic, pyloric), testes, thymus, thyroid gland, tongue, trachea, urinary bladder, uterus (horns and body), and vagina. Tissues were stained with hematoxylin and eosin prior to microscopic histopathology by a board-certified veterinary pathologist. Two rib bone marrow smears were also prepared.

Statistical Analysis

Statistical analysis was performed on body weight, body temperature, hematology, coagulation, clinical chemistry, organ weight, and toxicokinetics; a minimum of 3 animals per sex per group was required for analysis at each time point. Continuous variables were analyzed by comparing means among each of the 4 study groups using a 1-way analysis of variance (ANOVA) and Dunnett’s t test for significant results (P < .05) for parametric distributions or a Kruskal-Wallis test and Dunn’s test for significant results (P < .05) for nonparametric distributions. Repeated-measures ANOVA was used to analyze the toxicokinetics results at multiple time points.

Results

Clinical Observations

There were no deaths prior to necropsy, and no unscheduled euthanasia was required. No significant differences were noted between the groups with respect to body weight (mean change from predose to day 14 was 0.0 kg, 0.2 kg, −0.2 kg, and 0.4 kg for groups 1, 2, 3, and 4, respectively) or body temperature (mean decrease from predose to day 14 was 1.0°C, 2.0°C, 1.8°C, and 1.5°C for groups 1, 2, 3, and 4, respectively). There were no findings of interest noted during the ophthalmology examination.

Clinical Chemistry

Results for clinical chemistry are presented in Table 2. Findings of interest after 24 hours that were attributable to the drug included apparent dose-response elevations in liver function tests (LFTs), including ALT, AST, CK, and LDH. All clinical chemistry parameters had returned to within normal limits after 14 days.

Hematology and Coagulation

Results for hematology and coagulation are not presented as all findings were within normal limits, and there were no statistically significant differences in these parameters between the 4 groups after 24 hours or 14 days.

Urinalysis

All findings for urinalysis were within normal limits, and there were no statistically significant differences in these parameters between the 4 groups after 24 hours or 14 days.

Toxicokinetics

Urine phenol and lidocaine concentrations predose and after 2 hours, 8 hours, and 24 hours are shown in Figures 1 and 2. Results should be interpreted with caution because only 2/6 animals (33.3%) were able to produce a urine sample within 2 hours of drug administration.

Gross Pathology

None of the incidental findings at gross pathology were thought to be attributable to the drug.

Histopathology

Histopathogy test results for each study group at the supraspinous ligament L4–L5 injection site are presented in Table 3. Findings related to nerve fibers at the injection sites are presented in Table 4. Representative slides (2× and 10× magnification) of tissues taken from the supraspinous ligament of L4–L5 after 24 hours (panel A) and 14 days (panel B) in the low-dose and high-dose Proliferol groups are presented in Figures 3 and 4.

Discussion

In a survey of physicians experienced with administering similar drugs in humans for CLBP, the most commonly reported side effects were pain, stiffness, and bruising. ² Although similar symptoms were expected in test animals, especially those administered much larger doses than those used clinically in humans, no such findings were reported. This suggests that if such side effects were present in test animals, their severity was not sufficient to affect gait or trigger overt pain behavior (ie, guarding, vocalization, mutilation, restlessness, sweating, recumbency, depression, abnormal appearance). ^11,12

Findings of elevated ALT, AST, CK, and LDH levels after 24 hours in all Proliferol groups were likely due to a temporary hepatic insult from phenol, which is partially metabolized in the liver. Because there were no related liver findings on histopathology, changes in LFTs were not clinically meaningful. This is corroborated by a lack of reported hepatic AEs in a survey of practitioners using similar drugs in thousands of patients. ² As a precaution, physicians using similar drugs in humans may wish to use smaller doses in patients with known hepatic dysfunction or monitor LFTs and manage observed changes appropriately. ¹³

The toxicokinetics assessment of urinary lidocaine and phenol concentration indicates that these 2 components of Proliferol are both detected in the urine within 2 and 8 hours following drug administration, with only minimal amounts after 24 hours. This suggests the drug is rapidly metabolized and excreted in the urine. The small amounts detected in the placebo control group after 2 hours are possibly due to contamination because no lidocaine was administered to those animals. The difficulty experienced in obtaining voluntary urine samples after 2 hours suggests that this was insufficient recovery time after the drug administration for the animals to urinate freely.

A potential role for each of the drug ingredients in Proliferol, including dextrose, glycerin, phenol, and local anesthetic, was previously postulated. ^3,14 Individually, these ingredients are thought to produce a localized, controlled acute inflammatory response at the injection site; together, they are thought to act synergistically. A pilot study in humans that compared posterior sacroiliac ligament biopsies before and after treatment with prolotherapy using a similar drug reported increased cellularity due to fibroblasts and an increase in the average diameter of the examined fibers. ¹⁵ Although the soft tissue changes observed in the present study may give some cause for concern when taken at face value, they are consistent with the proposed mechanism of action for this drug.

The nerve fiber findings indicate that the acute inflammatory changes observed in surrounding tissues are reflected in adjacent peripheral nerve fibers. Although the clinical relevance is unknown, the neurolytic properties of phenol are likely not involved as no necrosis was detected. ¹⁴ Because intentional analgesic phenol injections are given in much higher concentrations (eg, 6%), it may be possible that lower concentrations are not sufficient to provide long-term pain relief. ¹³ Nevertheless, it is possible that peripheral nerve fiber changes short of necrosis could explain some of the improvements in pain reported by RCTs. ^16,17 Although concerns have been expressed that inadvertent intrathecal injection of this drug would result in serious AEs due to spinal cord injury, 2 case reports resolved without sequelae. ¹⁸

The proposed mechanism of action for prolotherapy suggests that the acute inflammatory changes observed at the injection sites are desired for the subsequent healing and connective tissue repair that is thought to occur. ^1,3,19,20 Given that the severity of changes observed in this study was of most concern at doses much greater than those used clinically, they are not likely to occur during routine clinical use in humans. However, the relationship between the acute inflammatory changes noted, the proposed mechanism of action of this drug by subsequent connective tissue repair, and the purported clinical benefits resulting from such repair are currently tenuous. Additional research is required to confirm or refute this hypothesis.

Overall, findings in this study are consistent with those reported in a previous pilot study in rats and swine that was performed with only 1 animal in each of 2 dose groups examined after 24 hours and 14 days (4 animals/species) with no control groups. ⁴ Study results are also similar to those reported in an acute toxicity study in rats in which the drug was administered by intramuscular injections given the small size of the test animals. ⁷ Both previous studies reported changes in LFTs after 24 hours that returned to normal after 14 days and varying degrees of inflammation and necrosis at the injection sites. Compared with the pilot study, the current study substantially increased the number of test animals from 4 to 48, had 3 dose groups rather than 2, increased the highest dose group from 5× to 10× low dose to maximize potential toxicity, added a saline control group to evaluate the effects of an invasive drug administration process, and monitored additional LFTs and acute inflammation markers to further define previously reported results. And unlike the previous study in rats where the drug was administered by intramuscular injection, the current study mimicked the drug administration used clinically in humans by intraligamentous injections in the lumbosacral area.

Collectively, these 3 studies suggest that the acute systemic toxicity of Proliferol is limited to temporary changes in LFTs due to hepatic insult, along with localized acute inflammatory changes at the injection sites. Findings noted in all 3 studies followed an apparent dose-response relationship, suggesting that the low dose tested, equivalent to the currently used clinical dose of 20 mL in humans for CLBP, is well tolerated and not associated with any clinically meaningful acute systemic or local toxicity.

Conclusion

Proliferol injections into the lumbosacral spine of Yucatan mini-swine at volumes of 0.25 to 2.5 mL/kg, equivalent to 20 to 200 mL in an 80-kg human on a drug volume to body weight basis, resulted in transient changes in LFTs after 24 hours, which returned to normal after 14 days. Injections of Proliferol also produced hemorrhage, inflammation, necrosis, and vascular changes in local tissues after 24 hours, with evidence of repair under way after 14 days. Although these findings are consistent with the proposed mechanism of action for this drug through the acute inflammatory connective tissue wound-healing cascade, their clinical relevance is uncertain. Clinical studies attempting to overcome the methodological limitations of previous studies using similar drugs are required to determine whether these types of injections are effective in the management of musculoskeletal disorders such as CLBP.